The study was part of a larger project approved by the Ministry of Health and Family Welfare and funded by the Swedish International Development Cooperation Agency (SIDA) with ID BGD10MWC. A mixed-methods case study format was used to describe the design and implementation of the telemedicine program. Qualitative analysis comprised document review, key informant interviews, and focus group discussions. Quantitative analysis employed descriptive statistics to compare service use trends before and after implementation.
Program documentation consisted of internal and external monitoring reports. Eighteen interviews were conducted with different types of program implementers. Ten participants were national and district-level managers and eight were midwives providing remote services. Two focus group discussions and eight additional interviews were held with pregnant and post-partum service recipients. Interviews and focus groups followed a standard questionnaire guide designed to gather information about the implementation process, including challenges encountered, solutions, and feelings and experiences with the program. Interviews and focus groups were held in Bangla, with notes kept in Bangla of each session. Content analysis drew out common themes across the qualitative data and provided context to the program documentation and service statistics.
Quantitative data were sourced from the national health information system (DHIS2) [11] and a separate Google Sheet used for tracking telemedicine service data. Telemedicine service statistics reported numbers of client visits for ANC, facility birth, PNC, FP and GBV screening, identification and referral. They were compared with DHIS2 statistics from the hospitals that introduced the telemedicine program both before and after the pandemic. Trends and differences were identified using data visualization tools to highlight notable patterns.
Program Description
Background
The telemedicine intervention arose from a partnership between the Bangladesh Directorate General of Nursing and Midwifery (DGNM) and UNFPA. The partnership supports the government in its introduction of an International Confederation of Midwives (ICM) standard midwifery cadre into the Bangladesh health system. Support to this new cadre of globally standard midwives had been ongoing for seven years prior to the pandemic with multi-donor funding, and by early 2020 had educated 6,116 diploma midwives and deployed 1,149 midwives into 342 sub-district hospitals. Most midwives have low-middle class backgrounds and all completed higher secondary education in Bangla medium schools prior to joining the three-year midwifery diploma program. Sub-district hospitals are public secondary-level care facilities that serve the lowest wealth quintile populations.
COVID-19 Telemedicine Guidance
The COVID-19 Pandemic UNFPA Global Response Plan provided the initial impetus for the telemedicine program. Under its umbrella, a Technical Brief for Maternity Services with addendum guidance documents for ANC and PNC were released in April and May of 2020. The guidance documents were designed to support the continuation of maternal health services following pandemic-associated service disruptions.
One key element of the Brief was the inclusion of ANC and PNC using remote contact where possible during the COVID-19 pandemic. Eight ANC visits are recommended according to WHO guidelines with the 1st, 4th, 6th and 8th visits in person and the others via phone or video contact. Figure 1 below shows the recommended schedule of ANC visits and care checklist for remote visits. The guide also included specific per visit clinical guidance.
Intervention Design
Initial discussions were held between UNFPA, DGNM, the Directorate General of Health Services and implementing partners about the program. Pandemic-related funding was sought and obtained from SIDA for implementation in five targeted districts: Dhaka, Cox’s Bazar, Bandarban, Noakhali and Sunamganj. Due to reports of increasing GBV during pandemic lockdowns [9, 10], it was determined that GBV screening and referral should be incorporated into the service. An orientation session was held to sensitize key stakeholders comprising district public health officers, project mentoring and monitoring officers, nursing supervisors, midwives, and hospital statisticians who would be involved in implementation. The early discussions drew attention to the lack of an existing system for scheduling client visits. Prior to this, clients determined when they visited facilities for ANC based on providers’ suggestions; subsequent visit appointments were not recorded at facilities. Only a handful of facilities used ANC cards in which visit appointments were documented. It was initially thought that clients would receive midwives’ mobile phone numbers and be invited to call if needed. While other telemedicine programs in Bangladesh are set up that way, later design discussions resulted in the determination that appointment scheduling, led by midwives, was necessary and in line with the UNFPA guidance. This decision initiated a significant change in standard procedure. Version 1 of a scheduling tool was developed and in November 2020 a three-day training of trainers was held to begin roll out.
The training brought together 90 midwives (20 in-person and 70 virtually) as well as nursing supervisors, district officers who oversee midwifery, and was led by the DGNM with support from UNFPA. It provided detailed guidance on organizing and carrying out telemedicine services, including GBV screening and referral. The GBV protocol had four components: 1) observation of signs and symptoms, 2) probing questions to determine a. whether a woman can safely speak freely and b. to create an opportunity for disclosure, 3) guidance for providing first line counseling and clinical services, and 4) making an appropriate referral following a referral guide. A smart phone was provided to each facility for midwives to share with monthly charges paid to facilitate phone calls and data entry for reporting (see Figure 3 for data flow). Using the visit schedule and checklist in Figure 1 and standard GBV protocol, midwives were instructed to follow these steps:
- At the first onsite (face-to-face) visit, determine the gestational age
- Based on the gestational age, decide when a phone call will be and schedule it within one month, one week or 2-5 days based on the gestational age
- Write patient's name and phone number in the scheduling tool for the agreed upon day
- Check the scheduling tool daily to identify who needs to receive a call
- Phone calls to be performed on the day they appear on the scheduling tool
- Midwives use the checklist and GBV protocol to guide phone calls
- At the end of each phone call, schedule the next on-site visit
Though midwives were instructed to attempt to reach women multiple times as needed, there was no specific guidance on how many calls should be attempted before considering the visit a no show.
Roll Out
In December 2020, telemedicine services were introduced in 13 of 36 (36%) sub-district hospitals in the five districts. Initially, all 36 hospitals in the selected districts were to adopt the intervention. This was reduced to 13 that had three or more midwives, the minimum staffing level needed to absorb the work. At the facility level, nursing supervisors oversaw quality. A supervision system was also established with monthly visits from mentors funded through UNFPA and embedded within district health offices. During these visits, mentors would review documentation of client visits, and discuss progress with the onsite midwives. Cross district supervision comprised meetings that served as a platform for discussing challenges and lessons learned. In July 2021, once the other 23 hospitals had at least 3 midwives employed onsite, they adopted the intervention as well. An additional 200 midwives were trained as part of the expansion.
Monitoring and Reporting
Consideration was given to whether telemedicine visits should be recorded in the usual register book. They were ultimately documented separately within the scheduling tool, capturing whether a planned visit occurred, the ANC/PNC visit number, and other services provided (e.g., FP or GBV). In-person visits were documented in the existing ANC register. This meant that telemedicine visits were not tracked in the government’s health information system. Instead, they were reported into a common Google sheet as a temporary measure until the government data system could include them. Ongoing program monitoring revealed the need to redesign the scheduling tool to an easier to manage format. Version 2 of the tool, designed as a standard appointment book, was well received as there was more space for recording scheduled appointments. Both versions of the tool are depicted in Figure 2.
Learnings
Qualitative
Qualitative analysis revealed both challenges and successes. Global guidance documents did not include considerations for facilities that needed to introduce a scheduling system. Yet, early conversations revealed that it was necessary to introduce one. This was a significant practice adjustment that had a learning curve. Once introduced, some midwives found it difficult to keep to the appointment schedule when the patient load was high. There were also challenges reaching patients at times who did not answer their phone or who had given their husband’s or mother-in-law’s phone number. Some women who were not directly reachable had restrictions on their access to phone use for ANC services. In addition, women who were reachable sometimes were reluctant to communicate about GBV when asked the screening questions. Successes included that the program enabled relinking lost to follow-up women to services, and facilitated critical check-ins with high-risk pregnant mothers. In addition, both midwives and pregnant women acknowledged that the program prevented unnecessary hospital visits and improved the provider-client relationship. It also raised awareness about pregnancy risks and midwives’ services, and encouraged facility births.
Quantitative
Maternity care service use data were compared to examine change over time from before the pandemic to after a year of telemedicine implementation. Total numbers of women seen at the 36 intervention facilities (both in-person and remotely) were looked at in 2019, 2020 and 2021. Across 31 of the 36 facilities, most services increased from pre-pandemic levels. For ANC, percent increase varied by visit number as follows: ANC 1 (31%), ANC 2 (28%), ANC 3 (18%), ANC 4 (25%) (Figure 4). Vaginal births increased from pre-pandemic levels by 19%, while PNC 1 increased by 20% and PNC 2 saw a 1% decrease (Figure 5). There was also a doubling of identified cases of postpartum hemorrhage and eclampsia (Figure 6). Five facilities were excluded as a result of the number of employed midwives at those facilities reducing from six to three during the implementation period, and subsequent declines in service use that affected the overall results. In addition, while GBV service data showed that GBV screening occurred during 88% of telemedicine services, fewer than 100 cases were identified and referred.
These results show that (apart from PNC 2) service utilization surpassed not just the declines experienced at the onset of the COVID-19 pandemic, but also the pre-pandemic levels that reflected system-wide gaps.