You are on page 1of 25

Measuring the Quality of Care

Assessment of Lao PDR Health Sector:

Luang Prabang
Kirthi Ramesh Disclaimer: The views expressed in this
Social Development Specialist, Asian Development Bank paper/presentation are the views of the author and
do not necessarily reflect the views or policies of the
Asian Development Bank (ADB), or its Board of
Azusa Sato Governors, or the governments they represent. ADB
does not guarantee the accuracy of the data included
Health Specialist, Asian Development Bank in this paper and accepts no responsibility for any
consequence of their use. Terminology used may not
necessarily be consistent with ADB official terms.

John Peabody MD, PhD, FACP

President, QURE Healthcare
Professor, University of California, San Francisco and Los Angeles

Mary Tran, MS
Director, QURE Healthcare
• Mission objectives
• Quality of Care Assessment
– Methodology
– Results

© QURE 2017

Mission Objectives (September 2018)
• Obtain a sample assessment of the
quality of clinical care in Lao PDR:
Luang Prabang
• Assess vignettes as an effective
measurement tool to assist providers in
elevating their care and improving
patient outcomes nationwide

© QURE 2017

Improving the Quality of Healthcare

Access Quality
Structural Clinical
Inputs Practice

Factors that The sum of clinical Effects of

affect the actions by providers that healthcare on
context in which translate inputs into patients or
care is delivered outputs populations

Quality Improvement Leads to Better Health

© QURE 2017
How Can You Measure and Improve Clinical
Practice: Clinical Performance and Value (CPV)
• Taking a history
• Conducting a physical
1• Measure: Physician cares for examination
patient vignettes, providing their • Ordering tests
responses to:
• Making a diagnosis
• Providing treatment

• Performance benchmarks
2• Feedback: Physician engaged to peers and to the
with customized feedback, evidence base
receiving information on: • Unneeded workup
• Personalized education
• Links to references
© QURE 2018 6.
5 sources of data used
1. Lao’s District Health Information System 2
2. Qualitative interviews with hospital and
district administrators
3. Provider Surveys
4. Facility Surveys
5. CPV Vignettes for hypertension, diabetes
mellitus and children’s diarrhea

© QURE 2017

We Designed 3 CPV Patient Vignettes
Case Type Patient Presentation and Key Assessment Points
Mr. Bouphavanh, 52 years old, presents with unexpected weight loss
1. Recognize and institute cost-effective work-up of newly diagnosed diabetes and its
2. Emphasize medical nutrition therapy as first-line treatment for newly diagnosed diabetes
Diabetes 3. Prescribe appropriate pharmacologic treatment for diabetes (metformin as first line)
4. Estimate and manage overall cardiovascular risk
5. Emphasize primary preventative care for patients with diabetes including screening,
vaccination and counseling
Mr. Kommandam, 66 years old, presents with shortness of breath
1. Recognize and initiate management of acute decompensated heart failure (HF)
Chronic including hospitalization;
Obstructive 2. Institute appropriate pharmacologic treatment for uncontrolled hypertension
Pulmonary 3. Recognize the patient has co-existing subclinical hypothyroidism
Disease 4. Estimate and manage overall cardiovascular risk
5. Emphasize primary preventative care for patients with COPD including screening,
vaccination and counseling
Keophothong, 5 years old, is brought in by his mother looking weak
1. Recognize severe dehydration in a child with diarrhea
2. Utilize cost-effective diagnostic tests to evaluate severity and etiology of diarrhea
Pediatric 3. Administer appropriate non-pharmacologic and pharmacologic treatment
Diarrhea for diarrhea with dehydration
4. Prescribe antibiotic therapy for a patient with acute gastroenteritis when necessary
5. Identify preventive strategies for diarrhea including those on sanitation, hygiene and
safe food practices

© QURE 2018
57 Providers Across 6 Facilities in Luang
Prabang Completed the CPV Vignettes
Baseline Provider Characteristics
Number of Providers 57

Female 35 (61%)
Age 38.1+11.4

Provider Type
Advanced Nurse 1 (2%)
General Practice Physician 44 (77%)
Specialty Physician 4 (7%)
Did not State 7 (12%)

Facility (6)
Lao-Friend Hospital (Pediatrics) 10 (18%)
Luang Prabang Provincial Hospital 20 (35%)
Nan District Hospital 7 (12%)
Nambak District Hospital 8 (14%)
Ngoy District Hospital 5 (9%)
Xieng Ngeun District Hospital 7 (12%)
© QURE 2017
Overall Quality of Care Scores Varied, Showing a
Bell Shaped Distribution
CPV Score Distribution (n=57) Average Score:

25 Pediatric Aggregate performance

Number of Observations

Adult of previous QURE low-

20 middle income country
15 (LMIC) CPV studies

10 (45%-65%)

0 10 20 30 40 50 60 70 80 90 100
Overall CPV Score

© QURE 2017
Highest Scores in History and Physical; Most
Variation in Diagnosis and Treatment
Lao Mission 2: CPV Score by Domain (132 CPVs)


60% 57% 48%
44% 43%

Total History Physical Workup Diagnosis Treatment

28% of cases had at least one perfect domain score

© QURE 2017
Opportunity to Reduce Unneeded Testing,
Lower Costs and Improve Diagnostic Workup
Average No. of Unneeded Top 3 Unneeded Tests Ordered
Tests Ordered
Diabetes Diabetes
Sputum test (9)
0.8 2D echo/TTE (7)
HIV test (6)
Diarrhea Diarrhea
TB test (6)
0.3 HbA1c (3)
Blood typing (2)
Hypertension Hypertension
Abdominal ultrasound (6)
0.5 Sputum test (4)
Albumin (4)

Note: © QURE 2017

• Estimated costs come from 2016 Medicare national average payments
• Data includes unneeded labs, imaging and diagnostic procedures 13
Average Scores for Pediatric Diarrhea Higher
than for NCD Cases

Diarrhea Diabetes Hypertension

Overall Overall Overall
Facility N N N
(%) (%) (%)
Lao Friend Hospital 10 49.7 -- -- -- --
Luang Prabang Provincial Hospital -- -- 21 39.9 20 46.0
Nambak District Hospital 8 43.6 8 40.3 8 39.0
Nan District Hospital 7 52.0 7 34.1 7 38.9
Ngoy District Hospital 5 62.6 5 42.4 5 47.4
Xieng Ngeun District Hospital 7 54.2 7 39.1 7 39.5

Total Cases 37 48 47
Average Score 51.4% 39.2% 42.9%

© QURE 2017
Provider Survey: What is high quality healthcare?
No. Characteristics % (n=57)
1 Evidence based/Effective treatment 19%
2 Good health/no illness 19%
3 Preventive and primary care 18%
4 Better outcomes 16%
5 Patient centered 11%
6 Accessible/comprehensive 11%
7 Good reception 9%
8 Good medical history 9%
9 Right diagnosis 9%
10 Good hygiene and sanitation 5%
11 Modern equipment/technology 4%
12 5G1S 4%
13 Quick recovery 4%
14 Specialty referrals 2%
15 Sufficient staff 2%
© QURE 2017
16 Ethical 2% 15
Provider Survey: What are the top barriers to
providing high quality healthcare?
No. Characteristics % (n=57)
1 Lack of provider education 46%
2 Having necessary equipment 42%
3 Low patient cooperation 19%
4 Low patient education 14%
5 Access to medicines 12%
6 Lack of guidelines 11%
7 Language barriers 11%
8 Patient ethnic beliefs 7%
9 Staffing shortage 7%
10 Too many patients to see 5%
11 Access to remote areas 4%
12 Prevention 4%
13 Lab skills 4%
© QURE 2017
14 Self-preservation 2% 16
We Also Completed 15 interviews with
District or Hospital Administrators
• Administrative and clinical stakeholders believe that improving skillsets of
providers is of primary importance to improving the quality of healthcare in Lao.
This ranked higher than obtaining healthcare equipment and resources in
discussions - “What good is having equipment if our doctors don’t know how to
use it?”

• Awareness of and use of clinical guidelines (e.g. WHO PEN guidelines) is


• District and Provincial hospitals offering only some ad hoc continuing medical
education. Provincial and District hospital administrators report rotating
providers quarterly or semi-annually through courses funded by the provincial

• Providers expressed strong enthusiasm for receiving additional training and

support in our discussions directly with them and in their discussions with their
© QURE 2017
Improve Practice Change Through Serial
CPV® Measurement and Feedback

CPV® Case CPV® Vignette Individual Group Result Clinical Practice

Writing Administration Feedback Report Outs Change

© QURE 2017

Thank You
CPVs have an Extensive Peer-Reviewed
Clinical Validation Record in the Literature

Results show that

CPVs document
actual practice
nearly as well as
patients (at a much
lower cost) and
better than chart

© QURE 201821
Process is Most Proximal to Better
• Process measures are under our • Outcomes are affected by
control factors outside of health care
• Some providers deliver better clinical delivery
care than others • Outcome measures offer no
• Changing clinical behaviors is possible insight on how to improve
with good data and a gentle touch patient care

Diabetes HTN

Blood sugar
screening/testing rates Blood pressure screening rates

Examples: Treatment with hypoglycemics &

Treatment with anti-hypertensives

Lower rates of strokes and Heart Attacks

Fewer strokes and heart attacks &
and Kidney Failure Rate of patients with Kidney Failure
healthy nondiabetic blood sugar

© QURE 2018
Drill Down: Diarrhea Care
Caring for Keophothong, 5
• 100% (n=37) identified the primary diagnosis (acute gastroenteritis), BUT
only one identified the specific etiology of the disease (Vibrio cholerae)

– On avg., only 24.7% (S.D. 14.1%) of necessary tests were ordered,

which included stool analysis for ova and parasites and stool culture

– Only 5% ordered antibiotics (likely driven by absence of the specific

etiologic diagnosis)

• More positively, 86% of providers ordered oral rehydration solution for the

• Once the patient was stabilized and able to be discharged home, 62% of
providers offered appropriate preventive care services in the form of
counseling on good feeding practice, proper hygiene, food handling, etc.

© QURE 2017

Drill Down: Hypertension Care
Caring for Mr. Kommandam, 66
• Only 21% (n=47), ordered the necessary diagnostic 2D
echocardiogram and only 31.9% performed a neuro exam

• 62% of providers made the primary diagnosis of decompensated heart


• 88% (41 of 47) of providers correctly admitted the patient to the

– But 21 unnecessarily admitting to the ICU (a much higher and more expensive
level of care than is needed for this patient)

• 74% ordered a diuretic but only 40% of providers correctly started an

ACE inhibitor, 15% ordered a statin and only 2% ordered aspirin or a
beta blocker

© QURE 2017

Drill Down: Diabetes Care
Caring for Mr. Bouphavanh, 52
• Great workup: 92% (n=48) ordered either a fasting or random blood
sugar for this patient and nine providers ordering an HbA1c test

• 92% correctly identified the primary diagnosis of new-onset type 2

diabetes mellitus

• BUT 62% unnecessarily admitted this patient to the hospital (medical

therapy and discharge home with a follow-up would have been

• Only 50% of providers placed the patient on metformin that is

recommended as first line drug therapy by WHO PEN and DM

© QURE 2017