Background : Maternal health remains challenging for many parts of the world. It is estimated in addition to 529,000 maternal deaths worldwide, 9.5 million women had pregnancy-related morbidity and 1.4 million of them survived life-threatening complications (so called obstetric near-miss). In general, obstetric near-miss cases go through similar life-threatening events as in maternal death cases. The large incidence of near-miss cases provides the opportunity to investigate maternal health problem through obstetric near-miss to complement maternal death investigation which requires a large sample.

Objective : This analysis was aimed to describe the characteristics of near-miss cases.

Methods : This is a retrospective secondary analysis of maternity admissions (pregnancy, delivery, and post-partum) in 2005-2006 collected from two public hospitals in Banten Province, Indonesia (n=1,806). Maternal deaths were excluded from our analysis. The near-miss definition was determined by an expert panel who reviewed the criteria used in previous studies and adjusted to the local context.

Results : About 67% of near-miss cases were near-miss at admission which showed that these cases arrived at the hospitals with life-threatening conditions. The majority of the near-miss cases came from rural/remote areas (65.8%), were uninsured or covered by government/worker compensation insurance (65%), and assisted by traditional birth attendants before being taken to hospital (38.6%). Those characteristics were significantly higher in near-miss than non-near-miss cases (p<0.05). Also, a higher proportion of near-miss cases were grand multiparity (23.9%) and had non-obstetric complications (16.1%) compared to non-near-miss (12.6% and 1.3%, respectively). Women from the rural/remote areas and the low-economic group in Indonesia were still disproportionately affected by maternal health problem. The analysis supports the evidence that characteristics of near-miss cases are consistent with those of maternal death cases in general, suggesting the feasibility to study near-miss cases as a proxy of maternal death.