The BetterBirth Study collected more than 204 million quantitative data points and a wealth of qualitative data over the course of the study, from November 2014 to August 2017. While many of our methods have been described in depth in multiple publications, this appendix serves as a brief reference on all data sources referred to in this report, both published and unpublished.
Seven-day maternal mortality, perinatal mortality, and self-reported severe maternal morbidity outcomes
Number of participants or facilities
Total sample of 157,689 enrolled women with births across 120 facilities (60 intervention and 60 comparison facilities). Of these, final seven-day outcomes were obtained for 157,145 cases.
Description
All women who were registered for childbirth at a BetterBirth facility were included in the sample, excluding those who were referred in, those who were managed for abortion, or those who did not consent to participate. Outcomes were obtained either through a follow-up call or in-home visit from the eighth day to the 43rd day after delivery.
Direct observation of childbirth care
Number of participants or facilities
- Baseline: 10 facilities (five intervention, five control); 603 births observed at one or more pause points
- After two months of coaching: 30 facilities (15 intervention, 15 control); 2,563 births observed at one or more pause points
- After six months of coaching: 10 facilities (five intervention, five control); 897 births observed at one or more pause points
- Four months after the end of the intervention: 30 facilities (15 intervention, 15 control); 2,325 births observed at one or more pause points
Description
Independent observers (trained nurses) recorded birth attendants’ adherence to birth practices using a standardized data-collection form and included 60 observations at each of the following observation points: 1) on admission, 2) just before pushing, 3) one minute after delivery, 4) within one hour of delivery—for a total of 240 observations per facility for each round of data collection. Birth practices at the time of discharge were not observed, because of the uncertain nature of time and place of discharge in these facilities. Data were collected only during daytime hours because of security concerns.
Maternal death qualitative review
Number of participants or facilities
98 out of 149 maternal deaths in the study
Description
Of the 149 maternal deaths that occurred in the study, routine follow-up call recordings with families for 98 cases were located, analyzed, and had complete data related to circumstances surrounding the location of a woman’s death. We performed qualitative coding of the recordings and created summary statistics for coded variables.
Perinatal verbal autopsy
Number of participants or facilities
Of the 7,445 perinatal deaths, we identified 161 perinatal deaths (106 early neonatal deaths, 55 stillbirths) where the delivery was observed (i.e., we had data on adherence to Safe Childbirth Checklist practices). Due to twins, the total number of interviews with mothers was 158.
Description
We conducted in-person verbal and social autopsy interviews with these families to document the circumstances of the perinatal death. Descriptive statistics were generated for type and cause of death and referrals.
Birth attendant interviews
Number of participants or facilities
A total of 33 interviews were conducted in 12 intervention facilities (four low-mortality sites, four medium-mortality sites, four high-mortality sites).
Description
We conducted in-depth semi-structured interviews with birth attendants at selected facilities after the completion of the intervention to glean experiences with the Checklist.
Medical officer in charge/childbirth quality champion interviews
Number of participants or facilities
A total of 47 interviews with the medical officer in charge and the childbirth quality champion(s) were conducted in 20 intervention facilities.
Description
Semi-structured interviews were completed at intervention facilities that had recently phased out of the study. Sites were chosen based on coach and coach team leader feedback, so interviews could be evenly distributed across high- and low-performing facilities. Interviews were conducted with both the head of facility (medical officer in charge or medical superintendent) and the nominated childbirth quality champion. If a facility had more than one childbirth quality champion, all were asked to participate.
BetterBirth staff focus groups
Number of participants or facilities
- Direct observers’ focus group: one focus group discussion, seven participants
- Coach focus group: one focus group discussion, six participants
Description
Informal focus groups were held with BetterBirth Study staff. The discussions focused on study staff perceptions of successes, challenges, and other significant experiences regarding the BetterBirth intervention and study (both to understand quality improvement from an implementation perspective, and to understand the context of the intervention within the study facilities).
Patient Satisfaction Survey
Number of participants or facilities
149,268 patients (75,540 intervention and 73,728 control) answered the satisfaction questions.
Description
As part of the follow-up phone call or visit to assess seven-day outcomes, women were additionally asked two satisfaction-related questions concerning the care received during childbirth.
Checklist Utilization Survey
Number of participants or facilities
969 healthcare workers were surveyed.
Description
Anonymous surveys were conducted to capture health care worker opinions about ease of usability of the Checklist and inclination of health care workers to adopt the Checklist as part of their practice. Each intervention facility administered the survey three times: eight weeks after the start of the intervention, and then six and 12 months after the initial survey.
Cost effectiveness analysis: Time motion
Number of participants or facilities
1,559 measurements from 10 intervention facilities were observed.
Description
Observations were done to obtain granular data on how much time it takes to complete Checklist-related activities.
Cost effectiveness analysis: Time use
Number of participants or facilities
61 surveys of staff nurses from 15 intervention facilities were conducted.
Description
A survey was conducted to capture health-care worker perception of time spent with the WHO Safe Childbirth Checklist in relation to other clinical duties.
Cost effectiveness analysis: Work sampling
Number of participants or facilities
More than 1,300 hours of observations from 20 intervention and control facilities were collected.
Description
Observations were done to obtain a snapshot of how health workers’ time during their work shifts is divided among activities in which they are directly interacting with patients, engaging in direct patient care activities but not interacting with patients, and engaging in work activities that do not pertain to specific patients.