Modeling the Risk Factors of Miscarriage Using Survival Analysis Techniques

Background: Miscarriage, also known as spontaneous abortion, is a significant adverse outcome of pregnancy. The risk factors associated with the transition from a normal pregnancy to a complete miscarriage before 28 weeks of gestational age have not been exhaustively established. Use of logistic regression to assess the factors associated with spontaneous abortion excludes the longitudinal and incompleteness aspects of miscarriage data. However, miscarriage is dynamical process where time until it occurs may be of interest.

Objectives: This paper modeled the risk factors associated with miscarriage using survival analysis, estimated and compared survivorship of levels of categorized variables, fitted proportional hazards and accelerated failure time models and compared their inferential capacity and their efficacy in identifying risk factors.

Methods: A retrospective study was conducted for pregnant women who were enrolled for antenatal care in Kakamega County General Teaching and Referral Hospital (KCGTRH) in Kakamega county, western Kenya. Pregnant women with recognized pregnancy and enrolled in the pre-natal care during the period from 1 January, 2019 up to 31 October, 2020 was recruited into the study. For descriptive analysis and estimation of survival functions, the study used Kaplan-Meier (K-M) and chi-squared test for independence. Comparison of survivorships in categorized variables was done using Kaplan-Meier curves and log-rank test. The Cox proportional hazards (PH) model and parametric models were used to analyze miscarriage data. All analysis were carried out using R software and SPSSv20. The level of significance was 5%.

Results: Of the total sample 248 mothers (4.1%) miscarried, while 5729 (95.9%) were censored. The significant factors identified by log rank test were ethnicity (P=.000), levels of education (P=.048), place of residence (P=.000), employment status (P=.004), malaria status (P=.000) and UTI status (P=.000). The covariates in categorized form found significant by log rank were number of previous stillbirths (P=.000) and number of ANC visits (P=.000). The factors ethnicity, place of residence, malaria status, number of previous miscarriages, number of previous stillbirths and number of ANC visits were identified as the risk factors associated with miscarriages using cox model, parametric proportional hazards model and accelerated failure time models. The study found equivalent hazard ratios for among Cox model, parametric proportional hazards (PH) models and accelerated failure time (AFT) models.

Conclusion: From the findings of this study it can be concluded that the Gompertz proportional hazards (PH) regression model demonstrates a more favorable level of conformity to the data in comparison to the Cox model and accelerated failure time (AFT) models and there is association between certain explanatory variables and the time to miscarriages. In this paper survival analysis was used to analyze miscarriage data.