A Retrospective Analysis To Determine If Obstructive Sleep Apnea During Pregnancy is a Predictor of Length of Stay in Women Who Delivered in a Military Treatment Facility

Abstract

Background Trends in hospital stays from 1997 to 2013 indicated that childbirth is the most common reason for hospitalization among women of childbearing age. The average maternal length of stay (LOS) increased dramatically following delivery of a baby from 0.1 percent during 2003-2008 to 1.2 percent during 2008-2012. The exact cause(s) for this rise in the average LOS is unknown but it is likely a complex and multi-factorial one. A plausible explanation may be the increased frequency and severity of several maternal factors that contributed to unanticipated delivery-related outcomes resulting in a prolonged LOS following delivery of a baby. In two large retrospective cross-sectional studies of obstructive sleep apnea (OSA) during pregnancy, the authors reported women who had been diagnosed with OSA were more likely to be older and have an extended LOS following delivery of their baby. Purpose and Research Aims The primary aim of this study was to determine if OSA was an independent predictor of LOS and secondarily to determine what maternal and fetal clinical outcomes were predictors of LOS following delivery of their baby among women who delivered at a military treatment facility (MTF). Methods A retrospective cohort study of all women (N = 305,001) who gave birth at a MTF from 2008 to 2014. Multivariate logistic regression models between LOS and OSA and selected maternal comorbidities, adverse pregnancy outcomes, maternal, clinical conditions, and fetal outcomes were built. Cases were identified using ICD-9 codes while controlling for demographics and adjusting for confounders but not adjusting for intermediate variables according to the study’s conceptual model. Conceptual Model A biological model of sleep disturbance and inflammation served as the study’s conceptual model. Findings The presence of OSA increased the odds of a prolonged LOS by nearly 2.5 times. All maternal, pregnancy-related outcomes, cesarean delivery, gestational hypertension, gestational diabetes, preeclampsia, postoperative wound; maternal clinical conditions, acute renal failure and pulmonary edema and the fetal outcome, preterm delivery, were predictors of a prolonged LOS. Implications for Research Research must be ongoing to elucidate the gender differences of OSA in women and to develop a female-specific screening instrument.