. 2017 Sep 14;12(9):e0184432. doi: 10.1371/journal.pone.0184432

Table 3. Effect estimates of multivariate association adjusted for affordability, education and need or perceived need for skilled care for childbirth (study details available in S2 Table. Summary of 57 included studies).

Study	Settings	(Reference)
Distance to the nearest facility
De Allegri et al. 2011	Rural Burkina Faso	≤5km	>5km	0.035^**
Johnson et al. 2015	Rural Ghana	≤8km	>8km	0.74 (0.62–0.88)^**
Lwelamira and Safari 2012	Rural Tanzania	<5km	5-10km	0.87 (0.73–1.04)	>10km	0.62 (0.47–0.81)^*
Nakua et al. 2015	Rural Ghana	≤5km	6-10km	0.32 (0.13–0.74)^*	11-15km	0.40 (0.11–1.46)
Mageda et la. 2015	Rural Tanzania	<5km	5-10km	0.43 (0.5–1.7)	≥10km	0.43 (0.3–0.8)^**
O’Meara et al. 2014	Rural Kenya (multi-site)
	- Butula	Every one kilometre increment: 1.33 (1.00–1.69)^*
	- Bunyala	Every one kilometre increment: 0.80 (0.60–1.07)
	- Teso North	Every one kilometre increment: 1.14 (0.79–1.64)
	- Bungoma East	Every one kilometre increment: 1.18 (0.93–1.51)
Ndao-Brunblay et al. 2014	Rural Tanzania	Every one kilometre increment: 0.89 (p-value = 0.085)
Distance to health facilities equipped to provide skilled care for childbirth
Mpembeni et al. 2007	Rural Tanzania	0-5km	6+km	0.25 (0.16–0.37)^***
De Allegri et al. 2015	Rural Burkina Faso	≤6km	7km	(0.01–0.30)^*
Moran et al. 2006	Rural Burkina Faso	<22.8km	≥22.8km	0.39 (0.202–0.759)^*
Mills et al. 2008	Urban and rural Ghana	0-9km	10-19km	0.54 (0.37–0.79)^*	20+km	0.31 (0.23–0.43)^*
Magadi et al. 2000	Urban and rural Kenya	<5km	5-10km	0.47^*	>10km	0.38^*
Gage 2007	Rural Mali	<1km	1-4km	0.526 (0.277–1.001)	5-9km	0.491 (0.277–0.871)^*	10-14km	0.418 (0.212–0.825)^*	15-29km	0.403 (0.209–0.779)^***	30+km	0.623 (0.262–1.480)
Okafor 1991	Rural Nigeria	Every one kilometre increment: -0.097^*
Lohela et al. 2012	Rural Malawi	Every one^$ kilometre increment: 0.90 (0.87–0.93)^***
Kruk et al. 2015	Rural Tanzania	Every one kilometre increment to dispensary (equipped to provide maternity care): 0.93 (0.84–1.04)
		Every one kilometre increment to primary health clinics (equipped to provide maternity care): 1.07 (0.97–1.19)
		Every one kilometre increment to hospital (higher-level and provide maternity care): 0.40 (0.26–0.63)^***
Hounton et al. 2008	Rural Kenya	Every one kilometre increment to health centre (lower-level and usually led by a nurse) for <7.5km: 0.77 (0.75–0.79)^***
		Every one kilometre increment to health centre (lower-level and usually led by a nurse) for ≥7.5km: 0.97 (0.95–0.98)^***
		Every one^$ kilometre increment to hospital (higher-level and the main source of surgical care): 0.97 (0.97–0.99)^***
Joharifard et al. 2012	Rural Rwanda	Every one kilometre increment (up to 14): 0.909 (0.608–1.907) INSIGNIFICANT
Kitui et al. 2013	Urban and rural Kenya	Adjusted effect of distance to the nearest HF offering maternity care is insignificant (results not presented in original study) INSIGNIFICANT
Anyait et al. 2012	Mostly rural Uganda	Adjusted effect of distance to the nearest HF offering maternity care is insignificant (results not presented in original study) INSIGNIFICANT
Nuwaha and Amooti-kaguna 1999	Mostly rural Uganda	Adjusted effect of distance to the nearest HF offering maternity care is insignificant (results not presented in original study) INSIGNIFICANT
Travel time
Wado et al. 2013⁺	Urban and rural Ethiopia	<60min	>60min	0.55 (0.34–0.89)^*
Hailu et al. 2014⁺	Urban and rural Ethiopia	<60min	≥60min	0.3 (0.11–0.87)^*
Gebru et al. 2014⁺	Ethiopia	≤60min	>60min	0.249 (0.143–0.434)^**
Abikar et al. 2013	Kenya	<60min	>60min	0.26 (0.08–0.81)^*
Van Eijk et al. 2006⁺	Rural Kenya	<60min	60min	0.58 (0.33–1.05)	>60min	0.36 (0.18–0.75)^*
Spangler and Bloom 2010⁺	Rural Tanzania	<30min	30-60min	0.45 (0.31–0.64)^***	≥60min	0.26 (0.18–0.38)^***
Kawakatse et al. 2014⁺	Rural Kenya	<20min	21-40min	0.547 (0.536–0.558)	41-60min	0.533 (0.515–0.554)	>60min	0.403 (0.282–0.573)^***
Masters et al. 2013⁺⁺	Rural Ghana	Every one hour increment: 0.801 (0.69–0.93)^***
Teferra et al. 2012⁺	Urban and rural Ethiopia	Adjusted effect of walking time is insignificant (results not presented in original study) INSIGNIFICANT
Amano et al. 2012⁺	Urban and rural Ethiopia	Adjusted effect of walking time is insignificant (results not presented in original study) INSIGNIFICANT
Nuwaha and Amooti-kaguna 1999⁺	Mostly rural Uganda	Adjusted effect of motorized travel time is insignificant (results not presented in original study) INSIGNIFICANT

Reverse associations shown if original results were presented with greater distance or travel time as reference category or unskilled care for childbirth as the outcome of interest.

^$ Association shown was converted from a 10-km increment by raising the original estimate to the power of 0.1.

⁺ Walking time;

⁺⁺ Motorized travel time;

* p<0.05;

** p<0.01;

*** p<0.001