Background
The WHO Maternal Near Miss (MNM) approach was developed to evaluate and improve quality of obstetric care worldwide. This study aimed to study the incidence of MNM and quality of care at a district hospital in rural Rwanda by applying this approach.

Methods
A facility based, prospective cohort study conducted at a district hospital in rural Rwanda between June 2013 and December 2014. Subjects were followed from time of admission to discharge or death.

Results
In 3979 deliveries, 3827 singletons and 152 twins pairs were born. Among the 4131 neonates, there were 3994 live births and 137 stillbirths. Ninety-nine women suffered severe maternal outcome (SMO): 86 maternal near misses and 13 deaths. This adds up to a maternal near miss ratio of 21.5 per 1000 live births (95 % CI 17.3-26.5), a maternal mortality ratio of 325 per 100 000 live births (95 % CI 181–543) and a mortality index of 13.1 % (95 % CI 7.3–21.9). Hemorrhage (n = 49, 57 %) and hypertensive disorders (n = 27, 31.4 %) were the commonest MNM conditions. Eclampsia (n = 4/13; 30.7 %) was the leading cause of maternal mortality, while sepsis/peritonitis following cesarean section (n = 2/6; 33.3 %) had the highest mortality index. Seventy-seven out of 99 SMO cases (77.9 %) were referred from other facilities with critical conditions and 28 out of 99 SMO cases (28.3 %) were admitted into the Intensive Care Unit. Several indicators such as administration of oxytocin, magnesium sulfate and antibiotics were found to be suboptimal.

Conclusions
MNM is common at district level in Rwanda. The MNM approach enabled us to identify shortfalls in clinical practice and the referral system.