FIELD ACTION REPORT

Management of Preeclampsia, Severe Preeclampsia,

and Eclampsia at Primary Care Facilities in Bangladesh
Anna Williams,a Marufa Aziz Khan,b Mohammed Moniruzzaman,a Sk Towhidur Rahaman,a
Imteaz Ibne Mannan,c Joseph de Graft-Johnson,d Iftekhar Rashid,e Barbara Rawlinsf

Program introduction, including cascade training, to screen for severe preeclampsia and eclampsia and initiate
treatment with magnesium sulfate was somewhat successful. Challenges included inconsistent adherence to the
national protocol, data quality, and some issues with supplies and equipment.

ABSTRACT
Introduction: Eclampsia-related conditions are the second leading direct cause of obstetric deaths in Bangladesh. Efforts to prevent such
deaths in low- and middle-income countries are increasingly focused on task shifting at the primary care level to enable frontline provi-
ders to screen and initiate treatment for women with preeclampsia, severe preeclampsia, and eclampsia (PE/SPE/E). The MaMoni
Health Systems Strengthening project (funded by the United States Agency for International Development) implemented a magnesium
sulfate intervention at primary care facilities in 4 Bangladesh districts in 2016 and 2017.
Methods: The project trained frontline providers through a cascade approach from the national to the union level. A PE/SPE/E patient
algorithm, digital blood pressure machines, and eclampsia kits with magnesium sulfate were supplied to service providers at each facil-
ity. We conducted a retrospective record review of facility-level data to assess the degree to which newly trained frontline providers
adhered to a protocol that incorporated the use of magnesium sulfate for SPE/E in primary care settings.
Results: In total, 283 women were found to have PE/SPE/E. Fifty-four percent were managed according to the protocol. The required
supplies were present at each facility, but some issues existed with regard to availability and functionality of blood pressure
apparatuses.
Discussion: Challenges related to recordkeeping and service quality limited the analysis. Frontline providers need refresher trainings,
ongoing supervision, properly calibrated blood pressure devices, and performance monitoring support in order to improve screening
and management of PE/SPE/E in primary care facilities.

INTRODUCTION of frontline providers, and systems challenges that lead

P reeclampsia, severe preeclampsia, and eclampsia to delays in women receiving necessary treatment.2–4 In
(PE/SPE/E) are hypertensive disorders of pregnancy Bangladesh, eclampsia-related conditions are the second
that contribute significantly to global maternal and peri- leading direct cause of obstetric deaths and lead to 24%
natal mortality.1 Marked by high blood pressure (BP) of all maternal deaths.5 Over 1,000 women die each year
and the presence of albumin in urine, preeclampsia is a in Bangladesh due to PE/SPE/E. As in many low- and
risk factor for the potential development of severe pre- middle-income countries, most pregnant women who
eclampsia or full-blown eclampsia and should be moni- develop PE/SPE/E in Bangladesh do not get diagnosed
tored. Management of SPE/E poses a challenge in or treated. They either do not access the health system
low- and middle-income countries due to a lack of basic at all, are not screened properly, or do not receive timely
supplies, health worker shortages, limited competencies treatment due to delays in (1) making the decision to
seek care, (2) being transported to receive care, and
(3) actually receiving the required treatment at the care
a
Save the Children, Dhaka, Bangladesh. site where it is available.6,7
b
Pathfinder, Dhaka, Bangladesh.
c
Jhpiego, Kabul, Afghanistan.
d

e
Save the Children, Washington, DC, USA. PE/SPE/E Detection and Management at the
United States Agency for International Development/Bangladesh, Dhaka,
Bangladesh. Primary Care Level: A Global Priority
f
Jhpiego, Washington, DC, USA. In recent years, global efforts to reduce eclampsia-
Correspondence to Anna Williams (annacw@gmail.com). related deaths have focused on task shifting, or enabling

Global Health: Science and Practice 2019 | Volume 7 | Number 3 457

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

frontline health workers to identify women with hypertensive disorders of pregnancy currently
PE/SPE/E and initiate management of the disor- exists.11 It has been further tested in community
der.8,9 Calcium supplementation is recommended settings through the Community Level Interventions
for preventing preeclampsia when dietary intake for Preeclampsia clinical trials in India, Pakistan,
of calcium is low, while antihypertensive drugs Mozambique, and Nigeria (https://clinicaltrials.gov/
may be necessary for women with PE. Magnesium ct2/show/NCT01911494). The results are forthcom-
sulfate (MgSO4) is recommended by the World ing and are expected to make a valuable contribu-
Health Organization (WHO) to manage SPE/E tion to the evidence base on effective intervention
among pregnant women. In settings where admin- strategies for identifying and managing PE/SPE/E
istering a full MgSO4 regimen (which includes a at the community level in low-resource settings.
“loading dose” followed by scheduled maintenance Currently, though, evidence is lacking on the
Evidence is doses) is not possible, WHO recommendations in- effectiveness of program interventions focused on
lacking on the clude providing the initial MgSO4 loading dose diagnosis, management, and referral of women
effectiveness of (via intramuscular injection and/or intravenous with eclampsia-related conditions at primary-
program drip) and immediately transferring the individual level health facilities and in communities. This
interventions for to a higher level of care.10 To implement this strat- topic is of particular interest in Bangladesh follow-
the diagnosis, egy, frontline health workers in low- and middle- ing changes within the past few years in the na-
management, income countries need to have access to BP gauges, tional PE/SPE/E protocol and recent program
and referral of urine dipsticks, and MgSO4, and need to be trained efforts. This article examines service delivery data
women with to screen all pregnant women >20 weeks of gesta- from 35 primary care facilities that received support
eclampsia-related tion for elevated BP, urine albumin, and the pres- for providing screening and pre-referral treatment
conditions. ence of any danger signs. If SPE/E is identified, the with MgSO4 as part of their standard maternal
workers need to administer a MgSO4 loading dose health services. The facilities were supported by
and facilitate a timely referral of the woman to a the MaMoni Health Systems Strengthening project
higher-level health facility. (MaMoni HSS) to improve the quality and reach of
Although the inputs are standard, numerous their maternal and newborn health services by us-
obstacles may be encountered when rolling this ing a range of evidence-based interventions, in-
service out in low- and middle-income countries. cluding the introduction of PE/SPE/E screening
Weak health systems may have inadequate ser- and management following national guidelines.
vice delivery protocols, provider skills, systems
for supportive supervision, availability of essential
supplies (such as BP gauges, urine dipsticks, and PROJECT DESCRIPTION
injectable MgSO4), and collection and use of mon- MaMoni HSS was a large maternal, newborn, and
itoring data. Barriers to ensuring women with PE/ child health (MNCH) project in Bangladesh that
SPE/E are identified and optimally managed likely was funded by the United States Agency for
vary from setting to setting, yet they are expected International Development (USAID) between
and must be addressed to further reduce global 2013 and 2018. Its maternal health interventions
maternal mortality. focused on strengthening public-sector services
The research group that developed the Pre- from the community level to secondary-level re-
eclampsia Integrated Estimate of Risk (fullPIERS) ferral facilities to provide quality antenatal care
model for high-income, tertiary care settings also (ANC) during labor and delivery, newborn care,
developed the miniPIERS model for providers to and postnatal care (PNC) including postpartum
use in primary care settings. The miniPIERS is a val- family planning (Figure 1). The project also
idated model for identifying women at increased worked at the national level to support the
risk of adverse maternal outcomes associated with Ministry of Health and Family Welfare (MOHFW)
hypertensive disorders of pregnancy. It relies on a to develop a maternal health strategy and stan-
simple assessment of maternal demographics (ma- dard operating procedures (SOPs). The SOPs in-
ternal age, parity, and gestational age), signs (BP corporated a complete package of evidence-based
and proteinuria), and symptoms (headache, visual practices and interventions adopted by MOHFW
disturbance, chest pain, difficulty breathing, upper for implementation through the public sector ser-
abdominal pain, nausea, vomiting, and vaginal vice delivery system at various levels of care. The
bleeding with abdominal pain). The miniPIERS project’s other work at the national level included
study found that using the model in resource- the development of various guidelines, protocols,
limited settings has the potential to significantly im- training materials, and job aids for the roll-out of
prove care where minimal or no monitoring of the interventions across the country.

Global Health: Science and Practice 2019 | Volume 7 | Number 3 458

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

FIGURE 1. MaMoni HSS Project’s Maternal Health Interventions in Bangladesh

Abbreviations: ANC, antenatal care; CEmONC, comprehensive emergency obstetric and neonatal care; HSS, Health Systems
Strengthening; PE/SPE/E, preeclampsia, severe preeclampsia, or eclampsia.

Bangladesh Health System FWVs receive 18 months of training after complet-

Although used in Bangladesh since 1998, MgSO4 ing secondary school and provide MNCH care
was not available for treating SPE/E at the primary services, including family planning, delivery, and
care level via antenatal services prior to 2016 be- immunization. They are the lead providers of ANC
cause of a lack of a standard protocol and uncer- services at UH&FWCs. A medical officer, who is a
tainties about the skill level and competence of doctor with at least 5 years of professional medical
primary care providers. A basic overview of the education, serves at some, but not all UH&FWCs
health system in Bangladesh is necessary back- (Figure 1).12 Across the public health system, the
ground to understand the intervention we discuss health care workforce has been described as being
here. The secondary and tertiary levels of the in crisis due to a shortage of trained providers, in-
health system comprise subdistrict (locally re- cluding FWVs; an inappropriate skill mix; and in-
ferred to as Upazila) health complexes, maternal equitable distribution.13 FWVs are critical frontline
and child welfare centers, district hospitals, and providers staffing nearly 5,000 UH&FWCs around
various teaching and specialist hospitals. At the the country.14 While they are recognized and
primary care level are union health and family counted within national health surveys as medical-
welfare centers (UH&FWCs), union sub-centers, ly trained providers, serious gaps in their provision
and community clinics. UH&FWCs (which are of maternal health services have also been
the focus of this article) are typically staffed by documented.15,16
1 subassistant community medical officer (SACMO),
1 to 2 family welfare visitors (FWVs), and 1 or more PE/SPE/E Case Detection and Management
support staff. SACMOs have participated in a 3-year by Frontline Providers in Bangladesh
medical training course following secondary school Following a pilot test conducted in 2013 and
and provide basic primary health care services. 2014, the National Technical Committee of the

Global Health: Science and Practice 2019 | Volume 7 | Number 3 459

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

Directorate General of Family Planning (under address eclampsia at the community level within
MOHFW) endorsed a protocol for the identifica- Bangladesh and in other low-resource settings.
tion and pre-referral management of severe pre-
eclampsia and eclampsia at union-level facilities
by the FWVs and SACMOs. The protocol recom- METHODS
mended that all pregnant women receive at least
4 quality antenatal check-ups and that measure- Facility Selection
ment of BP, urinalysis for proteinuria, and screen- The 45 UH&FWCs initially selected to receive the
ing for the presence of SPE/E danger signs should PE/SPE/E intervention were chosen with consid-
be done at every antenatal, intrapartum, and eration for having relatively high ANC coverage,
postnatal service visit. All women identified with having a resident FWV around the clock, having a
severe preeclampsia or eclampsia (see case defini- referral facility at an accessible distance, and hav-
tions in Figure 2) were to be given a loading dose ing outreach services with comparatively strong
intramuscular injection of MgSO4 and then re- performance. All facilities were located in 4 dis-
ferred to the nearest comprehensive emergency tricts (Figure 4) where the project focused on
obstetric and neonatal care (CEmONC) facility. strengthening primary- and secondary-level pub-
For women with preeclampsia, the protocol speci- lic services to provide a complete package of
fied that they should be referred to a nearby evidence-based MNCH interventions, including
CEmONC facility for treatment with antihyper- family planning and nutrition.
tensive drugs and monitoring. Primary care provi-
ders in Bangladesh are not authorized to prescribe
or administer antihypertensive drugs to pregnant Initiation of the Intervention
women. A pictorial algorithm (Figure 3) was de- To initiate the PE/SPE/E intervention, sensitiza-
veloped by MOHFW together with development tion meetings were held with relevant district-
partners as a job aid for frontline providers at and subdistrict-level health officials from the
UH&FWCs to guide them through triage and man- 4 districts together with representatives from
agement of women with PE/SPE/E in line with the MOHFW, members of the Obstetrical and
We conducted a national protocol. Subsequently, MaMoni HSS se- Gynaecological Society of Bangladesh (OGSB),
record review of lected 45 UH&FWCs for early implementation of and MaMoni HSS program managers. A baseline
facility-level data this protocol as part of the larger set of maternal survey carried out across all 45 facilities consisted
on PE/SPE/E health interventions under the project. of a retrospective record review of 6 months of
services at 35 This article presents findings from a record facility-level data (covering the period October
primary care review of facility-level data on PE/SPE/E services 2015 to March 2016) to identify documented
facilities in at 35 of the 45 UH&FWCs, and it additionally cases of preeclampsia and eclampsia and how
Bangladesh. provides recommendations for future efforts to they were managed. The data source for the base-
line was a new MNCH patient register that the
project had earlier worked with MOHFW to dis-
tribute and train providers on how to use, as a re-
placement to using 4 separate registers to capture
FIGURE 2. Case Definitions of Preeclampsia, Severe Preeclampsia, and the same information. In addition, a routine ser-
Eclampsia According to National Protocol, Bangladesh vice delivery point survey conducted quarterly by
the project was used to check whether MgSO4 and
BP apparatuses were present at each UH&FWC.
A memorandum of understanding was devel-
oped between MaMoni HSS and OGSB in order
to roll out training for FWVs and SACMOs in the
targeted facilities. OGSB developed the training
materials and carried out a cascade training ap-
proach in which national-level expert trainers
established a group of district-level master trainers
who then replicated the training at the communi-
ty level for FWVs and SACMOs as new facilities
adopted the intervention. Service providers at
Abbreviation: dBP, diastolic blood pressure. secondary-level referral facilities also received an
orientation from OGSB on the intervention to

Global Health: Science and Practice 2019 | Volume 7 | Number 3 460

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

FIGURE 3. Pictorial Algorithm for the Management of Women With PE/SPE/E Developed by the Ministry of
Health and Family Welfare, Bangladesh

Abbreviations: BP, blood pressure; MgSO4, magnesium sulfate; PE/SPE/E, preeclampsia, severe preeclampsia, or eclampsia.

Global Health: Science and Practice 2019 | Volume 7 | Number 3 461

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

FIGURE 4. Map of Bangladesh Showing the Intervention Focus Districts

prepare them to receive and manage the referred learned to identify PE/SPE/E based on the case
cases (Figure 5). identification criteria in Figure 2. They were
taught how to administer a loading dose of
MgSO4 via intramuscular injection and refer iden-
Provider Training tified SPE/E cases to the nearest CEmONC facility.
Two-day competency-based trainings for frontline Each participant was provided with a digital BP
providers at the 45 facilities were carried out be- machine and a laminated copy of the patient algo-
tween March and May of 2016. In the trainings, rithm (that included both the pictorial version and
providers were taught how to check BP using a 1-page text description of the algorithm).
both digital (Microlife brand) and manual BP Eclampsia kits that consisted of 4 preloaded vials
cuffs, conduct a urine protein analysis, and screen of injectable MgSO4 were purchased by the project
all women for danger signs. Training participants and allocated to UH&FWCs based on a rough

Global Health: Science and Practice 2019 | Volume 7 | Number 3 462

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

FIGURE 5. Cascade Training Model

Abbreviations: FWV, family welfare visitor; SACMO, sub-assistant community medical officer.

estimate of possible eclampsia incidence. Incidence visit was to be added to women’s initial record. A
estimates were produced following a 2-step pro- second record was to be created in the supplemen-
cess. First, an initial calculation was made of the tal reporting form (hereafter referred to as a “pa-
crude birth rate in each UH&FWC catchment area tient linelist”) only for women diagnosed with
using data from the 2011 Bangladesh census. SPE/E. The purpose of the patient linelist was to
Then, calculations of PE and E incidence for each provide condensed essential SPE/E reporting in-
catchment area were made based on estimates of formation to MaMoni HSS. This form did not in-
PE and E incidence (PE 2.8% of live births and clude variables related to ANC, intrapartum, or
eclampsia 2.3% of PE) in developing countries PNC services, nor did it record information about
published by EngenderHealth in a 2007 report.17 referrals for women with PE. However, it captured
Based on these projections, FWVs were provided outcome information not recorded in the MNCH
with a monthly supply of MgSO4, which they registers, such as whether referrals were complet-
restocked periodically from subdistrict drug storage ed and maternal and newborn outcomes.
facilities when conducting general inventory
restocking as part of their regular work. Urine test
tubes and strips for measuring albumin were al- Complementary Program Inputs
ready available at all primary care facilities through The project’s other maternal health interventions—
the existing supply chain. to increase ANC coverage and quality, raise aware-
Upon completion of the training, providers be- ness at the community level about PE/SPE/E dan-
gan screening for PE/SPE/E at their facilities and ger signs and the newly available services, improve
managing SPE/E with a pre-referral loading dose the quality of CEmONC services at referral facilities,
of MgSO4. Services were documented in the facil- and monitor and improve the availability of essen-
ity’s MNCH register by FWVs in fields designed to tial drugs and supplies—complemented the efforts
capture key details of ANC, intrapartum, and PNC to improve PE/SPE/E identification and manage-
services. A single initial record was supposed to be ment at UH&FWCs. As part of its overall scope, the
created in the MNCH register for all women when project also carried out monthly monitoring and
they received ANC, intrapartum, and/or PNC ser- supervision visits at selected facilities with district-
vices. At each new visit (e.g., second or third ANC and subdistrict-level managers. Facilities at differ-
visit, intrapartum care following ANC, or PNC fol- ent levels of the health system that were supported
lowing intrapartum care) information about that by the project’s various MNCH interventions

Global Health: Science and Practice 2019 | Volume 7 | Number 3 463

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

(Figure 1) received these supervision visits, includ- fields across all 3 of these services were closely ex-
ing some of the 35 UH&FWCs included in this amined and cleaned to establish uniformity of the
article. During these visits, a standard monitoring presentation of key information. This clean-up pri-
checklist was completed to facilitate review of a marily consisted of ensuring that all instances of
broad range of service quality issues, including treatment with magnesium sulfate were written as
around PE/SPE/E. Additional monitoring and “MgSO4” and creating coded columns for PE/SPE/E
supervision was carried out between January cases and for referred cases. In addition to the gener-
and July 2017 by OGSB together with MaMoni ation of the point estimate, cases with a documented
HSS managers and local-level health officials to PE/SPE/E diagnosis were also compared with cases
specifically assess PE/SPE/E service provision at with only the indications of PE/SPE/E documented
8 UH&FWCs—1 high-performing and 1 low- but not the actual diagnosis. The final point estimate
performing UH&FWC in each focus district. merged the findings from the analysis of both the
MNCH registers and the patient linelists.
Ethical approval for this analysis was granted
Data Analysis
from the Johns Hopkins School of Public Health
A secondary analysis of data from MNCH registers
Institutional Review Board as well as the
and patient linelists covering the period from
Bangladesh Medical Review Council’s National
September 2016 to August 2017 was carried out
Research Ethics Committee.
to develop a point estimate of correct initial man-
Other analyses were also carried out to assess
agement of PE/SPE/E identified during ANC,
facility readiness and provider competency.
intrapartum, and PNC visits at 35 of the 45 inter-
Results from a quarterly service delivery point sur-
vention facilities. The 10 facilities not included in
vey managed by MaMoni HSS were reviewed to
the analysis were no longer adequately staffed or
verify the presence of MgSO4 and BP machines at
had structural problems that prevented them
each of the 35 UH&FWCs covering the periods
from providing consistent ANC services during
January–March 2016 and July–September 2017,
this time. Photocopies of MNCH registers were
as well as just prior to and at the end of the period
made by FWVs, transported to Dhaka in sealed
of analysis. A short questionnaire was completed
boxes, and entered into an Excel spreadsheet for
by field-level MaMoni HSS staff in April 2018 to
analysis. Patient linelists were already kept in
check for the presence of the laminated algorithm,
password-protected electronic files in the MaMoni
test tubes, and urinalysis strips at each UH&FWC.
HSS Dhaka office. The analysis was designed to
This questionnaire also double-checked for the
generate descriptive statistics summarizing key
presence of MgSO4 and BP machines. Both of these
variables that reflect compliance with the PE/SPE/
datasets were used to ascertain facility readiness.
E screening and management protocol.
Provider knowledge was assessed by analyzing
The variables analyzed included the proportion
results from pre- and post-training questionnaires
of women screened for PE/SPE/E, the proportion
with 32 items that checked providers’ knowledge
with indications of PE/SPE/E, and the percentage
of the PE/SPE/E competencies covered in the train-
of those identified who received a loading dose of
ing. Qualitative analysis consisted of reviewing the
MgSO4 and referral to a higher level of care.
reports from the OGSB-led supervision visits, as
Screening was determined by looking at relevant
well as reports from the project’s joint supervision
variables across ANC, intrapartum, and PNC visit
visits and quarterly reports that had been submitted
records. Key variables in ANC records included dia-
to USAID to gather contextual information to in-
stolic blood pressure (dBP) and proteinuria, as well
form the program description and discussion.
as open text fields for capturing pregnancy danger
signs, patient “complaints and disease,” and provid-
er “treatment and advice.” Intrapartum records in- RESULTS
cluded check boxes for blurred vision, severe Results from the baseline revealed that providers
headache, and convulsions, as well as a general had documented 3 cases of PE and 2 cases of
“delivery complications” field and write-in fields eclampsia between October 2015 and March
for treatment and referral information. BP and uri- 2016. None of the women with documented
nalysis are required during intrapartum care, but eclampsia were treated with MgSO4. Analysis of
these variables were not available in the intrapar- the service delivery point dataset revealed that pri-
tum records. PNC records captured dBP and gener- or to the intervention, MgSO4 was not present at
al write-in fields for complaints and disease and for any of the UH&FWCs. All but 4 had BP machines.
treatment and advice, but not urinalysis. Write-in These machines were assumed to be manual BP

Global Health: Science and Practice 2019 | Volume 7 | Number 3 464

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

gauges, which are provided to facilities through whose diagnoses were reported to MaMoni HSS.
the national supply chain, although the type of Records from only 51 of these women also
machine was not indicated in this dataset. In the appeared in the MNCH register.
second service delivery point survey covering Analysis of the MNCH registers revealed that
July–September 2017, all 35 facilities had MgSO4 most pregnant women (9,898, 74%) were be-
and BP machines. The questionnaire completed tween 20 and 29 years of age (Table 2). A total
in April 2018 showed that all 35 facilities had the of 8,462 (65%) pregnant women received just
laminated PE/SPE/E algorithm and urinalysis test 1 ANC consultation during pregnancy, while
tubes and strips. At that time, 33 of the 35 facilities 2,358 (18%) received 3 or more (Table 3).
reported having a BP machine. The 2 that did not Both a dBP and a proteinuria reading were
have a BP machine noted that the FWVs were us- recorded at over 90% of ANC visits. Among the
ing their own personal BP machines in the facili- 5,833 PNC visits documented in the MNCH regis-
ter, dBP was recorded 98% of the time. Across
ties. An additional 3 facilities (which reported
both datasets, 283 women were identified as
having BP machines) noted that they were using
having PE/SPE/E—52 preeclampsia, 214 severe
manual machines that were not giving correct
preeclampsia, and 17 eclampsia (Figure 6). An ad-
readings. Two facilities reported stock-outs of
ditional 250 women without PE/SPE/E had docu-
MgSO4 at the time they completed the question- mented hypertension (dBP≥90). In the MNCH
naire (Table 1). On average, providers correctly register, 53% of PE/SPE/E cases had been identi-
answered 18 out of 32 questions (or 57%) on the fied by a provider, meaning that a written diagno-
pretest and 26.25 out of 32 (82%) on the posttest. sis of PE/SPE/E or treatment with MgSO4 was
The following summary of the facility-level data recorded. The rates at which the clinical manage-
comprises analyses of both the MNHC register data ment of identified cases followed the national pro-
with ANC, intrapartum, and PNC service records tocol are depicted in Figure 6. Overall, the records
and the patient linelists with only women who had indicated that providers adhered to the protocol
an SPE/E diagnosis. Missing records appeared to be for 54% of women with PE/SPE/E (153 women).
common in both datasets. For example, 25 women Adherence to the protocol was lowest for PE—
who had SPE/E documented by an FWV in the only 15% of women with PE were referred.
MNCH register did not appear in the patient linelist. Adherence was highest for eclampsia, with a load-
Likewise, 88 women with SPE/E were reported to ing dose of MgSO4 being administered and a refer-
MaMoni HSS via the patient linelist, but their ral being made to a CEmONC facility for 94% of
records did not appear in the MNCH register. women with eclampsia.
Across both datasets, 13,346 women were seen Among the PE/SPE/E cases that were not re-
for ANC, intrapartum, and/or PNC services at the ferred, 74 women (62%) were admitted to a
35 UH&FWCs between September 2016 and UH&FWC for delivery. Among women with SPE/E
August 2017. The MNCH registers contained who were referred, referral compliance was only
records of 13,031 ANC visits, 3,641 intrapartum documented for about half (79, 52%). Of this group,
visits, and 5,833 PNC visits. The patient linelists 59 (75%) complied with their referral, while 20
contained records of 139 women with SPE/E (25%) did not.

TABLE 1. Facility Readiness Survey Results, July–September 2017

Readiness Indicators Availability in the Facilities

PE/SPE/E pictorial algorithm Present at all 35 facilities

Test tube for albumin test Present at all 35 facilities
Urine strip for albumin test Present at all 35 facilities
BP machine  2 facilities did not have a BP machine; FWVs used their personal BP machines instead.
 3 facilities reported having a BP machine that gave incorrect readings.

MgSO4 Missing at 2 facilities

Abbreviations: BP, blood pressure; FWV, family welfare visitor; MgSO4, magnesium sulfate. PE/SPE/E, preeclampsia, severe pre-
eclampsia, or eclampsia.

Global Health: Science and Practice 2019 | Volume 7 | Number 3 465

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

by a provider versus identified in the analysis based

TABLE 2. Age Distribution of Women Who Received only on dBP, proteinuria, and/or a danger sign, fur-
ANC, Delivery Services, and/or PNC (N=13,346) ther highlighted the data quality challenges.
Two possible explanations could account for
Mother’s Age No. (%)
the lack of record duplication expected between
<20 1,728 (13) the MNCH register and the patient linelists. First,
frontline providers may have sometimes estab-
20–24 5,802 (43)
lished a record for a woman in the MNCH register
25–29 4,096 (31) but not the linelist (and vice versa). Second, some
≥30 1,564 (12) of the relevant MNCH registers may not have been
shared from all 35 facilities. Both of these scenari-
Not recorded 156 (1)
os could potentially explain why 88 women with
Abbreviations: ANC, antenatal care; PNC, postnatal care. SPE/E were reported to MaMoni HSS in the pa-
Source: Maternal, newborn, and child health register. tient linelists but not found in the MNCH register.
Another quandary was that the prevalence of
PE was much lower than that of both hyperten-
TABLE 3. Distribution of Pregnant Women by Total sion and SPE. Typically, a declining pattern would
Number of ANC Consultations Received (N=13,031) be present, with the highest numbers of women
having hypertension, fewer women having PE,
ANC Visits No. (%) and many fewer women having SPE/E. This pat-
tern not being apparent in our dataset is likely
1 8,462 (65) explained by inaccurate measuring or recording
2 2,211 (17) of BP and/or proteinuria (in addition to the issues
3 1,787 (14) with missing data already described). BP measure-
ment and/or urinalysis may not have been done at
≥4 571 (4)
all (despite a reading having been recorded) or
Abbreviation: ANC, antenatal care. may have been done incorrectly. Further, BP mea-
Source: Maternal, newborn, and child health register. surements may have been rounded up or down
when recorded. Rounding up could have skewed
Among the women who were referred, type the results toward higher numbers of SPE cases
of delivery was recorded for 147. Of these, than there actually were. For women with a dBP
118 (80%) had a vaginal birth, while 29 (20%) measurement of ≥90, if a negative proteinuria
had a cesarean delivery. Newborn outcomes were measurement was recorded but the test was not
actually done, records may have fallen into the
recorded for 154 (54%) of all PE/SPE/E cases.
category of hypertension, rather than PE. Anecdotal
Among those with recorded newborn outcome in-
evidence from both project staff and local researchers
formation, 150 (97%) of deliveries resulted in a
studying FWV skills in detecting and managing PE/
live birth. There were 3 stillbirths and 1 newborn
SPE/E supports that any of these may be realistic sce-
death. In all 4 of these cases, the mother had been
narios. While not documented thoroughly enough to
referred to a higher of level of care.
be fully substantiated, these scenarios are also sup-
ported by some published literature.3,18
DISCUSSION The data quality challenges highlighted the need
We aimed to generate a point estimate of correct for ongoing mentoring, support, and refresher train-
diagnosis and initial management of PE/SPE/E by ing for frontline workers. This need is also apparent in
frontline providers at primary care facilities in the low rate of adherence to the PE/SPE/E protocol
Bangladesh. We found that frontline providers and is emphasized in the broader literature on this
managed slightly over half of women with PE/ topic as well.3,5,8,10 Furthermore, while FWVs’ adher-
SPE/E in line with their training. Significant chal- ence to the standard protocol requires more intensive
We found that lenges were noted with regard to the quality of the monitoring and guidance, improvements in record-
frontline providers data. This issue stood out in the analysis because keeping are also essential for accurate tracking of ser-
managed slightly large numbers of women with SPE/E were present vice quality and case management. In an ideal
over half of in the MNCH register but missing from the patient scenario, a rapid feedback loop would exist in which
women with PE/ linelists, and vice versa. The range of missing infor- service data would be regularly consolidated and
SPE/E in line with mation in the MNCH register, as well as the disparity summarized, and trends and issues discussed and
their training. between women whose diagnosis was documented addressed directly with FWVs and SACMOs.

Global Health: Science and Practice 2019 | Volume 7 | Number 3 466

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

FIGURE 6. Client Flow Chart

Abbreviation: ANC, antenatal care; BP, blood pressure; HSS, Health Systems Strengthening; MgSO4, magnesium sulfate; MNCH,
maternal, newborn, and child health; PE/SPE/E, preeclampsia, severe preeclampsia, or eclampsia; PNC, postnatal care.

Other challenges are the availability of MgSO4 primary care settings in low- and middle-income
and functioning BP machines. MgSO4 being on the countries. Findings from program experience in
government’s essential drug list is an advantage, but Bangladesh indicate that intensive inputs are re-
if MOHFW cannot fund its availability at the prima- quired to introduce and maintain quality of PE/
ry care level, then its supply is dependent on donor SPE/E service delivery in primary care facilities.
funds and is not sustainable. Finally, more attention The findings also demonstrate that delivering
should be paid to ensuring that frontline providers competency-based training together with the pro-
have access to functioning BP apparatuses and use vision of essential supplies (i.e., BP machines, a vi-
them correctly. The project’s scope was limited to sual job aid, and injectable MgSO4), supportive
providing replacement devices to facilities where supervision, and complementary program inputs
providers reported problems with the functionality at the national, community, and secondary care
of the digital BP machine they had been provided levels are effective interventions to begin to enable
with. However, maintaining proper calibration of frontline providers to comply with PE/SPE/E
aneroid devices and ensuring correct measurement screening and management protocols. In addition
techniques are fundamental challenges, particularly to these inputs, well-functioning BP apparatus-
in low-resource settings. es, routine monitoring of facility-level data, and
ongoing performance management are also
critical for providing and monitoring quality ser-
CONCLUSION vices. These findings can contribute to strength-
Community-based management of PE/SPE/E is an ening community-level PE/SPE/E interventions
important maternal health intervention that is be- in Bangladesh and in other low-resource prima-
ing tested in a variety of community-level and ry care settings.

Global Health: Science and Practice 2019 | Volume 7 | Number 3 467

Management of Preeclampsia and Eclampsia in Bangladesh www.ghspjournal.org

Acknowledgments: Joby George (Save the Children Bangladesh) and 9. Danmusa S, Coeytaux F, Potts J, Wells E. Scale-up of magnesium
Setara Rahman (Jhpiego Bangladesh) provided critical technical and sulfate for treatment of pre-eclampsia and eclampsia in Nigeria.
managerial supervision of the production of this article. Int J Gynaecol Obstet. 2016;134(3):233–236. CrossRef.
Medline
Funding: This article was made possible by the generous support of the 10. World Health Organization (WHO). WHO Recommendations
American people through the United States Agency for International for Prevention and Treatment of Pre-eclampsia and Eclampsia.
Development (USAID), under the terms of the Leader with Associate
Geneva: WHO; 2011. https://apps.who.int/iris/bitstream/
Cooperative Agreement No. AID-338-LA-13-00004. The contents are
the responsibilities of the authors and do not necessarily reflect the views handle/10665/44703/9789241548335_eng.pdf. Accessed
of USAID or the United States Government. August 1, 2019.
11. Payne BA, Hutcheon JA, Ansermino JM, et al.; miniPIERS Study
Competing Interests: None declared. Working Group. A risk prediction model for the assessment and tri-
age of women with hypertensive disorders of pregnancy in low-
resourced settings: the miniPIERS (Pre-eclampsia Integrated Estimate
REFERENCES of RiSk) multi-country prospective cohort study. PLoS Med. 2014;11
1. Rawlins B, Plotkin M, Rakotovao JP, et al. Screening and manage- (1):e1001589. CrossRef. Medline
ment of pre-eclampsia and eclampsia in antenatal and labor and
12. Ahmed SM, Alam BB, Anwar I, et al. Bangladesh Health System
delivery services: findings from cross-sectional observation studies in
Review. Vol. 5 No. 3. Manila: World Health Organization,
six sub-Saharan African countries. BMC Pregnancy Childbirth.
Regional Office for the Western Pacific; 2015. http://www.searo.
2018;18(1):346. CrossRef. Medline
who.int/entity/asia_pacific_observatory/publications/hits/hit_
2. Kinney MV, Smith JM, Doherty T, Hermida J, Daniels K, Belizán JM. bangladesh/en/. Accessed August 1, 2019.
Feasibility of community level interventions for pre-eclampsia: per-
13. Ahmed SM, Hossain MA, RajaChowdhury AM, Bhuiya AU. The
spectives, knowledge and task-sharing from Nigeria, Mozambique,
health workforce crisis in Bangladesh: shortage, inappropriate skill-
Pakistan and India. Reprod Health. 2016;13(1):125. CrossRef.
mix and inequitable distribution. Hum Resour Health. 2011;9(1):3.
Medline
CrossRef. Medline
3. Goldenberg R, Jones B, Griffin J, et al. Reducing maternal mortality
from preeclampsia and eclampsia in low-resource countries—what 14. Human Resources Management (HRM) Unit, Ministry of Health and
should work? Acta Obstet Gynecol Scand. 2015;94(2):148–155. Family Welfare (MOHFW). HRH Datasheet-2014. Dhaka,
CrossRef. Medline Bangladesh: MOHFW; 2015. http://www.mohfw.gov.bd/index.
php?option=com_docman&task=doc_download&gid=7306&lang=
4. Salam RA, Das JK, Ali A, Bhaumik S, Lassi ZS. Diagnosis and man-
en. Accessed July 5, 2019.
agement of preeclampsia in community settings in low and middle-
income countries. J Family Med Prim Care. 2015;4(4):501–506. 15. Ministry of Health and Family Welfare (MOHFW). Bangladesh
CrossRef. Medline Essential Health Service Package. Dhaka, Bangladesh:
MOHFW;2016. http://www.mohfw.gov.bd/index.php?option=
5. National Institute of Population Research and Training (NIPORT),
com_docman&task=doc_download&gid=9484&lang=en.
International Centre for Diarrhoeal Disease Research, Bangladesh
Accessed July 5, 2019.
(icddr,b), and MEASURE Evaluation. Bangladesh Maternal Mortality
and Health Care Survey 2016: Preliminary Report. Dhaka, 16. Talukder MN, Rob U, Khan AKMZU, Noor FR, Roy S, Noor AF.
Bangladesh, and Chapel Hill, NC: NIPORT, icddr,b, and MEASURE Union Health and Family Welfare Centers in Chittagong and
Evaluation; 2017. https://www.measureevaluation.org/resources/ Munshiganj: Are They Ready to Provide 24-Hour Normal Delivery
publications/tr-17-218. Acessed August 13, 2019. Services? Dhaka: Population Council; 2015. https://www.
6. Warren C, Hossain S, Ara Nur R, Sultana K, Kirk KR, Dempsey A. popcouncil.org/uploads/pdfs/2015RH_UHFWCs-Chittagong
Landscape Analysis on Pre-eclampsia and Eclampsia in Bangladesh. Munshiganj.pdf. Accessed August 1, 2019.
Washington, DC: Population Council; 2015. http://www. 17. O’Hanley K, Kim T, Tell K, Langer A. Balancing the Scales,
endingeclampsia.org/wp-content/uploads/2017/04/FINAL- Expanding Treatment for Pregnant Women with Life Threatening
LandcapeAnalysis-Bangladesh-USAID-GA31-EndingEclampsia- Hypertensive Conditions in Developing Countries. New York:
Mar2016.pdf. Accessed August 1, 2019. EngenderHealth; 2007. https://www.engenderhealth.org/wp-
7. Biswas A, Anderson R, Doraiswamy S, et al. Timely referral saves the content/uploads/imports/files/pubs/maternal-health/
lives of mothers and newborns: Midwifery led continuum of care in engenderhealth-eclampsia-report.pdf. Accessed August 13, 2019.
marginalized teagarden communities—a qualitative case study in 18. Sultana K, Dempsey A. Landscape Report on Pre-eclampsia and
Bangladesh. F1000 Res. 2018;7:365. CrossRef. Medline Eclampsia in Bangladesh. Washington, DC: Population Council,
8. Firoz T, Sanghvi H, Merialdi M, von Dadelszen P. Pre-eclampsia in 2017. http://www.endingeclampsia.org/wp-content/uploads/
low and middle income countries. Best Pract Res Clin Obstet 2017/08/Providers_Bangladesh_FINAL.pdf. Accessed August 1,
Gynaecol. 2011;25(4):537–548. CrossRef. Medline 2019.

Peer Reviewed

Received: April 9, 2019; Accepted: July 20, 2019; First Published Online: September 16, 2019

Cite this article as: Williams A, Khan MA, Moniruzzaman M, et al. Management of preeclampsia, severe preeclampsia, and eclampsia at primary care
facilities in Bangladesh. Glob Health Sci Pract. 2019;7(3):457-468. https://doi.org/10.9745/GHSP-D-19-00124

© Williams et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly cited. To view a
copy of the license, visit http://creativecommons.org/licenses/by/4.0/. When linking to this article, please use the following permanent link: https://
doi.org/10.9745/GHSP-D-19-00124

Global Health: Science and Practice 2019 | Volume 7 | Number 3 468