Table 1. Selected model parameters: Incidence and mortality of pregnancy and delivery-related complications, and impact of interventions.
| Parameter | Hemorrhage | Obstructed Labor | Hypertensive Disorders | Sepsis | Unsafe Abortion |
| Incidence and mortality | |||||
| Probability of event [15],[29],[31]–[44] | 0.114 | 0.047 | 0.035 | 0.050 | 0.128a |
| Range | 0.051–0.228 | 0.030–0.074 | 0.025–0.050 | 0.043–0.060 | 0.050–0.250 |
| Probability of morbidity [28],[30],[45]–[48] b | 0.008 | 0.022 | 0.001 | 0.400 | 0.120 |
| Range | 0.006–0.010 | 0.018–0.026 | 0.001–0.001 | 0.320–0.480 | 0.096–0.144 |
| CFR [7],[49]–[51] | 0.010 | 0.007 | 0.017 | 0.013 | 0.003 |
| Adjusted CFRc | 0.023 | 0.019 | 0.021 | 0.028 | 0.009 |
| Range | 0.007–0.030 | 0.005–0.025 | 0.012–0.027 | 0.009–0.036 | 0.002–0.012 |
| Attributable mortality [29] d | 46.2% (9%–73%) | 14.1% (3%–52%) | 13.7% (0%–18%) | 17.4% (0%–20%) | 8.6% (0%–20%) |
| Model-projected attributable mortality | 40.6% | 16.8% | 12.3% | 20.4% | 9.8% |
| Impact of interventions | |||||
| Decreased incidence [34],[37] e | 50%,75%e | — | — | 25%, 50%f | — |
| Rangef | 25%–91% | — | 25%–50% | 0%–60% | 0%–100% |
| Decreased CFR [7],[17],[50],[52]–[64] g | 75% | 95% | 59% | 90% | 98%a |
| Rangef | 60%–90% | 76%–100% | 45%–95% | 63%–93% | 50%–100% |
Incidence of elective abortion is 0.170, and 75% are assumed to be unsafe in the base case [15],[29],[39]–[44] . Case fatality rate (CFR) of safe abortion is 0.000006, representing a 98% reduction in mortality [50],[62]–[64]. For more details on abortion-related assumptions, see Text S1. Incidence of miscarriage (not shown) is 0.150 [69],[70].
Specific examples of nonfatal complications include Sheehan's Syndrome following maternal hemorrhage, fistula resulting from obstructed labor, neurological sequelae from eclampsia, pelvic inflammatory disease (PID). Not shown but included are the risk of infertility from PID (0.086) and the risk of severe anemia following maternal hemorrhage (0.09) [23],[45],[46].
CFRs were adjusted based on complication severity (e.g., life threatening complications requiring cEmOC) and underlying severity of anemia [71]. See Text S1.
Estimates for distribution of causes of maternal mortality for India are from India overall estimates from Khan et al. [29], based on the entire Asia region, as well as other data to establish a range for sensitivity analysis [14],[15]. Cause-specific proportions sum to 66%, reflecting approximately 33% indirect causes, although this varies from 15% to 35% in different studies. Estimates shown reflect adjustment of data from Khan et al. [29] such that a distribution is shown for the 66% of direct causes, to compare to model output. Further, anemia was reported to be responsible for 15% of deaths and was assumed to exert mortality impact on direct causes through severity of PPH, sepsis, and unsafe abortion.
Incidence of sepsis reduced by 50% with SBA and clean delivery in birthing center, bEmOC, and cEmOC; and reduced by 25% with SBA and clean delivery at home [37]. Incidence of maternal hemorrhage reduced by 50%–75% depending on expectant versus active management of labor; we assume for the status quo, all cEmOC facilities provide active management, 50% of bEmOC facilities provide active management, and birthing centers/health centers provide expectant management only [34]. Exploratory analyses that estimate the impact of community-based provision of oral misoprostol in birthing centers and at home assume a 25% to 50% reduction in PPH [22],[72]. For each baseline estimate, sensitivity analysis was conducted across a plausible range based on literature review; references are documented in the Text S1.
For each baseline estimate, sensitivity analysis was conducted across a plausible range based on literature review; references and assumptions are documented in the Text S1.
Estimates shown represent average reduction in case fatality rate provided complications necessitating surgery (e.g., cesarean section), blood transfusion, intensive hemodynamic support are treated in cEmOC. Obstructed labor is managed using assisted vaginal delivery with forceps or vacuum and, if necessary, cesarean section; severe pre-eclampsia and eclampsia treated with intravenous hydralazine and magnesium sulfate, in addition to induction of labor or emergency cesarean section when required; sepsis treated with ampicillin, gentamycin, and metronidazole or equivalent regimen followed by an 8-d course of intramuscular gentamycin and oral metronidazole (see Text S1 for details) [7],[73].