Presentation from the dissemination of the Bangladesh Maternal Mortality and health Care Survey 2016. Dhaka, Bangladesh, November 22, 2017. United States Agency for International Development; UKaid; MEASURE Evaluation, the International Centre for Diarrhoeal Disease Research, Bangladesh; Bangladesh Ministry of Health and Family Welfare, and the Bangladesh National Institute of Population Research and Training
2. • BMMS 2016 is implemented by
• With technical assistance from
Organizations involved
BMMS: Bangladesh Maternal Mortality and Health Care Survey
3. • BMMS 2016 is funded by
Organizations involved
4. The BMMS provides an estimate of MMR
Measuring maternal mortality
Its importance
HPN: health, population and nutrition
SDG: sustainable development goal MMR: maternal mortality ratio
4th HPNSP:
121 deaths
per 100,000 LB
SDG:
70 deaths
per 100,000 LB
MMR targets for Bangladesh
2022 2030
8. BMMS 2016: field implementation
• Training
• Data collection
• Data collection agencies
9. BMMS 2016: quality control
• Quality control activities were done extensively by
20 independent teams from NIPORT and icddr,b
• Phase-specific discussions on fieldwork and detailed
debriefing sessions
• Feedback on fieldwork through computerized data
quality checks
• Fieldwork also monitored by MOHFW, USAID,
MEASURE Evaluation, NIPORT, icddr,b
42. Trends in socioeconomic indicators
among the lowest two wealth quintiles
4
22
1111
40
50
53
70
76
Electricity (national
grid)
Toilet (improved) Wall (non-kacha)
Percentage
BMMS 2001 BMMS 2010 BMMS 2016
43. Access to communication channels
52
6360
94
Exposed to either TV or radio
once a week
Household owns at least one
mobile phone
Percentage
BMMS 2010 BMMS 2016
46. Key measures of maternal
mortality
• Maternal mortality ratio (MMR)
Maternal deaths per 100,000 live births
• Pregnancy-related mortality ratio
Pregnancy-related deaths per 100,000 live
births
47. BMMS data sources
Household questionnaire
Household deaths in previous 3 years
Female deaths (13–49 years)
* At the time of death: whether pregnant, delivering or within
two months of delivery, or none of the above
Verbal autopsy interview (13–49 years)
* Causes of death identified by independent physician reviews
48. The verbal autopsy
Verbal autopsy questionnaire
• The BMMS 2010 VA questionnaire was
reviewed and revised based on:
Assignment of cause of deaths
• ICD-10 codes used to assign cause of
deaths
VA: verbal autopsy ICD: international classification of diseases
49. Definitions (ICD-10)
Maternal death
• Death of a woman while pregnant or within 42
days of termination of pregnancy … from any
cause related to or aggravated by the pregnancy
…, but not from accidental or incidental causes
Pregnancy-related death
• Death of a woman while pregnant or within 42
days of termination of pregnancy, irrespective of
the cause of death
52. *where physicians agreed that there was a maternal cause of death but could not agree whether it was
direct or indirect obstetric
The physician review process
1,524
adult female deaths
1,402 (92%)
deaths
122 (8%) deaths
Yes
66 (4%) deaths 56 (4%) deaths
Independent review by 2 physicians
Agreement on cause
Review by third physician
Agreement on cause
Expert committee review (10 cases*)
No
No
Yes
61. Causes of maternal deaths among women of
reproductive age (15–49 years) Bangladesh, 2016
Hemorrhage
31%
Eclampsia
24%
Obstructed or
Prolonged
Labor
3%
Abortion
7%
Other Direct
7%
Indirect
20%
Undetermined
8%
18% of maternal
deaths were due to
or associated with
NCDs
64. Maternal mortality ratios by timing of
death: Bangladesh, 2010 and 2016
35
18
142
48
13
135
During Pregnancy During delivery Postpartum
2010 2016
65. Measurement of maternal morbidity
Obstetric fistula (OF) and pelvic organ prolapse (POP)
• Used a screening tool in BMMS 2016 to identify reported
cases of OF and POP (3rd and 4th degree)
• Same tool used in a study population of 50,000 households
• Reported cases in the study population followed up and
clinically confirmed by physicians
• Adjustment factor derived from the study to be used to
estimate national prevalence of OF and POP from BMMS
2016
• The findings will be disseminated in January 2018
69. Utilization of maternal health services
• Antenatal care
• Delivery attendance by medically trained
providers
• Continuum of care
• Care seeking for maternal complications
• Delivery by C-section
• Out of pocket expenditure for delivery
70. Trends in Antenatal Care (ANC)
ANC: antenatal care
40
55
74
BMMS 2001 BMMS 2010 BMMS 2016
ANC from
Medically Trained Provider
12
23
37
BMMS 2001 BMMS 2010 BMMS 2016
4 + ANC
Percentage
Percentage
73. Deliveries in public, private, and NGO
facilities
Increased by
4 percentage
points
Increased by
18 percentage
points
Increased by
2 percentage
points
NGO: nongovernmental organization
6
10
143
11
29
1
2
4
BMMS 2001 BMMS 2010 BMMS 2016
Percentage
Public Private NGO
10
23
47
74. Deliveries in public, private, and NGO
facilities
6
10
143
11
29
1
2
4
BMMS 2001 BMMS 2010 BMMS 2016
Percentage
Public Private NGO
47
10
13% Upazila & above,
1% Union & below
Little information
on types of
facilities
NGO: nongovernmental organization
23
75. The poor-rich inequity in use of
health facilities for delivery
3
8
23
30
53
76
BMMS 2001 BMMS 2010 BMMS 2016
Percentage
Poorest Richest
Rich-poor
ratio 10 6.6
3.5
27%
point
45%
point
53%
point
76. Complete continuum of care (ANC + DC + PNC from
medically trained provider)
DC: delivery care
PNC: postnatal care
NO
maternity
care
21%
43%
77. Complete continuum of care (ANC +
DC + PNC from medically trained provider)
5
19
43
BMMS 2001 BMMS 2010 BMMS 2016
Percentage
78. There have been substantial
improvements in use of
medically trained attendants
and facilities for delivery.
Now we will examine care-
seeking behaviours for maternal
complications.
79. 53
68 67
BMMS 2001 BMMS 2010 BMMS 2016
Percentage
Sought any treatment
Care seeking for maternal complications
80. 53
68 67
BMMS 2001 BMMS 2010 BMMS 2016
Sought any treatment
Care seeking for maternal complications
81. Care seeking for maternal complications
16
29
46
BMMS 2001 BMMS 2010 BMMS 2016
Sought treatment from health facilitiesSought any treatment
53
68 67
BMMS 2001 BMMS 2010 BMMS 2016
Percentage
83. Proportion of facility deliveries
performed by C-section
65
35
39
83
All facilities Public sector NGO sector Private sector
Percentage
1 million
155,000
750,000
43,000
Number of
C-sections
per year
84. Median out-of-pocket expenditure for
deliveries
All amounts in Bangladeshi taka
1000
5000
6100
19000
Home Public sector NGO sector Private sector
Health Facility
85. Median out-of-pocket expenditure for
normal and C-section deliveries
1,000
3,000 2,600
6,800
12,400
15,500
20,000
Home Public NGO Private
Outofpocketexpenditure(BDT)
Normal delivery C-section
86. Source of funds for facility delivery, by
wealth status, 2016
68
33
28
6
85
15
20
2
Family fund Loan Gift from
relatives
Others
Percentage
Poor (bottom two quintiles) Non-poor (top three quintiles)
90. 322
194 196
0
50
100
150
200
250
300
350
400
1995 2000 2005 2010 2015 2020
MaternalMortalityRatio/100,000LB
BMMS
2001
Trends in Bangladesh
Maternal Mortality Ratio
BMMS
2010
BMMS
2016
• 2001 sample size
was 100,000 but
300,000 in 2016.
• CI’s in 2010 and
2016 do not
overlap with CI in
2001, so MMR is
significantly lower
than 2001 in later
surveys.
• Estimate for 2010
not significantly
different from 2016.
391
253
238
149
234
159
95%
Confidence
Intervals
Significant difference
91. Bangladesh ‘Hockey Stick’ MMR, ANC 4+,
Facility Delivery, 1990-2016
0
10
20
30
40
50
60
70
80
90
100
0
100
200
300
400
500
600
1990 2001 2010 2016
ANC4+,FacDelPercent
MaternalDeaths/100,000LB
ANC4+ FacDel MMR
• Steep decline in MMR
until 2010.
• Although use of ANC 4+,
and Facility Delivery
continued to increase.
• Why is there a plateau in
MMR when ANC & Facility
Deliveries are rising?
• Other factors must play a
role in the MMR trend.
92. MMRs stalling or fluctuating at high levels
0
100
200
300
400
500
600
700
800
900
1000
1990 95 2000 05 2010 15
Lesotho
• In Africa, HIV/AIDS has a big impact on MMR. Increases risk of
direct (haemorrhage, eclampsia) & indirect (e.g. TB) conditions.
• Ghana: MMR stalled ~300, though Facility Delivery still rising.
Low HIV/AIDS level, but poor quality of care in health services.
• Lesotho had up to 50% maternal deaths due to AIDS (2005), but
free maternal health services after 2005 shows MMR declining.
0
20
40
60
80
100
0
100
200
300
400
500
600
700
800
900
1000
1990 95 2000 05 2010 15
Ghana
FacDel MMR
93. MMRs stalling or fluctuating at medium levels
0
20
40
60
80
100
120
140
160
180
200
1990 1995 2000 2005 2010 2015
South Africa
0
10
20
30
40
50
60
70
80
90
100
0
20
40
60
80
100
120
140
160
180
200
1990 1995 2000 2005 2010 2015
Philippines
ANC 4+ FacDel MMR
• Philippines stalled 20 years, poor quality of care, but MMR decline
resuming as UHC reaches 80% after 2012, with free maternity care.
• South Africa: 1999-2009 Health Minister promoted ineffective ‘natural’
HIV cure, so many AIDS maternal deaths (67% in 2005) but from 2009
new Minister pushed Antiretroviral drugs for pregnant women.
No AIDS drugs
available
ARVs
94. MMRs stalling at low levels
0
10
20
30
40
50
60
1990 1995 2000 2005 2010 2015
Thailand
0
10
20
30
40
50
60
1990 1995 2000 2005 2010 2015
Cuba
• Cuba has very strong health system, but cannot get MMR down
below 40, ~ poor nutrition among women, and domestic violence.
• Thailand had HIV/AIDS (25% of maternal deaths in 2000), but from
2001 $1 UHC provides ~ free maternity services. HIV still 18% of MMR.
So MMR also affected by poor health services, lack of political will,
socio-economic factors, maternal nutrition, and other non-maternal
diseases like HIV and malaria.
95. Maternal Mortality Ratio and Facility Deliveries in 35
Sub-Saharan Africa, and South and South East Asia
R² = 0.068
0
100
200
300
400
500
600
700
800
20 30 40 50 60 70 80 90 100
MMR
Facility Delivery %
• Countries have very different
levels of MMR and Facility
Delivery.
• Modest trend of MMR decline
with increasing Facility Delivery.
• Can have high Facility
Deliveries (75%+) but MMR 2x
or 3x higher than Bangladesh
• No countries have lower MMR
than Bangladesh unless Facility
Deliveries more than 60%.
96. To achieve lower MMR, do C-section Rates
need to be higher than 31% (52 countries)?
• 80% have C-section rates
below 31, the Bangladesh
level.
• Most of these (75%) have
MMR below 20. Lowest
MMR at 10-15% C-Section.
• Answer is NO. No need for
a higher rate of C-sections
to achieve lower MMR.
• How can we stop or slow
the increase in C-sections?
R² = 0.3793
0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50 60
MMR
C-Section Rate (%)
97. • As C-sections rose(2000-2010), MMR stalled with increases in ICU admissions, blood
transfusions, hysterectomies, and maternal deaths.
• Since 2012, hospitals must show proof of need for C-sections (e.g., partogram) to
get reimbursed.
• Partnership between 150 hospitals and Govt has increased % vaginal births (fewer
C-sections), and reduced obstetric complications.
Brazil is an example of high C-section (CS) use,
but recently declining
38
56
0
10
20
30
40
50
60
0
20
40
60
80
100
120
1990 1995 2000 2005 2010 2015
C-sectionRate%
MMR
Brazil
C-sections MMR
98. Summary 1/2
• MMR can stall or at different levels, and for different reasons.
• Contributing factors: diseases like HIV/AIDS, malaria, weak health
systems, high costs, malnutrition, social factors like violence.
• Most not relevant to Bangladesh MMR plateau. What applies
most for Bangladesh is weak preparedness of health
facilities/providers to offer quality of care.
• May be other contributing factors, such as rising NCDs, and the
high rate of C-sections in ‘non-ready’ facilities may be having
unrecognized negative consequences.
99. Summary 2/2
In February 2011, we wrote in Implications from 2010 BMMS:
• “…quality of care is a concern”.
• “Health system problems persist, particularly with staffing, staff
attendance, logistics (medicines, blood), and skills.”
In 2014 BDHS Policy Briefs on Maternal Health:
• “Improving and Ensuring Quality of Care is Fundamental”.
In 2014 Bangladesh Health Facility Survey Policy Brief highlighted:
• “Bangladesh’s health facilities are not fully prepared to
provide quality health services”.
101. MMR: 576
MMR: 322
MMR: 194
Final hundred meters:
business as usual will not work
10 years
10 years
?? years
MDG5: 143
Malaysia: 31
Kerala: 81
2010 - 2016
102. Success is a lousy teacher, it
seduces smart people into
thinking they can’t lose.
It’s fine to celebrate success
but it is more important to
heed the lessons of failure.
- Bill Gates
103. Science Skill Implementation Survival
Utopia 1 1 1 = 1.00X X
Ideal .9 .9 .9 = 0.72X X
Attainable .8 .9 .5 = 0.36X X
Actual .8 .5 .5 = 0.20X X
Utstein Formula of Survival
X X =
Ref: Soroide E et al, The formula of survival in resuscitation, Resuscitation 2013 Nov, 84 (11): 1487-93.
104. 31% maternal
deaths are due
to hemorrhage
53% deliveries are
occurring at home,
17% Misoprostol coverage
24% maternal
deaths are due
to Eclampsia
28% facilities have
magnesium sulphate to
treat Eclampsia
Disjointed program focus
18% maternal
deaths are due
to or associated
with NCD
No screening,
management of NCD at
Primary Care level
105. Skilled providers are not available
46
20
Public (UHC+DH+MCWC) Private hospitals & clinics
Percentageoffacilitiesofferingnormal
deliveryserviceshaveatleastonestaff
trained
Ref: NIPORT, Bangladesh Health Facility Survey 2014
106. Science Skill Implementation
Survival
Utopia 1 1 1 = 1.00X X
Ideal .9 .9 .9 = 0.72X X
Attainable .8 .9 .5 = 0.36X X
Actual .8 .5 .5 = 0.20X X
Utstein Formula of Survival
X X =
107. Universal Health
Coverage
Essential
Service
Package
Coverage
Quality
Primary level: Preventive,
Promotive, Screening and
Primary care
Secondary+ level: Curative,
Advanced and Referral level
care
Service or clinical or procedural
Diagnostics
Communication
Physical and environmental
Equipment
Drugs and supplies
HR including skills, competencies
Infection prevention
Waste management
Record keeping and reporting
Management and administrative
Standard
Operating
Procedure
Regulatory
control
Quality of care: setting standards is key
Referral
Ref: Measure Evaluation,
Bangladesh Health Facility
Survey 2014 Policy Brief
108. Health care providers are not available
24 hours
39
89
7
62
55
All facilties
(excluding CC)
Public
(UHC+DH+MCWC)
Public (Union level) NGO Private hospitals &
clinics
Percentoffacilitiesprovidednormal
deliveryservices Ref: NIPORT, Bangladesh Health Facility Survey 2014
110. Appropriate care is not available
30
96
10
16
Public (UHC+DH+MCWC) Private Hospitals
Provided CS service Had all 9 CmOC signal functions
Percentageof
facilities
Ref: NIPORT, Bangladesh Health Facility Survey 2014
111. Universal Health
Coverage
Essential
Service
Package
Coverage
Quality
Primary level: Preventive,
Promotive, Screening and
Primary care
Secondary+ level: Curative,
Advanced and Referral level
care
Service or clinical or procedural
Diagnostics
Communication
Physical and environmental
Equipment
Drugs and supplies
HR including skills, competencies
Infection prevention
Waste management
Record keeping and reporting
Management and administrative
Standard
Operating
Procedure
Regulatory
control
Quality of care: setting standards is key
Referral
112. 6
10
143
11
29
1
2
4
BMMS 2001 BMMS 2010 BMMS 2016
Percentage
Consequences of unregulated market
164%
40%
Private dominance leads to–
• Deprivation of care for the poor
• Suppression of public sector
growth
• Substandard quality at high
price
Public sector should gear up, to –
• Cater to the need of the poor
• Attract people with high quaiity
of care
113. Home: 53% Facility: 47%
CS: 65%
of facility deliveries
15%
Life saving
CS Unnecessary
CSNeed life-saving
CS, Can’t have it
Impact of lack of quality regulation
114. Delay in decision
making to seek care
Delay in reaching
appropriate care
Delay in receiving
appropriate care
Household
Health
Care
Facility
The three delays
19%
13%
• 29% facilities have
24/7 providers
• Low levels of
readiness
• 24/7 CEmOC not
available
• CEmOC signal
functions not
available
Maternal deaths in transit
3% SBA coverage at
home deliveries
2010
2016
115. Way forward
• Improve overall implementation efficiency:
• ‘Comprehensive’ standard based quality of care
• Effective coverage of ‘key components’ by cause of
death
• Revisit quality of ‘program design’ – get basic concepts
and elements right
• Regulate market for quality standards, and
ensure accountability
• Incorporate quality parameters in routine
measurement
116. ANC
83% + MTP
74% + MTP
34%
15%
ANY
ANC
4
+All Components
ANC
+ MTP4 ANC
We should look at the right indicator