Maintaining rigor in research: flaws in a recent study and a reanalysis of the relationship between state abortion laws and maternal mortality in Mexico

Abstract

A recent publication [Koch E, Chireau M, Pliego F, Stanford J, Haddad S, Calhoun B, Aracena P, Bravo M, Gatica S, Thorp J. Abortion legislation, maternal healthcare, fertility, female literacy, sanitation, violence against women and maternal deaths: a natural experiment in 32 Mexican states. BMJ Open 2015;5(2):e006013] claimed that Mexican states with more restrictive abortion laws had lower levels of maternal mortality. Our s are to replicate the analysis, reanalyze the data and offer a critique of the key flaws of the Koch study. STUDY DESIGN: We used corrected maternal mortality data (2006-2013), live births, and state-level indicators of poverty. We replicate the published analysis. We then reclassified state-level exposure to abortion on demand based on actual availability of abortion (Mexico City versus the other 31 states) and test the association of abortion access and the maternal mortality ratio (MMR) using descriptives over time, pooled chi-square tests and regression models. We included 256 state-year observations. RESULTS: We did not find significant differences in MMR between Mexico City (MMR=49.1) and the 31 states (MMR=44.6; p=.44). Using Koch's classification of states, we replicated published differences of higher MMR where abortion is more available. We found a significant, negative association between MMR and availability of abortion in the same multivariable models as Koch, but using our state classification (beta=-22.49, 95% CI=-38.9; -5.99). State-level poverty remains highly correlated with MMR. CONCLUSION: Koch makes errors in methodology and interpretation, making false causal claims about abortion law and MMR. MMR is falling most rapidly in Mexico City, but our main study limitation is an inability to draw causal inference about abortion law or access and maternal mortality. We need rigorous evidence about the health impacts of increasing access to safe abortion worldwide. IMPLICATIONS: Transparency and integrity in research is crucial, as well as perhaps even more in politically contested topics such as abortion. Rigorous evidence about the health impacts of increasing access to safe abortion worldwide is needed.