Last September, I reviewed 2 studies on treatment of opioid withdrawal in newborns. The study today is a different type, looking at outcomes of maternal substance abuse utilizing administrative databases. It is perhaps not as "sexy" as the treatment studies, but no less important in dealing with this expanding health crisis.
Source: Hwang SS, Diop H, Liu C-l, et al. Maternal substance use disorders and infant outcomes in the first year of life among Massachusetts singletons, 2003-2010. J Pediatr 2017;191:69-75; doi.10.1016/j.jpeds.2017.08.045. See AAP Grand Rounds commentary by Dr. Benjamin Doolittle (subscription required).
In the study introduction, the authors point out an important difference in epidemiologic study design that is worth mention. Their study utilized 2 databases, the Massachusetts Pregnancy to Early Life (PELL) database and the Massachusetts Bureau of Substance Abuse Services (BSAS) Management Information System. PELL links birth and death records to maternal and infant hospital records, and the BSAS administers the state's substance abuse disorder prevention, treatment, and policy programs. For this study, the 2 databases were linked although unfortunately only through the year 2010, so the data presented are somewhat old. The study looked at the years 2003-2010.
Contrast this study with another type of epidemiologic study, a cross-sectional analysis. This study design, which the authors state is the predominant study type so far addressing maternal substance abuse effects on newborns, looks at a limited point in time and is less useful for looking at outcomes that may develop over time. (Of course, one could repeat the same cross-sectional study at different points of time, but that's a bit labor intensive and takes a long time to complete.) Having access to longitudinal data is more likely to inform future interventional (both preventive and therapeutic) studies by identifying useful outcome measurements and study populations.
In the current study, the unsurprising conclusions were that these children have poorer health outcomes both in the perinatal period and in the first year of life. The researchers acknowledge many limitations of the study, perhaps most importantly that their databases were not detailed enough to determine the quantity of drug exposure in mothers, not just for opioids but for other harmful drugs of abuse such as alcohol. So, attributing maternal opioid abuse as causal for infant outcomes becomes even more difficult than usual for an observational study such as this.
