Perinatal outcomes in Agincourt: 1995-2000
Duworko, James Tanu
Objective: The objective is to estimate the magnitude and determinants of perinatal mortality in Agincourt, and determine whether there is a difference in perinatal mortality rate between South Africans and self-settled Mozambicans. Design: Case-control study of 134 cases and 136 controls using longitudinal data drawn from the Agincourt dataset for the period 1995-2000 Methods: All cases were matched against a random selection of 136 controls. Odds ratios were used to assess risk, with p-values for trend where necessary. Logistic regression was used to determine independent effects of significant risk factors. Limitations of the study: Probable under-reporting of stillbirths and early neonatal deaths. Results: The Agincourt perinatal mortality rate is estimated as 13.4 per 1000 births (95%CI, 11.23- 15.8) with an increasing trend from 1995-2000 (X2 for trend 19.487, p-value <0.001). Delivery by a nurse attendant is a protective factor but not independently so. Multivariate analysis indicates that babies of women who never attended antenatal clinic during the index pregnancy are at higher risk of perinatal death (OR= 7.55; 95%CI, 2.03-28.05) compared to others whose mothers attended antenatal clinic at least four times. Women with history of perinatal death are at a higher risk of experiencing it again, compared with those without (OR =13.68; 95%CI, 1.43-130.82). The difference in perinatal mortality rate for South Africans (13.3) and former Mozambican refugees (11.8) is not statistically significant (p-value = 0.522). Conclusion: Perinatal mortality is rising; key risk factors are non-attendance for antenatal care by mothers, and previous perinatal death. There is no significant difference in perinatal mortality rate between South Africans and self-settled Mozambicans in Agincourt.
A research report submitted to the Faculty of Health Sciences, University of the Witwatersrand in partial fulfillment of the requirements for the degree of Master of Science in Medicine (Epidemiology and Biostatistics) Johannesburg, May 2014