Abstract：
BACKGROUND: While Malawi has made great strides increasing the number of facility-based births, maternal and neonatal mortality remains high. An intervention started in 2019 provided short-course training followed by year-long longitudinal bedside mentorship for nurse midwives at seven health facilities in Blantyre district. The intervention was initiated following invitation from the district to improve outcomes for patients during childbirth. This study examined the impact of the intervention on the reporting of obstetric and neonatal complications and related care.
METHODS: Patient level data were collected from the District Health Information System 2 database from intervention and non-intervention facilities. Bivariate analysis explored the impact of longitudinal bedside mentorship on select District Health Information System 2 variables at six-month intervals. Outcomes were then analyzed using nonlinear quantile mixed models to better account for the impact of time and clustering at the facility level.
RESULTS: Significant changes were found in the reporting of obstetric and neonatal complications over time at intervention facilities compared to non-intervention facilities. Intervention facilities showed statistically significant increases in the reporting of prolonged labor, pre/eclampsia, fetal distress, retained placenta, and premature labor. There was also a statistically significant decrease in the reporting of no complications in the multivariate model (95%CI: -0.8 to -0.2). In both the bivariate and multivariate models, the reporting of \'None\' significantly decreased (0.8 % median), while the reporting of prematurity (0.2 % median) and asphyxia (0.3 % median) both significantly increased. The missingness of data at intervention facilities decreased to almost zero compared to non-intervention facilities.
CONCLUSIONS: The increase in reported maternal and neonatal complications suggests improved early identification of complications at the facility level. The improved accuracy of patient data from intervention facilities shows the impact mentorship has on data quality which is crucial for the allocation of resources. By highlighting the apparent dose-response relationship of longitudinal bedside mentorship, this study will inform the broader use of mentorship in training programs. Future research is needed to explore the impact of longitudinal mentorship on quality of care.
METHODS: Patient level data were collected from the District Health Information System 2 database from intervention and non-intervention facilities. Bivariate analysis explored the impact of longitudinal bedside mentorship on select District Health Information System 2 variables at six-month intervals. Outcomes were then analyzed using nonlinear quantile mixed models to better account for the impact of time and clustering at the facility level.
RESULTS: Significant changes were found in the reporting of obstetric and neonatal complications over time at intervention facilities compared to non-intervention facilities. Intervention facilities showed statistically significant increases in the reporting of prolonged labor, pre/eclampsia, fetal distress, retained placenta, and premature labor. There was also a statistically significant decrease in the reporting of no complications in the multivariate model (95%CI: -0.8 to -0.2). In both the bivariate and multivariate models, the reporting of \'None\' significantly decreased (0.8 % median), while the reporting of prematurity (0.2 % median) and asphyxia (0.3 % median) both significantly increased. The missingness of data at intervention facilities decreased to almost zero compared to non-intervention facilities.
CONCLUSIONS: The increase in reported maternal and neonatal complications suggests improved early identification of complications at the facility level. The improved accuracy of patient data from intervention facilities shows the impact mentorship has on data quality which is crucial for the allocation of resources. By highlighting the apparent dose-response relationship of longitudinal bedside mentorship, this study will inform the broader use of mentorship in training programs. Future research is needed to explore the impact of longitudinal mentorship on quality of care.
摘要：
背景：虽然马拉维在增加设施出生人数方面取得了长足的进步，孕产妇和新生儿死亡率仍然很高。2019年开始的干预措施为布兰太尔地区的七个医疗机构的护士助产士提供了短期培训，然后进行了为期一年的纵向床边指导。干预是在该地区的邀请下启动的，目的是改善患者在分娩期间的结局。这项研究检查了干预措施对产科和新生儿并发症报告及相关护理的影响。
方法：患者水平的数据是从干预和非干预机构的地区卫生信息系统2数据库中收集的。双变量分析以六个月的间隔探讨了纵向床边指导对选定的地区健康信息系统2变量的影响。然后使用非线性分位数混合模型分析结果，以更好地说明时间和设施级别聚类的影响。
结果：与非干预机构相比，干预机构的产科和新生儿并发症报告随时间发生了显著变化。干预设施显示，长期分娩的报告在统计上显着增加，先兆/子痫，胎儿窘迫,保留胎盘,早产。多变量模型中无并发症的报告也有统计学上的显著下降(95CI:-0.8至-0.2)。在双变量和多变量模型中，“无”的报告显着下降（中位数为0.8%），而早产(中位数为0.2%)和窒息(中位数为0.3%)的报告均显著增加.与非干预设施相比，干预设施的数据缺失率几乎为零。
结论：报告的孕产妇和新生儿并发症的增加表明在机构层面对并发症的早期识别有所改善。来自干预设施的患者数据的准确性提高表明，指导对数据质量的影响对于资源分配至关重要。通过强调纵向床边指导的明显剂量反应关系，这项研究将为在培训计划中更广泛地使用导师制提供信息。未来的研究需要探索纵向指导对护理质量的影响。
方法：患者水平的数据是从干预和非干预机构的地区卫生信息系统2数据库中收集的。双变量分析以六个月的间隔探讨了纵向床边指导对选定的地区健康信息系统2变量的影响。然后使用非线性分位数混合模型分析结果，以更好地说明时间和设施级别聚类的影响。
结果：与非干预机构相比，干预机构的产科和新生儿并发症报告随时间发生了显著变化。干预设施显示，长期分娩的报告在统计上显着增加，先兆/子痫，胎儿窘迫,保留胎盘,早产。多变量模型中无并发症的报告也有统计学上的显著下降(95CI:-0.8至-0.2)。在双变量和多变量模型中，“无”的报告显着下降（中位数为0.8%），而早产(中位数为0.2%)和窒息(中位数为0.3%)的报告均显著增加.与非干预设施相比，干预设施的数据缺失率几乎为零。
结论：报告的孕产妇和新生儿并发症的增加表明在机构层面对并发症的早期识别有所改善。来自干预设施的患者数据的准确性提高表明，指导对数据质量的影响对于资源分配至关重要。通过强调纵向床边指导的明显剂量反应关系，这项研究将为在培训计划中更广泛地使用导师制提供信息。未来的研究需要探索纵向指导对护理质量的影响。
