Association between conflict and usage of maternal health services in Egypt: an uncontrolled before and after study

Background: United Nations’ data indicate that conflict-affected low- and middle-income countries contribute considerably to global maternal deaths. Usage patterns during conflict have not been examined in-depth and a cause-effect analysis is relevant for policy insights. This study assessed the association between the acute Egyptian conflict of 2011-2012 and maternal services usage and quality. Methods: An uncontrolled before-and-after study used data from the 2014 Egypt Demographic and Health Survey. The hierarchical nature of data was addressed using multi-level modelling. The pre-conflict sample included 2,569 births occurring from January 2009 to January 2011 and the peri-conflict had 4,641 births from February 2011 to December 2012. Results: After adjusting for potential socioeconomic confounders, conflict was not associated with antenatal service usage. Compared to pre-conflict births, peri-conflict births had lower odds of at least 24 hours post-delivery stay (8%; OR 0.92), delivery in public institutions (1%; OR 0.99), and institutional postnatal care (PNC) (1%; OR 0.99); and relatively higher odds of doctor-assisted deliveries (2%; OR 1.02), institutional deliveries (2%; OR 1.02), private institutional deliveries (3%; OR 1.03), and doctor-assisted PNC (2%; OR 1.02). During conflict, odds of doctor-assisted delivery increased more for oldest (>35 years) women (3%; OR 1.03); odds of receiving PNC decreased more for women aged 25-29 (8%; OR 0.92); and odds of skilled birth attendance increased more for rural women (2%; OR 1.02), primary-educated (4%; OR 1.04), employed (4%; OR 1.04), less poor (3%; OR 1.03) and middle household wealth quintiles (2%; OR 1.02). Odds of physician-assisted delivery increased more for rural women (3%; OR 1.03), primary-educated (5%; OR 1.05), employed (4%; OR 1.04), less poor and middle household wealth quintiles (3%; OR 1.03 each), and richest household wealth quintiles (2%; OR 1.02). Conclusions: Conflict-associated changes in maternal care were minimal, generally reflecting pre-conflict usage patterns. Association between acute conflict and maternal care indicated vertical equity had changed from recent trends. Policy intervention may help improving post-delivery stay. Maternal policy requires ‘specific in-built equity strategies’ to address unpredictable effect of conflict.