253: Predicting length of treatment for neonatal abstinence syndrome in methadone exposed neonates

Abstract

ObjectiveTo identify maternal variables predicting length of treatment for neonatal abstinence syndrome (NAS) in mehtadone exposed neonates.Study designRetrospective cohort study of neonates treated for NAS 2000-2006 whose mothers were on methadone maintenance at delivery. Square root transformation was used for length of treatment to normalize data. Mixed effects linear regression was used to model association between candidate variables and square root of length of treatment.Results177 neonates were identified. In univariate analysis the following variables did not predict length of treatment: weeks maternal methadone use (p=0.067), trimester of initial exposure (p=0.366), methadone dose at delivery (p=0.324), BMI (p=0.080), antidepressant use (p=0.431), benzodiazepine use (p=0.204), cigarette use (p=0.809), polysubstance use (p=0.185). The multivariate model considered the association between number of weeks exposed, trimester of initial maternal exposure, dose at delivery, gestational age at delivery and variables with p<0.15. In multivariate analysis gestational age at delivery was a significant predictor of the length of treatment (Table).Tabled 1Days of Neonatal Treatment (multivariate analysis)Adj. Meanp-valueNo. Weeks Exposed<1022.3.80011-2028.621-3029.8>3032.0Trimester Exposed127.0.851229.9327.2Dose at Delivery (mg)<5025.9.47151-10029.6101-15031.0151-20023.4>20030.7GA at Delivery23-3215.5<0.00133-3629.937-4242.0 Open table in a new tab ConclusionThe major predictor of length of treatment for NAS was gestational age at delivery: earlier gestational age required shorter treatment. ObjectiveTo identify maternal variables predicting length of treatment for neonatal abstinence syndrome (NAS) in mehtadone exposed neonates. To identify maternal variables predicting length of treatment for neonatal abstinence syndrome (NAS) in mehtadone exposed neonates. Study designRetrospective cohort study of neonates treated for NAS 2000-2006 whose mothers were on methadone maintenance at delivery. Square root transformation was used for length of treatment to normalize data. Mixed effects linear regression was used to model association between candidate variables and square root of length of treatment. Retrospective cohort study of neonates treated for NAS 2000-2006 whose mothers were on methadone maintenance at delivery. Square root transformation was used for length of treatment to normalize data. Mixed effects linear regression was used to model association between candidate variables and square root of length of treatment. Results177 neonates were identified. In univariate analysis the following variables did not predict length of treatment: weeks maternal methadone use (p=0.067), trimester of initial exposure (p=0.366), methadone dose at delivery (p=0.324), BMI (p=0.080), antidepressant use (p=0.431), benzodiazepine use (p=0.204), cigarette use (p=0.809), polysubstance use (p=0.185). The multivariate model considered the association between number of weeks exposed, trimester of initial maternal exposure, dose at delivery, gestational age at delivery and variables with p<0.15. In multivariate analysis gestational age at delivery was a significant predictor of the length of treatment (Table).Tabled 1Days of Neonatal Treatment (multivariate analysis)Adj. Meanp-valueNo. Weeks Exposed<1022.3.80011-2028.621-3029.8>3032.0Trimester Exposed127.0.851229.9327.2Dose at Delivery (mg)<5025.9.47151-10029.6101-15031.0151-20023.4>20030.7GA at Delivery23-3215.5<0.00133-3629.937-4242.0 Open table in a new tab 177 neonates were identified. In univariate analysis the following variables did not predict length of treatment: weeks maternal methadone use (p=0.067), trimester of initial exposure (p=0.366), methadone dose at delivery (p=0.324), BMI (p=0.080), antidepressant use (p=0.431), benzodiazepine use (p=0.204), cigarette use (p=0.809), polysubstance use (p=0.185). The multivariate model considered the association between number of weeks exposed, trimester of initial maternal exposure, dose at delivery, gestational age at delivery and variables with p<0.15. In multivariate analysis gestational age at delivery was a significant predictor of the length of treatment (Table). ConclusionThe major predictor of length of treatment for NAS was gestational age at delivery: earlier gestational age required shorter treatment. The major predictor of length of treatment for NAS was gestational age at delivery: earlier gestational age required shorter treatment.