Estimation Presentations WHO Blood

Pattern of Blood Usage
Estimate Blood Developed countries Developing countries

Requirements - complex medical and Limited diagnostic & treatment
facilities
Search for a surgical procedures
cardiac, vascular, complications during
Global Standard neuro, transplant pregnancy and childbirth
trauma care severe childhood anaemia,
often resulting from malaria
cancer chemotherapy
Dr Neelam Dhingra or malnutrition
Coordinator haematological
malignancies trauma
Blood Transfusion Safety
conflict, disasters, violence,
WHO, Geneva
road-traffic accidents
Emergency Trauma Care Maternal Mortality

Globally, >530 000 women die each year during pregnancy,
Worldwide, >100 million people sustain
childbirth or in PP period 99% of them in the developing world
injuries each year and >5 million die
from violence and injury 14 countries had MMRs of at least 1000, of which 13 are in the
RTAs are the 2nd leading cause of SSA, where the lifetime risk of maternal death is 1 in 73,
death and a leading cause of serious compared with 1 in 7300 in rich countries
injury for both sexes aged 529
Severe bleeding during delivery or after childbirth: commonest
Uncontrolled bleeding accounts for cause of MM, contributing up to 44% of maternal deaths in Africa,
>40% of trauma related deaths 31% in Asia and 21% in Latin America and the Caribbean
Capacity to provide safe blood
In most developing countries 50-80% of supplied blood is used for
transfusion - essential component of
obstetrics emergencies
Emergency Trauma Care Systems to
minimize death and disability in injured Blood transfusion: one of the eight signal functions of
patients Comprehensive Emergency Obstetric Care (EmOC) facilities
Blood Supply Estimating Blood Requirements
Role of blood transfusion services Important for planning a national blood programme
provide equitable access to safe blood/blood
Needs based programme - to satisfy the needs of a
products for all patients who need it
country's health care system, equitably and timely
adjust supply to actual needs (routine and emergency)
Shortages No global standard for estimating countries' needs for
various reasons (lack of donors or well organized blood/blood products, and estimates have to be made
donor programme, safety measures,) for each country and each region
periodic or continuous Plan blood collection and donor recruitment
crucial for patients when no alternative is available systematically, to avoid an excess or a shortage
1
Variables affecting demand and supply Demand and Supply
Geography, population and epidemiology

Epi Profile
Level and rate of development of health care Production
system
Prevention: e.g. anaemia, malaria
Diagnosis: e.g. haemophilia Pop Needs Clinical
Pop Demands Transfusion Demand
Treatment: e.g. advanced medical and
surgical procedures
Location and accessibility of health care facilities Clinical
Competence
Estimating blood needs
Country's Need for Blood and Blood

Products
Balance between demand and supply is

needed
Minimize wastage
Avoid blood shortages
Ensure appropriate use
Transfusion vol.45 Oct 2005 Supplement
Health Care System related Parameters

Geo-
Geo-population related Factors
Level & rate of development of health care
Size of the country
Accessibility of patients to health care
Geographical characteristics
Diagnostic and treatment facilities
Population Preventive public health measures
Size
Water, electricity, communication and transportation
Demographics age distribution, growth systems
Density/distribution in regions Hospitals
Epidemiology of diseases in the patients' population Number and location
(dependence on blood/blood products)
Total no. of hospital beds
Disasters (natural or man-made) No. of hospitals/beds for specialized complex care
New hospitals/ hospital beds
2
BTS related Parameters (1 of 2) BTS related Parameters (2 of 2)
Development and effectiveness of BTS to provide Degree of component preparation tailored to real need
safe blood/blood products to support regular and and resources of a country
specific transfusion needs What % of blood should be separated into
Number of blood centres and level of coordination components?
Assessment of services in each centre What to do with the excess plasma?
Options for Fractionation Contract / In-country

blood collected, blood processing, storage and
transport capacity Hospital blood stock management / inventory control
% of blood separated into components Future needs, including the feasibility of using recovered
plasma for fractionation
shelf-life of blood/blood components
Future need for apheresis and / or autologous programme
UK National Blood Service: overall efficiency

Clinical Blood Usage Parameters
NBS Donations 2005/06 NBS Total Losses 2005/06
>2.2million donors attending >2.2million donors attending
National guidelines on blood usage 100 20
90 18 0.27
Size of hospital (s) and number of patients Time-Expiring at
NBS
Time-Expiring at
NBS
2.23
80 16
Complete Donations
No. and kind of procedures, deliveries, anaemia pts 70
Not Validated
14 2.02
Complete Donations
Not Validated
Incomplete Bled
Per 100 Attendees

Per 100 Attendees
Donations
Clinical competence and experience of staff 60 12 Incomplete Bled
Donations
Attendees Not Bled
50 10
Training for hospital and blood bank staff Issued Red Cell
8
Attendees Not Bled
40
Annual blood usage review (past, present and future) -

13.14
30 6
hospitals, blood components 20 4
10 2
Different types of components needed
0 0
Michael Bowden 2006 Michael Bowden 2006
Complete donations not validated = testing losses (i.e. repeat reactives for microbiological markers,
abnormal test results in grouping) plus processing losses.
These figures do not include time-expiry/wastage in hospitals which was of the order of 2.2%in the
relevant period.
Historical Perspective (1 of 2) Historical Perspective (2 of 2)

Advance health care system
Joint study WHO/IFRCRCS(1986)
Need for cellular blood products can be met if number of units Average donation per 1000 population
donated annually correspond to 5% of population
Industrialized countries: 52
If at least 3% of the population is regular blood donors Middle-income countries: 10
(average annual donation of 1.5 to 2 ), all needs for cellular
Low-income countries: 1
products can be satisfied
Donation/hospital admission ratio was 0.44, 0.33 and
If health care is not fully operational 0.25
Need for blood should not relate to size of the population but 2% donor population may be sufficient
to other factors reflecting quality and extent of health service Blood requirement per bed per year 10-30
(Super-speciality - 30, Speciality - 20, General -10)
Management of Blood Transfusion Services, WHO, 1989
3
Donations/1000 population
80.0 Low HDI Medium HDI High HDI Method 1
Average 38.1
Number of donations / 1000 population
70.0 Average 2.3 Average 7.5

Range 0.4 7.5 Range 1.07 35.1 Range 4.92 68.0
60.0
Method 1: based on previous usage
50.0 Assess the number of units of blood used in a specified

75%
40.0 of global population
period in a defined geographical area or population
30.0
35%
of global blood collection
Analyse previous blood usage and requests for blood to
20.0 give an approximate indication of whether the demand
10.0 for blood is constant, increasing or decreasing
0.0
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
A total of 73 countries have donation rate of <1%

HDI values
(< 10 donations/1000 population) WHO-
WHO-BSI/GDBS 2007
Blood Usage
Method 2
120 countries report that a total of 51,400 hospitals
perform blood transfusions, serving a population of Method 2: based on acute hospital beds
around 3.6 billion
Suitable for countries with modern hospital services
Only 25 % hospitals performing transfusions in
Calculate 6.7 units of blood per acute hospital bed
developing countries and 33% hospitals in transitional
countries have a transfusion committee to monitor per year (WHO, 1971)
transfusion practices; as compared to 88% hospitals in
developed countries
Hospital Beds Model

Method 3
# hospital beds may provide an estimate of blood use Method 3: based on population
Limitations:
Used to estimate the number of units of blood needed to
growing trend to provide OPD treatment (reduced meet a countrys blood requirements over one year
relevance on # hospital beds as indicator of health
Calculation is based on 2% of population requiring blood
care)
per year
# hospital beds not dependent on size of population
(not an appropriate indicator for estimating and Can be used to calculate the blood requirements of
forecasting future demand) individual regions or districts within the country
hospital beds may be used for patients with complex

disorders with different levels of blood consumption
4
Population Based Model Method 3
Makes it possible to: Example
compare between countries or regions of For a country with a population of 10 million,

similar size calculate as follows:
project the trend in requirements in terms 10 000 000 x 2% = 200 000 units of blood per
of population trends year or approximately 3850 per week
paint a picture of the national situation A minimum of 100 000 donors will be needed if
each donor gives blood at least twice per year
Selecting a method to estimate

blood requirements
Method 1 is the most practical where there is a

constant supply of blood
Methods 2 and 3 can be useful where no data are

available or new established blood centre
5
Overview
The Global Burden of Disease
A brief introduction to the Global Burden of
approach to comparable
Disease project
international statistics
Issues in preparing comparable cross-national
statistics
Gretchen Stevens 1. Selecting indicators and metrics
Health Statistics and Informatics Department
2. Correcting for bias in available data
3. Estimating and communicating
uncertainty
Health Statistics and Informatics 3 February 2010 Health Statistics and Informatics
Global burden of Disease (GBD) GBD Principles

A standardized framework for integrating all Quantities of interest are total events or states at the
population level
available information on mortality, causes of
death, individual health status, and condition- Best available data used to make estimates
specific epidemiology to provide an overview of Corrections for major known biases to improve cross-
the levels of population health and the causes population comparability
of loss of health Comprehensive set of disease and injury causes
Consistent, comprehensive descriptive nothing is left out in principle
epidemiology No blanks in the tables, only wider uncertainty intervals
Common metric or summary measure Internal consistency used as a tool to improve validity
Health Statistics and Informatics Health Statistics and Informatics
Leading Causes of Mortality and Burden of Disease Overview

world, 2004
Mortality DALYs A brief introduction to the Global Burden of

% %
Ischaemic heart disease 12.2 Lower respiratory infections 6.2
Disease project
Cerebrovascular disease 9.7 Diarrhoeal diseases 4.8
Lower respiratory infections 7.1 Depression 4.3
Preparing comparable international
COPD 5.1 Ischaemic heart disease 4.1 statistics
Diarrhoeal diseases 3.7 HIV/AIDS 3.8
HIV/AIDS 3.5 Cerebrovascular disease 3.1
Tuberculosis 2.5 Prematurity, low birth weight 2.9
1. Selecting indicators and metrics
Lung cancer 2.3 Birth asphyxia, birth trauma 2.7 2. Correcting for bias in available data
Road traffic accidents 2.2 Road traffic accidents 2.7
Prematurity, low birth weight 2.0 Neonatal infections and other 2.7 3. Estimating and communicating
uncertainty
1
Three types of health statistics What is meant by comparable statistics?
Key elements:
Unadjusted statistics: derived directly from primary data
collection Quantities of interest are estimated at the
population level
Adjusted statistics: corrected for known biases
Predicted statistics: predicted using a statistical model,

Corrections for major known biases
includes both forecasts and
"farcasts" Estimates are made for every population
Adjusted and predicted statistics can be used for

national and international priority-setting
Three issues to consider when generating

comparable statistics Selecting health indicator and metrics
1. Selecting indicators and metrics Meaningful health indicator
Data are available or collectable

Disease and risk factor indicators are
3. Estimating and communicating uncertainty preferably comparable across diseases/risk
factors
Framework for Three issues to consider when generating

monitoring health systems comparable statistics
Inputs & processes Outputs Outcomes Impact
Intervention
Improved 1. Selecting indicators and metrics
Infrastructure; Coverage of health outcomes
ICT access & interventions & equity
services
Governance
Indicator
Financing
readiness Social and financial

domains
Health
workforce Prevalence risk risk protection 2. Correcting for bias in available data
Intervention behaviours &
Supply chain quality, safety factors Responsiveness
Information Efficiency
3. Estimating and communicating uncertainty
Data Population-based surveys
Administrative sources Facility assessments Coverage, health status, equity, risk protection, responsiveness
collection Financial tracking system; NHA
Databases and records: HR,
infrastructure, medicines etc.
Clinical reporting systems
Service readiness, quality, coverage, health status
Policy data
Civil registration
2
Bias in height and weight reported
Self-reported and corrected obesity
over the telephone
NHANES a national health examination survey
measures height and weight of a national sample
BRFSS a telephone health survey asks about height

and weight for state samples
Comparing these data:

Women underreport weight on the phone
Men overreport height on the phone
Ezzati et al. predicted unbiased state rates of overweight

and obesity using NHANES to correct BRFSS
Source: Ezzati, 2006
Three issues to consider when generating

comparable statistics Population-based hearing loss studies
1. Selecting indicators and metrics
3. Estimating and communicating

uncertainty
The boundaries and names shown and the designations used on this map do not
imply the expression of any opinion whatsoever on the part of the World Health
Organization concerning the legal status of any country, territory, city or area or of
its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted
lines on maps represent approximate border lines for which there may not yet be
full agreement.
WHO 2002. All rights reserved
Uncertainty in health estimates Predicted hearing loss prevalence

Western Europe Eastern Europe
Some sources of uncertainty:
Prevalence of moderate hearing loss
Input uncertainty (especially bias)

Model uncertainty (functional form, covariates)
Parameter uncertainty
(41+dB)
Estimating uncertainty in a consistent way across

diseases and risk factors has had limited success
Uncertainty is a major focus of the new round of

estimates
Age
3
Framework
Summary M&E of health systems strengthening
Inputs & processes Outputs Outcomes Impact
Choice of exposure metric may need to be Improved
flexible to meaningful and/or accommodate data Infrastructure;
ICT
Intervention
access &
Coverage of
interventions
health outcomes
& equity
availability services
Governance
Indicator
Financing
Health readiness Social and financial
domains
There is often a trade-off between data workforce
Intervention
Prevalence risk
behaviours &
risk protection
quality and population-based data Supply chain quality, safety factors Responsiveness
Modeling approaches can correct for missing or Data

Administrative sources Facility assessments Population-based surveys
Coverage, health status, equity, risk protection,
collection
biased data Financial tracking system; NHA
Databases and records: HR,
infrastructure, medicines etc.
Clinical reporting systems
responsiveness
Service readiness, quality, coverage, health status

Policy data
Civil registration
Assessing and communicating uncertainty is a Analysis Data quality assessment; Estimates and projections; In-depth studies; Use of research results;
continuing challenge & Assessment of progress and performance of health systems; evaluation
synthesis
Communication
Health Statistics and Informatics & use
Health Targeted and comprehensive reporting; Regular country review processes; Global reporting
Statistics and Informatics
Why calculate internationally Framework

comparable statistics? M&E of health systems strengthening
Global health can be characterized by:
Sparse, sometimes inconsistent data, especially Inputs & processes Outputs Outcomes Impact
where burden is highest Improved
Infrastructure; Intervention Coverage of health outcomes
Insufficient evidence-based priority-setting or ICT access &
services
interventions & equity
evaluation
Governance
Indicator
Financing
Health readiness Social and financial

domains Prevalence risk risk protection
workforce
Resource constraints Intervention
quality, safety
behaviours &
Responsiveness
Supply chain factors
Researchers can use data and experiences from
multiple settings to correct for biases in country
data or to make estimates when they are not
available
4
World Health Organization 12 April, 2010
Requirements of Blood & Blood

Global malaria burden
Components for Management of Malaria
4 species of human malaria parasites
Plasmodium falciparum, P. vivax, P. malariae, P. ovale
Dr Peter Olumese
Global Malaria Programme and few recent infections with the simiarn parasite P. knowlesi
WHO/HQ, Geneva
Estimated 243 (152-387) million malaria patients in 2008
WHO Experts' Consultation on Estimated 863 (610-1212) thousand malaria deaths in

Estimation of Blood Requirements 2008
03 February 2010
Geneva. 90% of deaths and 85% of cases occur in Africa south of
the Sahara -mainly among children under 5years of age
2 | Requirements of blood for management of malaria. Expert Consultation, Geneva, 03 Feb 2010
Malaria Control Technical Strategies

Clinical Disease and Epidemiology
..evidence-based actions
The nature of malaria clinical disease depends greatly on the
background level of the acquired protective immunity, a factor Early diagnosis and prompt treatment with effective
which is the outcome of the pattern and intensity of malaria medicines
transmission in the area of residence.
Where the transmission of malaria is stable,entomological Insecticide-treated nets (ITNs), Indoor Residual Spraying
inoculation rate [EIR] >10 per year), partial immunity to the clinical
disease and to its severe manifestation is acquired early in (IRS), and other vector-control methods
childhood.
severe manifestations mainly in the very young before acquisition of Intermittent preventive treatment in pregnancy (IPTp)
immunity
In areas of "unstable" malaria, the rates of inoculation fluctuate Emergency and epidemic preparedness and response
greatly over seasons and years. Entomological inoculation rates
are usually < 5 per year and often < 1 per year. This retards the
acquisition of immunity. Intermittent preventive treatment in infancy (IPTi)
all age groups (adults and children alike), are at high risk of progression to
severe malaria if untreated.
3 | Requirements of blood for management of malaria. Expert Consultation, Geneva, 03 Feb 2010 4 | Requirements of blood for management of malaria. Expert Consultation, Geneva, 03 Feb 2010
Severe malaria Malaria Anaemia

P. falciparum asexual parasitaemia and no other obvious cause of symptoms, the presence of one or
more of the following clinical or laboratory features classifies the patient as severe malaria
Anemia (hemoglobin level < 11 g/dL) remains one of the
Clinical features:
impaired consciousness or unrousable coma most intractable public health problems in malaria-
prostration, i.e. generalized weakness so that the patient is unable walk or sit up without assistance
multiple convulsions more than two episodes in 24 h endemic countries of Africa.
deep breathing, respiratory distress (acidotic breathing)
circulatory collapse or shock, systolic blood pressure < 70 mm Hg in adults and < 50 mm Hg in children

clinical jaundice plus evidence of other vital organ dysfunction
Haemoglobinuria
and has serious consequences as severe anemia

abnormal spontaneous bleeding (DIC)
pulmonary oedema (radiological)
(hemoglobin level < 5g/dL) is associated with an
Laboratory findings: increased risk of death
hypoglycaemia (blood glucose < 2.2 mmol/l or < 40 mg/dl)
metabolic acidosis (plasma bicarbonate < 15 mmol/l)
severe normocytic anaemia (Hb < 5 g/dl, packed cell volume < 15%)
Haemoglobinuria
hyperparasitaemia (> 2%/100 000/l in low intensity transmission areas or > 5% or 250
000/l in areas of high stable malaria transmission intensity)
hyperlactataemia (lactate > 5 mmol/l)
renal impairment (serum creatinine > 265 mol/l).
1
Severe Malaria Anaemia The burden of malarial anaemia
Hemoglobin less than 5g/dl (PCV less than 15%) It affects more than half of all pregnant women and
children less than five years old
Features of anaemic heart failure (even if PCV is >15%)
Each year in children < 5 years
In the presence of P.falciparum parasiteamia
1.4 - 5.7 million cases
190,000 - 974,000 deaths
Case fatality rate of severe anaemia (13.4 - 17.2%)
Highest mortality in infants
Antimalarial treatment
Any of the following antimalarial medicines are

Management of recommended
Severe Malaria
Artesunate (i.v. or i.m)
artemether (i.m.)
artemotil (i.m)
quinine (i.v. infusion or
i.m. injection). 3
Specific antimalarial treatment

Adjunctive therapy and supportive care
Other indications for blood and blood

Management of severe anaemia
products
The need for blood transfusion must be assessed with great care in each
individual child. Not only packed cell volume or haemoglobin concentration, but Disseminated intravascular coagulation, complicated by
also the density of parasitaemia and the clinical condition of the patient must be clinically significant bleeding, e.g. haematemesis or
taken into account.
melaena, occurs in fewer than 10% of patients.
In general, a packed cell volume of 12% or less, or a haemoglobin concentration
of 4 g/dl or less, is an indication for blood transfusion, whatever the clinical It is more common in non-immune patients with imported
condition of the child. malaria in the temperate zone
transfusion (10 ml of packed cells or 20 ml of whole blood per kg of body weight).
In children with less severe anaemia (i.e. packed cell volume 1318%, Hb 46 Treatment
g/dl), transfusion should be considered for high-risk patients with any one of the
following clinical features: transfuse fresh blood, clotting factors or platelets as required.
respiratory distress (acidosis);
impaired consciousness;
hyperparasitaemia (>20%).
The sicker the child the more rapidly the transfusion needs to be given.
2
Other indications for blood and blood

Challenges of blood transfusion
products
Hperparasitaemia Wastage
If parasitaemia exceeds 10% in severely ill patients, especially Non availability of paediatric blood bags
those deteriorating after optimal chemotherapy, exchange
transfusion with screened blood should be considered where Use of whole blood in place of packed cells or other specific
facilities are available blood components (e.g platelets)
Unnecessary transfusions
Exchange blood transfusion Inadequate diagnostics facilities
There have been many anecdotal reports and several series Blood storage facilities proper blood bank facilities
claiming benefit for exchange blood transfusion in severe
malaria but no comparative trials.
There is no consensus on whether it reduces mortality or how it
might work, so there is no global recommendation on its us
Comparison of trend patterns of inpatient malaria cases and deaths, by year, all ages, 2000/1-2006/7. Eritrea,
Rwanda, Sao Tome and Principe, Zambia, and Zanzibar.
12000 Eritrea 140 Rwanda
25000 450
Inpatient cases
Inpatient deaths 120 400
10000
20000 350
100
8000 300
80 15000
Deaths
Impact of the scale up malaria
Cases
250
Deaths
Cases
6000
60 200
10000
4000
control measures in Africa

150
40
5000 100
2000 20
50
0 0 0 0
2000 2001 2002 2003 2004 2005 2006 2000 2001 2002 2003 2004 2005 2006 2007
18,000 Sao Tome and Principe 350 Zambia Zanzibar

400,000 10,000 12000 450
16,000 9,000 400
300 350,000
10000
SME/MP/WHO Surveillance, 2008 14,000
250 300,000
8,000
7,000
350
12,000 8000 300
250,000 6,000
10,000 200
250
200,000 5,000 6000
8,000 150 200
4,000
150,000
6,000 4000 150
100 3,000
4,000 100,000 100
2,000
50 2000
2,000 50,000 1,000 50
12 April 2010
15 | 0 0 0 0 0
2 000 2 001 20 02 200 3 2004 2005 2006 2 007
0
2001 2002 2003 2004 2005 2006 2001 2002 2003 2004 2005 2006
Figure 2a. Malaria and non-malaria in- and out-patient cases, children <5 years old, January to Inpatient and outpatient indicators decline markedly, to low levels in 2006 and
October 2001-2007, Rwanda. LLIN = long-lasting insecticidial nets, ACT = artemisinin-based 2007, MRC research hospital and clinic, Fajara, Gambia
combination therapy medicines.
LLIN,
10000 ACT 14000 9000

70000 Inpatient Inpatient

s malaria malaria
sea

9000 8000
12000
60000
8000

cd cases deaths
em 7000
ess
ess

firn

74%
se 7000 10000 ac6000
50000 ac 100%
asc oc tn

tn

y-r iet

eti

ntie 6000

8000 ot ap5000 40000 ap
ta 5000 rao -in t-u
-pin

bla oa
riala4000

ir
ntie
30000
6000
riala 4000 ta am ala
aM3000
p- n-o3000 m
-n Outpatient
4000
uto 2000
N 20000
No
2000 ari slide
2000 ala 10000 Rainfall
1000 M 1000 positivity
0 0 0 0
rate (SPR) 73%
2001 2002 2003 2004 2005 2006 2007 2001 2002 2003 2004 2005 2006 2007
Year Year
Malaria out-patient laboratory-confirmed cases Non-malaria out-patient cases
Malaria in-patient cases Non-malaria in-patient cases
12 April 2010 12 April 2010
Source: Ceesay SJ et al. Changes in malaria indices between 1999 and 2007 in The Gambia: a retrospective analysis. Lancet 2008; 372:1545-54.
3
Combined Approved Interventions and the Impact Hemoglobin rises in 2005-2007 and blood transfusions in children
decrease to near zero in 2007, Sibanor, Gambia
Positivity
rate
Prevalence
rate
12 April 2010
19 | Requirements of blood for management of malaria. Expert Consultation, Geneva, 03 Feb 2010
Source: Ceesay SJ et al. Changes in malaria indices between 1999 and 2007 in The Gambia: a retrospective analysis. Lancet 2008; 372:1545-54.
Conclusions
The use of blood and blood products remains an

essential component of the management of severe
malaria
Estimating the requirement remains a challenge

especially now in the light of the reducing burden of
Thank
malaria due to rapid scale up of malaria control You
interventions
Urgent need to optimise the use of blood and blood

products especially in very young children with
severe malaria anemia.
4
Outline
When are blood and blood components used in maternity

Requirements of blood & care?
blood components in How often are blood and blood components required?
maternity care
What are the challenges in estimating requirements?
Matthews Mathai
Department of Making Pregnancy Safer
WHO Experts' Consultation on Estimation of Blood Requirements WHO Experts' Consultation on Estimation of Blood Requirements
1| Geneva Feb 3-5, 2010 2| Geneva Feb 3-5, 2010
Common indications Other indications
To correct hypovolemia resulting from acute haemorrhage In newborns

Antepartum and postpartum haemorrhage Top up transfusions in preterm infants
Ruptured ectopic pregnancy, uterine rupture and other genital As part of treatment of sepsis
lacerations For hypovolemia resulting from trauma
Post abortion Exchange transfusion
To treat severe anaemia To correct coagulation failure

With cardiac failure Placental abruption
Before or during surgery Retained products
Sepsis
Amniotic fluid embolism
Outline
When are blood and blood components used in maternity

care?
How often are blood and blood components required?
WHO analysis of causes of maternal death: a systematic review

WHO Experts' Consultation on Estimation of Blood Requirements Lancet
WHO Experts'367: 1066-1074,
Consultation 2006
on Estimation of Blood Requirements
1

Lancet 367: 1066-1074, 2006
Africa Asia LAC Developed

countries
Data sets 8 11 10 5
We need to
Maternal deaths 4508 16089 11777 2823
have a
Lancet 367: 1066-1074, 2006
Haemorrhage 33.9% 30.8% 20.8% 13.4%
(13.3-43.6) (5.9-48.5)
strategic
(1.1-46.9) (4.7-34.6)
Hypertension 9.1% 9.1% approach

25.7% to16.1%
(3.9-21.9) (2.0-34.3) achieve
(7.9-52.4) our(6.7-24.3)
Sepsis 9.7% 11.6%
goals7.7% 2.1%
(6.3-12.6) (0.0-13.0) (0.0-15.1) (0.0-5.9)

Lancet 367: 1066-1074, 2006 Estimated blood transfusion requirements
Country Year Maternal deaths MMR Haemorrhage Number per 1000 deliveries
DR Congo 1997 143 510 16%
Antepartum haemorrhage 7.26
Egypt 2000 585 84 30%
Postpartum haemorrhage 12.5
Senegal 2002 87 690 22%
Puerperal sepsis 8.0
Tanzania 1988 76 529 23%
South Africa 2002, 2003 3121 150 10%

Estimated global resources needed to attain universal coverage of maternal and
Zambia 1998 349 729 28% newborn health services
B Johns, K Sigurnbjrnsdottir, H Fogstad, J Zupan, M Mathai, T T-T Edejer
Zimbabwe 2001 92 695 19% Bull WHO 2007; 85: 256-263
MC W Africa 2001 55 334 33%
9| Geneva Feb 3-5, 2010 10 | Geneva Feb 3-5, 2010
WHO Global Surveys on Maternal and WHO Global Survey on Maternal and
Perinatal Health Perinatal Health
Facility based survey conducted between 2004 and 2008 Blood transfused in 1.58% of births
24 countries in Latin America, Africa and Asia Spontaneous vaginal delivery 1940/205303 (0.94)
Analysed mode of delivery, maternal and perinatal Operative vaginal delivery 198/7287 (2.72)
outcomes
Antepartum CS - no indications 14/1826 (0.77)
Varying caesarean delivery rates between and within Intrapartum CS - no indications 26/1063 (2.45)
regions and countries
Antepartum CS with indications 887/26876 (3.3)
Intrapartum CS with indications 1467/43815 (3.35)

11 | Geneva Feb 3-5, 2010 12 | Geneva Feb 3-5, 2010
2
WHO Global Survey on Maternal and

Perinatal Health Delivery complication: Excessive bleeding
% of live births in the last five years with excessive bleeding by region
40
Near miss study in LAC (n=97095 births) 35

34.4
32
34.6
% of live births in the last five years

30 28.3
25.7 25.1
25 22.3
Vaginal bleeding 22 in every 1000 births 20 17.5 16.4 17.2 18.4
15 13.2
10.7
Blood transfusion 13.8 in every 1000 births 10

3.8
7.3 7.2 7.4 6.8
5.4
7.4
Guatemala 1998-
Bangladesh 199
Bangladesh 199
Bolivia 1998
Tanzania 1996
Zambia 1996
Jordan 1997
Turkey 1998
Indonesia 1997
Indonesia 1994
Chad 1996-97
Egypt 1995
Yemen 1997
Philippines 1998
Brazil 1996
Colombia 1995
Republic 1996
Guatemala 1995
Nicaragua 1998
Peru 1996
Dominican
9-2000
6-97
99
Sub-Saharan Africa North Africa/West South & Southeast Asia Latin America & Caribbean
Asia/Europe
Source: Macro International Inc, 2010. MEASURE DHS STATcompiler. http://www.measuredhs.com, February 2 2010.
Outline Challenges
When are blood and blood components used in maternity Limited data
care? Amount
Blood or blood components
How often are blood and blood components required?
Estimated need versus actual transfusion requirements
Other issues
Epidemiology Malaria, helminthiasis, HIV
Recognising the need for blood transfusion
Intervention rates
Access to care
Anaemia in among pregnant women Proportion of pregnant women age 15-49 with anaemia. Anaemia
includes mild anaemia (haemoglobin count d/dl 10.0-10.9), moderate anaemia (haemoglobin count d/dl 7.0-9.9), and severe anaemic
(haemoglobin count d/dl below 7.0)
Gaps in skilled care In Bangladesh, only 18% of women
deliver with skilled professionals.
Wealthy women had 11 times higher
In China, women in the least affluent
access to skilled care than their poor
90 areas are twice as likely to deliver
counterparts.
80 without a trained health worker as
Mild anemia (%) Moderate anemia (%) Severe anemia (%)
% of preg nant w om en w ho are anae m ic
5.9 women in large cities.

70 5.1 1.5 4 2.3
2.6
60 2.7 1.2 2.7 2.2 2.7 3.4
50 44.2 38 35.4 0.7 3.3 1.5
39.8 38.2 30.6 28 26.7 1.9 1.5 0.2
32.7 30.6 0.7 0 0.4
40 32.8 30.3 0.6
27.7 18.1 15.6 0.5 0.1 0
31.4 27.4 22.9 19.1 12.5 11.6 5 0.3
30 16.3
18.8 14.7 3
19.5 14.2
20 16.9 13 0.2
33.2 5.8
26.9 30.3 27.3 30.6 31.2 30.8 27.4 28.8 28
25.7 24.6 25.8 22.7 23.4 22.5 22.7 23.5 26.5 21.8
10 18.8 19.6 20.9 17.7 19.7
16.5 13.1 14.7 15.4
0
Cong o Demo c ratic R epublic 2007
Burk in a Fas o 2003
Ca mbodia 2005
Ca meroon 2004
Mali 2006
Cong o (Braz z av ille) 2005
G uinea 2005
U ganda 2006
Niger 2006
G hana 2003
India 20 05-06
Tanz ania 20 04-05
H aiti 200 5-06
M adagas c ar 20 03-04
Nepal 2006
M oldov a 2005
J ordan 2007
Egy pt 2005
Ethiopia 2005
Hondu ras 200 5-06

Senegal 2005
Armenia 2005
Boliv ia 2003
L es otho 2004
Malawi 2004
Zimbab we 200 5-06
Az e rbaijan 2006
R wanda 2005
Swaz iland 200 6-07
In Ethiopia, only 6% of women deliver

with skilled professionals.
While Colombia has very high overall levels
of trained health workers, over a quarter of
5 - 20% the poorest will still deliver without skilled
21 - 40% professionals
41 - 60% Data source: proportion of births attended by a skilled health worker 2008 updates, WHO
The name as shown and the designations used in this map do not imply official endorsement off
61 - 80% acceptance by the United Nations.
81 -100 %
WHO Experts' Consultation on Estimation of Blood Requirements WHO Experts'
No data Consultation on Estimation of Blood Requirements
available
3
Percentage of births delivered by Caesarean section

Blood transfusion
% of births for which mothers received a blood transfusion
In Nigeria, while 70% of births occur in rural In Nepal, only 0.8% of poorest women have Nepal 2006
areas, only 1% of women in rural areas have access to C-section compared to 12% of the 3.5
access to C-section. most wealthy women.
3 2.9
2.5
2
1.4
% of births
1.5 1.3
1
1 0.8 0.8 0.7
0.6 0.5
0.4 0.3 0.4 0.4
0.5
0.1 0.1 0.1
0
Elsewhere
Richest
Health
Urban
Richer
Middle
Poorest
Poorer
6+
1
2 to 3
4 to 5
<20
20-34
35-49
Rural
facility
Place of Residence Wealth quintile Birth order Mother's age at
In Indonesia, women in urban areas are Less than 5 % delivery birth
three times more likely to have access to C- 5-15 %
section than their rural counterparts.
More than 15 %
No data available
Source: Macro International Inc, 2010. MEASURE DHS STATcompiler. http://www.measuredhs.com, February 2 2010.
Data source: Demographic and Health Survey
The name as shown and the designations used in this map do not imply official
WHO Experts' Consultation on Estimation of Blood Requirements
endorsement off acceptance by the United Nations. WHO Experts' Consultation on Estimation of Blood Requirements
4
Requirements of blood
and blood components for trauma care
To know blood requirements
Pablo Perel
How frequent is trauma?
How frequent is bleeding in trauma patients?
Which patients should receive transfusions?
Which blood components should they receive?
In addition To know blood requirements
CRASH-2 Trial How frequent is trauma?
Global injury-related Mortality Rankings of Deaths & DALYs: 1990 - 2020

Deaths DALYs
1990 rank 2020 rank 1990 rank 2020 rank
Road Traffic Injuries 9 6 9 3
Self Inflicted Injuries 12 10 17 14
Interpersonal Violence 16 14 19 12
War 20 15 16 8
Legend*
No data
120 - 131.1
95.0 - 119.9
70.0 - 94.9
The boundaries and names shown and the designations used on this map do not imply the expression of any 45.0 - 69.9
IfIf current
current trends
trends continue,
continue, road
road traffic
traffic and
opinion whatsoever on the part of the World Health Organization concerning the legal status of any country,
and
territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on
maps repre sent approximate border lines for which there may not yet be full agreement.
WHO 2002. All rights re served
Injury related mortality*, WHO Regions, 2000

intentional
intentional injuries
injuries will
will rank
rank in
in the
the 15
15 leading
leading
Africa Americas Eastern
Mediterranean
Europe South-East Asia Western Pacific
causes
causesof ofdeath
deathand
andburden
burdenof ofdisease.
disease.
HIC LMIC HIC LMIC HIC LMIC India Other LMIC HIC China Other LMIC
118.8 53.8 76.2 51.1 70.4 47.6 131.5 96.9 75.0 56.2 51.5 78.4
* Rate per 100 000 population High income countries Low/middle income countries
1
Ten most common causes of death in young people (10-24)
Reference: Patton GC, et al. Global patterns of mortality in young people: a systematic analysis of
population health data. Lancet. 2009 Sep 12;374(9693):881-92.
In-hospital trauma deaths Transfusion rates in trauma patients
Country Source Number of patients

Transfusion rate
Massive transfusion
United KingdomRegistry (TARN) 28,703 3.4% 0.5%
Registry
Germany 2,475 54% 17.1%
(DGU)
Exsanguination CNS injury
45% 41% US Trauma Centre 5,645 8% 3%
Israel Trauma Centre 986 33% 4.7%
4% 10%
Other Organ failure - Maegele, M Changes in transfusion practice in multiple injury between 1993 and 2006: a retrospective analysis
on 5389 patients from the German Trauma Registry. Transfusion Medicine. 19(3):117-124, June 2009.
- Como JJ et al Blood transfusion rates in the care of acute trauma. Transfusion. 2004 Jun;44(6):809-13.
Reference: Sauaia A et al. Epidemiology of trauma deaths: a reassessment. J Trauma 1995;38:185-193 - Soffer, D et al. Usage of Blood Products in Multiple-Casualty Incidents: The Experience of a Level I Trauma
Center in Israel. Archives of Surgery. 143(10):983-989, October 2008.

Recommendation 19
How frequent is bleeding in trauma patients? We recommend a target Hb of 7
to 9 g/dl (1C)
Crit Care. 2007;11(1):R17.
2
Red cell transfusions
Transfusion threshold Mortality in patients who declined blood transfusions
30 days mortality
Carson JL et al. Mortality and morbidity in patients with very low postoperative Hb levels who decline blood transfusion Transfusion. 2002
Jul;42(7):812-8.
Reference: Hill S Transfusion thresholds and other strategies for guiding allogeneic red blood cell
transfusion. Cochrane Database Syst Rev. 2002;(2): 14
Red cell transfusions To know blood requirements

Massive Bleeding
Blood components To know blood requirements

In stable patients (European Guideline)
FFP if PT > 1.5
Platelets if <50 x 109
In patients with massive bleeding

Plasma:RBC ratio (1:3)
Trauma exsanguination protocol (1:1)
(Cotton, BA et al. J. of Trauma-64(5):1177-1183, May 2008) Which patients should receive transfusions?
however, all of the studies are observational
(Snyder C The Relationship of Blood Product Ratio to Mortality: Survival Benefit or
Survival Bias? Journal of Trauma66(2):358-364, February 2009)
3
Frequency
Frequency
Point at which benefits of Point at which benefits of
transfusion exceeds harms transfusion exceeds harms
blood loss blood loss
Risk will be different according to the region

Frequency
Point at which benefits of

transfusion exceeds harms
blood loss
Number exposed to allogeneic blood

What do we know about interventions that RR (95% CI), random effects
can shift the curve ?

1)Hill S, Carless PA, Henry DA, Carson JL, Hebert
PPC, Henderson KM, McClelland B. Transfusion
thresholds and other strategies for guiding allogeneic
Cochrane reviews of blood sparing red blood cell transfusion. CDSR 2000 Issue 1.
1)Lin Y, Stanworth S, Birchall J, Doree C, Hyde C.
interventions (surgery) Recombinant factor VIIa for the prevention and

treatment of bleeding in patients without haemophilia.
CDSR 2007, Issue 2
1)Carless PA, Henry DA, Anthony DM. Fibrin sealant

use for minimising peri-operative allogeneic blood
transfusion. CDSR 2003 Issue 1.
1)Carless PA, Henry DA, Moxey AJ, O'Connell D,

Brown T, Fergusson DA. Cell salvage for minimising
perioperative allogeneic blood transfusion. CDSR
2006 Issue 4
1)Henry DA, Carless PA, Moxey AJ, O'Connell D,

Stokes BJ, McClelland B, Laupacis A, Fergusson DA.
Anti-fibrinolytic use for minimising perioperative
allogeneic blood transfusion. CDSR 2007 Issue 4.
Favours Favours
treatment control
4
Effects of TXA Effects of TXA
Blood units saved Re-operation

RR (95% CI)
TXA 1.1 (0.6-1.6) TXA 0.67 (0.41-1.09)
0 0.4 0.8 1.2 1.6 2.0 0 0.4 0.8 1.2 1.6 2.0
Blood units Antifibrinolytic better Antifibrinolytic worse
Effects of TXA Adverse effects of tranexamic acid
Mortality Outcome Tranexamic Acid

RR (95% CI) RR 95% CI
Myocardial Infarction 0.96 0.48-1.90
Stroke 1.25 0.47-3.31
TXA Deep venous thrombosis 0.77 0.37-1.61

0.60 (0.32-1.12)
Renal failure 0.73 0.16-3.32
No evidence of adverse effects for tranexamic acid

0 0.4 0.8 1.2 1.6 2.0
Antifibrinolytic better Antifibrinolytic worse
Rationale for the CRASH-2 Participating countries
Bleeding is a leading cause of trauma death
Antifibrinolytics reduce blood loss after surgery
A simple intervention like TXA could prevent thousands

of trauma deaths and transfusion associated infections
A large randomised controlled trial among trauma

patients with significant haemorrhage, of the
effects of antifibrinolytic treatment on death and
transfusion requirement
5
Results I Results II
Number of patients per country Characteristics of patients included
N %
Male 15,932 83.75
Gender
Female 3,091 16.25
<25 5,320 27.97
25-34 5,720 30.07
Age categories
35-44 3,573 18.78
>44 4,410 23.18
<1 1,791 9.41
Hours since injury 1 to 3 10,963 57.63
>3 6,269 32.95
Results III Results IV

Outcomes by subgroup Systolic blood pressure
Mortality and transfusion practices Mean blood units
Results VII Results VIII

Outcomes by subgroup Type of injury Outcomes by region
Mean blood units
Units Total
Region Mortality Transfusion
transfused patients
Africa 0.23 0.44 2.67 2,317
South East Asia 0.16 0.58 4.75 5,256
Western Pacific 0.15 0.72 12.30 74
Europe 0.13 0.33 10.24 2,127
Americas 0.15 0.51 7.65 5,425
Eastern Mediterranean 0.09 0.53 3.81 2,737
6
Results IX Tranexamic acid and blood loss
Outcomes and units transfused by region
according to subgroup Traumatic bleeding (CRASH-2 trial completed)
Systolic Blood Pressure Systolic Blood Pressure
Postpartum bleeding (WOMAN trial in progress)
AFRICA EUROPE
>89 76-89 <76 >89 76-89 <76 GI bleeding (in preparation)
Mortality 16 31 43 Mortality 7 14 42
Mean no of units 2.5 2.7 2.9 Mean no of units 8.8 11.4 12.4
SOUTH EAST ASIA AMERICAS If tranexamic acid also reduces blood loss in these conditions
Mortality 13 17 35 Mortality 9 19 32 then could have a substantial impact on global demand for
blood.
WESTERN PACIFIC EASTERN MEDITERRANEAN
Mortality 8 24 0 Mortality 4 9 24
www.crash2.lshtm.ac.uk www.woman.lshtm.ac.uk
7
World Health Organization 12 April 2010
Anaemia
A condition in which the number of red blood cells or their

oxygen-carrying capacity is insufficient to meet physiologic
Nutritional anaemias: physiological
demands. Such demands varies by age, gender, altitude, smoking,
and public health considerations
and pregnancy status.
Causes: diet reduced in nutrients/low bioavailability, chronic

Luz Maria de Regil, PhD, MSc
Epidemiologist, Micronutrients Unit inflammation, parasitic infections, and inherited disorders of
Department of Nutrition for Health and Development
haemoglobin structure.
WHO
Experts' Consultation on Estimation of Blood
Requirements
Geneva, 03-05 Feb, 2010
1| 2|
Anaemia Anaemia
Microcytic Normocytic Macrocytic Microcytic Normocytic Macrocytic
Iron Reticulocytes Megaloblasts Iron Reticulocytes Megaloblasts
Low High Low High Yes No Low High Low High Yes No
Sideroblasts Reticulocytes Sideroblasts Reticulocytes
No Yes Low High No Yes Low High
iron thalasse sidero aplastic haemolytic vitamin folic acid hepatic haemolytic iron thalasse sidero aplastic haemolytic vitamin folic acid hepatic haemolytic
deficiency mia anaemia disease B12 deficiency mia anaemia disease B12
blastic anaemia disease anaemia blastic anaemia disease anaemia
Leukaemia transfusion anaemia Leukaemia transfusion anaemia
anaemia reaction anaemia reaction
Adapted: Casanueva et al, 2008

3| 4|
Nutritional anaemias Iron deficiency anaemia

Prevalence of iron deficiency
120
Nutrient Effect on
100 w/o anaemia
Thiamine DNA synthesis Anaemia
80
Folic acid DNA and RNA synthesis
(%)
60
Vitamin B12 DNA and RNA synthesis
Vitamin E Lysis of membranes 40
Vitamin C Iron mobilization 20
Copper Ceruloplasmin: ferric to ferrous iron
0
Pyridoxine Hb synthesis: alcoholism 5 10 15 20 30 40 50 60 70 80
Iron Hb synthesis Prevalence of anaemia (%)
Source: WHO/ UNICEF / UNU. Iron deficiency anaemia: Assessment, prevention and
control, a guide for programme managers. Geneva ,WHO, 2001.
5| 6|
1
293 million children under 5 are anaemic Iron deficiency: health implications
Total body iron: 3-5 g
Adequate growth and development
! !
Immunity
!
!
Cognition and development

attention span, concentration, memory, learning ability
muscle function and manual dexterity, behaviour, social
interaction
Work productivity
Category of public health significance
(anaemia prevalence)
Normal (<5.0%)
Reproductive performance
Mild (5.0-19.9%)
Moderate (20.0-39.9%) Source: WHO/CDC. Worldwide prevalence of anaemia 1993- Raising of other metals levels
Severe (40.0%) 2005. WHO Global Database on Anaemia. Geneva, World Health
No Data Organization, 2008.
7| 8|
Iron deficiency: health implications Cut off points

Haemoglobin g/L
Population Anaemia Mild Moderate Severe

Children 6 59
months <110 100-109 70-99 <70
Children 5-11 years <115 110-119 80-109 <80
Children 12-14 years <120 110-119 80-109 <80
Non-pregnant women
(above 15 years
of age) <120 110-119 80-109 <80
Pregnant women <110 100-109 70-99 <70
Source: WHO Global burden
of disease, 2009
Men (above 15 years
of age) <130 110-129 80-109 <80
Source: WHO, VMNIS, 2010
9| 10 |
Altitude Smoking
In addition to a diet with low iron/poor bioavailability
Haemoglobin
Increase in Hb (mg/L)
Non smoker 0
Smoker + 0.3
-1 packet/day + 0.3
1-2 packets/day + 0.5
2 packets/day + 0.7
Source: WHO, VMNIS, 2010

Altitude (m)
11 | 12 |
2
Pregnancy Increased demands
Iron needs or losses

Physiologic condition
mg/day
Iron needs (mg / day )
Baseline Requirements 0.72

4 Fetus Menstruation 0.44
Breastfeeding 0.23
3
Mild intestinal parasitic infections
2 N. americanus 1.10
RBC
A. duodenale 2.30
1 Other parasitic infections
Fe losses 800mg T. trichuria (mild) 0.16
15 20 25 30 35 40
S. haematobium (severe) 2.10
Weeks of pregnancy Source: Viteri
Source: Stoltzfus y Dreyfuss, 1998
13 | 14 |
Intestinal losses Iron status indicators

12
Indicator Positive Normal Anaemia
Balance
10
Hb (g/L) >130 110-130 <110
Haemoglobin /(mg/dL)
8 Ferritin (g/L) >150 12-150 <12

RBC protoporphyrin 70 <70
6 (g/dL)
Plasma serum (mg/L) >175 11550 <40
4
Transferrin saturation (%) <16
2
STR International
0 Reference
0 1 - 1999 2000-3999 4000-5999 6000-7999 >7999 Reagent
eggs/g of fecal matter
Source: WHO/ UNICEF / UNU. Iron deficiency anaemia: Assessment, prevention and
Source: Stoltzfus y Dreyfuss, 1998
control, a guide for programme managers. Geneva ,WHO, 2001.
15 | 16 |
Treatment of anaemia WHO recommendations

Yes Low prevalence of Yes
Population level Nutrition counselling
anaemia in GAR
No
Intermediate Fortification
Yes
prevalence of Preventive supplementation
anaemia in GAR Nutrition counselling
No
Fortification
Individual level High prevalence of Yes
Therapeutic supplementation
anaemia in GAR
Nutrition counselling
No
Increased iron Yes Treatment of cause of anaemia

Diagnosis losses? Therapeutic supplementation
No
Belong to a No
vulnerable group? Nutrition counselling
Yes
Therapeutic supplementation
Nutrition counselling
Adapted: Casanueva et al, 2008
17 | 18 |
3
Anaemia
WHO-VMNIS
Microcytic Normocytic Macrocytic
Iron Reticulocytes Megaloblasts
Low High Low High Yes No
Sideroblasts Reticulocytes
No Yes Low High
iron thalasse sidero aplastic haemolytic vitamin folic acid hepatic haemolytic
deficiency mia anaemia disease B12
blastic anaemia disease anaemia
Leukaemia transfusion anaemia
anaemia reaction
19 | 20 |
Acknowledgements
Financial and/or technical support for the Micronutrients Unit
The Government of Luxembourg
US Centers for Disease Control and Prevention (CDC)
The Micronutrient Initiative (MI)
US Agency for International Development (USAID)
The Bill and Melinda Gates Foundation
21 |
4
Estimating National Blood Requirements
what should we measure?
1. Needs/demand for blood in the existing
health system
Estimating National Blood
2. Capacity and gaps of health system to
Requirements in Africa deliver services
Lawrence H. Marum, MD, FAAP, MPH Should a national blood service focus on estimating
Centers for Disease Control and
and meeting needs within existing health system
Prevention - Zambia OR expanding the number and distribution of
facilities with capacity for transfusion?
Figure 1: National Blood Transfusion Service (NBTS) collections per 1000

population in PEPFAR partner countries: 2003 and 2007
Drivers of demand for blood in Africa *Nigeria 0.0
0.1
Ethiopia 0.2
0.3
2003 Blood Collections
Haiti 1.0
1. Malaria associated anaemia (children * Tanzania 0.0

1.8
2.7
per 1000 Population
2007 Blood Collections
and mothers) Kenya 1.2

3.3
per 1000 Population
Rwanda 3.5
3.3
2. Emergency obstetrical services Uganda 3.8

4.3 WHO Recommended Target
Cote d'Ivoire 3.8
Post-partum haemorrhage Mozambique 3.4

4.8
5.4
Zambia 3.7
3. Surgical services Guyana
5.7
5.4
7.5
4. Trauma Namibia
Botswana 6.5
9.1
8.8
11.8
5. Medical treatment South Africa 16.6

17.7
0 2 4 6 8 10 12 14 16 18
HIV treatment associated anaemia Units of Whole Blood Collected per 1000 Population
* Nigeria and Tanzania had no NBTS in 2003
Zambia National Blood Transfusion Service: Transfusions/bed at different

progress in meeting national needs hospital levels - Zambia
2004 2009 16
14
Population of Zambia 10.5 million 12.5 million
12
Units blood collected 38,477 100,000 10 #1 08
(collections per 1000 population) (3.7) (8.0) #1 09
8
#2 08
Transfusion outlets/ blood centres 90 (90) 132 (9) 6 #2 09
4
Proportion voluntary donors 72% 100%
2
Discards 15.5% 9.2% 0

(HIV discards) (6.9%) (3.0%) Level 1 Level 2 Level 3
1
Proportion of paediatric University Teaching Hospital:
transfusions - Zambia changing uses of blood
40 1864 beds
35 Uses 27% of national
30 blood supply
25 #1 08
#1 09
Maternity 30%
Maternity
20
#2 08 Paediatric
Medical
Paediatric 17%
15 #2 09 Surgical
10
Other Medical 21%
5 Surgical 14%
0 Other: outpatient and
Level 1 Level 2 Level 3 special services 14%
Zambia HIV summary Progress in malaria control

12.5 million population; total fertility rate = 7 Deaths in hospitalized patients with malaria
Adult HIV prevalence (15-49 years) 14.3% decreased 60%
54% decrease in parasitemia (under 5 y/o)
>250,000 on ARVs (19,000 children)
69% decrease in severe anaemia
Primary first-line treatment changed from AZT-
3TC to TDF-FTC backbone regimes Distributing 400,000 nets per year
70% of those with CD4 < 250 Indoor residual spraying in 37 of 73 districts
40% of HIV+ pregnant women in 2010 (25,000) (over 1,000,000 homes annually)
will initiate AZT backbone regime (CD4<350) Artemesinin Combination Therapy 1st line and
Sulfadoxine-pyrimethamine in pregnancy
Blood Use at Macha Macha Hospital - All Blood

(rural mission/district hospital) Transfusions Childrens Ward
200
180
TB Ward
Non Proportion Maternity Ward
160
Childrens Childrens given Womens Ward
140
n u m b er p er m o n t h
Ward Ward Total to Children

Mens Ward
Year Transfusions Transfusions Transfusions 6 yrs 120
100
2000 588 219 807 73%
80
2001 512 267 779 66% 60

40
2002 336 258 594 57% 20
2003 393 305 698 56% 0

J ul
J ul
J ul
J ul
Oct
Oct
Oct
Oct
J an
J an
J an
J an
A pr
A pr
A pr
A pr
2000 2001 2002 2003
2
Macha Hospital - Non Children's Macha Hospital - Children's Ward
Ward Blood Transfusions Blood Transfusions
200 200
150 150
100 100
50 50
0 0
Jul
Jul
Jul
Jul
Jul
Jul
Jul
Jul
Apr
Apr
Apr
Apr
Apr
Apr
Apr
Apr
O ct
O ct
O ct
O ct
O ct
O ct
O ct
O ct
Jan
Jan
Jan
Jan
Jan
Jan
Jan
Jan
2000 2001 2002 2003 2000 2001 2002 2003
Macha Hospital Childrens Ward Malaria Diagnoses

Childrens Children's
500 Ward Ward Blood
450
A/L
Year Malaria Dx Transfusions
A/L
400 Introduced out of 2000 1479 588
as First stock
350 Line Rx for
Malaria
2001 1778 512
300
250
2002 1294 336
200 2003 1418 393
150
2004 423 155
100
2005 123 60
50
0
2006 565 225
JAN
APR
JAN
APR
JAN
APR
JAN
APR
JAN
APR
JAN
APR
JAN
APR
OCT
OCT
OCT
OCT
OCT
JUL
OCT*
JULY
JULY
JULY
JULY
JULY
JULY
2007 336 118

2000 2001 2002 2003 2004 2005 2006
Macha Hospital - Children's Ward Blood

Transfusions Blood Use at Macha
180 Conclusions:
160
140
Historically, over 50% of blood use is for
120 children under 6 years of age
100
80 Blood is primarily given during the peak
60 malaria transmission season
40
20 Hypothesis:
0
Control of malaria may lead to a significant
Ju l
Ju l
Ju l
Ju l
Ju l
Ju l
Ju l
Ju l
Ju l
Sept
O ct
O ct
O ct
O ct
O ct
O ct
O ct
O ct
O ct
Ju n e
Ja n
Ja n
Ja n
Ja n
Ja n
Ja n
Ja n
Ja n
Ja n
Apr
Apr
Apr
Apr
Apr
Apr
Apr
Apr
Apr
Apr
DJ ae nc
decrease in the need for blood at district

1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
level hospitals in sub-Saharan Africa
3
Emergency Obstetric Services Conclusions
1230 ANC facilities; 937 offer PMTCT Changing blood needs
Reduced paediatric transfusion w/ malaria control
132 transfusing facilities
Expansion of emergency obstetric services
Lack of trained nurse/clinical officers Increased chronic disease treatment (ARVs)
Only 60% of national health posts filled M&E needs
Rural retention schemes; retired nurses Comprehensive data on use of blood; vein to vein linkages
Prescriber information for monitoring and supervising
Expansion of C-section capacity to level 1 SmartDonor and SmartCare: national electronic records
hospitals and larger health centres many Address national transfusion coverage in Health System
that do not have transfusion capability Strengthening efforts how quickly can we safely expand?
4
Background : national health care
Canada
WHO Experts Group Consultation on Estimation Population:
33,968,200
of Blood requirements
Territory:
9,984,670 km
DIFFERENT MODELS OF BLOOD ESTIMATION

Quebec
A BLOOD CENTRE PERSPECTIVE Population:
7,828,879
Dr Francine Dcary Territory:

President and CEO 1,542,056 km
Carolina Sarappa
Hospitals:
Business Analyst
99
Hma-Qubec
Model:
Blood banks in
03-05 February 2010
hospitals
Blood Management System in Quebec Quebec statistics 2008-09
Blood units drawn: 245,938
RBC shipped to hospitals: 231,958
Inventory at HQ: 8 days
Outdating at HQ: 0.62%
Outdating in the hospitals: 1,4%
Comparison of RBC utilization with other

RBC utilization in hospitals
countries
Indications %
50
Surgery: Elective 25 45.5
Units/1000 population/year
40
Urgent 12 33.0
28.5
30
Obstetrics 1
20
Trauma 7
10
Hematological
43
disease & transplant 0
USA
France
QUEBEC
Others 12
Total 100
1
Demand Forecasting: Key Criteria to
Types of Forecasting Techniques
Consider
1. Informal: intuitive depending on individual experience and abilities 1. Item to be forecasted : RED BLOOD CELLS
2. Formal: Availability of RBC depends on :
2.1 Qualitative techniques Availability of blood donors

2.1.1 Delphi Quality of the donors
2.1.2 Market surveys Management of the RBC by the blood center (discard & outdate rate)
2.1.3 Life cycle analogy
2.2 Quantitative techniques Utilization of RBC depends on :

2.2.1 Auto Projection
2.2.2 Moving average Inventory management by client-hospitals
2.2.3 Exponential smoothing Stock rotation between hospitals
2.2.4 Causal Outdate rate target 2%
2.2.5 Regressive Depends distance from blood center
2.2.6 Box Jenkins
2.2.7 Econometric Age of the population
2.2.8 input-output Availability of data from hospitals
Key Criteria to Consider Key Criteria to Consider
2. Major challenges 3. Amount of historical data available

Donors Total number of RBC shipped :
Quality of donors deferral rate 18% By ABO Rh
Per year, month, week
Blood group distribution (ABO and Rh) in the population
Records exist on a daily basis at Hma-Qubec
Customer service and satisfaction
The right product at the right time Hospital utilization :
Historical data for sentinel hospitals
Maintaining adequate levels of Inventory Usage by diagnosis-related group (DRG)
No shortages of stock Outdate rate
Economic situation Events that affected the past demand

Must consider the health care system model because the economic situation may or Ex.: Letter requesting hospitals to decrease the expiry rate of RBC
may not influence your forecast.
Presence of a transfusion committee
Disaster planning
Capability of responding to emergencies
Key Criteria to Consider
4. Time allowed to prepare the forecast

Tied into the annual budgeting period
Also depends on the resources allocated to the forecast
Horizon:
Long term (5-10 yrs)
Short term (1-2 yrs)
Very short term (< 1yr)
Short-term forecasting
(1 to 2 years)
2
Quantitative techniques Results obtained
Nave method
Dt = Dt-1 + c
Forecast
Forecasting method % error
Where c is determined by historical data and /or expert judgement (12 months)
One of the simplest methods to use
Simple linear trend 219,406 Reference
For Hma-Qubec, this method works well since the demand is relatively stable.
Exponential smoothing 270,844 + 23.4 %
Exponential smoothing
Holt-Winters 231,376 + 5.5 %
This method is a special form of the weighted average and focuses on the most recent period.
Dt = aDt-1 + (1 - a)Dt-1 Holt-Winters
228,085 + 4.0 %
(Seasonality month)
Where a (the smoothing constant) is determined by trial and error
Holt-Winters
228,090 + 4.0 %
Fairly complex statistical methods are involved (Seasonality quarter)
Research on the factors that affect

Final model
hospital demand (2006)
A sample of 8 blood banks, accounting for 56,673/221,256 (24%) of the demand for The proposed model is a particular version of the ARIMA (Autoregressive Integrated
the year 2005-06, were interviewed. Moving Average) models.
The results indicate that the following variables have the most important effect on the
It is based on a chronological series related to the quantity of red blood cells
demand for red blood cells:
improvement of surgical and medical practices
distributed to hospitals for a 349 week period
medical and technological advancement from 11/22/1999 to 07/31/2007
substitute products and alternative treatments
education of hospital personnel
protocols of transfusion
For the following 18 week period, the absolute percentage errors range from 0.31%
expiration targets for blood products
to 10.61%
the mean error is 4.90% per week or 0.70% per day.
However, the findings appear almost impossible integrate into a regression model,
historical data remained the best indicator of the future demand for the purpose of
this study.
Demand Forecasting
ARIMA model Red Blood Cells Shipped to Hospitals
Actual Dem and Forecast % Error
250 000 10%
Observed 9.00%
5500
Forecasted
8%
240 000
5000
Distribution of Packed Red Cells
6.00% 6%
4.70% 231 958

4500
230 000 4%
227 581
223 723
221 659 221 256 2%
4000
223 100
220 000 0.84%
220 215
0%
3500
211 901
210 000 -1.83% -1.94% -2%
3000
-4%
200 000
-5.27%
-6%
0 100 200 300 400 -6.48%
Week
190 000 -8%
2001-2002 2002-2003 2003-2004 2004-2005 2005-2006 2006-2007 2007-2008 2008-2009
3
So how DO we forecast? Monitoring daily inventory
Demand Forecasting Date: 01-10-2010 Time: 6:00 am

2 forecasting techniques based on historical demand are currently used at Hma-Qubec Available 5 times a day via Intranet
Nave method
Arima model
TOTAL INVENTORY
MONTREAL & QUEBEC
Communication with hospitals RED BLOOD CELLS
MINIMAL
Creation of Hospital Relations Department (6 days)
LABELLED % TO LABEL %
99 hospitals
O- 500 673 135% 275 190%
A- 350 393 112% 179 163%
Our conclusion: B- 100 42 42% 49 91%
Although very efficient, the quantitative forecast should be coupled with expert AB- 45 103 229% 15 262%
judgement and increased communication efforts with hospitals in order to maximize
the accuracy of the forecast and effectively optimize the efficiency of the decision- O+ 1500 1855 124% 587 163%
making process.
A+ 1350 1813 134% 554 175%
Results in: B+ 300 546 182% 125 224%
Very short-term forecasting AB+ 115 346 301% 59 352%

TOTAL 4260 5771 135% 1843 179%
WEIGHTED INVENTORY*: 124% 163%
* Considers minimal stock for B+ and AB+
Inventory management Inventory management
Blood The committee meets on a weekly basis in order to:

Collection &
Donor Monitor inventory levels by product and by ABO/Rh (when applicable)
Transport Hospital
Recruitment Review the weeks activities:
Services
Success of blood drives
O negative in stock
Age of RBC inventory
Inventory of platelets collected by apheresis
Turn-around time at laboratory testing
Processing Shortage/surplus of personnel
Staffing INVENTORY COMMITTEE & Shipping Special cases and projects (ex.: TRALI, ISBT 128)
Weekly inventory report from sentinel hospitals (on average = 5 days)

Client-hospitals communicate active cases with blood product requirements
Adjust short term collection targets based on fluctuations in hospital demand and inventory
Business Hospital levels.
Analyst Relations Plan for holidays or other events
Product
Qualification
Days of Inventory per Week

from April 3rd 2006 to July 20th 2008 Further research
Days of Inventory Minim al Inventory Optim al Inventory
14
12
Long-term forecasting (5 10 years)
Forecast based on the population
pyramid
10
Drawbacks:
Does not consider the usage per
DRG
8
Does not take into account
medical and scientific advances
4
Apr-06 Sep-06 Feb-07 Jul-07 Dec-07 May-08
4
DIFFERENT MODELS OF BLOOD ESTIMATION
A BLOOD CENTRE PERSPECTIVE
Summary
1. Short-term forecasting (1-2 yrs): in a stable mature system, forecasting in best

done by simple techniques based on historical data, professional judgment and
constant communications with hospitals.
2. Long-term forecasting (5-10 yrs): for strategic planning, demographics and DRGs
are critical elements.
3. BUT: Very short term forecasting (daily, weekly and monthly): the most important
one since only this forecasting prevents shortage of blood and thus gives the best
service to the patients in need of blood.
5
Assessment of Needs for Blood Introduction
Products Pilot study of a model
The blood needs for a population could be
based on clinician experience defined as the sum of the needs for all patients
Elizabeth Vinelli, Rashid Salmi, Brian McClelland, Marco Pinel, with all diseases.
Rosa Kafati, Guillermo Guivobich, Juana Lozano, Ana del Pozo,
Christian Hertlein Since there are countless combinations of
disease, stage, co morbidity and intervention
that can put a patient at risk of needing a
Presented on behalf of the project group by Brian McClelland WHO
February 3, 2009
transfusion.
It is difficult to define the information required
and extract it from conventional registers of
clinical data.
Hypotheses 2
Hypotheses 1
The majority of blood needs are related to compensation of acute
or chronic anemia; therefore, if red cell needs were met, this
For a given period and given category of should be sufficient to meet the needs for platelets and plasma
since these can be separated from whole blood
disease, blood needs will be a function of:
Experienced specialist clinicians should be able to assign the
The number of individuals having the disease main groups of patients at risk of transfusion (GPRT) into broad
clinical categories in relation to transfusion needs
during the period
These are the categories that will have a major impact on the total
The proportion of them which develop a need for blood, either because they include
pathophysiological condition requiring many patients requiring a modest amount of blood, or because
they include a
transfusion smaller number of patients each requiring a large amount of
blood
Quantity of blood needed for each category
Disease groups other than the main GPRT should have a small
impact on blood needs .
Model Pilot study - Honduras

The model is based on estimates of four parameters
Country population 6.6 million (2004)
N size of the population;
P(Bi) frequency of the GPRT i in the population;
P(Ai) proportion of patients in the GPNB that will actually
Aimed to include the whole country
have anemia requiring transfusion of red blood cells;
i i average number of units needed per patient.
Included clinicians from hospitals that
The estimate of overall needs is given by provide coverage to 80% of the population
[N x P(Bi) x P(Ai) x i], i=1
1
Information sought from clinicians
List of Groups of patients needing blood
Formal consensus methods have been shown to
be helpful in developing initial estimates of Number/ of individuals in each group
patient populations in the absence of accurate during the defined period
clinical and epidemiological data. Proportion of these having a
pathophysiological condition requiring
blood
Quantity of blood needed for each
pathophysiological condition tha requires
blod
Stage 1:Nominal group concensus Stage 2:Survey

The participants took part in a structured two A survey of 35 clinician (different from the nominal
hour face to face meeting. group) indicated that the original categories were too
To define common ground and maximize areas broad which made estimating the basic parameters very
of agreement, participating clinicians were difficult, 31% of those surveyed felt that the original list
organized by broad clinical specilaties. was incomplete.
No usable information of prevalence was provided since
Experts were asked to discuss among clinicians felt they did not have enough information to
themselves and to list the top 3 clinical give a reasonable estimate.
conditions requiring red cell replacement.
Information on proportion of patients needing blood and
Through a series of scoring, voting and ranking units needed per patient was provided for all nine
methods a list of nine clinical conditions was original conditions.
obtained.
Stage 3:Face to Face Interviews Stage 4 Delphi round

To come to a formal agreement about how specific During the final phase to obtain consensus on the basic
the experts considered the selected clinical parameters an initial letter explaining the Delphi process
categories must be, face to face interviews with 121 and inviting participation was sent by electronic mail to
clinicians from 10 hospitals were conducted. 138 clinicians. 120 responded
By medical doctors that have been trained to Tables containing all the criteria collated from the
request the 3 basic parameters. previous exercise were included with the invitation. If
the clinician had participated in the previous round
Clinicians were asked to answer only on those his/her results were highlighted. Results from other
clinical categories they were familiar with. clinicians were also included but their names were kept
The consensus generated a larger list of clinical confidential. Each clinician was requested to agree with
conditions which were classified by ICD category their previous answers or to change in view of the
with the help of an expert from the local Pan groups response.
American Health Organization office. Averages were obtained for the proportion of patients
needing blood as well as the number of products that the
clinician estimated were necessary for the patient.
Final consensus is included in the following table.
2
Comparison with observational data
- blood request forms Data collected from patient charts
1376 blood request forms from the main reference hospital were
available for review, in 1335 there was a reason for the request
Demographic data: age, gender
68% of the reasons for transfusion matched one of the GPRT Hospital Stay: events, length of stay
Anemia: Hb and Hct at admission and before
The remaining 433 blood request forms fell into two categories: transfusion
unspecified anemia and others.
Presence of acute bleeding
100 transfusion requests were matched to discharge diagnosis. Only Number of patients for whom blood requested
patients who had been transfused were included in this sample. In Number of red cell units requested by the physician
67% of charts the discharge diagnosis coincided with one of the
GPRT. Anemia, diabetes and solid tumors accounted for most of the Number of patients transfused
remaining diagnosis. Number of red cell units transfused
Clinicians were able to select 67-68% of all the clinical conditions for
which patients were transfused during 2004.
Stage 1 result:
Limitations of chart review GPNB defined by Nominal Group
Burns
Neonates with pathology
Many charts were unavailable Gynaecologic and obstetric complications
Deceased patients could not be included Trauma
Orthopedic surgery
as their charts were not sccessible Cardiovascular surgery
Upper GI haemorrhage
Transfusions were poorly documented in Hematological and Hemato-oncological diseases
the charts Non-hematological cancers
Portion requiring
ICD Code transfusion:
Group Concensus Units requested Units transfused
1 Gastric Carcinoma C16 0.80 0.57 0.37
2 Cervical Carcinoma C53 0.72 0.86 0.64
3 Leukemias C91-
C91-C95 0.89 0.73 0.73
GPRT defined by the complete 4
5
Aplastic Anemia
Cardiovascular Surgery
D61 0.77
0.73
0.76 0.74
0.90 0.42
concensus process 6
7
Esophageal Varices
Peptic/ Doudenal Ulcer
I85
K25--K29
K25
0.96
0.45
0.86
0.72
0.76
0.62
8 Chronic Renal Insufficiency N18 0.90 0.65 0.44
9 Abnormal Uterine Bleeding N93.8-
N93.8-93.9 0.19 0.84 0.34
10 Incomplete Abortion 03--06
O 03 0.28 0.94 0.16
11 Postpartum Hemorrhage O 72 0.51 0.86 0.24
12 Premature babies P07 0.11 0.28 0.26
13 Prematurity and Sepsis P36 0.15 0.28 0.23
14 Fracture of the pelvis S 32.1-
32.1-32.8 0.54 0.60 0.38
15 Fracture of the femur S72 0.44 0.60 0.18
16 Myomas D25 0.17 1.00 0.34
17 Gun-
Gun-shot wound Y.24 0.65 0.80 0.22
18 Stabbings Y.28 0.42 0.51 0.19
3
CLINICAL ICD CONSENSUS REQUESTED OBSERVED
Participat Prevalence Prop Px Estimat Need
No. CONDITION CODE UNITS REQUIRED UNITS TRANSFUSIONS
No. CONDICION Clinicians 2004 Need Blood Units/Px
Units/ Px/Yr
Px /Yr
1 Gastric Carcinoma C16 3 4 3
1 Gastric Carcinoma 10 7.16E--05
7.16E 0.80 3
2 Cervical Carcinoma C53 3 5 4
2 Cervical Carcinoma 12 1.20E-
1.20E-04 0.72 3
3 Leukemias C91-
C91-C95 4 6 3
3 Leukemias 12 8.07E--05
8.07E 0.89 4
4 Aplastic Anemia D61 7 8 4
4 Aplastic Anemia 9 3.40E-
3.40E-05 0.77 7
5 Cardiovascular Surgery 4 3 3
5 Cardiovascular Surgery 3 4.50E-
4.50E-05 0.73 4
6 Esophageal Varices I85 4 6 5
6 Esophageal Varices 8 3.34E--05
3.34E 0.96 4
7 Peptic/ Doudenal Ulcer K25--K29
K25 3 4 3
7 Peptic/ Doudenal Ulcer 10 1.89E-
1.89E-04 0.45 3
8 Chronic Renal Insufficiency N18 2 5 3
8 Chronic Renal Insufficiency 3 1.95E--04
1.95E 0.90 2
9 Abnormal Uterine Bleeding N93.8--93.9
N93.8 2 3 2
9 Abnormal Uterine Bleeding 10 2.22E-
2.22E-04 0.19 2
10 Incomplete Abortion O 03-
03-06 2 2 2
10 Incomplete Abortion 6 1.11E-
1.11E-03 0.28 2
11 Postpartum Hemorrhage O 72 3 2 2
11 Postpartum Hemorrhage 11 2.03E--04
2.03E 0.51 3
12 Premature babies P07 0.30 ,10 0.097
12 Premature babies 14 2.65E-
2.65E-04 0.11 .3
13 Prematurity and Sepsis P36 0.30 .13 0.13
13 Prematurity and Sepsis 13 2.55E--04
2.55E 0.15 .3
14 Fracture of the pelvis S 32.1-
32.1-32.8 2 4 3
14 Fracture of the pelvis 19 2.96E-
2.96E-05 0.54 2
15 Fracture of the femur S72 2 3 2
15 Fracture of the femur 22 2.01E--04
2.01E 0.44 2
16 Myomas D25 2 3 2
16 Myomas 6 1.92E-
1.92E-04 0.17 2
17 Gun-
Gun-shot wound Y.24 3 3 2
17 Gun-
Gun-shot wound 13 1.80E-
1.80E-04 0.65 3
18 Stabbings Y.28 2 3 2
18 Stabbings 12 2.26E--04
2.26E 0.42 2
Verifying Clinical data by chart

Results: Prevalence
review
Through all the stages, clinicians were Based on the national database and
unable to provide any information on largest reference hospital database a
prevalence of the selected clinical sample was selected for chart review for
conditions. all 19 clinical conditions ( 5% of all cases
with a minimum sample of 50 charts
For patients with chronic conditions all
hospital admissions for the year 2004
were included
Data from Scottish Transfusion Epidemiology database

Results - verification
Hospital data
Other
Blood request forms: 1376 forms from the main 38%
reference hospital were available for review. Surgical
20%
In 41 the cause for transfusion was not documented.
Of the remaining 1335,

68% of the requests indicated conditions that
matched one of the final clinical groups Haematology
Solid
18%
Tumours
32% requests fell into two categories:
unspecified anemia and others of which
most were anemia, diabetes and solid tumors
4
Analysis Results: comparison of observed data with clinicians
Intraclass correlation coefficient was used to calculate the level of agreement concensus
between the clinician based need estimate versus the calculated need based CLINICAL ICD PREVALENCE MODEL REQUESTED OBSERVED
on units requested or units transfused. CONDITION CODE 2004 ESTIMATE UNITS * TRANSFUSIONS*
Gastric Carcinoma C16 7.10E-05 1169 1099 544
ICC calculated against units requested was 0.750 Cervical Carcinoma C53 1.19E-04 1769 3310 2097
Leukemias C91-C95 8.00E-05 1954 2207 1204
Aplastic Anemia D61 3.37E-05 1245 1325 679
ICC calculated against units transfused was 0.834 Cardiovascular Surgery 4.46E-05 894 826 381
Esophageal Varices I85 3.31E-05 872 1179 868
Peptic/ Doudenal Ulcer K25-K29 1.87E-04 1732 3889 2386
Overall ICC between clinician based estimates and Chronic Renal Insufficiency N18 1.93E-04 2381 4100 1756
hospital records data was 0.951 Abnormal Uterine Bleeding N93.8-93.9 2.20E-04 574 4049 1027
Incomplete Abortion O 03-06 1.10E-03 4230 14200 2417
Postpartum Hemorrhage O 72 2.01E-04 2108 2370 661
The intraclass correlation coefficient ranges from 0 and Premature babies P07 2.63E-04 60 51 45
1, a value of 1 indicates complete agreement Prematurity and Sepsis P36 2.53E-04 78 63 51

Fracture of the pelvis S 32.1-32.8 2.93E-05 217 470 232
Fracture of the femur S72 1.99E-04 1201 2048 491
Residual variability (1 - intraclass correlation coefficient) Myomas D25 1.90E-04 443 3414 886
Gun-shot wound Y.24 1.79E-04 2393 3269 540
is due to true variation between clinicians and Stabbings Y.28 2.24E-04 1292 2640 580
measurement error. TOTAL NEEDS 24612 50509 16846
Units available for transfusion in 2004: 37,000
Strengths Weaknesses
Clinicians were able to define a group of clinical Labour intensive, time consuming,costly
conditions to which 67-68% of all transfused patients Depends on the willingness of clinicians to participate,
could be assigned and on the adequacy of their knowledge of blood
Their estimates of the red cell transfusion requirements utilisation in their own specialties
showed reasonable comparability to those obtained from Clinicians could not estimate prevalence of the
hospital records. conditions identified, so there is dependence on an
Requires the investigators and blood services to consult additional data source
clinicians about the transfusion requirement for their Validation against other data sources depends on the
patients existence, availability and quality of hospital records and
Should engage clinicians to think about blood on availability of human resources to extract data from
requirements about the quality of data that could help to them
assess need, and the adequacy of the data currently Method does not take account of access to health care
available. Additional data on this would be essential to calculate
Anecdotal evidence from the project team that these blood requirements in any country where an important
approaches have been welcomed. proportion of the population lacks access to facilities in
This would seem to be supported by the high levels of which the availability of safe blood transfusion could
clinician participation in the present project benefit the outcomes of GPRT
Conclusions
Clinicians were able to define a group of clinical
conditions to which 67-68% of all transfused patients
could be assigned
Their estimates of the red cell transfusion requirements
showed reasonable comparability to those obtained from
hospital records.
Neither source of data can adequately identify the
adequacy or otherwise of the current supply levels.
To estimate the blood requirements for a given
population at a given time, data such as those
obtained in this study should be combined with
information on access to hospital care
5
Objectives
WHO Expert Consultation on Review parameters in health system and
clinical care which influence requirement of
estimation of blood blood and blood components
requirements Review of existing mechanisms/
methodologies and models of blood
estimation based on regional/country
experience
Day 1
Assess the feasibility and accordingly define
the steps in developing a simple model to
estimate blood needs
Gretchen Stevens
Neelam Dhingra
Global burden of disease database
New estimates due 2011
Overview of current situation Important principles in preparing cross-national
Little progress in the area of predicting blood statistics
requirements Selecting health indicator and metrics
Essential for planning Framework for monitoring health systems
Historical perspective Facility assessments
Various non-evidence based estimations Population-based surveys
Advanced healthcare systems Clinical reporting systems
Donations- 5% of population Correcting for bias in available data
3% of population regular blood donors Estimating and communicating uncertainty
Previous approaches Input uncertainty
Blood usage with different denominators Poor quality data
Per 1000 population
Model uncertainty
Per acute hospital bed
Parameter uncertainty
Peter Olumese Matthews Mathai
Declining falciparum malaria due to nets, Haemorrhage implicated in a high proportion

spraying and ACT of pregnancy-related deaths
Parallel decline in paediatric blood Existing surveys and data on incidence of
transfusions haemorrhage and blood transfusion
Seasonal demand requirements
Potential impact of paediatric blood packs on Access and equity issues
blood supply and safety Costs and planning tool already exists (to be
Evaluation and bulk procurement? (WHO) shared)
(Also protective needle cover) 9 signal functions of comprehensive obstetric
care
1
Pablo Perel
Luz Maria De Regil
Frequency of trauma
Low resolution data on mortality and DALYs (but no/little data on
incidence and morbidity) Global burden of disease database has data on
Increasing in importance anaemia but
Frequency of bleeding in trauma Variable Hb cut offs and definitions
Bleeding probably important cause of death Adjustments required for altitude (1g/dL per 1000m),
Reported transfusion rates differ widely smoking and gestation
Evidence base for transfusion of blood/products in Targeting of public health interventions depends on
trauma prevalence of anaemia in Groups at Risk
Weak/contradictory VMNIS (Vitamin and Mineral Nutritional Information
Interventions to reduce transfusions System
Well conducted clinical trials (e.g. CRASH-2) required Estimates of micronutrient deficiencies at national and
Clinical trials such as CRASH-2 may provide better regional levels
and higher resolution data on frequency of trauma Helminth infections?
and frequency of bleeding in trauma (WOMAN- PPH) Haemoglobinopathies?
Larry Marum Francine/Carolina

Zambia continues to make significant progress in
improving its national blood supply
Patterns of transfusion by hospital type Hema-Quebec: stable, mature and efficient
Tertiary facility
service
High (27%) and changing blood usage Demand defined as units shipped
Unknown and increasing burden of HIV-related High resolution historical data
transfusions Complex models do not seem to outperform
Malaria control and decreasing paediatric nave method
transfusions
Obstetrics- mismatches between pregnancy care and
Where supply meets demand prediction is
blood supply enhanced by inventory committee and good
Potential of SmartCare (patient/donor held electronic communication with hospitals
record)
Brian McClelland Additional thoughts

If RBC needs are met, the needs for other
components should also be met
Clear idea of what we are trying to estimate
Categorisation of GPRT by clinicians using Agreed definitions
consensus methods (multi-stage)
Access, appropriate prescribing
Comparison with observational data (evidence/consensus)
(limitations) Minimum data set
Prevalence/incidence difficult to define Existing data sets
Good statistical correlation between predicted Toolkit cf simple mathematical model
and observed data but large numerical
discrepancies Assessment and communication of
uncertainty of predictions
Engagement with clinicians
No measure of unmet demand (access)
2
Ideal: need Ideal: need
No Morbidity/Mortality
Donor selection / motivation Y Potential Donors Access 100%
(management) X units for transfusion, such that no patient experiences (at
no time) morbidity/mortality due to shortages in RBCs, X is
Donor deferral driving number of potential donors (Y)
Patient need
Donor screening (testing) -Define spatial and temporal specific drivers for transfusion, McClelland
- Malaria (ATLAS, Marum, Olumese, WHO)
Processing of blood products - Trauma (Perel)
- Nutrition (De Regil)
- Maternal Health (Mathai)
- etc
Inventory
- Define methodology (McClelland, Rao, GBD)
Patients X units for transfusion, such that no patient - Perfect transfusion practices
experiences (at no time) morbidity/mortality - Optimal use of alternatives (drugs, colloids, salvage, EACA etc)
due to shortages in RBCs
In fact: X is driving Y - No uncertainty and bias
Real: demand Real: use and demand How many

DALYs* lost
Morbidity/Mortality do we
Donor selection / motivation Y Potential Donors Access ?%
accept?
(management) X units for transfusion, such that current demand is
Use Demand
Anaemia (Fe def. etc), De Regil covered, Y (number of potential donors) and capacity is
Donor deferral driving X
Patients
+ TTI Marker, GDBS; WHO, Marum
Donor screening (testing) - Define spatial and temporal specific drivers for transfusion, McClelland and
others
Processing of blood products Losses in processing, Marum -Define methodology (McClelland, Rao, GBD)
- Expect difficult hurdles (Sarappa)
Outdating, Dcay/Sarappa
Inventory - Uncertainty (parameter (need for distributions,@Risk, R, Stevens), model)
- Bias (Rural vs Urban, etc)
Morbidity/Mortality, ?
Access (,,$)
- Incorporate transfusion practices (McClelland, Marum, )
Patients X units for transfusion, such that current - Availability and use of alternatives (drugs Oxytocin, EACA, etc)
demand is covered
DALYs = Disability Adjusted Life Years = The sum of years of potential life lost due to premature mortality and the years of productive life lost due to disability.
Y and/or capacity is driving X
Proposed definitions Proposed definitions

Demand: The amount of blood that would be transfused if Demand: The amount of blood that would be transfused if
all requests for blood were met (includes inappropriate all requests for blood were met (includes inappropriate
transfusions and unmet demand) transfusions and unmet demand)
Use: The amount of blood actually transfused (excludes Use: The amount of blood actually transfused (excludes
unmet demand) unmet demand)
Need: The amount of blood that would be used if all those Need: The amount of blood that would be used if all those
who needed a bood transfusion were recognised and had who needed a bood transfusion were recognised and had
blood appropriately prescribed i.e. includes unmet demand blood appropriately prescribed i.e. includes unmet demand
and excludes inappropriate transfusions and excludes inappropriate transfusions
1
If you have to forecast,
forecast often.
Edgar R. Fiedler,
The Three Rs of Economic Forecasting:
Irrational, Irrelevant and Irreverent
It is often said there are

two types of forecasts...
lucky or wrong.
Copyright 2003 FarWorks Inc.

Demand Forecasting at Control magazine,
Institute of Operations Management
Canadian Blood Services Good forecasters are not

smarter than everyone
else, they merely have
their ignorance better
organized. Anonymous
Yes, yes, I know that, Sidley everybody
knows that!... But look: Four wrongs
squared, minus two wrongs to the fourth
power, divided by this formula, do make a
right.
Tony Steed
Director, Market Knowledge & Donor Insight
1
2
Key Canadian demographics

CBS in the Canadian context
Health care in Canada is a provincial / territorial responsibility, with federal
Area: 10 million sq. km. (3.3 people government contributions
per sq. km.)
Canadian Blood Service (CBS) is Canadas national supplier of blood and
2 doctors and 4 hospital beds per blood products
1,000 population
Arms length, not-for-profit agency independent of government
High physician utilization rate
(About 80% of Canadian use family Exclusively serve 9 of 10 provinces and all 3 northern territories, who
physician services once a year; collectively fund CBS. (Quebec province operates its own blood service
58% more than twice a year Hema Quebec)
Health services utilization patterns, Regulator: Health Canada, a federal agency
for both individuals and for regions, Global budget; no charge to hospitals
are influenced by age, gender, self- Funding is based on annual estimates of blood requirements
rated health status, education
income, etc.
3 4
Operations overview Why forecast demand in the first place?

Supply Chain Management
Donor Clinic Component Inventory Hospital
Base Events Production & Testing & Issuing Shipments
Production
438,400 Testing 732
Donor Inventory Hospital
Customers R&D Customers Supply Demand
Education
916,000 WB units 4,525 employees 600,000 patient Supply Chain Management Enablers
41,000 17,000 volunteers transfusions a year Donor Recruitment Event Planning Production Planning Inventory Mgmt Customer Mgmt
Donor Retention Locations and Hours Discard Mgmt Inventory Protocol Product Disposition
plateletpheresis units 41 permanent Cost to recipient is
Donor Contact Appointment Booking Recall Mgmt Product Delivery Product Utilization
55,000 plasmapheresis collection sites fully covered under Donor Segmentation Donor Influx & Flow Order Fill Demand Estimates
units 12 manufacturing provincial / territorial Corporate Partners Staff & Volunteer Mgmt Integrated Systems
All units are freely centres government health
donated 3 blood-testing plans Demand Forecasts
2.16 WB donations per centres
donor The collection, manufacturing and delivery of blood products is a complex business,
supported by a host of enablers. Demand forecasts can provide the lead time necessary
to ensure that adequate supply is available to meet expected customer demand. It allows
CBS to maximize service delivery and stakeholder value.
5 6
1
What's behind the demand for blood? How is demand trending?
Total RBC Demand - 52 Week Moving Average O Neg RBC Demand - 52 Week Moving Average
840,000
100,000
830,000
820,000 95,000
810,000
800,000 90,000
790,000
85,000
780,000
770,000
80,000
760,000
750,000 75,000
7
7
05 -27
05 -27
7
06 -27
06 -27
06 -27
7
06 -27
7
07 -27
07 -27
07 27
07 -27
7
7
08 -27
-0 7
7
08 -27
7
09 -27
7
7
7
10 -27
3-2
6-2
9-2
2-2
3-2
6-2
9-2
2-2
3-2
6-2
9-2
2-2
3-2
6-2
9-2
2-2
3-2
6-2
9-2
2-2
20 3-2
20 -09-2
200 -05-2
20 -07-2
20 -11-2
20 -09-2
-2
20 -01-2
20 -05-2
200 7-2
20 1-2
20 -03-2
-2
20 -07-2
200 -09-2
1-2
20 -05-
20 -05
20 -07
20 -11
20 1
20 -03
20 -09
20 -01
20 -03
20 -07
200 -11
20 -03
20 -09
20 -01
200 -05
20 -11
-0
-0
5-0
-1
-0
-0
-0
-1
-0
-0
-0
-1
-0
-0
-0
-1
-0
-0
-0
-1
-0
-0
-1
-0
05
05
05
06
06
06
06
07
07
07
07
08
08
08
08
09
09
09
09
05
05
05
6
06
07
07
8
08
08
09
09
9
09
200
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
The longer-term underlying growth in total RBC demand has been in the order of 2% per
year, but this growth has not always been consistent. Note the slow down in the latter
half of FY0708, the rapid growth throughout FY0809 and the current decline in FY09/10.
Even O Neg growth has eased off in FY09/10 - slowing, but not declining. The
Average units of blood/blood products required per recipient = 4.6 units proportion of total RBC issues which are O Neg has grown from 10% at the start of
FY2004/05 to 11.3% this fiscal YTD
7 8
Who uses blood? Aging: A key influence on blood demand

3,000 200
3,000 200
2,700 180 2,700 180
2,400 160 2,400 160
Units Transfused per 1,000 Population

Units Transfused per 1,000 Population
2,100 140 2,100 140
1,800 120 1,800 120

Population (000)
Population (000)
1,500 100
1,500 100
1,200 80
1,200 80
900 60
900 60
600 40
600 40
300 20
300 20
0 0
0 0 0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+
0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ 2005 2013 2020 2030 Units Transfused per 1,000 FY0506 Blended
2005 Units Transfused per 1,000 FY0506 Blended

The proportion of the population aged 65 years or older is expected to grow from
Almost 57% of the blood transfused in Canada is used by recipients aged 65 years or older 14% in 2009 to 22% by 2030. In addition, the line representing the number of units
transfused per 1,000 population has been shifting upwards over time.
Source: BC Provincial Blood Coordinating Office (PBCO) 9 10
Collision course: Concentration of hospital demand

Demand meets supply Concentration of Demand
Demand Supply Top 100 Medical

Centres
Aging population Aging population

Increasing morbidity Health-cost
Top 50 Medical
Increasing medical containment Centres
procedures Creating new donors
Changing product mix New pathogens
Population growth Increasing deferrals Top 25 Medical
Ethnicity Regulatory issues Centres
Expanding role of the System capacity,

blood business flexibility, scalability 0% 10% 20% 30% 40% 50% 60% 70% 80%
Top 25 Medical Centres Top 50 Medical Centres Top 100 Medical Centres
Global competition for Technology deficit % of RBC Shipments 40% 60% 80%
fractionated products Skills deficit % of Medical Centres 5% 10% 20%
Supply management
Bloods competitive Concentration of hospital customer demand...
brand and profile
The Pareto Principle (80/20 rule) is alive and well at Canadian
Blood Services
11 12
2
Hospital shipments by day Smoothed hospital shipments by day
# of RBC Units Issued to Hospitals per Day # of RBC Units Issued to Hospitals per Day (7 Day Moving Average)
4,500 19,000
4,000 18,000
3,500 17,000
3,000 16,000
2,500 15,000
2,000 14,000
1,500 13,000
1,000 12,000
500 11,000
0 10,000
1
07 /27
07 /23
8
07 /14
07 /09
4
07 /30
07 /26
07 /21
7
08 /12
08 /07
0
08 /25
1
08 /16
08 /12
7
08 /02
08 /28
08 /24
9
09 /15
09 /10
5
09 /03
09 /29
4
09 /20
09 /15
09 /11
6
09 /01
09 /27
3
09 /18
4
20 /04/0
20 /06/1
20 /09/0
20 /12/1
20 /03/0
/3
20 /05/2
20 /08/0
20 /11/1
20 /02/0
20 /04/2
20 8/0
20 /10/2
2/1
1
07 /27
3
07 /18
4
07 /09
07 /04
0
07 /26
1
08 /17
2
08 /07
08 /04
08 /30
5
08 /21
6
08 /12
7
08 /02
8
08 /24
9
09 /15
0
09 /05
3
09 /29
4
09 /20
5
09 /11
6
09 /01
7
09 /23
4
20 /04/0
20 /05/2
20 /07/1
20 /09/3
20 /11/2
20 /01/1
20 4/2
20 /06/1
20 /08/0
20 /09/2
20 /11/1
20 1/1
20 /03/0
20 /04/2
20 /06/1
20 /08/0
20 9/2
20 /11/1
2/1
20 /04
20 /05
20 /07
20 /08
20 /09
20 /10
20 /11
20 /01
20 /02
20 /03
20 /04
20 /06
20 /07
20 /09
20 /09
20 /10
20 /12
20 /01
20 3
20 /03
20 5
20 /06
20 /07
20 /09
20 /09
20 /11
/0
/0
/0
/1
20 /04
20 /06
20 /08
20 /09
20 /10
20 /12
20 /02
20 3
20 /03
20 /05
20 /07
20 /09
20 /10
20 /12
20 /02
20 /03
20 /05
20 /07
20 /09
20 /10
/0
/0
/0
/0
/1
07
07
07
07
08
08
08
08
08
08
09
09
09
09
07
07
07
07
07
07
08
08
08
08
08
08
09
09
09
09
09
09
09
20
20
Smoothing daily shipments make it easier to identify that customer

The number of units shipped varies from one day to the next and
shipments have been growing over time and that there are troughs
typically ranges from between 500 and 3,500 units.
and peaks in demand surrounding the statutory holidays.
13 14
Distribution of demand by ABO Rh Demand growth rates by ABO Rh

Percentage of RBC Units Issued to Hospitals by ABO Rh Yr/Yr Growth in Demand 2009 vs 2008
2.0%
100%
90%
1.0%
80%
70% Total AB- 0.0%
Total B-
60% Total A-
-1.0%
Total O-
50%
Total AB+
40% Total B+
-2.0%
Total A+
30% Total O+
-3.0%
20%
10%
-4.0%
0% O+ A+ B+ AB+ O- A- B- AB- Blood
08 r
09 r
20 ay
07 v
20 ec
20 ay
08 v
09 c
20 ay
09 v
ec
07 l
20 ep
20 /Feb
08 l
20 ep
20 eb
09 l
20 ep
20 un
20 an
20 un
20 /Jan
20 un
07 t
08 t
09 t
07 r
08 r
09 r
20 ug
20 ug
20 ug
Group Rh
20 /Ju
20 /Ju
20 /Ma
20 /Ju
20 /Oc
20 /Oc
c
20 /Ap
20 /Ap
20 /Ap
20 /No
20 o
20 e
20 /No
/M
/O
/D
/N
/D
/D
/F
/M
/S
/M
/S
/M
/S
/J
/J
/J
/J
/A
/A
/A
07
08
09
07
07
07
08
08
08
08
08
09
09
09
09
07
08
09
20
20
20
Not all blood groups are in equal demand. Not all blood types grow at the same rate.
Almost 70% of all demand is concentrated in O Pos and A Pos. Demand for B+ and AB+ is declining; growth rates for Neg Rh are
higher than for Pos.
15 16
Customer demand forecasting process Why this process?

Preliminary
Top Down Top
Forecast:
Down Forecast:
Product level of detail as required Provides Top-Down and Bottom-Up perspectives
Feeds high level Budget sizing,
Collection & Production target setting
Updated Top
Down Forecast: Provides Base-Line and Overlay functionality
Utilizes one
two months
month ofofcurrent
currentyear
yeardata
data
Strategic/ Time Series

Preliminary
Quarterly updates
The past is generally a good predictor of the future
Executive Analysis
May Top Down
Direction
(Overlays)
(Historical
Data Trends)
Forecast
Lots of detailed hospital shipment data is readily
Top Down/
Bottom Up
Strategic/
Executive
Official
Demand
available (hospital level of detail; by product; daily,
Rationalization Rationalization Forecast
weekly, monthly, quarterly, annually)
Prepare & HCSR s
Distribute
Hospital
Canvass
Review
Worksheets
with Hospital
Preliminary
Bottom Up
Forecast
Comprehensive clinical driver data is not currently
Worksheets
Bottom Up Forecast:
Contacts
available (how many transfusions; what types of
RBC, Platelets & Plasma procedures; etc.)
Feeds detailed Budget distribution,
Ju
JuneJune
& Nov Collection & Production tactics Nov/Dec Clinical driver data available has not been particularly
Utilizes six months current year data
effective at forecasting variations in customer demand
(still need to forecast the clinical drivers)
Blend of Top-down (statistics based) and Bottom-up (customer
canvass based) forecasting techniques.
17 18
3
Model selection process Volatility of demand ...
12 Month Moving RBC Issue Yr/Yr Growth Rates Monthly RBC Issue Yr/Yr Growth Rates
4.0% 14.0%
Damped trend with multiplicative 3.5%

12.0%
10.0%
seasonality models because they 3.0%

8.0%
remain finite and flatten out over time 2.5%

6.0%
2.0% 4.0%
Weekly data model (R-squared = .45) 1.5%

2.0%
outperformed monthly data model (R- 1.0%

0.0%
squared = .26) 0.5%

-2.0%
-4.0%
Model with event adjustment variables 0.0% -6.0%
ct
ct
ct
ct
ct
ct
ct
ct
pr
ul
pr
ul
pr
ul
pr
ul
pr
ul
pr
ul
pr
ul
n
an
an
an
an
n
Ju
pr
ct
ct
ul
Ap
l
(R-squared = .83) outperformed
an
Ju
Ja
Ja
Ja
Ja
Ap
/O
/O
/O
/O
/O
/O
/O
/J
/J
/J
/J
/J
/J
/J
/A
/A
/A
/A
/A
/A
/A
Ja
O
/O
/J
/J
/J
/J
/J
/A
06/
/J
08/
08/
05
08
09
05
06
08
09
05/
07/
07/
08/
09/
05
06
09
05
06
08
09
07/
05
06
09
05
06
08
09
07/
07
07
08
09
05
06/
07
07
06
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
weekly model
Higher level model produced best fit
so ABO Rh distributions were made
using a Top Down Multiple Level Annualized year over year growth rates ranged from
model
3.8% in March 2009 to 0.2% in December 2009.
ABO Rh forecasts were further
distributed by Region Underlying trend is in the neighbourhood of 2% per
year, but exponential smoothing models will weight
recent data more heavily than past data.
19 20
Forecast model performance ... Inventory - the buffer between supply & demand
Daily Red Blood Cell Inventory Levels
20,000
35,000
19,000
32,500
18,000
30,000
17,000
27,500
16,000
25,000
15,000
22,500
14,000
20,000
13,000
17,500
12,000
15,000
11,000
12,500
10,000
10,000
1
1
-0
-0
8-0
0-0
-0
-0
-0
6-0
-0
0-0
-0
-0
-0
-0
-0
0-0
2-0
-0
04
06
12
02
04
08
12
02
04
06
08
02
7-0
7-1
8-0
8-1
9-1
9-1
7,500
7-
7-
7-
8-
8-
8-
8-
9-
9-
9-
9-
0-
0
1
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
5,000
Actual Budget View
1
01
01
01
01
01
01
01
01
1
-0
-0
0-0
-0
4-0
-0
4-0
-0
-0
1-0
-0
4-0
-0
-0
-0
1-0
7-
0-
7-
0-
1-
7-
0-
1-
04
07
01
01
07
10
04
04
07
10
-1
-0
-0
-1
-0
-0
-0
-1
-0
-0
-0
-1
-0
-0
-
-
04
04
04
05
05
05
05
06
06
06
06
07
07
07
07
08
08
08
08
09
09
09
09
10
The exponential smoothing model does a reasonable job of predicting future
Volatility in demand & supply can lead to unavoidable imbalances between the two.
demand - at least at an aggregate level of detail. The budget view remains
Inventory acts as a buffer to offset these imbalances, shrinking when demand outpaces
fixed for each fiscal period, but quarterly forecast updates are reviewed.
supply and growing when the opposite occurs. A minimum of 5 days of inventory on
hand for each ABO Rh type is recommended.
21 22
Demand forecasting - just the beginning ... When detailed demand data isn't available ...
Extraneous: RBC Units Issued per 1,000 population by fiscal period
Technology
Weather (Number)
Regulation
Trauma Q2 2009/10 Issued CAGR per
Labour Action 2004/05 2005/06 2006/07 2007/08 2008/09
Annualized CAGR 1,000 pop
Canada (Excl Quebec) 31.2 31.4 31.7 31.7 32.7 32.3 1.8% 0.7%
CBS Strategies/ Demand Forecast Comparison of Newfoundland and Labrador 37.7 37.7 36.9 37.3 35.9 33.5 -2.7% -2.4%
Goals/Tactics Actual Results to Prince Edward Island 25.5 28.2 26.5 28.2 30.9 26.8 1.4% 1.0%
Target Nova Scotia 35.3 34.8 35.9 33.3 33.7 32.8 -1.4% -1.5%
New Brunswick 30.7 31.3 32.8 32.4 33.4 32.4 1.0% 1.1%
Quebec 29.2 29.1 29.2 29.6 29.9 29.8 1.1% 0.4%
Ontario 31.4 31.3 31.2 30.9 32.3 32.3 1.5% 0.6%
Aging Population Hospital Expansion/ Manitoba 36.9 36.7 38.0 38.5 37.4 36.9 0.8% 0.0%
Contraction Saskatchewan 36.0 36.6 37.3 37.8 38.5 38.6 2.1% 1.4%
Alberta 30.4 31.7 32.1 31.6 32.2 31.0 3.0% 0.4%
Analysis of
British Columbia 27.2 28.1 28.8 30.5 31.0 31.0 3.8% 2.6%
Key Drivers Yukon Territory 26.4 25.8 30.1 26.1 27.9 26.3 1.5% -0.1%
Recommended Northwest Territories 21.6 19.2 17.4 17.2 17.3 19.1 -2.4% -2.4%
Actions Cause & Effect
Nunavut 10.1 9.7 9.2 8.4 9.6 9.4 -0.1% -1.6%
Formal Results
Review
Population statistics and forecasts are usually readily available. Relating hospital
demand to population growth will explain some, but not all of the growth in
Demand forecasting is just the beginning. Performance during the year must demand. Refining population estimates to weighted cohorts (e.g. by age band)
be closely monitored, analyzed and reviewed and lead to the development would likely yield improved demand forecasts. Provisioning for non-population
of appropriate corrective actions or adjusted goals and strategies. related overlays (e.g. increased health care funding) could also be considered.
23 24
4
Thank you! Questions?
Tony.Steed@blood.ca
25
5
Background Hong Kong
Area: 1,103 square km
Population: 7 million
Estimating blood demand Chinese 95%
people/ km2
Overall density: 6,300 people/
methodology practiced in HK Total health expenditure: 5.1% of GDP
Public health expenditure:
expenditure:
WHO Experts
Experts Consultation on 2.6% (2005/
(2005/06) (US$4.13 billion)
Estimation of Blood Requirements
Requirements Birth rate: 11.3 live births per 1000 population
03-
03-05 February 2010, WHO, HQ, Infant mortality rate: 1.8 per 1000 live births
Geneva
Life expectancy: male 79.3 yr, female 85.5 yr
Che-
Che-Kit Lin
Hong Kong Red Cross Blood Transfusion Service, Hospital Authority
Background - BTS Yr 2008 Blood Collection Statistics

Hong Kong Red Cross initiated the voluntary
non-
non-remunerated blood donation programme in WB collection = 206636 units
1952 Plasmapheresis = 2114 units
BTS established in 1984 Plateletpheresis = 868 units
Serving 20 public & 12 private hospitals Autologous WB = 130 units
Since 1991, BTS has become part of the public % of age eligible trade population donating =
hospital system (Hospital Authority) 3.3%
Public hospitals account for 90% of blood Donor deferral rate = 13%
consumption Average age of donor
Supply of blood components is free of charge Male = 39 , Female = 36
Hong Kong Red Cross Blood Transfusion Service, Hospital Authority Hong Kong Red Cross Blood Transfusion Service, Hospital Authority
Annual Blood Collection
1
Yr 2008 Blood Supply Statistics

WB derived Red Cell issued = 192,467 units;
Standard Red Cell (derived from 470ml of WB) issued
per 1000 population = 25.0 units
WB derived Plt issued = 127,152 units
One adult equivalent plt dose issued /1,000 population
= 4.54 units
WB derived FFP issued = 57,330 units
250-
250-300 mL equivalent FFP issued /1,000 population =
6.25 units
Average age of WB+RBC on issue = 14.7 days
Frozen Plasma For Fractionation Blood Inventory and Distribution

Management
Maintain optimal inventory levels in BTS and HBB
at all times with minimal expiration
First-in-
First-in-first-
first-out - based on the work of Cohen
and Pierskella 1975: application of management
science and mathematical inventory theory to
control blood inventory management in a regional
blood bank
(Cohen MA, Pierskella WP: Management policies for a regional
blood bank. Transfusion 1975;15:58-
1975;15:58-67)
2
Inventory in BTS and HBBs BTS WB/RBC inventory levels
BTS aims to maintain inventory of WB/RBC 5 Level Description Group O+ Group A+ Group B+ Group AB+
days of supply
I Desirable 1400 800 800 Not applicable
Stock holding of WB+RBC (all blood groups) at BTS as patient can
- 8.89 days (yr
(yr 2008 annual average) II Safe 1100 - 600 - 799 600 - 799 receive RBC of
any ABO group
Each HBB holds 1399
holds about 3-
3-5 days of hospital
III Action 600 - 400 - 599 400 - 599
consumption 1099
Inventory level in HBB is set out in the Blood IV Alarm 350 - 599 200 - 399 200 - 399
Supply Agreement
Agreement which is reviewed annually V Dangerous < 350 < 200 < 200
HBB Inventory Allocation Policy Electronic Blood Transfusion Network System

If BTS inventory dropped below safe level,
level, all HBB
inventories will be reduced on a pro-
pro-rata basis with Connect with blood banks in all public hospitals
hospitals
extra requests to be considered on case-
case-by-
by-case basis enable viewing of real-
real-time inventories in both
RBC Plt FFP BTS Commitment
BTS and HBBs
I I I Top up Clients stock as requested, provided that Clients RBC, The system also enables:
Platelet & FFP stocks do not exceed 110%, 100% and 100% of
agreed levels respectively. Blood ordering & shipment
II II II Top up Clients RBC, platelet & FFP stocks to at least 90%, 70% & Product tracking
70% of agreed levels respectively.
III Top up Clients RBC stock to 70% of agreed level. Checking of patients
patients previous cross match results
< III < II < II Further reduce stock supply and handle difference by case requests.
Client Satisfaction
250000 120 Blood Utilization Statistics
A n n u a l T a r g e r ( U n it )
200000 100
P ercentage
80
150000 Annual Target
60 Achievement (%)
100000
40 Client Satisfaction (%)
50000 20
0 0
P l a te le t
P l a te le t
P l a te le t
P l a te le t
P l a te le t
P l a te le t
P l a te le t
P l a te le t
P l a te le t
R e d c e ll
R e d C e ll
R e d C e ll
R e d C e ll
R e d C e ll
R e d C e ll
R e d C e ll
R e d C e ll
R e d C e ll
P la sm a
P la sm a
P la sm a
P la sm a
P la sm a
P la sm a
P la sm a
P la sm a
P la sm a
2001 2002 2003 2004 2005 2006 2007 2008 2009
3
Red Cells Utilization by Specialties in Public Hospitals
Year 2004 2005 2006 2007 2008
M&G 64279 65436 66460 71672 71577
Surgery 31873 31906 30124 31648 31533

Paediatrics 17597 18040 17860 18011 17510
Orthopaedics 11598 11374 11550 12746 13709
A&E 10949 11205 11091 11395 10490
O&G 7397 7350 7596 8610 8908
Oncology 5161 5167 5130 5191 5425
Chest 2438 1603 2892 2865 2940
Neurosurgery 1624 1784 1703 2147 1865
ICU 2735 1208 701 856 947
Other 5064 7852 4895 6175 8927
Total 160715 162925 161126 171316 173608
Platelet Utilization by Specialties in Public Hospitals FFP Utilization by specialties in Public Hospitals
Year 2004 2005 2006 2007 2008 Year 2004 2005 2006 2007 2008
M&G 46531 47504 53322 57346 57378 M&G 21393 21686 23316 22082 19906
Surgery 12665 14922 14577 15102 14120 Surgery 15020 15751 17069 17102 16849
Paediatrics 9374 10848 10940 11393 8403 Neurosurgery 1443 1504 1970 1829 2577
Oncology 8204 6913 6806 7914 7526 Orthopaedics 1731 1636 1651 1771 2252
Chest 2109 3072 2776 2852 2639 Chest 2164 2711 2277 2183 2157
Neurosurgery 1288 1680 1494 2139 2473 Paediatrics 2426 2209 1999 1888 1747
Orthopaedics 1519 1466 1582 1818 1939 ICU 5497 1960 961 1251 1618
O&G 1479 1317 1278 1384 1385 O&G 1003 932 936 1060 1178
ICU 4338 1960 605 929 1162 Oncology 1091 1413 1338 990 1121
A&E 245 373 404 338 310 A&E 333 373 479 405 441
Other 6922 6066 948 7893 10197 Other 742 1389 628 167 1609
Total 94674 96121 100191 109108 107532 Total 52843 51564 52764 51256 51455
WB+RC Expiration in Public Hospitals

Plt & FFP Expiration in Public Hospitals
2004 2005 2006 2007 2008

2004 2005 2006 2007 2008
WB+RC issued 159564 161660 160152 170625 172722 Platelet issued 88483 90641 94852 103197 100891
% of Platelet 6.54% 5.70% 5.33% 5.42% 6.18%

Total Expiration 1151 1265 894 691 886 Expiration
FFP issued 52843 51564 52764 51256 51411
Overall % of 0.72% 0.78% 0.60% 0.40% 0.51%
WB+RC % of FFP 0.37% 0.52% 0.23% 0.21% 0.33%
Expiration Expiration
4
Benchmarking Blood Utilization in Public Production planning
Hospitals
Estimate annual blood requirement
A territory-
territory-wide peer-
peer-to-
to-peer review of blood Review quarterly
quarterly demand of blood components
utilization and expiration by hospitals and various of different blood groups,
groups, blood collection, blood
clinical specialties. inventory and wastage and plan short term
BTS provides report every six months. adjustments.
Hospital Transfusion Committee has the responsibility
to review its hospital
hospitals performance and implements
Monthly communication with HBB on matters
improvement measures accordingly. that potentially affect short-
short-term demand
Weekly stochastic forecasting based on historical
demand and communication with HBB to plan
daily production of components
Estimation of Annual Blood Requirement Driving Factor
Annual production planning by causal

causal
forecasting
Review previous 12 month utilization
Determine
Determine driving factor
Determine the utilization trend factors
Determine factors
Hospitals
Hospitals Forecast of WB/RC Demand BTS Calculation of hospital WB/RC Demand
WB/RBC issued to hospitals for the

Commence the process in November each year
previous 12 month period from
Request each hospital to participate by November to October,
October,
completing a questionnaire to forecast demand
adjusting
adjusting for shortage of WB/RBC
for next year based on:
Previous 12 month utilization
stock replenishment during the period,
projected changes in demand due to factors such
hospitals
hospitals forecast increase or decrease
as organic growth, changes in level of services,
in blood demand for next year due to
changes in patient population served and patient expansion or contraction of services
demographics, etc. and population growth.
growth.
5
BTS Calculation of WB/RBC Demand BTS Calculation of WB/RBC Demand
for 2010-
2010-2011 for 2010-
2010-2011
= [WB/RBC issued in 2009 x population growth /
WB/RBC issued to clients in the period of
November 2008 to October 2009: 200,027 units Blood stock replenishment satisfaction in 2009]
+ Clients
Clients estimated potential increase/decrease
Average satisfaction rate of WB/RBC stock
replenishment during the period of November in WB/RBC demand for 10- 10-11
2008 to October 2009: 99.66%
Clients
Clients estimated potential increase/decrease in = (200,027 x 1.004 / 0.9966) + 20
annual WB/RBC demand = 20 units
HK population growth: 0.4%
= 201,532 units
Baseline of WB/RBC Demand

Demand for
2010/2011 Additional Factors that Need to be
Incorporated in the Final Calculation
Hospitals
Hospitals demand forecast : 205,682 units
Average production discard rate in the
BTS calculation : 201,532 units previous three years i.e. 2006-
2006-2008 = 6.63%
Since hospitals
hospitals forecast is greater than the BTS Average WB/RBC expiry rate in the previous
calculation,
calculation, it will be taken as the baseline three years, i.e. 2006-
2006-2008 = 0.013%
demand for further calculation to determine the
annual WB/RBC demand for 2010/2011.
2010/2011.
Final Estimation of Blood Demand for

Actual Annual Collection against Target
2010-
2010-2011
= Baseline demand / (1 - average production discard rate
- average WB/RBC expiry rate)
= 205,682 / (1 - 0.0663 - 0.00013)
= 220,318
Annual blood collection target for 2010-

2010-2011
= 220,000 2%
= 215,600 to 224,400 units
6
Long Term Forecast
Forecast Effect of the Silver Tsunami
Tsunami
In yr 2009, median age

Population of population is 40.9
65
Increase 25% in
next 25 years
with 12.8% aged
In yr 2033, 26.8% of
65

population will be
% Utilization of different age

age groups in
various specialties
MED SUR ORT ICU ONC PAED CTS Other*
<1m 100% 44.4% used by patients aged > 70
< 1 yr 1.1% 98.9% 57.3% used by patients aged > 60
1 10 1.8% 0.5% 90.2% 7.4%
11 20 5.0% 4.2% 1.8% 2.0% 1.4% 72.1% 0.9% 12.6%
21 30 20.8% 4.3% 2.3% 4.3% 1.1% 39.4% 1.1% 26.8%
31 40 31.2% 9.6% 5.1% 12.1% 2.8% 7.7% 1.2% 30.4%
41 - 50 40.4% 15.6% 4.8% 11.6% 6.4% 1.2% 20.0%
51 60 39.4% 20.6% 7.2% 10.1% 9.2% 1.7% 11.8%
61 70 48.0% 23.2% 5.2% 12.5% 4.7% 2.4% 4.0%
> 70 51.6% 23.2% 11.8% 7.2% 1.7% 1.0% 3.6% Unique patients 10542 11129 10980 11575 11464 12413
Hong Kong Red Cross Blood Transfusion Service, Hospital Authority Hong
Admission Kong
14515 Red Cross
15079 Blood14999
Transfusion
15790Service, Hospital16879
15801 Authority
Conclusion
Involving hospitals in forecasting short term demand
seemed to be an effective approach
To achieve maximum use of the precious gifts of life
from blood donors and to ensure all demands are met
timely,
timely, it is important to understand the complex
interrelations of supply and demand, factors that
impact upon them and have all parts and parties of
supply chain working together
Blood utilization is increasing
Aging of population will have significant impact on the
demand for blood supply
7
Needs Assessment:
How can we use this data
Epidemiology of Transfusion in
Scotland in UK or similar setting
Who gets transfused and why?
in resource restricted setting
WHO Workshop on Estimation of

needs for transfusion
Geneva February 3 5
2010
WHO Needs assessment Feb 1 WHO Needs assessment Feb 2

2010 2010
It should be part of our job to know

who gets the blood, why and how much Presentation
Develop a sustainable system providing
information about the clinical use of blood
Evolution and why
Estimate current requirements
AIMS
Predict future requirements
Methods
Identify variations in practice
Results
Encourage clinical review of blood use
What next
Understand demographic, socioeconomic and epidemiological influences
on blood use

2010 2010
Evolution
2000: STEP Introduction of tailored information system software
to enable the easy collection and analysis of transfusion data for prompt
Scottish Transfusion Epidemiology Project feedback to clinicians and Hospital Transfusion Committees. This would
involve the construction of a merged record that is a by-product of routine
clinical practice (held on the existing hospital patient administration system)
and data extracts from the local hospital laboratory system (and in some cases
operating theatre).
2005: STED
Scottish Transfusion
Epidemiology database
2010 2010
1
Why
Progressively uncovered problems in
data quality
2008 NHSS-AFB terminology and definitions for data items
NHS Scotland opacity of Progesa data
heterogeneity of hospital blood bank IT systems
Account for Blood heterogeneity of setting up [same system used in
different ways]
human resource for regular data extraction and QA
Etc etc

2010 2010
It should be part of our job to know

who gets the blood, why and how much
Develop a sustainable system providing

information about the clinical use of blood
Predict requirements
Identify variations in practice
Encourage clinical review of blood use
Understand demographic, socioeconomic and

epidemiological influences on blood use
Biggin K et al Transfusion 2009
2010 2010
Extract patient - specific transfusion data from 2002/03

2005/06 from majority of Health boards
STED
Link using Patient Identifiers with hospital inpatient and
Methods in outline day case records (Scottish Morbidity Record)
Prepare reports for clinical users, other stakeholders
Disseminate

2010 2010
2
Associating transfusion record with clinical
Structure of linked data Date rule
episodes Clinical rule
Patient had 6 admissions,28 procedure codes
Blood bank transfusion records Inpatient episode records 15 diagnosis codes and 4 transfusion episodes
January
RBC,Plt, 4 Procedure and 6 diagnosis fields
Patient Id RBC
Cyro,FFP
RBC, Plt, Record linkage Procedure and diagnosis fields Patient Id

Cyro,FFP
RBC
RBC,Plt, Date
Cyro,FFP Procedure and diagnosis fields
RBC,Plt,
Cyro,FFP RBC
December
RBC

2010 2010
Transfusion Records
Date of Transfusion
Non Health Service On line report on red
RBC Used Data sets cell use for surgical
Platelets used procedures
FFP used General Register Office
Death Records
Cryo used
Census Data
Other Health
Service data
sets Inpatient and Day
Patient case Records (SMR1)
Maternity
Date of admission
Neonatal
Date of discharge
Outpatients Cancer Registry
Procedures
Ward Watcher
(ICU) Diagnosis
Incidence Date Consultant responsible for
care
SCI Morphology/Stage/
Hospital of Treatment
referrals A&E Tumour size
PIS Hospital WHO Needs assessment Feb
Diagnosis 15 WHO Needs assessment Feb 16
Prescribing 2010 2010
Results Red cell use in coronary artery bypass graft

Transfusion for surgical procedures
Coronary artery bypass grafts
Red cell use for defined
4 Substantial reductions
surgical episodes (OPCS)
in one of the 3 cardiac
3.5
Utilises ~20% of total red cell use surgery units,
per year (~38,000 units)*
RBC units per procedure
3 associated with
intensified blood
Substantial reduction over period 2.5
saving initiatives
2003-2006 including cell salvage
2
Reduced variation among clinical 1.5

units
Variation between units
1
Cardiac: Coronary Artery reduced
Bypass Grafting 0.5
Vascular: Elective repair of
aortic aneurysm 0
Orthopaedic: Primary total hip Grampian Greater Lothian Grampian Greater Lothian
replacement
Glasgow Glasgow
2003 2006

2010 2010
3
Red cell use in aortic aneurysm repair Red cell use in primary total hip replacement
Elective repair of aortic aneurysm Primary total hip replacement
7 1.4
Substantial reductions Reductions in all health

1.2
6 across all health boards boards, some greater

than others.
1
5
4
variation variation in 0.8
practice between health Tayside increase is an
boards continues but 0.6 artefact of an identified
3
reduced anomaly with the source
2
0.4 data for 2006 and should
be ignored.
0.2
1
0
0
Fife
Tayside
Fife
Tayside
Greater Glasgow
Greater Glasgow
Ayrshire & Arran
Gram pian
Highland
Lanarkshire
Lothian
Ayrshire & Arran
Gram pian
Highland
Lanarkshire
Lothian
Borders
Western Isles
Borders
Western Isles
Ayrshire & Arran
Ayrshire & Arran

Fife
H ighland
Lothian
Tayside
Fife
H ighland
Lothian
Tayside
Grampian
Lanarkshire
Grampian
Lanarkshire
Greater Glasgow
Greater Glasgow
2003 2006 2003 2006

2010 2010
Example of STED report: red cell use in a medical condition lymphoid leukaemia
Total cases with first

Red cell use for patients with haematological diagnosis of C91 during
malignancies 1996 -2002
Total mortality 1996 -
2002
Conditions included (ICDM 10)

Lymphoma
Myeloma
Myeloid Leukaemia
Lymphoid Leukaemia
Other Leukaemias
Malignant immunoproliferative disease
Other & unspecified malignant neoplasms of lymphoid,
haematopoietic & related tissues
Utilises ~18% of total red cell use per year (~35,000 units)* Incidence Prevalence
Red cell use data
Annual
mortality
* STED Data for 2006 WHO Needs assessment Feb 21 WHO Needs assessment Feb 22
2010 2010
Red cell use for patients with Summary

diagnoses of malignant
tumours
Colon, rectosigmoid junction Other
Conditions described 38%
Bronchus & lung & rectum Cervix uteri Surgical
Prostate Corpus uteri 20%
Breast Liver & intrahepatic bile ducts
Stomach Thyroid & endocrine glands
o Small intestine
Ovary Eye, brain & other parts of
Oesophagus CNS
Kidney Testis
Pancreas Trachea Haematology
Solid
18%
Tumours
Utilises ~24% of total red cell use

per year (~45,000 units)*
* STED Data for 2006 WHO Needs assessment Feb 23 WHO Needs assessment Feb 24
2010 2010
4
Which clinical conditions do we still
have to label as other?
Demographics
patients who have multiple admissions
with several diagnoses
many have markers for gastrointestinal & AGE

liver disease.

2010 2010
The Scottish Population

Projected change in age structure
Population in Scotland Figure 4 The projected percentage change in age structure of Scotland's population,
2004-20311
6,000,000 100
+75%
5,000,000 Total Scotland 80
Population
4,000,000 Male 60
+39%
3,000,000 Female
40
2,000,000
20
1,000,000
0
0
-20 11%
-12% -
02
04
06
08
10
12
14
16
18
-15% -18%
20
20
20
20
20
20
20
20
20
-40
Year 0-15 16-29 30-44 45-59 60-74 75+

1
2004-based pr ojections Age

2010 2010
RBC Units Transfused per RBC Units Transfused

1,000Population by age/sex
Scotland (Excl FV) RBC units tx per
by age band, 2005/06
1,000 population by age/sex
Male Fema le RBC units transfused per 1,000 population by agegroup
Age ba nd 2003 2005 2003 2005
0-4 19.1 19.5 18.6 14.7
and sex (all Scotland , exluding FV)
5-9 3.2 2.8 2.7 2.8
10-14 3.9 4.4 3.0 3.7 300
RBC units per 1,000
15-19 9.7 7.3 7.8 7.3

250
20-24 6.5 10.5 10.5 9.0
population
25-29 10.2 10.6 13.4 13.3 200

30-34 12.0 10.4 20.0 20.3 Male
35-39 14.9 14.3 16.9 16.9 150
40-44 20.0 15.1 18.6 14.9 Female
100
45-49 28.2 22.8 20.7 21.5
50-54 42.5 32.4 28.3 27.1 50
55-59 54.1 54.8 39.4 40.4
60-64 83.4 76.4 55.7 47.7 0
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
>=85
65-69 117.9 117.2 86.0 78.9

70-74 179.9 151.0 112.1 112.1
75-79 212.0 213.7 131.7 135.2
Age group
80-84 241.2 237.0 172.2 162.6
>=85 277.5 273.3 188.1 179.2
Total 44.3 42.2 40.3 39.1

2010 2010
5
Population Change in Europe
RBC Projections for Scotland (excluding Forth Valley), 2006-2018
Figure 9 Projected Percentage Population Change in Selected European Countries, 2004-2031
230,000 30
RBC projection
20
220,000 (2005 HB rates
applied to
population 10
Percentage change
210,000 estimates)
0
RBC projection
200,000
(age/sex rates -10
applied to
190,000 age/sex -20
population
estimates)
-30
180,000
E stonia
E U25
E U15
Cyprus
Austria
England
Hungary
Italy
Portugal
P oland
Finland
Spain
Bulgaria
Latvia
Romania
Lithuania
Czech Republic
E U new members
Slovakia
Germany
Slovenia
Denm ark
Netherlands
France
Wales
M alta
Belgium
Ireland
Luxem bourg
Greece
Northern Ireland
Sweden
UK
Scotland
170,000
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Source: GAD (UK and constituent countries) and Euros tat. Note: Euros tat als o produce an alternative UK projection not shown here.

2010 2010
Red cell use/1000 population by health board
Orkney Islands
17.9
Shetland Islands
21.6
Socio-economic factors Western Isles

32.2 Grampian
36.0
Highland
31.2 Tayside
45.4*
Forth Valley
ND Fife
39.2
Greater Glasgow
57.3 Lothian
DEPRIVATION Lanarkshire
43.6
32.6 Borders
35.7
Ayrshire & Arran
33.8
Dumfries & Galloway Context: the population

31.7* demographics & clinical
2006 practice of each area.
(* 2005)
WHO Needs assessment Feb 33

Why?
WHO Needs assessment Feb 34
2010 2010
In patient/day case records per 1000 population RBC units transfused per 1,000 Inpatient/Daycase
records
RBC units tx per 1,000
Inpa tie nt/da ycase records
Inpatie nt/Da ycase re cords
per 1000 popn
He alth Board 2003 2005
Hea lth Boa rd 2003 2005 Argyll & Clyde 194.0 149.4
Argyll & Clyde 205.3 214.5 Ayrshire & Arran 141.1 138.2
Ayrshire & Arra n 236.5 247.5 Borde rs 169.9 161.6
Borde rs 196.2 217.4 Dumfrie s & Ga llow a y 143.8 146.6
Dumfrie s & Gallow a y 206.8 214.5 Fife 185.5 187.7
Fife 196.5 204.4 Grampia n 179.3 172.6
Gra mpia n 213.8 212.9 Grea te r Gla sgow 171.9 155.3
Gre ater Glasgow 340.4 349.5 Highla nd 127.4 114.9
Highla nd 261.5 269.4 La na rkshire 147.9 135.5
Lanarkshire 238.0 238.3 Lothia n 213.8 189.5
Lothia n 226.5 236.8 Orkne y 130.1 125.4
Orkne y 162.4 175.4 Shetla nd 113.2 18.6
She tla nd 163.6 192.7 Ta yside 200.2 183.3
Tayside 240.8 242.5 We ste rn Isle s 199.1 146.0
Scotla nd Ex cl FV 177.4 161.5
We ste rn Isles 249.6 229.5
Scotla nd Excl FV 244.8 251.5
2010 2010
6
Scottish Index of Multiple Deprivation 2006
" Deprivation takes many different forms in every known society. People
can be said to be deprived if they lack the types of diet, clothing, housing,
household facilities and fuel and environmental, educational, working and
social conditions, activities and facilities which are customary, or at least
widely encouraged and approved, in the societies to which they belong.
Townsend, P (1987) Deprivation, Journal of Social Policy 16 (1) pp 125-146
37 indicators
Income, Employment, Crime, Education, Health (including standardised
mortality ratios), Housing, Geographic access to services
SIMD 2006 divides Scotland into 6,505 data zones with median
population size of 769. These are ranked from 1 most deprived to 6,505
least deprived

2010 2010
Greater Glasgow: Scottish Index of Multiple Deprivation Alcoholic Liver Disease

2002/03 2005/06
(ICD10 K70)
Greater Glasgow Health Board Area Red Blood Cell Units Tra nsfused for patients with ALD*, 2002-2005
No Tx
No Tx patients with RBC/
Inpatient/ Platelet Patients with ALD per RBC Units Tx
ALD 1,000 Transfused Patient
SIMD 2006 Daycase RBC units units FFP units Cryo units Transfused Ayrshire & Arran 314 0.2 2,591 8.3
Quintile * episodes Transfused Transfused Transfused Transfused Patients Borders 31 0.1 369 11.9
Argyll & Clyde 396 0.2 3,762 9.5
1 (Least Deprived) 15.3% 14.0% 14.5% 11.8% 10.7% 13.5% Fife 227 0.2 2,302 10.1
2 14.2% 8.1% 8.1% 7.7% 6.3% 7.8% Greater Glasgow 1,135 0.3 9,443 8.3
Highland 136 0.2 1,079 7.9
3 19.0% 8.7% 9.5% 9.1% 10.6% 8.6% Lanarkshire 506 0.2 5,222 10.3
4 21.8% 16.6% 18.5% 14.4% 18.9% 16.2% Grampian 287 0.1 2,553 8.9
Orkney 12 0.2 116 9.7
5 (Most Deprived) 29.8% 52.6% 49.4% 57.1% 53.5% 53.8% Lothian 640 0.2 6,341 9.9
A categorisation which divides the population of Greater Glasgow Health Board Tayside 214 0.1 2,014 9.4
Western Isles 24 0.2 178 7.4
into five equal categories based on the range of SIMD scores so that 20% of
Dumfries & Galloway 53 0.1 583 11.0
the population falls into each quintile (population weighted). Quintile 1 is the
Shetland 5 0.1 20 4.0
least deprived, quintile 5 the most deprived. Scotland (excl FV) 3,980 0.2 36,573 9.2
* with any diagnosis of ICD10 K70 in their clinical history

2010 2010
Greater Glasgow
ALD by SIMD quintile From now on
Alcoholic Liver Disease in Greater Glasgow, 2005
No Patients RBC Units RBC per NHS Scotland Account for blood: data
Tx with ALD Transfused in patient Tx warehouse development with automated
SIMD Quintile in 2005 2005 (2005) harvest of standardised data from hospital
1 (Least depived) 12 57 4.8
2 10 113 11.3 blood bank systems.
3 18 118 6.6
4 39 242 6.2
5 (Most deprived) 173 1,264 7.3 Finally it is adequately funded
Greater Glasgow 252 1,794 7.1
No Patients Transfused who have a history of Alcoholic Liver Disease and
were transfused in 2005

2010 2010
7
Needs Assessment: Needs Assessment:
How can we use this data - in UK or How can we use this data - in resource
similar setting restricted setting

2010 2010
8
Aggregated Blood Report from
AABB (ARC and ABC facilities)
National Blood Collection and Utilization US Blood Centers Estimated Days of Supply
and 16.0
January 29, 2010
14.2 O+
Blood Availability and Safety Information System 14.0
O-
Estimated Days
12.0
(BASIS) 10.0
A+
A-
8.0 6.7 6.8 6.9 7.2
B+
6.0 4.6
3.1 4.0 B-
4.0
Jerry A. Holmberg, Ph.D,
Ph.D, MT(ASCP)SBB 2.0
AB +
Senior Advisor for Blood Policy 0.0

AB -
Office of the Secretary Blood Groups/Types
Office of Public Health and Science
Aggregated Blood Report from Estimated US Blood Supply

AABB (ARC and ABC facilities) Jan 29, 2010
Estimated
Hospital Estimated
Estimated Quantity In Blood Centers and Hospital Blood Estimated Blood Inventory: US Blood
January 29, 2010 Groups/ Days of Distribution constant 6 Supply
140,000
Type Supply Inventory day supply (RBC)
Estimated In Blood Center 6.7 119,359 226,416
120,000 O+ 107,058
Quantity of Blood
307,450
100,000
O- 3.1 10,332 19,721 30,053
80,000
60,000 Fixed estimate In hospital A+ 6.8 108,336 95,788 204,124
40,000 based on assumption of
constant 6 day supply, i.e., A- 4.6 13,056 16,904 29,960
20,000 281,731
0
B+ 6.9 29,331 25,356 54,687
O+ O- A+ A- B+ B- AB + AB - B- 4.0 3,724 5,635 9,359
Blood Type
AB+ 14.2 19,949 8,452 28,401
AB-
AB- 7.2 3,364 2,817 6,181
Total 307,450 281,731 589,181
Impression BASIS Report

The blood supply is adequate to meet needs with
just over 6 days supply of O Positive blood
and just over 3 days supply of O Negative blood
available at blood centers across the nation.
Hospital inventory shortage reporting in the Blood

Availability and Safety Information System
(BASIS) continue to reflect instances of RBC
and platelet shortages; however, alternate
source purchases remain relatively low.
Platelet inventories continue to appear tight as a

national aggregate. Represents approximately 95 Sentinel Hospitals reporting consistently
1
BASIS Report BASIS Hospital Red Cell Supply
(All Group/Types)
Represents approximately 95 Sentinel Hospitals reporting consistently Represents approximately 95 Sentinel Hospitals reporting consistently
BASIS Hospital Platelet Supply BASIS Shortages
BASIS Shortages BASIS Shortages
2
National Blood Collection and
Utilization Survey Overview
Introduction
Methods
Key Findings
Biovigilance
Hospital Costs of Blood
Collection and Processing
Experience
Transfusion Experience
Cellular Therapy Products
Historical Perspectives
Current Issues in Transfusion and
Transplantation
www.hhs.gov/bloodsafety
Whole Blood and

2006 Key Findings: Donors
Red Blood Cell Apheresis
Estimated Collections by Blood Centers and Hospitals 12,142,000 donors presented
Type of 2006 2004 Per Cent
RBC Blood Hospital Total Total Difference 9,554,000 allogeneic donors
Collection ()
Center 2,726,000 first-
first-time donors (28.5%)
WB 13,486,000 665,000 14,151,000 14,087,000 0.5%
Allogeneic 6,828,000 repeat donors (71.5%)
WB 242,000 93,000 335,000 463,000 - 27.6%*
Autologous
Repeat donors provided 11,697,000
WB Directed 47,000 22,000 70,000 117,000 - 40.3%* donations 1.7 donations/donor
RBC 1,603,000 16,000 1,619,000 836,000 93.7% *
Apheresis
Total 15,378,000 796,000 16,174,000 15,503,000 4.3%
May not be a appreciable increase due to blood centers were not weighted in 2004
Significant Difference (*) from 2004 to 2006
95% Confidence Intervals calculated
Whole Blood and

2006 Key Findings
Red Blood Cell Apheresis
30,044,000 components transfused
Estimated Collections by Blood Centers and Hospitals
14,650,000 red cells
Type of 2006 2004 Per Cent
10,388,000 platelet concentrate eq. RBC Blood Hospital Total Total Difference
Collection Center ()
4,010,000 plasma
Total 15,378,000 796,000 16,174,000 15,503,000 4.3%
993,000 cryoprecipitate
Rejected on 137,000 14,000 151,000 274,000 - 44.9%*
Testing
Available 15,241,000 782,000 16,023,000 15,299,000 5.2%
Supply

3
Red Blood Cell Transfusions WB and RBC Recipients
Estimated Transfusions by Blood Centers and Hospitals
Type of 2006 2004 Per Cent 3.0 units per recipient (unweighted
(unweighted))
RBC Difference
Transfusion
Blood Hospital Total Total
()
8,275,000 allogeneic units (incl
(incl directed)
Center
2,740,000 recipients
Allogeneic 716,000 13,262,000 13,978.000 13,728,000 1.8%
(not 2004: 2.7 units per recipient
directed)
Autologous 7,000 182,000 189,000* 271,000 - 30.3%* Extrapolation of ratio of
transfused/recipient
Directed 0 126,000 126,000 132,000 - 4.6%
Pediatric 5,000 352,000 357,000* 59,000 504.8%*
Estimated 5 M recipient
6.6% decrease in transfusion recipients
Total 729,000 13,921,000 14,650,000 14,191,000 3.2% (compared to 2004)
Platelets Transfused Platelet Dose for Transfusions

60
50
Estimated Transfusions by Blood Centers and Hospitals
Per Cent 40
Type of 2006 2004 Per Cent of 30
Product Blood Hospital Total Total Difference Hospitals 20 2001
Transfused Center ()
10 2004
WBDP
WB Derived 223,000 1,073,000 1,296,000 1,537,000 -15.7% 0 2006
Apheresis WBDP
Platelets 5 or 7 9 >10
17%
less
Apheresis 411,000 8,681,000 9,092,000 8,343,000 9.0%
Platelets (1,515,000) Dose of Platelet
Concentrate
Total 634,000 9,754,000 10,388,000* 9,881,000 5.1%
Parenthesis notation is apheresis units including splits Apheresis

Significant Difference (*) from 2004 to 2006 83%
Outdate of Red Cells

Plasma Type Transfused Component Processed/ Total Per Cent
produced Outdate Outdate
WB/RBC 16,745,000 401,000
Cryo-
Allogeneic 252,000 63%
Reduced
Pediatric Plasma Autologous 131.000 33%
1% 5%
FFP Directed 5,000 1%
Jumbo
Plasma Whole Blood 13,000 3%
2% Plasma w/i 24 hours
Jumbo Plasma Whole Blood

Directed 3%
Plasma w/i 24 1%
FFP Pediatric Allogeneic
hours 77% Autologous
15% Cryo- Reduced Directed
Plasma Autologous Allogeneic Whole Blood
33% 63%
4
Current Issues in Transfusion Trends in WB and RBC Collections
Blood inventory shortages for non-
non-surgical procedures 18.0
16.0
2006 2004
14.0
13.5% (231/1707) reported at 16% (257/1604) reported at
Millions of Units
least one day shortage least one day shortage 12.0
Mean number of days was 22 Mean number of days was 19.27 10.0
8.0 Total
Six (6) hospitals reported 365 Eight (8) hospitals reported 365
Allogeneic
days the blood needs were not days the blood needs were not 6.0
Autologous
met. met.
4.0
Number of days regular or
2.0
standing order was incomplete
0.0
44,910 total days (estimated?)
1989 1992 1994 1997 199 2001 2004 2006
On any given day, 123 Survey Year
hospital did not have their
standing order met.
Trends in Estimated Rates of Blood

Trends in WB and RBC Collections
Collection and Transfusion in the
and Transfusions
17
US, 1980-
1980-2006
100.0
2006: 84.1
16 Collections 150K 90.0
Transfusions
1.3 M
Units/1000 US Population
15
Millions of Units
Available Collections 80.0
14 70.0
60.0
13
2006: 48.9
50.0
12
40.0
11
30.0 Transfusions 95% Cl/1000 pop. (all ages)
10 Collections 95%CI/1000 pop. (ages 18-64)
20.0
1989 1992 1994 1997 1999 2001 2004 2006 1980 1982 1984 1986 1987 1989 1992 1994 1997 1999 2001 2004 2006
Survey Year
Survey Year
Comparison in Developed In Developed Countries

Countries Estimation based on Population
US England Australia Denmark Sweden Age and sex distribution similar except
Units of 48.75 (2001) 44.9 28 54.08 45.3 (1996-
20020
(1996- more men transfused with platelets and
RBC per 48.9 (2006) 58.6 (2000-
(2000-
1000 2002) plasma
Population
Recipient <41 18.8% <40 14.4% <40 15.4% <39 9.4% <39 9.8% More blood used in older population
Age 41-
41-65 27.8% 40-
40-70 38.4% 40-
40-70 36.7% 40-
40-59 18.2% 40-
40-59 15.1%
>65 53.3% > 70 47.2% > 70 47.9% > 60 72.4% > 60 75.2% Cardiovascular surgery predominated
Gender M 48.5% M 50.4% M 52.5% M 53.2% M 52.9%
highest use
F 51.5% F 49.6 F 47.5% F 46.8% F 47.1%
Kamper-Jorgensen Transfusion 2009; 49:888-894

Cobain Transfusion Medicine 2007, 17, 10-15
5
Blood Systems
American Red Cross
United Blood Services
2007 2007
870,000 collects that serve 500 hospitals in 18 states 6,332,000 collections that served over 2500 hospitals
1/3 of continental US throughout US
Estimation of blood needs Roll up of individual hospital utilization data on the most recent
recent
12 months
Roll up of individual hospital distribution for most recent 60
Application of Sales and Operation Planning (SNOP) to create
months (5 yrs) distribution is assumed ~ to transfusions
a 12 month projection
(return policy)
Application of statistical software package (Decision Pro) Based on distribution, assumption that distribution ~ transfusion
transfusion
Annual survey of hospitals to determine new changes in
Unfilled orders are tracked solely for customer satisfaction
services that might change utilization
Forward looking estimates do not currently take into
consideration planned changes in hospital services (e.g. Since 2009, review of unemployment statistics which seem to
track blood utilization figures due to loss of health insurance
cardiac surgery)
coverage
C:T ratio is not used as BSI collects no transfusion data
Information obtained via phone interview by Karen Lipton Information obtained via phone interview by Karen Lipton
American Red Cross

The ARC business planning process consist of
the following elements:
Monthly Sales and Operations Planning Process
Monthly Manufacturing Review
Weekly Master Production Scheduling
Daily Inventory Management Process
All these different processes work off the same
set of base numbers through a series of
connected information systems.
Information obtained via phone interview by Karen Lipton
6
2006 Blood collections per 1000 population
Estimating Blood Requirements National Blood Services only
Ethiopia
Nigeria
Haiti
Tanzania
Mozambique
Kenya
Uganda
Rwanda
Cote d'Ivoire
Zambia
Guyana
Blood Safety Namibia

Botswana
South Africa
Geneva, Switzerland 0 2 4 6 8 10 12 14 16 18
February 4, 2010 Units collected / 1000 population
Zambia Blood Donations 2003-

2003-08 Moving Target--
Target--Population
--Population
110000
100000
90000
80000
Units collected
Family/Replacement
70000
donors
60000 Voluntary non-
50000 remunerated donors
40000 Regular donors
30000
20000
10000
0
2003 2004 2005 2006 2007 2008
Infrastructure and Collections in 11 2006 Blood collections per 1000 population

countries without 100% NBTS coverage National Blood Services only
Country Centers Centers NBTS Units NBTS Units Ethiopia
2003 2007 2003 2007 Nigeria
Botswana 2 2 11,583 22,230 Haiti

Tanzania
Cote dIvoire 2 2 67,780 92,009
Mozambique
Ethiopia 4 12 17,208 32,442 Kenya
Guyana 1 5 4,008 5,475 Uganda
Haiti 5 17 8,711 17,094 Rwanda

Cote d'Ivoire
Kenya 5 6 41,869 123,787 Zambia
Mozambique 111 140 67,105 115,071 Guyana
Nigeria 0 9 0 16,987 Namibia
Tanzania 0 7 0 109,471 Botswana

South Africa
Uganda 5 5 103,000 130,000
Zambia 90 9 40,616 68,056 0 2 4 6 8 10 12 14 16 18
MMWR November 28, 2008/57(47):1273-1277 Units collected / 1000 population
1
Coverage
2006 Blood collections per 1000 population
All collections (Red
(Red =
= Hospital
Hospital collected/tested)
collected/tested)
Ethiopia
Nigeria
Haiti
Tanzania
Mozambique
Kenya
Uganda
Rwanda
Cote d'Ivoire
Zambia
Guyana
Namibia
Botswana
South Africa
0 2 4 6 8 10 12 14 16 18
Units collected / 1000 population
Table 11: Table 12: Estimated number of blood

Number of units of blood collected in 2006 donations in 2006
Type of facility Number of units of blood Groups # of # of Average Total # Total

donations (%) units sites per site of donations
Primary 1,742 (1) visited facilities
Primary 1,742 10 174.2 952 165,838
Secondary 87,7575 (50.4)
Secondary 87,757 54 1622 944 1,534,122
Tertiary 84,505 (48.6)
Tertiary 84,505 48 1760 48 84,505
Total 174,004 (100)
Total 174,004 112 1,784,465
Data Sources
Rapid assessment
Surveys
Ministry Data
2
Stepwise approach-
approach-infrastructure Stepwise approach
approachclinical issues
Organization Composition
Hospital based? Whole blood
Regional? Pediatric units
National? Components
Incorporate Ministry, Private, NGO, Faith based? RBCs,
RBCs, FFP, platelets
Usage vs demand Clinical guidelines
Family replacement vs Volunteer Physician familiarity
familiaritypractice
Transition
Transitionutilization increase due to previous Training on new component options
unmet demand Laboratory
Data management capability--
capability--inventory
--inventory Baseline values
Manual/paper Monitoring
Electronic
Summary Comprehensive Blood Safety
Processing
Collection Utilization
Communication TTI Prevention (MDG 6)

TTI Prevention
partnerships (MDG 8) Maternal Health (MDG 5)
(MDG6)
TTI Prevention (MDG6) Child Health (MDG 4)
Quality Systems
Work Force Trauma (MVA)
Work Force
Development HIV/AIDS Care and Rx
Development
Health Care Worker Quality Systems
Health Care Worker
Safety Patient Safety
Safety
IT systems Work Force
IT System
Quality Systems Development
Waste Management
Health Education Worker Safety
Waste Management IT System
Waste Management
3
Objectives
Country and National Blood Transfusion
Blood Supply and Demand: Service background
Georgetown, Guyana Investigation background
November, 2007 Case definitions
Methods
Results
Sridhar Basavaraju, MD Shortage calculations
Medical Officer
Centers for Disease Control and Prevention Lessons learned
Atlanta, USA
Guyana Guyana National Blood Transfusion

Service (NBTS)
National borders:
Venezuela, Brazil, Total Collections, 2007: 5,475 units (7.3 units
Suriname, Caribbean Sea per 1,000 population)
Population: 750,000 (30% 90% units distributed: whole blood or packed
live in Georgetown) red cells
Gross National Product per 80% of collections: distributed to Georgetown
capita: <$1,000 Public Hospital Corporation (GPHC)
60% of collections: voluntary, non-
remunerated donors
Study Background
December, 2007: NBTS review suggests 60%
of all blood orders unmet
GPHC-NBTS opinion differences
Reports of delayed surgeries Case Definitions
Inappropriate requests
Blood returned unused
Field investigation to determine:
Was enough blood collected?
Was there a true shortage of blood?
Case Definition Case Definition
Unit: An individual blood product WB, PRC, Order: A unit is requested by GPHC by
FFP, PLT, Cryo submitting a written blood request form.
Issued: Unit given by NBTS to ward for the Rejected: The order by GPHC ward is refused
purpose of transfusion by the NBTS due to a problem with
Filled: Unit is prepared by NBTS and ready to information provided to NBTS
be issued. A unit must be filled before it is Unused: Issued unit is not transfused into a
issued. Not all filled units are issued patient by ward.
Case Definition Case Definition

Returned: An issued unit is returned to NBTS Unit not required: An order which is stated by
unused by ward ward to be no longer necessary. NBTS will not
Reissued: A returned unit is given by NBTS to fill this order
ward for the purpose of transfusion Shortage: An order not filled by NBTS as the
Expired: A unit is no longer fit to be transfused specific blood product is not available.
as > 35 days have lapsed since being filled
Blood Request Form
Methods
Methods Methods
Audit of NBTS data for November 2007 Blood Request Form Files
3 Logbooks in NBTS Issued
Book 1: Book of Cross matched issued blood Filled and not picked up by ward
Book 2: Book of daily requests Rejected orders
Book 3: Book of returned units Orders no longer required
Comparison between logbooks and Blood
Request Form files
Units Ordered by GPHC November, 2007
Packed Red 1130

Cell
Whole Blood 41
Results
Platelet 2
Fresh Frozen 133
Plasma
Cryoprecipitate 22
Total 1328
GPHC NBTS
Units Issued to GPHC November, 2007 Packed 1130 Packed 431
Red Cell Red Cell
Packed Red Cell 431 Whole 41 Whole 22
Whole Blood 22 Blood Blood
Platelets 2 Platelets 0
Platelets 0
Fresh 133 Fresh 59
Fresh Frozen Plasma 59 Frozen Frozen
Plasma Plasma
Cryoprecipitate 59 Cryo.
Cryo. 22 Cryo.
Cryo. 59
Total 571 Total 1328 Total 571
Overview of Issued and Non-issued units
Issued & not

returned : 427
Issued units: Expired: 6
571
Unaccounted: 19
Returned
unused:
144
Returned unused &
reissued in December,
Shortage Calculation
2007: 3
Total units
ordered:
Returned unused &
1328 Reason not issued
reissued in November,
Actual
2007: 116*
shortage:
Request rejected: 52 (168-
21 116)
Shortage before
Not Issued Units: Filled & not accounting for returned
757 issued: 86 & reissued units: 168
Request
cancelled: 482
Georgetown Public Hospital Corporation

Blood Unit Requests: November 2007 Shortage Calculation
Total Units Ordered: 1328 Shortage
(Calculated Shortage) 52
Units undelivered: 757 Units Delivered: 571
* 100 = 8.7%
Actual Shortage (Calculated Demand) 543 + 52
Cancelled 482 Returned Unused 144
168
Filled, 86 Reissued 116 Guyana total collections, 2007: 7.3 units per 1,000 population
uncollected - 116
Rejected 21 Expired/Unaccounted 25 WHO recommendation: 10-20 units per 1,000 population per year
52 units
Shortage 168 Transfused 543
Limitations Lessons Learned

Limited data: 1 month only (November, 2007) 7.3 units collected per 1,000 population
Seasonal disease/injury trends resulted in an 8.7% shortage
Variations in blood collection patterns Blood collection requirements
Data represents GPHC recommendations should include capacity to
Other local hospitals not included conduct transfusions
Not nationally representative Recommendations should account for stock
NBTS collections only included management (to reduce wastage)
Acknowledgements
Guyana NBTS
Clem McEwan
Olwyn John
Bonita Richards
CDC Questions/Discussion
Claudette Harry
LaMar Hasbrouck
Nicolette Henry
Lawrence Marum
John P. Pitman
The findings and conclusions in this presentation are those of
the author(s) and do not necessarily represent the views of the
Centers for Disease Control and Prevention
Challenges
Paucity of relevant empirical data in Africa

Level of Resources to Data on current capacity of BTS to collect, test, process and distribute
blood. What proportion of the current need is met?
Ensure Safe Blood in Africa Population in need of Safe Blood epidemiology of disease conditions
that require Safe Blood transfusion
Current access of this group to health services and potentially to Safe
Blood when needed
Obtaining available cost for Safe Blood strategy, particularly in the
presence of multiple funding sources..
Pamela Rao - Forecasting scale-up costs for national programs in the
Associate Director, Global Health and Development Strategies
absence of data almost impossible..
Social & Scientific Systems, Inc.
Hence, a model-based approach.
Working Group 2SSS and AABB
..and it became 2-7-10 24/7 job !!
Comparative Analysis of Methods Strengths of RDM Approach
WHO Method 3, Module 1 (Requirement of blood Relatively simple modeling approach

units = ~2% of population)
Resource estimation for individual HIV/AIDS intervention rather
than resource optimization across several interventions
WHO ManualCosting Blood Transfusion Services,
1998/2001 Population-based estimates therefore, addresses the
magnitude of safe blood problem in the country
GOALS ModelCosting guidelines for HIV/AIDS strategies
Health system performance access to health services and
RDM/UNGASS Model potentially access to safe blood
Allows for cost of scaling-up therefore, addresses unmet need
Internationally accredited3 pieces of work (WB-MAP,

UNGASS, & Tx & Care)
Principal Steps Involved in the Model RDM Adapted for BTS
Establishing the size of populations that are at greatest need

and will benefit from Safe Blood strategy Target Group
Proportion of Target group that has access to health

services Potential Target Group (PTG). Using PTG to
determine current and future levels of coverage.
Estimating costs using project level cost data to scale up

programs to desired coverage levels and or feasible
coverage levels.
1
Factors Considered in Choosing Sample Countries BTS Composite Indicator
Strength of blood program, country-specific health HDI (life expectancy, literacy, GDP)
system performance indicator, Human Development
Index (HDI) Health System Performance Indicator (WHR 2000)
Size of population, HIV prevalence, Malaria status BTS indicator

Policy variables (25%)
Geographic representation (East, South, West Africa) Operations variables (50%)
Quality systems variables (25%)
Representation of 4 organizational models for BTS
Size of Population, HIV Prevalence, and Geographic Representation

Malaria Status
Nigeria
Kenya
Cote dIvoire
Mozambique
Representation of 4 Organizational Models for BTS Sample Countries for Varied Analysis
Centralized/national: One national center controls the Geographic Organizational

services for the whole country with or without regional centers Country Capacity Location Model
Regionalized: Regional centers operate with some autonomy, Centralized/
Cote d Ivoire Rel. strong W. Africa
with degree of national control National
Hospital-based: Each hospital runs its own services with or Kenya Medium E. Africa Regionalized
without coordination at national level
Mozambique Weak S. Africa Mixed
Mixed: Combination of hospital-based and some regional and
national coordination (incomplete coverage)
Nigeria Weak W. Africa Hospital-based
2
Providing a Model to Estimate the Level of Resources for Safe Blood Sources to Identify the Target Groups
1. Identifying target group requiring Country-specific literature review of published

and grey literature
Safe Blood
Preliminary analysis of ongoing large study on
use of blood in Cote d Ivoire
MOH applications and TA providers baseline

assessment report, if available
Conditions Requiring Blood Transfusion in Rank Order Target Groups Requiring Safe Blood in Africa
Pediatric anemia (severe 8g/dL.), maternal iron Children under 5

deficiency, iron-deficient diets, malaria, intestinal
helminthes, low-birth weight babies (<2,500g) Pregnant women
Maternal severe anemia (WHO Criteria: Hemoglobin Population with severe adult anemia (women and men)
concentration <7.0 g/dl)
% of population undergoing elective surgery
Postpartum hemorrhage
Shock/trauma/burns (accidents etc.) % of population undergoing trauma, shock and burns
Elective surgery Others

Adult anemia (elderly 64 years GI bleeds etc.)
Others
Providing a Model to Estimate Level of Country-Specific Anemia Rates

Resources for Safe Blood in Africa
Cte
1. Identifying priority target group requiring Safe Blood dIvoire Kenya Mozambique Nigeria
2. Estimating the size of priority target groups 1
Children <5 65.5% 60.0% 79.6% 69.2%
Pregnant
45.9% 46.7% 52.3% 51.7%
women
2
Women 40.0% 40.0% 40.0% 40.0%
Men3 30.0% 30.0% 30.0% 30.0%
Source:
1
UNICEF. State of the Children 2006
2
IFPRI- Harvest plus
3
WHO/NHD/01.3. Iron deficiency anemia: Assessment, prevention, and control
3
Estimating the Target Group Size Likely to Be Transfused Size of the Target GroupPopulation-Based Estimates
Severe Anemia Rates as proportion of those anemic
Cte
Anemia in children < 5 years 10% dIvoire Kenya Mozambique Nigeria
Children <5 1,708,240 3,340,200 2,362,528 14,414,360

Anemia in pregnant women
(underestimatedcalculated from 20%
annual number of live births) Pregnant women 30,339 617,374 402,187 2,751,991
Low birth weight (based on annual Low birth weight 112,370 132,200 115,350 745,220
number of live births)
20%
Shock/trauma/
Shock/trauma/burns/elective 295,373 746,300 500,300 2,862,678
15% burns/surgery
surgery
Women 1,615,600 2,963,200 1,816,400 11,362,800
Anemia in women (15-64 years)
less live births, year 2004
3%
Men 1,446,000 2,739,000 1,491,000 10,473,000
Anemia in men (15-64 years) 2% Total population

16.94 32.98 19.11 125.74
of country (in millions)
Triangulating Population-Based Target Size

Size of Target Group-Conservative EstimatesWhy?
Data with Demand for Blood by Target Group
Age group: 515 not covered

Literature review
There seems to be significant proportion of men
receiving blood in CImore research is required
CI blood use study
Population attributable fraction of adult malaria due to
HIV with HIV prevalence of 8% among adults is about MOH applications, if available
5% for clinical malaria
Population attributable fraction of adult malaria due to

HIV with HIV prevalence of 30% among adults is
about 35% for clinical malaria
Shock, trauma, surgery is likely to be more
Overall number could be 1025% more
Proportion of Total Demand for Safe Blood Justifications des Demandes de Produits Sanguins par Types
Among Target Groups de Services: Priodes Fvrier-Juillet 2004 et 2005
5060% children under 5

1520% maternal
Small % low-birth weight babies
25 % adult anemia
1216% elective surgery, shock/trauma/burns
Remaining for other causes
4
Triangulating Target Group Data With Providing a Model to Estimate Level of
Available Country-Specific Data on dd Resources for Safe Blood
Proportions of PTGs requiring transfusions
Cte 1. Identifying target groups
Kenya Mozambique Nigeria
dIvoire
2. Estimating size of target groups in Africa
Children <5 45.47% 45.16% 47.33% 46.19%
Pregnant women 16.15% 16.70% 16.12% 17.64% 3. Estimating total PTG that may require blood
transfusion (i.e., proportion of the target
Low birth weight 5.98% 3.58% 4.62% 4.78% group with access to health care services)
Shock/trauma/
11.79% 15.14% 15.04% 13.76%
burns/surgery
Women 12.90% 12.02% 10.92% 10.92%
Men 7.70% 7.41% 5.97% 6.71%
Total population
16.94 32.98 19.11 125.74
Proxy Indicators to Estimate Target Groups Access to

Health ServicesComposite Indicators Access to Health Services
For children For Shock/Trauma/Burns, Surgery, and Adult Anemia

Proportion of children who had completed the final
immunization dose for diphtheria, pertussis, tetanus (DPT3) Access to hospitals from published literature
% of children <5 with fever receiving anti-malarial drugs
% of children < 5 with acute respiratory infections (ARIs) seen by a

health worker
For pregnant women

% of pregnant women who receive antenatal care4 visits
% of women giving birth who were attended by skilled health
personnel at birth
Composite Indicators Used Country-Specific Estimates of PTGs
Cte
Cte Kenya Mozambique Nigeria
Kenya Mozambique Nigeria dIvoire
dIvoire
Children <5 75,163 165,674 111,748 441,079
Children <5* 44.00% 49.60% 47.30% 30.60%
Pregnant women 40,929 72,233 46,332 290,610
Pregnant women* 67.45% 58.50% 57.60% 52.80%
Low birth weight 15,159 13,114 10,912 45,607
Low birth weight* 67.45% 49.60% 47.30% 30.60%
Shock/trauma/burn
9,304 17,687 14,409 85,880
s/surgery
Shock/trauma
21.00% 15.80% 19.20% 20.00%
burns/surgery** Women 10,178 14,046 10,462 68,177
Women** 21.00% 15.80% 19.20% 20.00%
Men 6,073 8,655 5,725 41,892
Men** 21.00% 15.80% 19.20% 20.00%
Total PTG 156,806 291,409 199,588 973,245
Total population
16.94 32.98 19.11 125.74
* Composite Indicator
** Hospital Utilization Indicator
5
Providing a Model to Estimate Level of
ResultsMagnitude of the Problem Resources for Safe Blood
1. Identifying the TGs in Africa
2. Estimating the size of TG that may require blood transfusion
3. Developing country-specific estimates of PTG with health

access needing transfusion
4. Determining average requirement of blood for each PTG

to estimate requirement of blood for PTG
Average Number of Blood Units Per Case Total Number of Blood Units Required for PTG in a year
Cte
Number of dIvoire
Kenya Mozambique Nigeria
Target group blood units
Children <5 37,581 82,837 55,874 220,540
Children <5 0.5
Pregnant women 102,321 180,582 115,830 726,526
Pregnant women 2.5 Low birth weight 7,579 6,557 5,456 22,804
Shock/trauma/
Low birth weight 0.5 27,913 53,062 43,226 257,641
burns/surgery
Shock/trauma/ Women 25,446 35,114 26,156 170,442

3
burns/surgery
Men 12,146 17,310 11,451 83,784
Women 2.5
Total blood units 212,986 375,462 275,993 1,481,737
Men 2 Total population

16.94 32.98 19.11 125.74
Magnitude of the Problem Base Year 2004 Providing a Model to Estimate Level of
Resources for Safe Blood
2. Estimating the total TG that may require blood transfusion
3. Estimating the size of PTG with health access needing

transfusion
4. Determining average requirement of blood for each PTG

to estimate requirement of blood for PTG
5. Projecting growth over the next years
6
Growth Rates Adjusted Growth Factor
Preliminary Base Growth Factor was calculated by assuming It has been observed that factors other than per capita
that the growth factor for all countries for Sub-Saharan Africa
(SSA) would range from a minimum of 0.1 to a maximum of income influence performance of health systems and
0.2. potential expansion of BTS.
The minimum growth factor of 0.1 was assigned to the
average per capita income of SSA countries with lower per
To adjust the base growth factors for other factors, we
capita (per <$750) the maximum to per capita income >
$4,000. estimated regression equations expressing the coverage for
DPT as a function of GNI per capita and calculated the ratio
Growth factors for all countries between the lowest income
and highest income groups were calculated by linear of observed coverage to predicted coverage. This ratio as a
interpolation between 0.1 and 0.2 based on their per capita proportion was used to adjust the base growth factor.
income relative to the average per capita income of the lowest
group to the average of the highest group.
Projecting Expansion of Blood Transfusion

BTS Growth Factor
Services Capacity
We assumed the number of safe blood units provided would Assumptions:

increase in 2005 by 30% in Kenya and by a growth rate in
the other three countries in the same ratio to 30% as each Changes to BTS occur slowly over timeMax. 30%
countrys adjusted growth factor is to Kenyas adjusted increase in capacity
growth factor.
Total need for safe units of blood (base year) will not
The BTS growth factor (G) is defined as the proportion of change during projection period i.e. increase in
unmet need newly covered in 1 year. If At is the proportion access to health care will potentially increase dd for
of unmet safe blood need that has been met in year t: At = Safe Blood
At-1 + (1 - At-1) * G.
Providing a Model to Estimate Level of Resources for Safe Blood

Determining Unit Cost of Safe Blood

1. WHO cost projections by Working Group 1
2. Estimating the total TG that may require blood transfusion
2. Current cost of Safe Blood from the current
3. Estimating the size of PTG with health access needing operating budget in Cote dIvoire
transfusion
3. Consultation and personal communication with
4. Determining average requirement of blood for each PTG to Emergency Plan TA providers
estimate requirement of blood for PTG
4. Literature review published and unpublished
5. Projecting growth over the next years documents
6. Applying unit cost of safe blood to estimate level of

resources
7
Cost of 1 unit of whole blood (recurrent) - CI
Blood Transfusion Costs
Recruitment donor (communication, vehicles etc. food for

the donors)
BLOOD TRANSFUSION SERVICE
Costs Allocated by Activity Collection (materials, poches , cold chain..)
Total Costs
Lab testing (HIV, Hep B, Hep C, Syphilis, - labels, cost of additional test kits
Blood Donor Blood testing & Blood Storage &

Blood Collection Processing ( production)
Recruitment Processing Distribution
Capital costs Capital costs Capital costs Capital costs Storage & Cold chain, Distribution
+ + + +
fixed cost - QA, informatics, administration, logistics,
Recurrent costs Recurrent costs Recurrent costs Recurrent costs
incineration, maintenance buildings and equipment
Personnel salaries.
Total recurrent cost per unit of safe blood (21500
CFA = $ 40
Country-Specific Unit Cost (recurrent) Applied in the Model

Assumptions Scenario 1
The requirement of safe blood for potential target group

(with access to HS) as projected with base year (2004) BTS
Country Cost capacity will remain constant
Cote dIvoire $40
The number of safe blood units provided would increase in
2005 by 30% in Kenya, and by a growth rate in the other
Kenya $30 three countries in the same ratio to 30% as each countrys
adjusted growth factor is to Kenyas adjusted growth factor
Mozambique $30
The increase in safe blood units in absolute number from
Nigeria $25 year 2004 to 2005 will remain constant for the projected
years
WHO $20 Recurrent cost of safe unit of blood in CI=$40, Kenya &
Mozambique=$30 and Nigeria=$25
WHO economic $45-50
CNTS, CI $50
Assumptions Scenario 2
Results Table Scenario 1
The requirement of safe blood for potential target group

(with access to HS) as projected with base year (2004) BTS
capacity will remain constant
The number of safe blood units provided would increase in
2005 by 30% in Kenya, and by a growth rate in the other
three countries in the same ratio to 30% as each countrys
adjusted growth factor is to Kenyas adjusted growth factor
The same proportion of the unmet need for BTS in each
country that was met in 2005 would be met in each year of
the remainder of the projection period. This constant
proportion is called the BTS growth factor
Recurrent cost of safe unit of blood in CI=$40, Kenya &
Mozambique=$30 and Nigeria=$25
8
Level Resources needed for 4 countries (2006-2010)
Results Table Scenario 2
Cost of Blood (at $40/unit) per year
$12
$11
$10
$9
(in millions)
$8 Cote d'Ivoire
US Dollar
$7 Kenya
$6
$5 Nigeria
$4 Mozambique
$3
$2
$1
$0
2006 2007 2008 2009 2010
Year
Capital Cost Budgeting

Total CostsRecurrent + Capital Costs
Projected 2010 capacity e.g. 189,000 units of safe blood

Existing capacity 89,000 = 100,000 additional.
Assuming the existing BTS has reached its optimum capacity

and any additional capacity will require capital investment, we
estimated $1m of capital investment for a regional bank with a
capacity of 20,000 units.
$ 1m includes building, 2 vehicles, cold chain equipments, lab

equipments, beds, centrifuges, informatics and training cost.
Total CostsRecurrent + Capital Costs Factors to consider in other scenarios
Increase in population over 2006 -2010 will increase dd

Projected Expenditures for Cote d'Ivoire NBTS Unprecedented international efforts for HIV/TB/Malaria will
result in health system strengthening that will increase
Millions access to health care services, and potentially dd for
$8 blood.
$7
$6
Increase in anemia rates in a high prevalence HIV population
$5 where ART strategy has been implemented in large
$4 scale.
$3
Increase on clinical Malaria cases in high prevalence HIV
$2
$1 populations
$- Success of Malaria efforts may decrease demand for safe
2006 2007 2008 2009 2010
115,704 134,057 152,409 170,761 189,113 blood.
Operating Budget Capital Investments
9
Points for discussion
Points for Discussion and Recommendations
Increase in HIV among adult population will decrease the potential supply
There is a desperate need for complete and accurate data for better of safe blood and increase the dd for blood due to increase in clinical
planning, budgeting, and forecasting resources required to ensure malaria among HIV positive (Mozambique) and anemia resulting from ART
Safe Blood in African countries. (Botswana)
Strengthening Safe Blood supply in isolation without an

understanding of levels of development in health system in the
country should be avoided. Strengthening should be coherent to the
level of sophistication in the overall health system.
Major international efforts such as Roll Back Malaria, Safe

Motherhood Initiative, and Family Planning can help avoid risks in
transfusion by decreasing demand for Safe Blood in Africa.
10
Demand: The
Need: An estimation of the amount of blood to amount of blood that
meet the transfusions requirements of the would be transfused
population according to current guidelines, best if all prescriptions for
blood were met.
practices and policies.
Demand may be
appropriate or
inappropriate
practices
Need
Population Need
Tools for Estimation of Blood
Current Health System Needs

Demand
Needs
Unmet Demand
Unpresecriped needs
Development Index
Use
Appropriate
Use: The actual amount of
blood currently transfused
(use may be appropriate or Inappropriate
inappropriate.
BTS Tools
He eng osp
St S/H ity
alt the ital
m
BT pac
r
hS n
te
Ca
ys
ys
ed t S
tem
Ne rren
SE
s
/U
Cu
De ply
nd
p
ma
Su
Population Needs
Clinicians Clinical Use/

Evidence Base PH
Leadership
Data Sets Needed Resources Needed
1
Process Forward
2
Objective: Need
The amount of blood that would be used if
To provide resources that will assist all those patients who could benefit* from
authorities to estimate the current and transfusion were recognised and if blood
was prescribed according to appropriate
future blood transfusion needs for patients guidelines.
treated in their health systems.
*All those patients who could benefit implies 100% access to health
services but where access to the health system is restricted to a
part of the population, need is in effect limited to those who have
access.
Demand Need, use and demand

This term may reflect numerous additional Quantity to met need
factors such as the reliability of the blood Optimum prescribing and use
supply and delivery system, relationships [Conditioned by access to care]
between clinic and blood bank etc,etc Use
Observed quantity actually transfused OR
Surrogate for transfused eg delivered
Demand
Observed quantity that is requested from blood
providers
Target population composed of Identifying target population 1

Groups of patients at risk of transfusion GPART Case definitions could be built around
[Groups of patients needing blood,GPNB]
Age /gender
Patients at risk of transfusion Over 80
Under 5 in high incidence malaria zone
Defined as:
Combinations of diagnosis and intervention
HIV on HARTT
Diagnosis
Patients with conditions that may require blood transfusion, ICD 10
but for whom transfusion may be avoided by preventive Intervention
or alternative interventions OPCS
Specialty based
Surgery, paediatrics
DRG
Geography or Institution
1
Identification
of target populations 2 Size of the target populations
Incidence, Prevalence
From existing sources in country or countries
with similar demography etc
Sources of epidemiological data
Primary data collection Burden of disease data
Professional concensus Review the RAO model in detail
Published or grey literature
Burden of disease databases
New systematic reviews may be needed
Data on blood use for target Data sets and definitions for describing
populations hospital blood use
Is supply sufficient?
From existing sources in country or countries Georgetown model,
with similar demography etc
Primary data collection Which conditions receive how much blood?
Professional concensus Zambia model
Published literature
How many patients per time period?
Examples of practical guidance, Worked example

training materials
Maternity
Sources of existing data where to look
Haemorrhage
Practical issues in accessing data
Recording Other deliver complications
Analysing Puerperal sepsis
Uncertaintyimportance of factoring in Maternal anaemia
Estimates how to develop and use
Worked examples
2
Maternity
Data Items
Deliveries/year
Number with APH, PPH
Proportion with MH that bleed in hospital or reach
hospital alive and potentially resuscitatable OR
Number of MH patients who each hospital
Proportion who need transfusion
Units of blood used per patient with MH who is
transfused
Number of patients having C section
Proportion needing blood
Quantities of blood needed

Estimation Presentations WHO Blood

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Estimation Presentations WHO Blood

Загружено:

Авторское право:

Доступные форматы

Pattern of Blood Usage

Estimate Blood Developed countries Developing countries

Emergency Trauma Care Maternal Mortality

Blood Supply Estimating Blood Requirements

Geography, population and epidemiology

Estimating blood needs

Country's Need for Blood and Blood

Balance between demand and supply is

Transfusion vol.45 Oct 2005 Supplement

Health Care System related Parameters

Assessment of services in each centre What to do with the excess plasma?

Options for Fractionation Contract / In-country

UK National Blood Service: overall efficiency

Per 100 Attendees

Annual blood usage review (past, present and future) -

hospitals, blood components 20 4

Historical Perspective (1 of 2) Historical Perspective (2 of 2)

70.0 Average 2.3 Average 7.5

50.0 Assess the number of units of blood used in a specified

A total of 73 countries have donation rate of <1%

Hospital Beds Model

hospital beds may be used for patients with complex

Makes it possible to: Example

compare between countries or regions of For a country with a population of 10 million,

Selecting a method to estimate

Method 1 is the most practical where there is a

Methods 2 and 3 can be useful where no data are

Global burden of Disease (GBD) GBD Principles

Health Statistics and Informatics Health Statistics and Informatics

Leading Causes of Mortality and Burden of Disease Overview

Mortality DALYs A brief introduction to the Global Burden of

Health Statistics and Informatics Health Statistics and Informatics

Predicted statistics: predicted using a statistical model,

Adjusted and predicted statistics can be used for

Health Statistics and Informatics Health Statistics and Informatics

Three issues to consider when generating

1. Selecting indicators and metrics Meaningful health indicator

Data are available or collectable

Health Statistics and Informatics Health Statistics and Informatics

Framework for Three issues to consider when generating

Inputs & processes Outputs Outcomes Impact

readiness Social and financial

Health Statistics and Informatics Health Statistics and Informatics

BRFSS a telephone health survey asks about height

Comparing these data:

Ezzati et al. predicted unbiased state rates of overweight

Health Statistics and Informatics Health Statistics and Informatics

Three issues to consider when generating

1. Selecting indicators and metrics

2. Correcting for bias in available data

3. Estimating and communicating

Health Statistics and Informatics Health Statistics and Informatics

Uncertainty in health estimates Predicted hearing loss prevalence

Input uncertainty (especially bias)

Estimating uncertainty in a consistent way across

Uncertainty is a major focus of the new round of

Health Statistics and Informatics Health Statistics and Informatics

Modeling approaches can correct for missing or Data

Service readiness, quality, coverage, health status

Why calculate internationally Framework

Health readiness Social and financial

Health Statistics and Informatics Health Statistics and Informatics

Health Statistics and Informatics Health Statistics and Informatics

Requirements of Blood & Blood