Between November 2004 and September 2011, Li et al. (2014) conducted a study to determine the risk factors for mortality rates in expectant women in the Republic of Tanzania. The study involved a prospective cohort study in HIV care and treatment clinic settings in Dar es Salaam city. Data were collected for all women registered and visiting those health facilities. Logistic regression was applied in the determination of mortality risk factors among the subjects. It predicted that an observation fell into one of the two dichotomous dependent variables based on one or more independent variables. In the study, the dependent variable was mortality, while independent variables were pregnancy and HIV status. In 2004, there were 18,917 expectant mothers with 13% of them having multiple pregnancies.
During the study period, 363 maternal deaths occurred giving a ratio of 1,729 per 100,000 live births. The findings were in contrast to an initial report by the World Health Organization of a mortality rate of 210 maternal deaths for every 100,000 live births. The study found that women who were on ARVs before their pregnancies had decreased mortality rates of 55% at confidence intervals of 0.29-0.70 (Li et al., 2014). Mothers who had been taking ARVs for a longer period had a lower risk of mortality (about 8%) for every month they took the drugs. The study concluded that mortality rates were high in expectant mothers, and there was a need to initiate antiretroviral therapy early enough and provide them with nutritional interventions.
This study was well formulated with respect to the study design and analysis as was evidenced by differentiating patients based on ARV use. Moreover, the subjects were closely monitored for a specified period to determine how their variations affected the mortality rates. A major drawback of the study was the cost and workload involved in making follow-ups on many study subjects. Logistic regression, in this case, was appropriate since the responses from the expectant mothers were not continuous but discrete for each mother.
From the analysis of this study, it can be concluded that regression and correlation are paramount in public health research involving complex relationships, for example, qualitative and quantitative measurements.
Reference
Li, N., Matchi, E., Spiegelman, D., Chalamilla, G., Hertzmank, E., Sando, D., & Fawzi, W. (2014). Maternal mortality among HIV‐infected pregnant women in Tanzania. Acta Obstetricia et Gynecologica Scandinavica, 93(5), 463-468.