Академический Документы
Профессиональный Документы
Культура Документы
UN Photo/Eskinder Debebe
1799
Editorial
Getty Images
Comment
Published Online
November 18, 2010
DOI:10.1016/S01406736(10)62106-X
Corbis
1801
Comment
3
4
5
6
7
9
10
Comment
3
4
5
6
7
9
10
Comment
Getty Images
1803
Comment
AH was Chair of the Task Force on Climate Change Mitigation and Public
Health. We declare that we have no conicts of interest.
1
2
3
4
5
6
10
11
Getty Images
Comment
4
5
8
9
10
11
12
13
14
Published Online
November 5, 2010
DOI:10.1016/S01406736(10)61998-8
See Articles page 1829
1805
Comment
Oldenbeuving AW, de Kort PL, Jansen BP, Kappelle LJ, Roks G. A pilot study
of rivastigmine in the treatment of delirium after stroke: a safe alternative.
BMC Neurol 2008; 8: 34.
van Eijk MMJ, Roes KCB, Honing MLH, et al. Eect of rivastigmine as an
adjunct to usual care with haloperidol on duration of delirium and
mortality in critically ill patients: a multicentre, double-blind,
placebo-controlled randomised trial. Lancet 2010; published online Nov 5.
DOI:10.1016/S0140-6736(10)61855-7.
Sprung CL, Annane D, Keh D, et al, for the CORTICUS Study Group.
Hydrocortisone therapy for patients with septic shock. N Engl J Med 2008;
358: 11124.
Russell JA, Walley KR, Singer J, et al, for the VASST Investigators.
Vasopressin versus norepinephrine infusion in patients with septic shock.
N Engl J Med 2008; 358: 87787.
Ouimet S, Riker R, Bergeron N, Cossette M, Kavanagh B, Skrobik Y.
Subsyndromal delirium in the ICU: evidence for a disease spectrum.
Intensive Care Med 2007; 33: 100713.
Marquis F, Ouimet S, Riker R, Cossette M, Skrobik Y. Individual delirium
symptoms: do they matter? Crit Care Med 2007; 35: 253337.
Comment
8
9
10
11
12
13
14
Published Online
November 9, 2010
DOI:10.1016/S01406736(10)62046-6
See Articles page 1838
1807
Comment
8
9
10
11
12
13
14
Published Online
November 9, 2010
DOI:10.1016/S01406736(10)62046-6
See Articles page 1838
1807
Comment
4
5
8
9
10
11
1808
Comment
4
5
8
9
10
11
1808
Comment
1809
Comment
10
1810
Comment
10
1810
Comment
5
6
7
8
1811
Comment
Getty Images
Getty Images
World Report
World Report
AP
Barbara Fraser
www.thelancet.com Vol 376 November 27, 2010
World Report
World Report
Tom Partt
www.thelancet.com Vol 376 November 27, 2010
Perspectives
WEBVIDEO Caption: High Society
at Wellcome Collection
Exhibition
Enjoying the high lifedrugs in history and culture
High Society
Wellcome Collection, London,
UK, showing until Feb 27, 2011.
http://www.wellcomecollection.
org/whats-on/exhibitions/
high-society.aspx
Joanna Bourke
j.bourke@bbk.ac.uk
1817
Perspectives
Book
Debating drug policy and the path to change
1818
Virginia Berridge
Virginia.Berridge@lshtm.ac.uk
Perspectives
On Reection
A Chinese puzzle
Jeremy Laurance
J.Laurance@independent.co.uk
Stephen Pincock
stephen.pincock@journalist.co.uk
1819
Perspectives
1820
Perspectives
Danielle Ofri
Department of Medicine, New York University School of Medicine,
New York, NY 10016, USA
Further reading
Ofri D. Incidental ndings:
lessons from my patients in the
art of medicine. Boston, MA:
Beacon Press, 2006.
Ofri D. Medicine in translation:
journeys with my patients.
Boston, MA: Beacon Press, 2010.
Ofri D. Singular intimacies:
becoming a doctor at Bellevue.
Boston, MA: Beacon Press, 2009.
1821
Obituary
Geo Watts
geo@scileg.freeserve.co.uk
Correspondence
zhao_jc120@126.com
AFP/Getty Images
Submissions should be
made via our electronic
submission system at
http://ees.elsevier.com/
thelancet/
1823
Correspondence
Is antiretroviral therapy
modifying the HIV
epidemic?
Julio Montaner and colleagues
(Aug 14, p 532)1 present data from
British Columbia, Canada, on the
association between rates of HIV
diagnoses, coverage of antiretroviral
therapy (ART), and average HIV viral
load. They are incorrect in describing
their study as a population-based
cohort study. It is an ecological study,
because the data on exposure (ART
and viral load) and outcome (HIV
diagnoses) do not come from the same
couples. This is a crucial distinction
because population associations
found in ecological studies often fail
to reect individual-level biological
eects.2
Montaner and colleagues do not
consider the most important confounding eectnamely, the rate
of unsafe injecting in injecting drug
users. Montaners group has reported
data which show that policy changes
led to a greater than 50% decline in
syringe borrowing, and reductions in
HIV incidence in British Columbia.3
Reduction in unsafe injecting has been
acknowledged as an important reason
1824
Correspondence
Authors reply
We acknowledged in our paper that
ours was an ecological study, and as
such the results could not be taken as
denitive proof of causality. We also
indicated that the association between
increasing coverage of highly active
antiretroviral therapy (HAART) in HIVinfected individuals who met contemporary treatment guidelines and
decreasing yearly new HIV diagnoses
occurred against a background of
increased yearly HIV testing, as well
as improved risk ascertainment due
to mandatory HIV reporting, and
increased rates of sexually transmitted
infections.
Notably, our results were internally
reproducible. We recorded decreases
in yearly new HIV diagnoses during
two distinct periods of HAART expansion, which were separated by a stable
period of HAART use. The latter
two periods were characterised on a
prospective basis and were entirely
consistent with the predictions of our
previously published mathematical
models.1 Finally, we were able to
relate the reductions in community
plasma viral load during HAART
expansion with the decrease in new
HIV diagnoses, providing a plausible
biological mechanism to account
for the association, as proposed by
others.2
We also acknowledged that the use
of yearly new HIV diagnoses represented a limitation of the study;
however, it is also clear that there is
no widely accepted gold standard to
estimate HIV incidence, particularly
in population-wide studies. In this
www.thelancet.com Vol 376 November 27, 2010
*Julio S G Montaner,
P Richard Harrigan, Thomas Kerr,
Evan Wood, Patricia Daly
jmontaner@hivnet.ubc.ca
British Columbia Centre for Excellence in HIV/AIDS,
Vancouver, BC V6Z 1Y6, Canada (JSGM, PRH, TK,
EW); and Vancouver Coastal Health Authority,
Vancouver, BC, Canada (PD)
Correspondence
Authors reply
We acknowledged in our paper that
ours was an ecological study, and as
such the results could not be taken as
denitive proof of causality. We also
indicated that the association between
increasing coverage of highly active
antiretroviral therapy (HAART) in HIVinfected individuals who met contemporary treatment guidelines and
decreasing yearly new HIV diagnoses
occurred against a background of
increased yearly HIV testing, as well
as improved risk ascertainment due
to mandatory HIV reporting, and
increased rates of sexually transmitted
infections.
Notably, our results were internally
reproducible. We recorded decreases
in yearly new HIV diagnoses during
two distinct periods of HAART expansion, which were separated by a stable
period of HAART use. The latter
two periods were characterised on a
prospective basis and were entirely
consistent with the predictions of our
previously published mathematical
models.1 Finally, we were able to
relate the reductions in community
plasma viral load during HAART
expansion with the decrease in new
HIV diagnoses, providing a plausible
biological mechanism to account
for the association, as proposed by
others.2
We also acknowledged that the use
of yearly new HIV diagnoses represented a limitation of the study;
however, it is also clear that there is
no widely accepted gold standard to
estimate HIV incidence, particularly
in population-wide studies. In this
www.thelancet.com Vol 376 November 27, 2010
*Julio S G Montaner,
P Richard Harrigan, Thomas Kerr,
Evan Wood, Patricia Daly
jmontaner@hivnet.ubc.ca
British Columbia Centre for Excellence in HIV/AIDS,
Vancouver, BC V6Z 1Y6, Canada (JSGM, PRH, TK,
EW); and Vancouver Coastal Health Authority,
Vancouver, BC, Canada (PD)
Correspondence
ptha@glo.regionh.dk
Danish Headache Centre, University of Copenhagen,
Glostrup Hospital, 2600 Glostrup, Denmark
1
1826
Jennifer Kyne
jennifer.kyne@mhra.gsi.gov.uk
Medicines and Healthcare Products Regulatory
Agency, London SW8 5NQ, UK
1
Improving surgery
service delivery in
context
Luke Funk and colleagues (Sept 25,
p 1055)1 estimate that 2 billion people
lack access to surgery worldwide.
This estimate conrms anecdotal
experiences of substantial unmet
surgical need by those who work with
colleagues in low-income countries.
As Paul Myles and Guy Haller note
in their accompanying Comment,2
although the availability of pulse
oximetry was used as a proxy for access
in Funk and colleagues study, many
other factors such as human resource
constraints (numbers of skilled
surgical and anaesthesia providers),
health-system factors (public vs
private insurance, cost of care), and
other patient-driven factors aect
the actual delivery of care. Given the
complex interplay of these factors on
the ground, it is dicult to identify the
rate-limiting step to service delivery
for any given context.
A substantial (and unknown)
proportion of patients with surgically
correctable disorders do not reach
care at all. Specically, trauma
accounts for the greatest surgical
burden in poor countries, and most
deaths have been shown to occur in
the prehospital setting of countries
with no emergency system.3 In one
such setting (Uganda), we developed
and implemented a basic, contextappropriate prehospital curriculum
for lay rst-responders (police, taxi
drivers, public ocials) and showed
that this can improve care at minimal
cost. This intervention was coupled
with hospital-based trauma training,
and such an approach might also be
applicable in other austere settings.4
www.thelancet.com Vol 376 November 27, 2010
Correspondence
*Doruk Ozgediz,
Jacqueline Mabweijano,
Cephas Mijumbi, Sudha Jayaraman,
Michael Lipnick
dozgediz@hotmail.com
University of Toronto, Toronto, ON M5G 2J8,
Canada (DO); Makerere University, Kampala,
Uganda (JM, CM); University of California at San
Francisco, San Francisco, CA, USA (SJ); and Global
Partners in Anesthesia and Surgery, San Francisco,
CA, USA (DO, JM, CM, SJ, ML)
1
2
3
Cancer funding in
developing countries:
the next health-care
crisis?
We welcome the call to expand
cancer control in low-income and
middle-income countries (LMIC) by
Paul Farmer and colleagues (Oct 2,
p 1186).1 The Global Fund to ght
AIDS, Tuberculosis and Malaria was
formed in 2002 after a call in The
Lancet by Amir Attaran and Jerey
Sachs2 for a special funding stream for
HIV/AIDS. Can we expect the same for
cancer?
In less developed countries alone,
more people develop (71 million) and
die from cancer (48 million) each year3
www.thelancet.com Vol 376 November 27, 2010
The printed
journal
includes an
image merely
for illustration
Reuters
Funk and colleagues study underscores the urgent need for resources to
implement and assess interventions
that reduce global disparities in access
to essential surgical care, which should
be seen as a basic human right.5
1827
Correspondence
Published Online
November 15, 2010
DOI:10.1016/S01406736(10)62108-3
1828
Ami Schattner
amiMD@clalit.org.il
Department of Medicine, Kaplan Medical Center,
POB 1, Rehovot 76100, Israel
1
Compassionate
optimism
Many non-clinical factors can interfere
with the pure, evidence-based process
of physicians decision making.1
They include diverse personal factors
such as physicians reactions to fear
of litigation and poor tolerance of
common clinical uncertainty, but also
overcondence and hidden nancial
incentives.2,3 A previously unreported
personal factor aecting objective
evidence-based decision making
by physicians might be termed
compassionate optimism.
Compassionate optimism might
occur when a physicians empathy and
compassion towards his or her patient
causes erroneous interpretation of
clinical data to favour a more benign, less
ominous diagnosis, contrary to the facts
and clinical probability. For example,
a clinician who strongly sympathises
with a young mother who has osteolytic spine lesions might adhere to a
false belief that they represent a treatable granulomatous disease,4 whereas
Department of Error
Alwan A, MacLean DR, Riley LM, et al. Monitoring
and surveillance of chronic non-communicable
diseases: progress and capacity in high-burden
countries. Lancet 2010; 376: 186168In the
Summary and line 7 of the Mortality section of
this Series (Nov 27), the proportion of deaths
occurring in people younger than 70 years
should have been 47%. This correction has
been made to the online version as of Nov 15,
2010, and to the printed version.
Fernald LCH, Gertler PJ, Neufeld LM. 10-year
eect of Oportunidades, Mexicos conditional
cash transfer programme, on child growth,
cognition, language, and behaviour: a
longitudinal follow-up study. Lancet 2009;
374: 19972005In this Article (Dec 12,
2009), the last sentence of the Findings
section in the Summary should have read: An
additional 18 months of the programme
before age 3 years for children aged 810 years
whose mothers had no education resulted in
improved child growth of about 15 cm
assessed as height-for-age Z score ( 023
[0023044] p=0029), independently of cash
received. This correction has been made to
the online version as of Nov 26, 2010.
Articles
Summary
Background Delirium is frequently diagnosed in critically ill patients and is associated with adverse outcome. Impaired
cholinergic neurotransmission seems to have an important role in the development of delirium. We aimed to establish
the eect of the cholinesterase inhibitor rivastigmine on the duration of delirium in critically ill patients.
Methods Patients (aged 18 years) who were diagnosed with delirium were enrolled from six intensive care units in
the Netherlands, and treated between November, 2008, and January, 2010. Patients were randomised (1:1 ratio) to
receive an increasing dose of rivastigmine or placebo, starting at 075 mL (15 mg rivastigmine) twice daily and
increasing in increments to 3 mL (6 mg rivastigmine) twice daily from day 10 onwards, as an adjunct to usual care
based on haloperidol. The trial pharmacist generated the randomisation sequence by computer, and consecutively
numbered bottles of the study drug according to this sequence to conceal allocation. The primary outcome was the
duration of delirium during hospital admission. Analysis was by intention to treat. Duration of delirium was censored
for patients who died or were discharged from hospital while delirious. Patients, medical sta, and investigators were
masked to treatment allocation. Members of the data safety and monitoring board (DSMB) were unmasked and did
interim analyses every 3 months. This trial is registered with ClinicalTrials.gov, number NCT00704301.
Findings Although a sample size of 440 patients was planned, after inclusion of 104 patients with delirium who were
eligible for the intention-to-treat analysis (n=54 on rivastigmine, n=50 on placebo), the DSMB recommended that the
trial be halted because mortality in the rivastigmine group (n=12, 22%) was higher than in the placebo group (n=4, 8%;
p=007). Median duration of delirium was longer in the rivastigmine group (50 days, IQR 27142) than in the
placebo group (30 days, IQR 1093; p=006).
Interpretation Rivastigmine did not decrease duration of delirium and might have increased mortality so we do not
recommend use of rivastigmine to treat delirium in critically ill patients.
Funding ZonMw, the Netherlands Brain Foundation, and Novartis.
Introduction
Delirium is an acute neuropsychiatric syndrome,
characterised by disturbances of attention and other
cognitive functions.1 Delirium is very common in patients
in intensive care units, with reported frequencies up
to 80%,2,3 and is associated with increased complications,
length of hospital stay, and mortality.35 Antipsychotics
and benzodiazepines have become the standard drug
treatment for delirium in critically ill patients,6 although
few placebo-controlled trials have been done for
antipsychotics7,8 and none for benzodiazepines.9
Furthermore, haloperidol might lead to extrapyramidal
side-eects and ventricular arrhythmia,10 whereas
benzodiazepines might even induce delirium.11
Several lines of evidence suggest that impaired
cholinergic neurotransmission has an important role in
the development of delirium.12 Serum anticholinergic
activity was found to be increased in patients with
delirium,13 and drugs with anticholinergic eects can
www.thelancet.com Vol 376 November 27, 2010
Methods
Patients
Articles
6724 patients
4 withdrawn by family
Procedures
54 included in intention-to-treat analysis
12 died
6 died
1830
4 died
7 died
Articles
Rivastigmine (n=54)
Placebo (n=50)
Age (years)
680 (114)
700 (122)
Male sex
38 (70%)
29 (58%)
APACHE II score
203 (89)
196 (79)
SOFA score
56 (23)
55 (31)
26 (23)
23 (23)
RASS score
+2
12 (22%)
9 (18%)
+1
20 (37%)
14 (28%)
3 (6%)
7 (14%)
10 (19%)
11 (22%)
9 (17%)
9 (18%)
Admitting specialty
Internal medicine
17 (31%)
15 (30%)
General surgery
21 (39%)
16 (32%)
13 (24%)
18 (36%)
Neurology or neurosurgery
Emergency admission to intensive care unit
3 (6%)
1 (2%)
46 (85%)
32 (64%)
33 (61%)
36 (72%)
Sensory handicap
Visual
Auditory
7 (13%)
12 (24%)
IQCODE score
515 (63)
529 (59)
11 (20%)
6 (12%)
Smoking
15 (28%)
10 (20%)
5 (9%)
5 (10%)
Benzodiazepines
7 (13%)
6 (12%)
Antipsychotics
6 (11%)
3 (6%)
Analgesics
5 (9%)
Psychiatric diagnosis
Drug use before admission to intensive care unit
120 (58360)
120 (65225)
Data are mean (SD), number (%), or median (IQR). APACHE=acute physiology and chronic health evaluation.
SOFA=sequential organ failure assessment. RASS=Richmond agitation sedation scale. IQCODE=informant
questionnaire on cognitive decline in the elderly. *Data were missing for three patients on rivastigmine.
Articles
Rivastigmine (n=54)
Placebo (n=50)
p value
Primary outcome
Delirium duration (days)
50 (27142)
30 (1093)
006
40 (20160)
25 (1058)
006
60 (35115)
60 (30215)
095
95 (48118)
80 (1090)
029
109 (51327)
Secondary outcomes
Total dose of study drug (mL)
64 (45225)
012
Fixation days
10% (0043)
10% (0023)
090
0 (01)
0 (01)
067
23 (2031)
20 (1825)
00040
030
Median of mean use of psychotropic drugs for individual patients (mg per day)
Haloperidol
32 (2656)
31 (1853)
Lorazepam
02 (0006)
00 (0005)
028
Propofol
00 (001043)
00 (00237)
022
69/659 (10%)
16/459 (3%)
<00001
590/659 (90%)
443/459 (97%)
<00001
15 (930)
8 (317)
<00001
Hospital
29 (1755)
25 (1739)
006
Mortality
During treatment with study drug
12 (22%)
4 (8%)
007
At 90 days of follow-up
18 (33%)
11 (22%)
014
Data are median (IQR) or number (%), unless otherwise indicated. DSI=delirium severity index. RASS=Richmond
agitation and sedation scale.
Statistical analysis
On the basis of an average duration of delirium in
patients in intensive care of 34 days (SD 19),28 at 80%
power with =005, 220 patients per group and
440 patients in total were needed to detect a clinically
signicant reduction in the average duration of delirium
of 15% or half a day.
The primary outcome was compared across the two
study groups in an intention-to-treat analysis with a
Mann-Whitney U test and Cox regression analysis.
Duration of delirium was censored for patients who died
or were discharged from hospital while delirious to avoid
underestimation of the duration of delirium in these
patients. We used Cox regression analysis because it is
more appropriate for analysis of censored data than is
linear regression analysis of logarithmically transformed
data. Hazard ratios (HRs) were adjusted for sex and
emergency admissions to the intensive care unit at
baseline. The primary outcome was also analysed in
subgroups of patients meeting three endpoints: end of
delirium (CAM-ICU negative for 48 h), discharge from
hospital, or death during treatment with the study drug.
Mortality rates were compared between the two groups
with a Kaplan-Meier curve and log-rank statistics.
Secondary outcomes were compared across the two study
groups in an intention-to-treat analysis with Mann-Whitney
U tests. Delirium severity was assessed as the cumulative
DSI score in each patient divided by number of days during
which the patient had delirium. Disease severity was
estimated as the cumulative SOFA score in each patient
divided by the number of study days. We calculated p values
with Mann-Whitney U tests or tests as appropriate.
After interruption of the study, a post-hoc exploratory
analysis of several clinical indicators was done to identify
an underlying biological mechanism. To this end, standard
haemodynamic indicators (heart rate, arterial blood
pressure, and urine production), which were used as
surrogate variables that could represent cardiac output,
and indicators of infections (number of leucocytes and
www.thelancet.com Vol 376 November 27, 2010
Articles
Duration
of study
(days)
Dose of
study drug
(mL)*
Autopsy
10
Rivastigmine
Yes
22
15
34
075
Rivastigmine
Yes
70
30
13
075
Rivastigmine
Chronic renal failure and abdominal aortic aneurysm for which endograft
placement was done, but became infected with Staphylococcus aureus,
resulting in septic shock; persistent endograft leakage resulting in death
Yes
Female
69
16
12
Placebo
Yes
Patient 5
Male
77
19
10
225
Rivastigmine
No
Patient 6
Male
78
26
15
Rivastigmine
No
Patient 7
Female
59
20
14
075
Rivastigmine
Yes
Patient 8
Male
82
19
18
Placebo
No
Patient 9
Male
74
11
13
11
Rivastigmine
No
Patient 10 Male
88
26
11
075
Placebo
No
Patient 11 Male
57
40
12
075
Rivastigmine
No
Patient 12 Male
55
23
16
15
Placebo
No
Patient 13 Female
75
17
075
Rivastigmine
Yes
Patient 14 Female
83
12
13
225
Rivastigmine
No
Patient 15 Female
62
38
18
15
Rivastigmine
No
Patient 16 Female
80
15
Rivastigmine
No
Sex
Age
(years)
APACHE II
score
Patient 1
Female
66
26
Patient 2
Male
74
Patient 3
Male
Patient 4
SOFA
score*
APACHE=acute physiology and chronic health evaluation. SOFA=sequential organ failure assessment. *At 24 h before death.
Results
The DSMB recommended that the trial be halted after
the fourth interim analysis and inclusion of 109 patients
1833
Articles
100
Placebo
Rivastigmine
Survival (%)
80
60
40
20
0
0
See Online for webappendix
Patients at risk
Placebo 50
Rivastigmine 54
30
60
Time (days after inclusion)
44
40
41
37
90
39
36
Discussion
We have shown that in critically ill patients, rivastigmine
did not decrease duration of delirium when it was added
to standard treatment with haloperidol (panel).
Furthermore, rivastigmine was associated with a more
severe type of delirium, longer stay in the intensive care
unit, and higher mortality than was placebo, although the
dierence in mortality between treatment groups was not
signicant. The DSMB advised that the trial should be
halted prematurely because of increased mortality in the
rivastigmine group compared with the placebo group.
Other important arguments were the fact that
demographic and clinical characteristics of the treatment
groups were well balanced at baseline, and rivastigmine
was not associated with any benecial eects. The
investigators had no counter arguments to the advice of
the DSMB so the trial was halted prematurely.
This large randomised placebo-controlled trial of a
cholinesterase inhibitor is one of the very few randomised
controlled trials of treatment for delirium in critically ill
patients.7,8,29,31 This trial benets from the multicentre
design, which contributes to generalisability of our
ndings. Although we excluded 141 patients with no
informed consent and 139 for other reasons before
randomisation, the other reasons included inability to
speak Dutch or English, expected stay in the intensive
care unit of less than 48 h, and logistical problems.
Therefore, we believe that our study population can be
www.thelancet.com Vol 376 November 27, 2010
Articles
Interpretation
In our trial, the cholinesterase inhibitor rivastigmine had no
benecial eect for treatment of delirium in critically ill
patients, and might have increased mortality. These results,
combined with the ndings of previous studies, do not
support the use of cholinesterase inhibitors to treat delirium
in critically ill patients.
Articles
Contributors
MMJvE and AJCS conceived and designed the study with input from
MAK, WAvG, RCvdM, and JK. MMJvE, MLHH, MAK, AK, MvdJ, PES,
JK, and AJCS undertook the trial and collected data. KCBR supervised
all statistical analyses and was responsible for the post-hoc data
analysis. MMJvE, KCBR, and AJCS interpreted the data. MMJvE and
AJCS drafted the report with input and editing from all authors.
Participating centres
Department of Intensive Care Medicine, University Medical Centre,
Utrecht, Netherlands (MMJvE [coordinator], JK [director], AJCS
[principal investigator], n=60 patients included in analysis); Department
of Intensive Care, Medical Centre Alkmaar, Alkmaar, Netherlands
(MLHH [investigator], n=11); Department of Intensive Care, Medical
Centre Leeuwarden, Leeuwarden, Netherlands (MAK [investigator],
n=11); Department of Intensive Care, Diakonessenhuis Utrecht,
Utrecht, Netherlands (AK [investigator], n=8); Department of Intensive
Care, Erasmus MC-University Medical Centre, Rotterdam, Netherlands
(MvdJ [investigator], n=8); and Department of Intensive Care, Gelre
Hospitals (Lukas site), Apeldoorn, Netherlands (PES [investigator], n=6).
DSMB members
R J van Marum (Department of Geriatrics, Jeroen Bosch Ziekenhuis,
Den Bosch, Netherlands [chair]); I van der Tweel (Department of
Biostatistics, Julius Centre for Health Sciences and Primary Care,
University Medical Centre, Utrecht, Netherlands); P Eikelenboom
(Department of Neurology, Academic Medical Centre, and Department
of Psychiatry, VU Medical Centre, Amsterdam, Netherlands);
J M A Sitsen (Utrecht Institute for Pharmaceutical Sciences, Utrecht
University, Utrecht, Netherlands); E de Jonge (Department of Intensive
Care Medicine, Leiden University Medical Centre, Leiden, Netherlands);
and D Zandstra (Department of Intensive Care Medicine, Onze Lieve
Vrouwe Gasthuis, Amsterdam, Netherlands).
14
Conicts of interest
RCvdM has received payment and travel expenses for presentations and
attendance at committee meetings from the following professional
organisations: the Netherlands Society of Psychiatry, the Netherlands
Society of Intensive Care, the Health Care Inspectorate, and the
Academic Medical Centre, University of Amsterdam. All other authors
declare that they have no conicts of interest.
Acknowledgments
We thank all patients and their legal representatives for participation in
this trial; all research nurses in participating centres; medical and nursing
sta in participating centres who cared for patients and collected data;
data managers who participated in this trial, especially Frank Leus and
Joost Schotsman; the pharmacists in all participating centres for their
help; Michel Bots and Ardine de Wit for their contribution to the design
of the trial; and DSMB members. This study was funded by ZonMw, the
Netherlands organisation for health research and development (grant
number 80-82305-98-08109), the Netherlands Brain Foundation (grant
number 2008(1).30), and Novartis, Basel, Switzerland (unrestricted grant).
References
1
American Psychiatric Association. Diagnostic and statistical manual
of mental disorders, 4th edn. Washington, DC: American
Psychiatric Association, 2000.
2
Ouimet S, Kavanagh BP, Gottfried SB, Skrobik Y. Incidence, risk
factors and consequences of ICU delirium. Intensive Care Med 2007;
33: 6673.
1836
10
11
12
13
15
16
17
18
19
20
21
22
23
24
25
Articles
26
27
28
29
30
31
32
33
34
1837
Articles
Summary
Lancet 2010; 376: 183845
Published Online
November 9, 2010
DOI:10.1016/S01406736(10)61997-6
See Comment page 1807
Department of Medical
Microbiology, University of
Nairobi, Nairobi, Kenya
(R T Lester MD, A Kariri BSc,
S Karanja BSc, E Ngugi PhD,
L J Gelmon MD, J Kimani, MBChB);
Department of Medical
Microbiology, University of
Manitoba, Health Sciences
Centre, Winnipeg, MB, Canada
(R T Lester, T B Ball PhD,
L J Gelmon, J Kimani,
Prof F A Plummer MD); Division
of Infectious Diseases,
Department of Medicine,
University of British Columbia,
Vancouver, BC, Canada
(R T Lester); School of
Kinesiology and Health
Sciences, Department of
Psychology, York University,
York, ON, Canada (P Ritvo PhD);
Faculty of Health Sciences,
University of Ottawa, Ottawa,
ON, Canada (E J Mills PhD);
Department of Global Health,
University of Washington,
Seattle, WA, USA
(M H Chung MD); Department of
Economics (W Jack DPhil) and
Georgetown Public Policy
Institute (J Habyarimana PhD),
Georgetown University,
Washington, DC, USA;
Collaboration for Outcome
Research and Evaluation,
Faculty of Pharmaceutical
Sciences, University of British
Columbia, Vancouver, BC,
Canada (M Sadatsafavi MD,
M Najafzadeh MSc,
C A Marra PharmD); University
of Nairobi Institute of Tropical
and Infectious Diseases,
Nairobi, Kenya
(B Estambale MBChB);
Department of Clinical
Epidemiology and Biostatistics,
McMaster University, Hamilton,
ON, Canada (L Thabane PhD);
1838
Background Mobile (cell) phone communication has been suggested as a method to improve delivery of health
services. However, data on the eects of mobile health technology on patient outcomes in resource-limited settings
are limited. We aimed to assess whether mobile phone communication between health-care workers and patients
starting antiretroviral therapy in Kenya improved drug adherence and suppression of plasma HIV-1 RNA load.
Methods WelTel Kenya1 was a multisite randomised clinical trial of HIV-infected adults initiating antiretroviral
therapy (ART) in three clinics in Kenya. Patients were randomised (1:1) by simple randomisation with a random
number generating program to a mobile phone short message service (SMS) intervention or standard care. Patients
in the intervention group received weekly SMS messages from a clinic nurse and were required to respond within
48 h. Randomisation, laboratory assays, and analyses were done by investigators masked to treatment allocation;
however, study participants and clinic sta were not masked to treatment. Primary outcomes were self-reported ART
adherence (>95% of prescribed doses in the past 30 days at both 6 and 12 month follow-up visits) and plasma HIV-1
viral RNA load suppression (<400 copies per mL) at 12 months. The primary analysis was by intention to treat. This
trial is registered with ClinicalTrials.gov, NCT00830622.
Findings Between May, 2007, and October, 2008, we randomly assigned 538 participants to the SMS intervention (n=273)
or to standard care (n=265). Adherence to ART was reported in 168 of 273 patients receiving the SMS intervention
compared with 132 of 265 in the control group (relative risk [RR] for non-adherence 081, 95% CI 069094; p=0006).
Suppressed viral loads were reported in 156 of 273 patients in the SMS group and 128 of 265 in the control group,
(RR for virologic failure 084, 95% CI 071099; p=004). The number needed to treat (NNT) to achieve greater than
95% adherence was nine (95% CI 50295) and the NNT to achieve viral load suppression was 11 (582273).
Interpretation Patients who received SMS support had signicantly improved ART adherence and rates of viral
suppression compared with the control individuals. Mobile phones might be eective tools to improve patient
outcome in resource-limited settings.
Funding US Presidents Emergency Plan for AIDS Relief.
Introduction
Health programmes that use mobile communication
technologies are emerging with the aim of strengthening health systems.13 The United Nations Joint
Programme on HIV/AIDS (UNAIDS) and WHO have
added wireless communication technologies to their
strategic plans.4,5 However, at present no published
clinical trial has reported the use of mobile health
technologies to improve patient-centred outcomes in
developing countries.
Present eorts to control the HIV/AIDS pandemic
include treatment with antiretroviral therapy (ART),
targeted prevention strategies, and treatment as
prevention measures (ie, prevention of HIV spread by
treating HIV positive people and thereby reducing the
risk of onward transmission).6,7 However, widespread
progress at controlling the pandemic is restricted by poor
infrastructure and increasing health-system costs.
Articles
Methods
Patients
Patients initiating ART were recruited from three
dierent HIV clinics that are involved in intense ART
provision scale-up. The University of Nairobi Pumwani
Clinic serves a very low-income population in Nairobi13
and the Coptic Hope Center for Infectious Diseases
operates out of a faith-based hospital located in a higherincome area of Nairobi.14 The Kajiado Clinic is a
government health centre in a large rural district. We
chose these three locations because they should represent
the regional diversity of health settings.
Patients were eligible for study participation if they
were over 18 years old, initiating ART for the rst time,
and able to access a mobile phone on a near-daily basis
and communicate via short message service (SMS).
People who did not own mobile phones were eligible if
they had shared access (with corroborative agreement by
the phone owner), and illiterate patients were eligible if
assisted by a literate partner. Participants used existing
mobile phone services; phones and network airtime
credit were not provided.
Patients provided written or verbal informed consent at
enrolment in a language they understood. The study
protocol was approved by the University of Manitoba and
Kenyatta National Hospital ethics review boards.
This trial is registered with ClinicalTrials.gov,
NCT00830622.
Procedures
Antiretroviral drugs were provided by the government of
Kenya with support from the US Presidents Emergency
Plan for AIDS Relief (PEPFAR), and consisted primarily
www.thelancet.com Vol 376 November 27, 2010
43 ineligible
39 had inadequate
phone access
4 declined participation
538 randomised
7 withdrew
265 assigned to
standard care
3 withdrew
273 included in
primary analysis
265 included in
primary analysis
1839
Articles
Women
Age (years)
177 (65%)
174 (66%)
Clinic
University of Nairobi Pumwani Clinic
120 (44%)
131 (49%)
117 (43%)
92 (35%)
Kajiado Clinic
36 (13%)
42 (16%)
1609 (1414, 10 15220)
WHO stage
1
52 (23%)
67 (29%)
59 (25%)
101 (44%)
103 (43%)
9 (4%)
13 (5%)
459 (105)
62 (26%)
483 (096)
Language literacy
English only
Kiswahili only
Both
Other only
1 (0%)
2 (1%)
48 (18%)
36 (14%)
213 (79%)
215 (81%)
9 (3%)
12 (5%)
Education
10 (4%)
14 (5%)
Primary
None
108 (40%)
86 (32%)
Secondary
106 (39%)
124 (47%)
49 (18%)
41 (16%)
Post-secondary
Monthly income
<2000 KES (<US$1/day)
72 (29%)
64 (28%)
114 (47%)
98 (43%)
46 (19%)
61 (27%)
13 (5%)
7 (3%)
Owns
239 (88%)
225 (85%)
Shares
34 (12%)
40 (15%)
Residence status
Rural
51 (19%)
50 (19%)
Urban
222 (81%)
215 (81%)
Data are number (%), mean (SD, range), median (IQR), mean (SD). Percentages do not add up to 100% in some cases
because of rounding. KES=Kenyan shillings. *Data missing for one patient in the SMS group and four in the control
group. Data missing for 44 patients in the SMS group and 28 in the control group. Data missing for 21 patients in
the SMS group and 25 in the control group; for 19 patients in the SMS group and ten in the control group, baseline
viral load was below the limit of detection (400 copies per mL). Log of viral load for these patients was given as
log10(400). Data missing for two patients in the SMS group. Data unavailable for 28 patients in the SMS group and
35 in the control group.
Statistical analysis
We calculated that a sample size of at least 534 would be
required to detect a 10% improvement in adherence, with
80% power and 005 level of signicance.20,21 Demographic
and covariate information were recorded at baseline
(month 0) and at scheduled visits at 6 and 12 months. Selfreported adherence to ART was assessed by a standardised
questionnaire at each follow-up visit. Study sta
maintained a study register to record all SMS responses
and other mobile phone communications with patients.
Patients dened as lost to follow-up were those unable to
be traced within 3 months of the study end date.
Detailed description of the analysis methods can be
found in the trial protocol.20 Briey, we analysed the
www.thelancet.com Vol 376 November 27, 2010
Articles
Results
Between May, 2007, and October, 2008, we enrolled
581 participants (gure 1). Consecutive enrolment was
attempted; however, one site enrolled alternate patients
into separate studies. After screening, 39 patients were
www.thelancet.com Vol 376 November 27, 2010
SMS group
(number [%])
Control group
(number [%])
RR (95% CI)*
p value
Primary outcome
Intention-to-treat analysis
Self-reported adherence (>95%)
168 (62%)
132 (50%)
081 (069094)
0006
Viral suppression
(<400 copies per mL)
156 (57%)
128 (48%)
085 (072099)
004
Complete-case analysis
Self-reported adherence
168 (91%)
132 (91%)
100 (094107)
094
Viral suppression
156 (75%)
128 (66%)
088 (077100)
0047
53 (19%)
61 (23%)
124 (082189)
031
Loss to follow-up
17 (6%)
27 (10%)
169 (091323)
0094
Mortality
Secondary outcomes
25 (9%)
30 (11%)
127 (072222)
042
Withdrawal
7 (3%)
3 (1%)
226 (059867)
034||
Transfer out
4 (1%)
1 (0%)
025 (019217)
038||
Percentages do not add up to 100% in some cases because of rounding. *For non-adherence or virologic failure.
273 patients in the SMS group and 265 in the control group. Because the intention-to-treat analysis classed all patients
with missing data as non-adherent or having viral failure, the number of adherent patients and number of patients with
viral suppression are the same here as in the intention-to-treat analysis. 185 patients in the SMS group and 145 patients
in the control group. 208 patients in the SMS group and 194 patients in the control group. ||Fishers exact test.
Articles
OR (95% CI)
p value
Viral load
OR (95% CI)
p value
Sex
Sex
Women
Men
112/177 (63%)
070 (046107)
56/96 (58%)
049 (027088)
070
Women
101/177 (57%)
066 (043100)
55/96 (57%)
080 (045142)
Men
073
Mobile phone
Mobile phone
Shares
19/34 (56%)
053 (021133)
Owns
149/239 (62%)
064 (044093)
086
Shares
22/34 (65%)
036 (014094)
Owns
134/239 (56%)
078 (054112)
036
Clinic
Clinic
Pumwani
82/120 (68%)
064 (038108)
Pumwani
75/120 (63%)
054 (033089)
Kajiado
16/36 (44%)
Kajiado
20/36 (56%)
Coptic
70/117 (60%)
059 (034102)
Coptic
61/117 (52%)
105 (061181)
WHO stage
WHO stage
1
36/52 (69%)
070 (032153)
31/52 (60%)
094 (044198)
45/67 (67%)
054 (026111)
42/67 (63%)
058 (028117)
54/101 (53%)
062 (036108)
56/101 (55%)
058 (033100)
7/9 (78%)
024 (004166)
5/9 (56%)
128 (023719)
Rural
30/51 (59%)
051 (023112)
Urban
126/222(57%)
075 (052110)
Overall
174/265 (66%)
070 (050098)
002
006
Residence
Residence
Rural
29/51 (57%)
051 (023112)
Urban
139/222 (63%)
065 (044095)
Overall
168/273 (62%)
062 (044088)
0125
025
05
1
Favours
SMS
group
081
0125
025
Favours
control
group
05
1
Favours
SMS
group
057
2
Favours
control
group
08
07
06
05
04
03
02
Sawa (ne)
No response
Shida (problem)
01
0
0
6
Months
10
11
12
Discussion
This study shows that mobile health innovations can
improve HIV treatment outcomes. Patients who received
the SMS support were more likely to report adherence to
ART and were more likely to have their viral load
suppressed below detection levels than patients who
received the standard care alone.
The primary analysis classed all-cause attrition as
treatment failures. Thus, the higher follow-up rates and
www.thelancet.com Vol 376 November 27, 2010
Articles
Articles
1844
Articles
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
1845
Articles
Summary
Lancet 2010; 376: 184652
Published Online
October 27, 2010
DOI:10.1016/S01406736(10)61195-6
See Comment page 1808
Department of Health, London,
UK (N Sachedina MBBS,
Prof L J Donaldson MD); and
National Patient Safety
Agency, London, UK
(L J Donaldson)
Correspondence to:
Prof Liam J Donaldson, National
Patient Safety Agency,
48 Maple Street, London
W1T 5HD, UK
liam.donaldson@npsa.nhs.uk
Background Young people (aged 018 years) have been disproportionately aected by pandemic inuenza A H1N1
infection. We aimed to analyse paediatric mortality to inform clinical and public health policies for future inuenza
seasons and pandemics.
Methods All paediatric deaths related to pandemic inuenza A H1N1 infection from June 26, 2009, to March 22, 2010
in England were identied through daily reporting systems and cross-checking of records and were validated by
conrmation of inuenza infection by laboratory results or death certicates. Clinicians responsible for each
individual child provided detailed information about past medical history, presentation, and clinical course of the
acute illness. Case estimates of inuenza A H1N1 were obtained from the Health Protection Agency. The primary
outcome measures were population mortality rates and case-fatality rates.
Findings 70 paediatric deaths related to pandemic inuenza A H1N1 were reported. Childhood mortality rate was
6 per million population. The rate was highest for children aged less than 1 year. Mortality rates were higher for
Bangladeshi children (47 deaths per million population [95% CI 17103]) and Pakistani children (36 deaths per million
population [1864]) than for white British children (4 deaths per million [36]). 15 (21%) children who died were
previously healthy; 45 (64%) had severe pre-existing disorders. The highest age-standardised mortality rate for a preexisting disorder was for chronic neurological disease (1536 per million population). 19 (27%) deaths occurred before
inpatient admission. Children in this subgroup were signicantly more likely to have been healthy or had only mild
pre-existing disorders than those who died after admission (p=00109). Overall, 45 (64%) children had received
oseltamivir: seven within 48 h of symptom onset.
Interpretation Vaccination priority should be for children at increased risk of severe illness or death from inuenza.
This group might include those with specied pre-existing disorders and those in some ethnic minority groups. Early
pre-hospital supportive and therapeutic care is also important.
Funding Department of Health, UK
1846
Introduction
Methods
Study population
Mandatory daily reporting systems were established for
all suspected and conrmed deaths from pandemic
inuenza A H1N1 in England. Further deaths were
identied through cross-checking of records held by the
Regional Directors of Public Health and by the Health
Protection Agencys inuenza reference centres. For all
reported cases, a member of the Chief Medical Ocers
clinical team contacted the responsible senior physician
to obtain further details. A death was conrmed as related
to pandemic inuenza A H1N1 if there was laboratory
evidence of H1N1 infection or if H1N1 infection (or
synonym) was recorded on the death certicate. Further
details of the method of death ascertainment have been
reported previously.5
Study design
All validated deaths in children aged less than 18 years
were extracted from this dataset. The clinician responsible
for each individual child was interviewed by telephone by
www.thelancet.com Vol 376 November 27, 2010
Articles
60 000
Number of deaths
Midpoint case estimates
40 000
6
5
30 000
4
3
20 000
2
10 000
50 000
7
1
0
22 29 6 13 20 27 3 10 17 24 31 7 14 21 28 5 12 19 26 2 9 16 23 30 7 14 21 28 4 11 18 25 1 8 15 22 1 8 15 22
July,
August,
September,
October,
November,
December,
January,
February,
March,
June,
2009
2009
2009
2009
2009
2009
2010
2010
2010
2009
Figure: Estimated incidence of pandemic inuenza A H1N1 cases and conrmed deaths in children in England, by week
Data for case estimates provided by the Health Protection Agency.
Statistical analyses
Case-fatality rates were calculated for every age-group
with Health Protection Agency midpoint case estimates.
The upper and lower estimates by age were used to
calculate lower and upper estimates for the case-fatality
rates, respectively. A 95% condence interval was
calculated around these estimates, to account for the
uncertainty around the number of cases. The case-fatality
1847
Articles
Number
of deaths
< 1 year
151 (34635)
14 (626)
14 years
15
33 (9114)
6 (310)
59 years
22
17 (554)
8 (512)
1015 years
15
10 (335)
4 (27)
1617 years
26 (6102)
7 (313)
017 years
70
19 (751)
6 (58)
Table 1: Case-fatality rates for pandemic inuenza A H1N1 infection and population mortality rates by
age in England from June 26, 2009, to March 22, 2010
Number of deaths
Fever
60 (86%)
Cough
47 (67%)
Dyspnoea
46 (66%)
Dehydration
18 (26%)
17 (24%)
Diarrhoea
15 (21%)
Nausea or vomiting
14 (20%)
Seizure or convulsion
5 (7%)
Headache
4 (6%)
Sore throat
4 (6%)
Results
A total of 70 paediatric deaths related to pandemic
inuenza A H1N1 occurred in England during the study
1848
Articles
6 (214)
p value
7 (2512)
09761*
Ethnic origin
White British
Asian or Asian British (any)
12 (63%)
25 (49%)
6 (32%)
21 (41%)
07547
1 (5%)
3 (6%)
9 (47%)
8 (16%)
00109
9 (47%)
19 (37%)
05842
Pre-hospital antivirals
2 (10%)
1 (2%)
2 (125)
10 (55165)
<00001*
1 (0252)
2 (0375)
04533*
01770
Data are number of patients (%) unless otherwise stated. *Mann-Whitney U test; 2 with Yates correction; Fishers test.
Table 3: Characteristics of children presenting with rapid deterioration leading to death before, or at the
point of, hospital admission (early death) and of those who died after hospital admission (late death)
37 (53%)
7 (212)
13 (19%)
4 (15)
Home oxygen
12 (17%)
4 (0548)
10 (14%)
12 (9512)
Pulmonary hypertension
6 (9%)
14 (06511)
Tracheostomy
6 (9%)
6 (2251025)
4 (6%)
11 (814)
Home ventilation
3 (4%)
11 (5611)
17 (24%)
5 (114)
12 (17%)
Cardiomyopathy
3 (4%)
16 (07710)
16 (1216)
41 (59%)
6 (1510)
33 (47%)
6 (29)
Gastro-oesophageal reux
30 (43%)
7 (898)
Unsafe swallow
24 (34%)
7 (389)
38 (54%)
7 (298)
Cerebral palsy
22 (31%)
7 (49)
Epilepsy
22 (31%)
8 (612)
7 (10%)
6 (2713)
4 (6%)
5 (7%)
5 (7%)
14 (816)
1 (1%)
16*
Other
75 (1427)
16 (1316)
Articles
Deaths
33
392
16
72
232 (132377)
35
31
1536 (9882242)
Immunosuppression
38
166 (66343)
449 (851333)
Diabetes mellitus
None
13
10 125
167 (89268)
0 ()
13 (0722)
Table 5: Age-standardised mortality rate for deaths related to pandemic inuenza A H1N1 for children
aged between 6 months and 18 years with specic pre-existing disorders
Discussion
Our report of the 70 deaths in children in England related
to pandemic inuenza A H1N1 has shown that mortality
disproportionately aects ethnic minorities and those
with pre-existing disorders. Many deaths occurred before
hospital admission, especially in healthy children or
those with only mild pre-existing disorders.
The overall childhood mortality rate for pandemic
inuenza A H1N1 reported here (six per million
population) is close to that for the Netherlands (ve per
million population for children under 14 years9) but lower
than that in Argentina (11 per million population).3 A high
disease burden was reported in the southern hemisphere
because the pandemic coincided with the inuenza
season in that region. Delayed presentation and
unrecognised illness might have contributed further to
1850
Articles
Articles
Contributors
LJD planned this study, oversaw its design, and contributed to writing
the paper. NS planned and designed the data collection process, obtained
and analysed all data, and drafted the manuscript.
Conicts of interest
LJD was the Chief Medical Ocer for England from 1998 to May, 2010.
In this role he advised the government on public health policy, including
the management of the pandemic. NS supported him in this task from
2009 to 2010. Both authors declare that they have no additional conicts
of interest.
Acknowledgments
We thank the many clinicians in England who supplied the clinical
information about their patients; Matthew Mak, Emma Stanton,
Paul Rutter, and Tara Fajeyisan for their work in the validation of all
deaths; Oliver Mytton for assistance with data analysis; and Iain Yardley
for assistance in editing the manuscript. This work was done as part of
the public health response to pandemic inuenza in England. The costs
of the study were small and met from the Chief Medical Ocers budget
within the Department of Health without the need for additional funding.
References
1
Pitman RJ, Melegaro A, Gelb D, Siddiqui MR, Gay NJ,
Edmunds WJ. Assessing the burden of inuenza and other
respiratory infections in England and Wales. J Infect 2007;
54: 53038.
2
Miller E, Hoschler K, Hardelid P, Stanford E, Andrews N,
Zambon M. Incidence of 2009 pandemic inuenza A H1N1
infection in England: a cross-sectional serological study.
Lancet 2010; 375: 110008.
3
Libster R, Bugna J, Coviello S, et al. Pediatric hospitalisations
associated with 2009 pandemic inuenza A (H1N1) in Argentina.
NEJM 2010; 362: 4555.
4
CDC. Surveillance for pediatric deaths associated with 2009
pandemic inuenza A (H1N1) virus infectionUnited States,
AprilAugust 2009. MMWR Morb Mortal Wkly Rep 2009; 58: 94147.
5
Donaldson LJ, Rutter PD, Ellis BM, et al. Mortality from pandemic
A/H1N1 2009 inuenza in England: public health surveillance
study. BMJ 2009; 339: b5213.
6
Saklad M. Grading of patients for surgical procedures.
Anaesthesiology 1941; 2: 28184.
7
Oce for National Statistics. Mid-2008 population estimates:
England and Wales. http://www.statistics.gov.uk/statbase/product.
asp?vlnk=15106 (accessed Feb 25, 2010).
8
Oce for National Statistics. Mid-2007 population estimates
ethnicity tables: England and Wales. http://www.statistics.gov.uk/
statbase/product.asp?vlnk=14238 (accessed Feb 25, 2010).
9
van t Klooster TM, Wielders CC, Donker T, et al. Surveillance of
hospitalisations for 2009 pandemic inuenza A(H1N1) in the
Netherlands, 5 June31 December 2009. Eurosurv 2010;
15: pii=19461.
10 Pitman RJ, Melegaro A, Gelb D, Siddiqui MR, Gay NJ, Edmunds WJ.
Assessing the burden of inuenza and other respiratory infections in
England and Wales. J Infect 2007; 54: 53038.
1852
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Articles
Summary
Background More than 23 million children died in India in 2005; however, the major causes of death have not been
measured in the country. We investigated the causes of neonatal and child mortality in India and their dierences by
sex and region.
Methods The Registrar General of India surveyed all deaths occurring in 200103 in 11 million nationally
representative homes. Field sta interviewed household members and completed standard questions about events
that preceded the death. Two of 130 physicians then independently assigned a cause to each death. Cause-specic
mortality rates for 2005 were calculated nationally and for the six regions by combining the recorded proportions for
each cause in the neonatal deaths and deaths at ages 159 months in the study with population and death totals from
the United Nations.
Findings There were 10 892 deaths in neonates and 12 260 in children aged 159 months in the study. When these
details were projected nationally, three causes accounted for 78% (079 million of 101 million) of all neonatal deaths:
prematurity and low birthweight (033 million, 99% CI 031 million to 035 million), neonatal infections
(027 million, 025 million to 029 million), and birth asphyxia and birth trauma (019 million, 018 million to
021 million). Two causes accounted for 50% (067 million of 134 million) of all deaths at 159 months: pneumonia
(037 million, 035 million to 039 million) and diarrhoeal diseases (030 million, 028 million to 032 million). In
children aged 159 months, girls in central India had a ve-times higher mortality rate (per 1000 livebirths) from
pneumonia (209, 194226) than did boys in south India (41, 3056) and four-times higher mortality rate from
diarrhoeal disease (177, 162193) than did boys in west India (41, 3055).
Interpretation Five avoidable causes accounted for nearly 15 million child deaths in India in 2005, with substantial
dierences between regions and sexes. Expanded neonatal and intrapartum care, case management of diarrhoea and
pneumonia, and addition of new vaccines to immunisation programmes could substantially reduce child deaths in India.
Funding US National Institutes of Health, International Development Research Centre, Canadian Institutes of Health
Research, Li Ka Shing Knowledge Institute, and US Fund for UNICEF.
Introduction
Yearly child mortality rates in India have fallen between
17%1 and 23%2 in the past two decades. Despite this
decrease, the United Nations (UN) estimates that about
235 million children died in India in 2005. This gure
corresponds to more than 20% of all deaths in children
younger than 5 years worldwide, which is more than in
any other country.1,3 Large dierences in overall child
survival between Indias diverse regions have been
previously documented.4,5 However, no direct and nationally
representative measurement of the major causes of death
in neonates (<1 month) and at ages 159 months has been
done,6 and how these causes of death vary across Indias
regions is unknown. Social preference for boys is strong,
as noted by widespread selective abortion of female fetuses7
and by lower immunisation rates in girls.8 The
consequences of boy preference on child mortality remain
undocumented. Understanding of the causes of child
death might, therefore, help to guide the use of widely
practicable interventions for neonatal and child survival.3,9
Most deaths in India, including of children, are not
medically certied since most occur at home, in rural
www.thelancet.com Vol 376 November 27, 2010
Methods
Study setting and procedures
Details of the design, methods, and preliminary results of
the Million Death Study have been previously published.1115
1853
Articles
North
194 million births
013 million deaths
Jammu and
Kashmir
Central
1028 million births
111 million deaths
East
619 million births
054 million deaths
Himachal
Pradesh
Punjab
Haryana
Uttarakhand*
Delhi
Sikkim
Rajasthan*
Uttar Pradesh
Assam*
Nagaland
Bihar*
Meghalaya
Jharkhand*
Tripura
Manipur
Mizoram
Ch
ha
tti
sg
arh
Madhya Pradesh*
Gujarat
West
Bengal
Orissa*
Northeast
099 million births
009 million deaths
Maharashtra
West
344 million births
021 million deaths
Andhra Pradesh
Kamataka
South
446 million births
027 million deaths
Tamil Nadu
la
Kera
Figure 1: Yearly number of livebirths and deaths in children aged 04 years in India, by region, 2005
*These lower-income states are known as the Empowered Action Group plus Assam (EAGA) states.
Articles
Results
Of the 24 841 child deaths surveyed, 93% (23 152) were
double-coded by physicians and included in the study
(table). Reasons for exclusion were missing information
about age or sex (n=191), and non-legible forms,
Girls
Total
Rural area
Died in a
Two coders
health facility immediately
agreed
Boys
Girls
Total
Boys
Girls
Total (99%CI)*
2012
1619
3631
3265
988
2381
130
108
120
185
142
327 (309345)
Neonatal infections
1544
1339
2883
2694
346
1804
103
94
99
145
123
268 (253286)
1219
854
2073
1869
631
946
80
59
70
113
77
190 (176206)
316
243
559
502
118
251
20
16
18
28
21
49 (4258)
Congenital anomalies
213
146
359
304
139
202
14
10
12
20
13
33 (2842)
Diarrhoeal diseases
175
162
337
318
26
227
12
12
12
17
15
32 (2640)
Tetanus
149
115
264
255
14
180
13
10
12
18
14
32 (2639)
Injuries
27
20
47
43
15
02
01
02
5 (38)
414
325
739
665
147
329
27
25
24
39
33
72 (6181)
6069
4823
10 892
6335
(582%)
401
335
369
568
440
Pneumonia
1542
1890
3432
3146
404
2546
112
160
135
159
210
369 (348390)
Diarrhoeal diseases
1184
1532
2716
2480
293
2146
89
134
111
126
176
302 (283323)
Measles
308
450
758
687
64
374
25
42
33
36
56
92 (79104)
418
433
851
772
142
490
30
35
32
42
46
88 (77100)
Other causes
All causes
9915
(910%)
2417
(222%)
1008
159 months
Injuries
400
357
757
689
91
673
29
29
29
42
38
80 (6892)
Malaria
262
325
587
562
43
354
17
24
20
24
32
56 (4765)
Meningitis/encephalitis
232
209
441
396
94
183
19
19
19
27
25
52 (4362)
Nutritional diseases
141
201
342
303
18
190
11
19
15
16
25
41 (3451)
38 (3147)
147
213
360
324
50
117
11
18
14
15
23
143
182
325
298
43
120
10
15
12
14
19
33 (2741)
Other causes
847
844
1691
1490
218
882
64
72
69
91
95
186 (170203)
5624
6636
12 260
11 147
(909%)
1460
(119%)
8075
(659%)
417
567
489
592
745
1337
11 693
11 459
23 152
21 062
(910%)
3877
(167%)
14 410
(622%)
818
902
858
1160
1185
2345
All causes
04 years
All causes
Total livebirths (2005): 273 million; 142 million boys, 131 million girls. Mortality estimates exclude stillbirths, cancelled reports (ie, not coded), and children with missing information about sex or age. The
percentage of deaths that could not be coded was 65% in boys, 56% in girls, 80% in urban areas, and 58% in rural areas. *99% CIs are provided for the causes of death but not for the UN totals for all-cause
child deaths.For prematurity: mortality rate for boys 95 (99% CI 89101), estimated total deaths 135 000; for girls 74 (6980), 97 000 deaths; and for both 85 (8189), 232 000 deaths. Low birthweight:
mortality rate for boys 35 (3139), estimated total deaths 50 000; for girls 34 (3039), 45 000 deaths; and for both 35 (3238), 95 000 deaths. These two conditions are combined because of the diculty in
dierentiating them in verbal autopsies. Infections category includes neonatal pneumonia, sepsis, and CNS infections (about 2000 deaths every year). For neonatal pneumonia: mortality rate for boys
60 (99% CI 5565), estimated total deaths 85 000; for girls 58 (5363), 76 000 deaths; and for both 59 (5563), 161 000 deaths. For sepsis: mortality rate for boys 42 (3847), 60 000 deaths; for girls
36 (3240), 47 000 deaths; and for both 39 (3642), 107 000 deaths. These three conditions are combined because of the diculty in dierentiating them in verbal autopsies. Sample-weighted percentage
of deaths: 871% occurred in a rural area and 166% occurred in a health facility.
Table: Causes of death in neonates and at ages 159 months in this study and estimated national totals
1855
Articles
India
(235 million deaths; MR=858)
Pneumonia (16%)
Prematurity and
low birthweight (14%)
Birth asphyxia and
birth trauma (8%)
Neonatal deaths
43%
Other infectious
diseases (11%)
Neonatal
infections* (12%)
Tetanus (1%)
Other (17%)
Other (7%)
(13%)
Boys
(116 million deaths; MR=818)
Diarrhoeal
diseases
(1%)
Girls
(119 million deaths; MR=902)
Pneumonia (14%)
Pneumonia (18%)
Prematurity and
low birthweight (12%)
Prematurity and
low birthweight (16%)
Other infectious
diseases (9%)
Neonatal deaths
49%
Neonatal deaths
37%
Other infectious
diseases (13%)
Neonatal
infections* (10%)
Tetanus (1%)
Other (17%)
Neonatal
infections* (13%)
(11%)
Diarrhoeal
diseases
Tetanus (2%)
Other (7%)
(1%)
Higher-income states
(071 million deaths; MR=610)
Other (6%)
Other (17%)
(15%)
Lower-income states
(164 million deaths; MR=1041)
Pneumonia (12%)
Pneumonia (17%)
Prematurity and
low birthweight (11%)
Prematurity and
low birthweight (20%)
Other infectious
diseases (10%)
(1%)
Diarrhoeal
diseases
Neonatal deaths
48%
Birth asphyxia and
birth trauma (10%)
Other (21%)
Other infectious
diseases (13%)
Neonatal
infections* (13%)
Tetanus (2%)
Neonatal
infections* (8%)
Other (6%)
Other (15%)
Other (9%)
(9%)
Diarrhoeal
diseases
(1%)
(14%)
Diarrhoeal
diseases
(2%)
Figure 2: Causes of death in children aged 04 years in India, by sex and by state income, 2005
MR=mortality rate in children younger than 5 years. *Includes neonatal pneumonia, sepsis, and CNS infections. Lower-income states are Assam, Bihar, Chhattisgarh,
Jharkhand, Madhya Pradesh, Orissa, Rajasthan, Uttarakhand, and Uttar Pradesh; higher-income states are the remaining 26 states/union territories.
1856
Articles
West
(021 million deaths; MR=601)
Pneumonia (13%)
Pneumonia (8%)
Prematurity and
low birthweight (23%)
Other infectious
diseases (10%)
Prematurity and
low birthweight (20%)
Other infectious
diseases (12%)
Neonatal deaths
50%
Neonatal deaths
46%
Birth asphyxia and
birth trauma (10%)
Neonatal
infections* (6%)
Other (24%)
Other (19%)
Neonatal
infections* (9%)
(8%)
Diarrhoeal
diseases
(10%)
Diarrhoeal
diseases
(1%)
East
(054 million deaths; MR=873)
North
(013 million deaths; MR=656)
Pneumonia (15%)
Prematurity and
low birthweight (13%)
Other infectious
diseases (7%)
Other (9%)
Other (7%)
(1%)
Prematurity and
low birthweight (12%)
Pneumonia (18%)
Other infectious
diseases (12%)
Neonatal
infections* (12%)
Neonatal
infections* (12%)
Other (19%)
Other (17%)
Other (9%)
(12%)
(1%)
Diarrhoeal
diseases
(14%)
Northeast
(009 million deaths; MR=916)
Pneumonia (16%)
Tetanus (1%)
Other (6%)
(1%)
Diarrhoeal
diseases
Central
(111 million deaths; MR=1079)
Prematurity and
low birthweight (10%)
Pneumonia (17%)
Prematurity and
low birthweight (12%)
Neonatal
infections* (9%)
Other infectious
diseases (12%)
Neonatal
infections* (12%)
Other (6%)
Tetanus (2%)
Other (15%)
Other (7%)
Other (16%)
(15%)
Diarrhoeal
diseases
(1%)
(13%)
Diarrhoeal
diseases
(2%)
1857
Articles
Study deaths
200103
Total deaths
(thousands)
in 2005 (99% CI)
(7196)
(78108)
(96111)
(115136)
(118135)
(133158)
16
10
65
56
130
50
3631
120 (115124)
(1419)
(811)
(5969)
(5161)
(121139)
(4654)
327 (309345)
Neonatal infections*
South
160
West
226
North
212
Northeast
184
East
798
Central
1303
Subtotal (B)
38
51
76
79
100
145
2883
(3146)
(4361)
(6589)
(6694)
(93108)
(136154)
17
17
15
8
62
149
99 (94103)
(1421)
(1521)
(1317)
(79)
(5867)
(140158)
268 (253286)
58
59
64
72
75
81
Subtotal (C)
2073
70 (6674)
190 (176206)
Total (A+B+C)
8587
289 (283292)
785 (773797)
(5068)
(5069)
(5871)
(5987)
(6488)
(7489)
26
20
40
7
14
83
10
(2230)
(1724)
(3644)
(69)
(1217)
(7691)
15
Figure 4: Mortality rates for the three leading causes of neonatal death in India, by region, 2005
*Includes neonatal pneumonia, sepsis, and CNS infections.
Discussion
More than three-fths of all 23 million child deaths in
India in 2005 were from ve causes: pneumonia,
prematurity and low birthweight, diarrhoeal diseases,
neonatal infections, and birth asphyxia and birth trauma.
Each of the major causes of neonatal deaths can be
prevented or treated with known, highly eective
and widely practicable interventions such as improvements
in prenatal care, intrapartum care (skilled attendance,
emergency obstetric care, and simple immediate
care for newborn babies), postnatal family-community
care (preventive postnatal care, oral antibiotics, and
management of pneumonia),24 and tetanus toxoid immunisation.25 Concern has been raised that neonatal death
rates in India are not falling fast enough.9 However, our
results suggest that almost half of Indias neonatal deaths
are caused by birth asphyxia and birth trauma, sepsis,
pneumonia, and tetanusmost of which can be avoided
by increases in delivery and postnatal care.26
The substantial regional dierences in cause-specic
mortality, even in girls (webappendix p 12), could indicate
the existence of some underlying social, behavioural, or
biological risk factors for child deaths.10 However, at ages
159 months, girls in every region die more commonly
than do boys, and inequities in access to care, rather than
biological or genetic factors, are a more plausible
explanation for these recorded dierences between
sexes.7,8 Household surveys4,27 show little dierence
between sexes in the rate of respiratory symptoms and
diarrhoeal disease, whereas our study and previous
analyses28 have shown substantial sex dierences in
mortality. Integrated management of child illnesses
increases care seeking for illnesses,29 and reduces child
deaths,30 but in India, boys use such programmes more
than girls.9 Fewer girls than boys are vaccinated in health
facilities.8 However, outreach programmes that visit
households immunise a greater proportion of girls than
do facility-based vaccination programmes.9 Addition of
vaccines against pneumonia (pneumococcal conjugate,
Haemophilus inuenzae type B) and diarrhoeal diseases
(rotavirus) to outreach home-based immunisation
programmes would reduce child deaths and narrow the
gap in child mortality in India between sexes.9,31
www.thelancet.com Vol 376 November 27, 2010
Articles
Study deaths
200103
Total deaths
(thousands)
in 2005 (99% CI)
Pneumonia
South
West
North
Northeast
East
Central
129
262
304
281
860
1596
47
76
96
142
153
180
Subtotal (A)
3432
135 (130141)
369 (348390)
Boys/south*
Girls/central*
66
896
41 (3056)
209 (194226)
9 (713)
103 (95111)
49
59
81
126
133
145
18
26
16
79
13
150
22
27
20
15
97
188
(3859)
(6488)
(84109)
(120167)
(141164)
(170191)
(1726)
(2230)
(1621)
(1217)
(88102)
(175196)
Diarrhoeal diseases
176
West
153
South
250
North
720
East
227
Northeast
1190
Central
(3959)
(4972)
(6993)
(116137)
(111158)
(135155)
(1420)
(2232)
(1418)
(7285)
(1116)
(139159)
Subtotal (B)
2716
111 (105116)
302 (283323)
Boys/west
Girls/central
84
697
41 (3055)
177 (162193)
7 (510)
87 (8095)
6148
246 (239252)
Total (A+B)
671 (653688)
5
10
15
20
25
Figure 5: Mortality rates for the two leading causes of death in children aged 159 months in India, by
region, 2005
*Boys from the south region have the lowest mortality rate for pneumonia by sex and region (at age
159 months), and girls from the central region have the highest. Boys from the west region have the lowest
mortality rate for diarrhoeal diseases by sex and region (at age 159 months), and girls from the central region
have the highest.
1859
Articles
Acknowledgments
The Registrar General of India established the SRS in 1971, has
continued it ever since, and is collaborating with several of the authors
on the ongoing Million Death Study. External funding is from the
Fogarty International Centre of the US National Institutes of Health
(grant R01 TW0599101), Canadian Institute of Health Research
(CIHR; IEG-53506), International Development Research Centre
(Grant 102172), Li Ka Shing Knowledge Institute and Keenan Research
Centre at St Michaels Hospital, University of Toronto, and the US Fund
for UNICEF (via a grant from the Bill & Melinda Gates Foundation for
CHERG; subgrant 50140). PJ is supported by the Canada Research Chair
programme. SKM is a Fellow of the Pediatric Scientist Development
Program. The opinions expressed in this paper are those of the authors
and do not necessarily represent those of the Government of India. We
thank Joy Lawn, Colin Mathers, Mikkel Oestergaard, Prem Mony, and
Alvin Zipursky for comments; and Maya Kesler, Brendon Pezzack,
Chinthanie Ramasundarahettige, Peter Rodriguez, and
Wilson Suraweera for data support.
References
1
UN Population Division. World population prospects (2008
revision). April, 2009. http://esa.un.org/peps/peps_interpolateddata.htm (accessed June 14, 2010).
2
Rajaratnam JK, Marcus JR, Flaxman AD, et al. Neonatal,
postneonatal, childhood, and under-5 mortality for 187 countries,
19702010: a systematic analysis of progress towards the
Millennium Development Goal 4. Lancet 2010; 357: 19882008.
3
Black RE, Cousens S, Johnson HL, et al. Global, regional, and
national causes of child mortality in 2008: a systematic analysis.
Lancet 2010; 375: 196987.
4
International Institute for Population Sciences (IIPS) and Macro
International. National Family Health Survey (NFHS-3), 200506:
India. Mumbai: IIPS, 2008.
5
Registrar General of India. Sample Registration System. New Delhi:
Oce of the Registrar General of India, 2004.
6
Baqui AH, Darmstadt GL, Williams EK, et al. Rates, timing and
causes of neonatal deaths in rural India: implications for
neonatal health programmes. Bull World Health Organ 2006;
84: 70613.
7
Jha P, Kumar R, Vasa P, Dhingra N, Thiruchelvam D,
Moineddin R. Low male-to-female sex ratio of children born in
India: national survey of 11 million households. Lancet 2006;
367: 21118.
8
Corsi DJ, Bassani DG, Kumar R, et al. Gender inequity and
age-appropriate immunization coverage in India from 1992 to 2006.
BMC Int Health Hum Rights 2009; 9 (suppl 1): S3.
9
Jha P, Laxminarayan R. Choosing health: an entitlement for all
Indians. May, 2009. http://cghrindia.org/images/choosing-health.
pdf (accessed Sept 10, 2010).
10 Jha P. Avoidable mortality in India: past progress and future
prospects. Natl Med J India 2002; 15 (suppl 1): 3236.
11 Jha P, Gajalakshmi V, Gupta PC, et al. Prospective study of one
million deaths in India: rationale, design, and validation results.
PLoS Med 2006; 3: e18.
12 Registrar General of India, Centre for Global Health Research.
Causes of death in India, 20012003: Sample Registration System.
New Delhi: Government of India, 2009.
13 Dhingra N, Jha P, Sharma VP, et al, for the Million Death Study
Collaborators. Adult and child malaria mortality in India: a
nationally representative mortality survey. Lancet 2010; published
online Oct 21. DOI:10.1016/S0140-6736(10)60831-8.
14 Jha P, Jacob B, Gajalakshmi V, et al. A nationally representative
case-control study of smoking and death in India.
N Engl J Med 2008; 358: 113747.
15 Jha P, Kumar R, Khera A, et al. HIV mortality and infection in
India: estimates from nationally representative mortality survey of
11 million homes. BMJ 2010; 340: c621.
1860
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Series
The burden of chronic, non-communicable diseases in low-income and middle-income countries is increasing. We
outline a framework for monitoring of such diseases and review the mortality burden and the capacity of countries
to respond to them. We show data from WHO data sources and published work for prevalence of tobacco use,
overweight, and cause-specic mortality in 23 low-income and middle-income countries with a high burden of noncommunicable disease. Data for national capacity for chronic disease prevention and control were generated from a
global assessment that was done in WHO member states in 200910. Although reliable data for cause-specic
mortality are scarce, non-communicable diseases were estimated to be responsible for 234 million (or 64% of the
total) deaths in the 23 countries that we analysed, with 47% occurring in people who were younger than 70 years.
Tobacco use and overweight are common in most of the countries and populations we examined, but coverage of
cost-eective interventions to reduce these risk factors is low. Capacity for prevention and control of noncommunicable diseases, including monitoring and surveillance operations nationally, is inadequate. A surveillance
framework, including a minimum set of indicators covering exposures and outcomes, is essential for policy
development and assessment and for monitoring of trends in disease. Technical, human, and scal resource
constraints are major impediments to the establishment of eective prevention and control programmes. Despite
increasing awareness and commitment to address chronic disease, concrete actions by global partners to plan and
implement cost-eective interventions are inadequate.
Introduction
Non-communicable Diseases
and Mental Health
(A Alwan MD,
Prof D R MacLean MD);
Department of Chronic
Diseases and Health
Promotion (L M Riley MSc);
Tobacco Free Initiative
(E Tursan dEspaignet PhD,
D Bettcher PhD); and Mortality
and Burden of Disease,
Department of Health
Statistics and Informatics
(C D Mathers PhD,
G A Stevens PhD), World Health
Organization, Geneva,
Switzerland
Key messages
Low-income and middle-income countries are
undergoing a rapid rise in the burden of
non-communicable diseases with major adverse
social, economic, and health outcomes
We provide a surveillance framework to quantify and
monitor non-communicable diseases and their
determinants
Prevalence of tobacco use in men and overweight in men
and women are high in many of the 23 low-income and
middle-income countries with high rates of
non-communicable disease
Age-specic death rates for non-communicable diseases
are higher in the 23 low-income and middle-income
countries than it is in high-income countries
Health-system capacity in the 23 countries with a high
burden of non-communicable diseases, including
surveillance, is inadequate to reduce the disease burden
and urgently requires strengthening
Disease surveillance should be integrated into national
health information systems and regular monitoring of
actions to prevent non-communicable diseases is needed
Correspondence to:
Dr Ala Alwan,
Non-communicable Diseases
and Mental Health, World Health
Organization, 20 Avenue Appia,
1211 Geneva, Switzerland
alwana@who.int
1861
Series
100
Men
Women
Prevalence (%)
80
60
40
20
Ar
ge
Ba nti
ng na
lad
es
h
Br
az
i
Bu l
rm
a
C
DR hina
Co
ng
o
Eg
y
Et pt
hi
op
ia
In
In dia
do
ne
sia
Ira
M n
ex
ic
Ni o
ge
Pa ria
k
Ph ista
ilip n
pi
ne
Po s
lan
d
So Rus
ut sia
hA
fr
T h ic a
ail
an
Tu d
rk
Uk ey
ra
Vi ine
et
na
m
Figure 1: Estimated prevalence of daily tobacco use in adults aged 15 years or older, 2005
Data from WHOs report on the global tobacco epidemic.6 No data for tobacco use were available for Colombia.
Methods
As part of the implementation of its non-communicable
diseases action plan,10 WHO established an expert
epidemiology reference group in 2009 to help to develop a
national disease-surveillance framework. Components and
the core indicators were adopted on the basis of existing
WHO techniques and data sources that are used for
periodic reporting (every few years) of health issues, such
as WHO updated mortality estimates, global tobacco
surveillance system,12 the STEPwise approach to chronic
disease risk factor surveillance,13 Global NCD InfoBase,14
and other relevant databases.
WHO periodically estimates deaths globally, regionally,
and nationally by cause.15 We draw on provisional
estimates for the year 2008, which are presently under
country consultation. These estimates use standard
methods16 that draw on updated estimates of all-cause
mortality,17 the latest available death-registration data,18
WHO and UNAIDS programme estimates for some
specic causes (tuberculosis, HIV-1/AIDS, and malaria),
updated estimates of cause-specic deaths in children
younger than 5 years,19 and estimates of cancer deaths by
site for 2008 from the IARC Globocan database.20
For 12 of the 23 countries we assessed, we calculated
mortality rates from recent death-registration data, which
were projected to the year 2008 if 2008 data were not
1862
Series
100
Men
Women
Prevalence (%)
80
60
40
20
0
Ar
g
Ba enti
ng na
lad
es
h
Br
az
i
Bu l
rm
a
C
Co hina
lo
m
DR bi
Co a
ng
o
Eg
Et ypt
hi
op
ia
I
In ndia
do
ne
sia
Ira
M n
ex
i
Ni co
ge
Pa ria
Ph kista
ilip n
pi
ne
Po s
lan
d
So Ru
ut ssia
hA
f
Th rica
ail
an
Tu d
rk
Uk ey
ra
Vi ine
et
na
m
Figure 2: Estimated prevalence of overweight (body-mass index of 25 or more) in adults aged 15 years or
older, 2005
Ukraine
Russia
South Africa
Egypt
Nigeria
Poland
Thailand
Burma
Indonesia
DR Congo
India
Bangladesh
Argentina
Brazil
Pakistan
Philippines
China
Ethiopia
Turkey
Mexico
Vietnam
Cardiovascular diseases and diabetes
Cancers
Chronic repiratory diseases
Other chronic diseases
Iran
Colombia
All 23 countries
0
100
200
300
400
500
600
700
800
Figure 3: Death rates from non-communicable diseases per 100 000 adults aged 1569 years in
23 high-burden countries
Risk factors
Series
Mortality
Deaths from chronic disease in the 23 high-burden
countries accounted for 234 million (41%) of global
deaths from all causes and 80% of deaths from noncommunicable diseases in all low-income and middleincome countries. Deaths from non-communicable
diseases in people aged younger than 70 years in the
23 countries accounted for 47% of all non-communicable
disease deaths in these countries and for 71% of such
deaths in people younger than 70 years globally.
Integrated NCD policy
Developed
Operational
Alcohol
Yes
No
No
Yes
Yes
Yes
Bangladesh Yes
Yes
Yes
Yes
Yes
Yes
Brazil
Yes
Yes
Yes
Yes
Yes
Yes
Burma
Yes
No
Yes
Yes
Yes
Yes
China
No
NA
No
No
No
No
DR Congo
Yes
No
Yes
Yes
Yes
Yes
Egypt
Yes
Yes
NR
Yes*
NR
Yes*
Ethiopia
Yes
No
Yes
Yes
Yes
Yes
India
Yes
Yes
Yes
Yes
Yes
Yes
Indonesia
Yes
Yes
Yes
Yes
Yes
Yes
Iran
No
NA
No*
Yes*
Yes*
Yes*
Mexico
Yes
Yes
Yes
Yes
Yes
Yes
Nigeria
Yes
Yes
Yes
Yes
Yes
Yes
Pakistan
No
NA
No*
No*
No*
No*
Philippines
Yes
Yes
Yes
Yes
Yes
Yes
Poland
Yes
Yes
Yes
Yes
Yes
Yes
Russia
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Thailand
Yes
Yes
Yes
Yes
Yes
Yes
Turkey
No
NA
Yes*
Yes*
Yes*
Yes*
Ukraine
No
NA
No*
No*
No*
Yes*
Vietnam
Yes
Yes
No
No
No
No
Overall
17/22 (77%)
19/22 (86%)
Argentina
Data were reported to WHO as part of the NCD country capacity assessment in 2010. Colombia did not respond to the
capacity assessment, so no data were available. NCD=non-communicable disease. NA=not applicable. NR=not
reported. *Country did not report risk factor as part of their integrated NCD policy, but did report a standalone policy.
Table 1: Presence of an integrated non-communicable disease policy, operational status, and inclusion of
specic risk factors for 22 countries with high burdens of non-communicable diseases
1864
Age-specic death rates were higher in many lowincome and middle-income countries than they were in
high-income countries. For the 14 countries with death
registration data for cause of death, the overall agestandardised death rates from non-communicable
diseases were 711 per 100 000 for men (58% higher than
that for men in high-income countries in 2008) and
508 per 100 000 for women (69% higher).
Of the 23 countries shown in gure 3, death rates from
chronic diseases in people aged 1569 years were highest
in eastern European countries such as Ukraine and
Russia, and in some countries of the African continent
such as South Africa, Egypt, and Nigeria. Although agespecic death rates for most non-communicable diseases
are projected to decline with increasing rates of economic
development, the ageing of the populations of these
countries will lead to a substantially increased overall
number of deaths. Overall mortality from noncommunicable diseases for people younger than 70 years
is projected to rise from 108 million in 2010 to
154 million in 2050 for the 23 countries we examined.
Whereas deaths from infectious disease in people
younger than 70 years are projected to decline by around
2% per year during the next 40 years, the number of
deaths from cardiovascular disease are projected to
increase by 07% per year, and from cancer by 11% per
year in high-burden countries.
Country capacity
Of the 23 high-burden countries we selected to investigate,
only one country (Colombia) did not participate in the
2010 assessment. In the 2010 survey, all countries, apart
from Poland, reported the existence of a unit, branch, or
department in their Ministry of Health with responsibility
for non-communicable disease. However, of these
countries, only 17 reported having an integrated policy,
strategy, or action plan in place, which were reported as
being operational in only 13 countries (table 1). In this
context, operational means that the country believes that
the policy is being actively implemented. With respect to
policies, strategies, or plans of action to address individual
diseases or risk factors, 18 countries reported plans to
address cancer (country data not shown), 19 countries
reported having a tobacco control policy, and only
18 countries reported policies or programmes for diet or
physical activity (table 1).
For national surveillance systems for non-communicable
diseases, most countries had funding for surveillance,
monitoring, and assessment (table 2). 20 countries reported
that mortality data were contained in their national health
reporting system, but only ten of these countries stated
that their mortality data were population-based (table 2)
and only seven reported regular provision of reliable data
for cause-specic mortality to WHO. With respect to risk
factors for non-communicable diseases, 17 countries
reported that their national health-reporting system
detailed risk factors, and 11 of these systems included
www.thelancet.com Vol 376 November 27, 2010
Series
Mortality
Cancer registry
Risk factors
Yes*
Yes
Yes*
Bangladesh
No
Yes
Yes
Brazil
Yes*
Yes*
Yes*
Argentina
Burma
Yes
Yes
Yes
China
Yes*
Yes
Yes*
DR Congo
Yes
No
No
Egypt
Yes*
Yes
Yes*
Ethiopia
Yes
No
No
India
Yes
Yes
Yes
Indonesia
Yes
Yes
Yes
Iran
Yes*
Yes*
Yes*
Mexico
Yes*
Yes
Yes*
Nigeria
Yes
Yes
Yes*
Pakistan
No
No
No
Philippines
Yes
Yes*
Yes*
Poland
Yes*
Yes*
Yes*
Russia
Yes*
Yes
No
South Africa
Yes*
Yes
Yes*
Thailand
Yes
Yes
Yes
Turkey
Yes
Yes*
Yes*
Ukraine
Yes*
Yes
Yes
Vietnam
Yes
Yes
No
Overall
20/22 (91%)
19/22 (86%)
17/22 (77%)
Data were reported to WHO as part of the NCD country capacity assessment in 2010.
Colombia did not respond to the capacity assessment, so no data were available.
NCD=non-communicable disease. *Population-based data.
Dedicated NCD
oce in Ministry
of Health
Cancer policy
2000
2010
2000
2010
Operational
in 2010
2000
2010
Operational
in 2010
Argentina
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Bangladesh
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Brazil
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Burma
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
China
Yes
Yes
Yes
No
No
Yes
Yes
Yes
DR Congo
Yes
Yes
No
No
NA
No
No
NA
Egypt
No
Yes
No
NR
NA
No
Yes
Yes
Ethiopia
Yes
Yes
No
Yes*
No
No
Yes*
No
India
Yes
Yes
NR
Yes*
Yes
Yes
Yes*
Yes
Indonesia
No
Yes
No
Yes
Yes
No
Yes
Yes
Iran
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Mexico
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Nigeria
Yes
Yes
No
Yes*
Yes
No
Yes
No
Pakistan
No
Yes
No
No
NA
Yes
No
NA
Philippines
Yes
Yes
Yes
Yes*
Yes*
Yes
No
NA
Poland
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Russia
Yes
Yes
Yes
Yes*
Yes
No
Yes*
Yes
Thailand
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Turkey
Yes
Yes
No
Yes
No
No
Yes
No
Vietnam
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Overall
14/20
(70%)
19/20
(95%)
11/20
(55%)
16/20
(80%)
13/20
(65%)
13/20
(65%)
17/20
(85%)
14/20
(70%)
Data were reported to WHO as part of NCD country capacity assessments in 2000 and 2010. Colombia, South Africa, and
Ukraine did not respond to the capacity assessment in either 2000 or 2010, so no data were available. NA=not applicable.
NCD=non-communicable disease. NR=Country did not respond to this question. *Integrated NCD policy (an integrated
policy addresses more than one risk factor or more than one NCD, in this case cardiovascular disease or cancer).
Table 3: Countries reporting key elements of non-communicable disease prevention capacity, 2000
and 2010
Series
Number of
countries (%)
Article
Criteria
20
4 (17%)
Protection from 8
tobacco smoke
3 (13%)
Oers of help to 14
quit
2 (9%)
4 (17%)
50% (average of the front and back of the cigarette pack),
including pictures or pictograms and appropriate characteristics
such as specic health warnings mandated, appearing on
individual packages and on any outside packaging and labelling
used in retail sale, describing specic harmful eects of tobacco
use on health; warnings are large, clear, visible, legible, rotate,
and are written in all principal language(s) of the country
Enforced bans
13
5 (22%)
Raised taxes
1 (4%)
Monitoring of
tobacco use
Table 4: Number of countries with a high burden of chronic, non-communicable diseases that had
implemented demand-reduction measures in December, 2008
Discussion
Our analysis shows that, in high-burden countries, the
capacity to eectively deal with the existing and projected
burden of non-communicable diseases is inadequate.
Chronic diseases are a leading health and developmental
challenge.32 The situation is especially serious in lowincome and middle-income countries that are undergoing
a rapid rise in premature mortality and an increasing
burden on their health systems despite availability of
cost-eective interventions.33 Such diseases are
1866
Series
Conclusions
This dearth of capacity, which includes important gaps in
surveillance, is a major challenge to global health
development. The negative eects on health and
socioeconomic development are increasingly understood
and require a strong response from all countries. Recent
developments, including the UN General Assembly
resolution on non-communicable diseases, call for a high
priority to be given to these diseases, not only within the
health sector, but also within the national and global
development agendas.39 It calls for a high-level meeting
of the General Assembly on non-communicable diseases
in September, 2011, with the participation of heads of
state and government. The resolution requests the UN
Secretary General to submit a report on the global status
of these diseases. One area of emphasis during the
discussions leading to the high-level meeting in 2011 will
therefore cover the need for strengthening of global
monitoring of trends and approaches, for integration of
non-communicable disease surveillance into national
www.thelancet.com Vol 376 November 27, 2010
1867
Series
17
18
19
20
21
22
23
24
25
26
27
28
29
1868
30
31
32
33
34
35
36
37
38
39
40
Viewpoint
1869
Viewpoint
Viewpoint
4
5
10
11
12
13
14
15
16
17
18
19
20
1871
Case Report
A hearty sneeze
Harry A Roselle, Marc Herman
Lancet 2010; 376: 1872
Department of Medicine
(H A Roselle MD) and
Department of Radiology
(M Herman MD), Englewood
Hospital and Medicine Centre,
New Jersey, USA
Correspondence to:
Dr Harry A Roselle, Department
of Medicine, Englewood
Hospital and Medical Centre,
309 Engle St, Englewood,
New Jersey 07631, USA.
rosellesrhj@aol.com
1872