Investigation of Predictors of Newborn Screening Refusal in a Large Birth Cohort in North Dakota, USA.

Objectives The objective of this study was to identify maternal and provider predictors of newborn screening (NBS) refusal in North Dakota between 2011 and 2014. Methods Records of 40,440 live resident births occurring in North Dakota between 2011 and 2014 were obtained from the North Dakota Department of Health and included in the study. Factor-specific percentages of NBS refusals and 95% confidence intervals were computed for each predictor. Since the outcome is rare, multivariable Firth logistic regression was used to investigate maternal and provider predictors of NBS refusal. Model goodness-of-fit test was evaluated using the Hosmer-Lemeshow test. All analyses were conducted in SAS 9.4. Results Of the 40,440 live births, 135 (0.33%) were NBS refusals. 97% of the refusals were to white women, 94% were homebirths, and 93% utilized state non-credentialed birth attendants. The odds of NBS refusals were significantly higher among non-credentialed birth attendants (p < 0.0001), homebirths (p < 0.0001), and among those that refused Hepatitis B vaccination (HBV) at birth (p = 0.047). On the other hand, odds of NBS refusals were significantly (p < 0.0001) lower among women that had more prenatal visits. Conclusions for Practice This study provides preliminary evidence of association between NBS refusal and provider type, home births, and HBV refusal. Additional studies of obstetric providers, home births and women are needed to improve our understanding of the reasons for NBS refusal to better deliver preventive services to newborns.