Factors associated with maternal survival in a low-resource setting

Maternal mortality is a global concern and a consensus has been reached that the health of mothers and children is an important indicator of national health and the socio-economic development of countries. In 2015 the reduction of maternal mortality was adopted in the global and national initiatives for improving maternal health under Sustainable Development Goals(SDG). Despite the commitment set out in SDGs maternal mortality remains high, Zimbabwe is among the 40 countries in the world with high maternal mortality rate. There was need to analyse the causes for such high maternal mortalityhence the need for this research.The objective of the study was to determine the factors associated with maternal survival of pregnant mothers at Mpilo Central Hospital. The results of the study could help reduce maternal mortality. Data were entered into Stata12.1 statistical package. Univariate statistics were performed and presented as frequencies and percentages for categorical variables. We used Variance Inflated Factor (VIF) to test for multi-collinearity for all the candidate explanatory variables. Binary logistic regression was used to calculate the probability of maternal death given different variables. A p value < 0.05 was taken as statistically significant. 95% Confidence Interval (CI) was used. The Cox proportional hazard model was used on the factors which were found to be statistically significant in maternal mortality to analyze if they have an effect on the survival time of patients. The p-value of our model was <0.0001 which identified that the model fits the data and can be used to predict survival. Kaplain Meir survival curves were used to compare survival of patients with respect to their booking status. Using binary logistic regression, the following were statistically significant associated with maternal death: age (OR 3.434, 95% CI 1.508-7.819, p=0.003), education level (OR 0.114, 95% CI 0.018-0.724, p=0.021), booking status (OR 29.547, 95% CI 8.016-111.290, p=0.001), ANC visits (OR 2.549, 95% CI 1.732-3.750, p = 0:001), PPH (0R 3.302, 95% CI 1.291-8.447, p=0.013), PIH(OR 0.010, 95% CI 0.001-0.112, p=0.001), APH (OR 3.941, 95% CI 1.371-11.329, p=0.011), sepsis (OR 5.358, 95% CI 1.792-16.022, p=0.003), retroviral infection (OR 8.466, 95% CI 1.921-37.320, p=0.005), anaemia (OR 5.647, 95% CI 1.491-21.383, p=0.011), other mode of deliveries which includes vacuum and forceps (OR 21.751, 95% CI 2.305-205.276, p=0.007) and miscarriages (OR 6.995, 95% CI 1.813-26.987, p=0.005). Using Cox regression, the following were found to be statistically significant; patients who did not book their pregnancies (HR 5.196, 95% CI 2.342-11.527, p=0.001), PPH (HR 1.790, 95% CI 1.097-2.919, p=0.020), PIH (HR 0.071, 95% CI 0.010-0.515, p=0.009), APH (HR 2.153, 95% CI 1.345-3.447, p=0.001) and ANC visits (HR 1.379, 95% CI 1.039-1.739, p=0.006)

孕产妇死亡率（Maternal mortality）是全球备受关注的公共卫生议题，学界已达成共识：母婴健康是衡量国家卫生水平与各国社会经济发展的重要指标。2015年，降低孕产妇死亡率被纳入可持续发展目标（Sustainable Development Goals, SDG）框架下的全球及各国改善孕产妇健康行动之中。
尽管可持续发展目标已作出相关承诺，但孕产妇死亡率仍居高不下，津巴布韦是全球40个高孕产妇死亡率负担国家之一。明确此类高孕产妇死亡率的成因具有重要现实意义，本研究由此开展。
本研究旨在明确姆皮洛中心医院（Mpilo Central Hospital）孕妇的孕产妇生存状况相关影响因素，研究结果或可为降低孕产妇死亡率提供决策参考。
研究数据录入Stata12.1统计分析软件。对分类变量采用单变量统计分析，以频数及百分比进行结果呈现。使用方差膨胀因子（Variance Inflated Factor, VIF）对所有候选解释变量进行多重共线性检验。采用二分类Logistic回归计算不同变量下的孕产妇死亡概率，以p值<0.05作为统计学显著性的判定标准，并采用95%置信区间（Confidence Interval, CI）进行区间估计。
针对在孕产妇死亡率分析中筛选出的具有统计学显著性的影响因素，本研究采用Cox比例风险模型分析其对患者生存时间的影响。本模型的p值<0.0001，表明模型拟合度良好，可用于患者生存预测。采用Kaplan-Meier生存曲线对比不同产检建档状态下患者的生存情况。
通过二分类Logistic回归分析，以下因素与孕产妇死亡具有统计学显著性关联：年龄（优势比OR=3.434，95%CI：1.508~7.819，p=0.003）、受教育程度（OR=0.114，95%CI：0.018~0.724，p=0.021）、产检建档状态（OR=29.547，95%CI：8.016~111.290，p=0.001）、产前保健（Antenatal Care, ANC）随访次数（OR=2.549，95%CI：1.732~3.750，p=0.001）、产后出血（Postpartum Hemorrhage, PPH）（OR=3.302，95%CI：1.291~8.447，p=0.013）、妊娠高血压疾病（Pregnancy-Induced Hypertension, PIH）（OR=0.010，95%CI：0.001~0.112，p=0.001）、产前出血（Antepartum Hemorrhage, APH）（OR=3.941，95%CI：1.371~11.329，p=0.011）、败血症（OR=5.358，95%CI：1.792~16.022，p=0.003）、逆转录病毒感染（OR=8.466，95%CI：1.921~37.320，p=0.005）、贫血（OR=5.647，95%CI：1.491~21.383，p=0.011）、包括吸引产与产钳助产在内的其他分娩方式（OR=21.751，95%CI：2.305~205.276，p=0.007）以及流产史（OR=6.995，95%CI：1.813~26.987，p=0.005）。
通过Cox回归分析，以下因素具有统计学显著性：未进行孕期建档的患者（风险比HR=5.196，95%CI：2.342~11.527，p=0.001）、产后出血（HR=1.790，95%CI：1.097~2.919，p=0.020）、妊娠高血压疾病（HR=0.071，95%CI：0.010~0.515，p=0.009）、产前出血（HR=2.153，95%CI：1.345~3.447，p=0.001）以及产前保健随访次数（HR=1.379，95%CI：1.039~1.739，p=0.006）。