The All Babies Count Initiative: Impact of a Health System Improvement Approach on Neonatal Care and Outcomes in Rwanda

Hema Magge; Evrard Nahimana; Jean Claude Mugunga; Fulgence Nkikabahizi; Elisabeth Tadiri; Felix Sayinzoga; Anatole Manzi; Merab Nyishime; Francois Biziyaremye; Hari Iyer; Bethany Hedt-Gauthier; Lisa R. Hirschhorn

doi:10.9745/GHSP-D-20-00031

Figures & Tables

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
FIGURE.
All Babies Count Initiative Design Implemented in 2 Districts in Rwanda
Abbreviations: ABC, All Babies Count; LS, learning session; MOH, Ministry of Health; QI, quality improvement.

View popup

TABLE 1.

Selection of Maternal and Newborn Health Core Improvement Collaborative Indicators and Outcome Measures Used in Evaluating an All Babies Count Initiative Implemented in Rwanda

	Indicator	Numerator	Denominator	Data Source	Facility
Antenatal care^a	1. ANC standard 4 visit coverage	Number of women who delivered in a HF who had 4 ANC visits	Total number of health facility deliveries PLUS total number of referrals of laboring women from HC to hospital PLUS home deliveries	HMIS	HCs
Delivery services	2. Percentage of pregnant women with facility delivery	Total number of health facility deliveries	Total number of home deliveries PLUS total number of health facility deliveries	Community HMIS	HCs
	3. Time to cesarean delivery for emergency (from determination of need at hospital to time of delivery)	Average time to cesarean delivery for emergency (from determination of need by doctor at hospital to time of cesarean delivery incision)		Chart review^b	Hospital
	4. Antenatal steroids for preterm labor	Number of women with preterm labor <34 weeks treated with dexamethasone	Number of women with preterm labor (<34 weeks)	Chart review	Hospital (HCs added later)
Postnatal care	5. Percentage of babies with immediate skin-to-skin after delivery	Number of babies placed immediately skin to skin	Number of babies born vaginally	Chart review	HCs and hospital
Postnatal care	6. Danger sign assessment within 24 hours	Newborns checked for danger signs in postpartum ward within 24 hours	Total deliveries MINUS stillbirths macerated MINUS stillbirths fresh MINUS death at birth of live born babies	HMIS	HC and hospitals
Outcome	7. Neonatal unit case fatality	Number of deaths in neonatal unit (<28 days)	Number of admissions to neonatal unit (<28 days)	Chart review	Hospital
	8. District-wide neonatal mortality	HMIS neonatal deaths PLUS death at birth of live born babies PLUS number of neonatal deaths at community	Live births PLUS number of home deliveries	HMIS	HCs and hospital
	9. Facility neonatal mortality among preterm/LBW^c	Hospital deaths due to prematurity PLUS HC deaths due to prematurity	Hospital LBW/non-preterm PLUS hospital preterm PLUS HC LBW/non-preterm PLUS HC preterm	HMIS	HCs and hospital
	10. Facility neonatal mortality among non-preterm/LBW^d	Hospital deaths of all causes except prematurity PLUS HC deaths of all causes except prematurity	Hospital and HC live births MINUS (hospital LBW/non-preterm PLUS hospital preterm PLUS HC LBW/non-preterm PLUS HC preterm)	HMIS	HCs and hospitals
	11. Facility stillbirths (macerated and fresh)^c	Stillbirths macerated PLUS stillbirths fresh	Total deliveries	HMIS	HCs and hospital

Abbreviations: ANC, antenatal care; HC, health center; HF, health facility; HMIS, health management information system; LBW, low birth weight.
↵a Rwinkwavu health center did not have delivery services, so the denominator was changed to number of community HMIS health facility deliveries at district hospital plus number of community HMIS home deliveries.
↵b Chart review indicators were newly introduced at the start of the intervention.
↵c Prematurity defined as gestational age ≤37 weeks, and low birth weight defined as birth weight < 2500 g per World Health Organization standard definition.
↵d Stillbirths defined in national HMIS data dictionary as a baby born with no signs of life at or after 22 weeks gestation and with birth weight greater than or equal to 500 g. Stillbirth analysis restricted to facility level given the absence of community-based recording of stillbirths.

View popup

TABLE 2.

Change in Maternal and Newborn Health Quality of Care Indicators in All Babies Count Initiative Implemented in 2 Districts in Rwanda

Indicator	Aggregate				South Kayonza			Kirehe
	Baseline	Endpoint	P Value	Median Difference (IQR)	Baseline	Endpoint	P Value	Baseline	Endpoint	P Value
	Median (IQR)	Median (IQR)	P Value	Median Difference (IQR)	Median (IQR)	Median (IQR)	P Value	Median (IQR)	Median (IQR)	P Value
1. Percentage of deliveries where mothers had 4 standard ANC visits	13.7 (6.7, 44.2)	30.4 (15.1, 43.1)	.04	13.6 (−3.3, 26.4)	42.4 (19.9, 47.3)	36.4 (13.5, 66.3)	.78	10.2 (2.8–26.8)	29.6 (18–34.2)	.02
2. Percentage of pregnant women delivering in facilities	89.6 (86.3, 94.9)	92.6 (86.8, 95.8)	.25	7.8 (−2.8, 5.1)	92.2 (87.2, 95.3)	95.2 (93.0, 96.3)	.33	88.4 (86.3–96.7)	91.0 (86.2–94.6)	.36
3. Percentage of babies who are provided immediate skin-to-skin after birth	53.6 (0, 80.9)	97.4 (96.4, 99.3)	<.001	43.6 (17.7, 95.7)	91.2 (75.9, 100)	99.6 (96.7, 100)	.18	4.3 (0, 53.6)	97.2 (96.1, 98.8)	<.001
4. Percentage of newborns checked for danger signs within 24 hours of birth	46.6 (31.1, 96.7)	98.7 (96.4, 100)	<.001	47.7 (−1.4, 67.1)	52.3 (33.8, 98.6)	100 (93.8, 100)	.06	45.7 (15.8, 82.1)	98.6 (97.4, 98.8)	.003
5. Average hospital time to emergency cesarean delivery (minutes)					167	50	—	82	61	—
6. Percentage of women with preterm labor who are treated with antenatal steroids	0 (0, 0)	41.7 (0, 100)	.32	—	0 (0, 0)	16.7 (0, 33.3)	—	0 (0–0)	75 (0, 100)	.32
7. Percentage of facilities with at least 2 MNH clinically trained staff	100	100	—	15.1 (4, 29.8)	100	100	—	100	100	—
8. Percent availability of essential medications for MNH care	61.2 (45.0, 77.8)	81.8 (72.7, 81.8)	<.001	31.2 (19.5, 37.8)	35.0 (25.0, 45.0)	83.3 (75.0, 87.5)	.01	66.7 (55.6, 77.8)	77.8 (72.7, 81.8)	.05
9. Percent availability of functioning equipment essential for MNH care	55.6 (48.2, 61.1)	86.6 (77.8, 88.9)	<.001	—	55.6 (47.2, 56.6)	88.9 (88.9, 94.4)	.01	55.6 (48.2, 63.9)	81.1 (77.8, 88.6)	<.001
10. Patient satisfaction^a: average satisfaction with ANC	—	—	—	—	2.8 (SD: 1.57)	3.1 (SD: 1.53)	.11	3.3 (SD: 1.62)	3.6 (SD: 1.52)	.01
11. Patient satisfaction: average satisfaction with maternity care	—	—	—	—	2.4 (SD: 1.67)	2.3 (SD: 1.160	.95	3.5 (SD: 1.59)	3.4 (SD: 1.52)	.61

Abbreviations: ANC, antenatal care; IQR, interquartile range; MNH, maternal and newborn health; SD, standard deviation.
↵a Patient satisfaction scores on a Likert scale: 1=excellent; 2=very good; 3=good; 4=fair; 5=poor.

View popup

TABLE 3.

Change in Neonatal Mortality Across 2 Districts in Rwanda Where All Babies Count Initiative Was Implemented

Indicator	Baseline	Endpoint	aOR (95% CI)^a
Neonatal mortality (deaths/1,000 live births)
Aggregate	30.1 (103/3,426)	19.6 (57/2,902)	0.54 (0.36, 0.83)
Southern Kayonza	35.4 (55/1,553)	18.5 (24/1,295)
Kirehe	25.6 (48/1,873)	20.5 (33/1,607)
Facility neonatal deaths in preterm infants/1,000 preterm and LBW live births
Aggregate	198.8 (32/161)	100.6 (18/179)	0.47 (0.25, 0.90)
Southern Kayonza	290.3 (18/62)	134.3 (9/67)
Kirehe	141.4 (14/99)	80.4 (9/112)
Facility neonatal deaths in non-preterm babies/1,000 non-preterm and LBW live births in district
Aggregate	10.4 (36/3,446)	7.5 (24/3,181)	0.60 (0.36, 1.02)
Southern Kayonza	7.9 (11/1,387)	5.9 (7/1,177)
Kirehe	12.1 (25/2,059)	8.5 (17/2,004)
Facility stillborn rate (macerated and fresh)/total per 1,000 births
Aggregate	23.4 (84/3,590)	28.8 (99/3,436)
Southern Kayonza	20.7 (29/1,398)	34.5 (44/1,274)	1.71 (1.06, 2.75)
Kirehe	25.1 (55/2,192)	25.4 (55/2,162)	0.90 (0.61, 1.32)

Abbreviations: aOR, adjusted odds ratio; CI, confidence interval; LBW, low birth weight.
↵a Odds estimates from models stratified by district only provided if test for interaction between district and time in regression was statistically significant at 0.05 level.

View popup

TABLE 4.

Change in Neonatal Mortality in All Babies Count Intervention Area in 2 Districts Compared With the Rest of Rural Rwanda

	Pre-intervention (Deaths/1,000 Live Births)	Post-intervention (Deaths/1,000 Live Births)	Per 1,000 Change
(1) Southern Kayonza/Kirehe (ABC)	30.1 (103/3,426)	19.6 (57/2,902)	−10.4
Southern Kayonza	35.4 (55/1,553)	18.5 (24/1,295)
Kirehe	25.6 (48/1,873)	20.5 (33/1,607)
(2) Southern Kayonza/Kirehe (HMIS)	22.3 (87/3,896)	15.6 (59/3,785)	−6.7
Southern Kayonza	28.7	25.3
Kirehe	18.3	10.7
(3) Rural Rwanda comparison districts^a	13.4 (834/62,382)	15.9 (958/60,225)	2.5
	Difference
Difference 1 − 3	−13.0
Difference 2 − 3	−9.2

Abbreviations: ABC, All Babies Count; HMIS, health management information system.
↵a Bugesera, Gakenke, Gatsibo, Gicumbi, Gisagara, Huye, Kamonyi, Karongi, Muhanga, Musanze, Ngoma, Ngororero, Nyabihu, Nyagatare, Nyamagabe, Nyamasheke, Nyanza, Nyaruguru, Rubavu, Ruhango, Rulindo, Rusizi, Rutsiro, Rwamagana Districts HMIS reported data.