Table 4.
Estimates for the risk of neonatal abstinence syndrome (NAS) based on the duration of prescription opioid use during pregnancy, Medicaid data 2000-07
| Population | Total NAS cases/sample size | Absolute NAS risks per 1000 deliveries (95% CI) based on duration of prescription opioid use | Relative risks (95% CI) for long term v short term prescription opioid use | |||||
|---|---|---|---|---|---|---|---|---|
| Long term* | Short term* | Long term | Short term | Unadjusted | Adjusted† | |||
| Whole cohort | 810/34 111 | 895/256 494 | 23.7 (22.1 to 25.4) | 3.5 (3.3 to 3.7) | 6.81 (6.19 to 7.48) | 2.05 (1.81 to 2.33) | ||
| Study groups based on hierarchical NAS risk factors‡ | ||||||||
| Group 1: Documented history of opioid drug misuse/dependence | 471/2139 | 514/2677 | 220.2 (200.8 to 241) | 192.0 (175.8 to 209.3) | 1.15 (1.03 to 1.28) | 1.26 (1.09 to 1.45) | ||
| Group 2: Documented history of alcohol/non-opioid drug misuse | 153/4973 | 166/23 580 | 30.8 (26.1 to 36) | 7.0 (6.0 to 8.2) | 4.37 (3.52 to 5.43) | 4.90 (3.34 to 7.2) | ||
| Group 3: Exposure to other psychotropic medications§ in the third trimester | 92/7027 | 56/27 696 | 13.1 (10.6 to 16.1) | 2.0 (1.5 to 2.6) | 6.48 (4.65 to 9.02) | 3.81 (2.44 to 5.95) | ||
| Group 4: Documented smoking history | 26/3966 | 36/24 546 | 6.6 (4.3 to 9.6) | 1.5 (1 to 2) | 4.47 (2.7 to 7.39) | 3.71 (1.61 to 8.55) | ||
| Group 5: No history ofour above listed f risk factors | 68/16 006 | 123/177 995 | 4.2 (3.3 to 5.4) | 0.7 (0.6 to 0.8) | 6.15 (4.57 to 8.26) | 5.67 (3.07 to 10.47) | ||
NAS=neonatal abstinence syndrome, CI=confidence interval.
*At least 30 days of prescription opioids dispensed during pregnancy was defined as long term use and <30 days as short term use.
†Propensity score (PS) matching of long term users with short term users was used to adjust for confounding factors including age, geographic region, race, pain condition diagnoses, calendar year of delivery, and other NAS risk factors. Risk ratios derived using generalized estimating equations with log link and exchangeable working correlation matrix to account for clustering owing to inclusion of mothers with multiple deliveries in the cohort.
‡Five mutually exclusive groups were created hierarchically based on the presence of four risk factors. Everyone in group 1 had a history of opioid misuse/dependence; everyone in group 2 had a history of alcohol/non-opioid drug misuse but no opioid misuse or dependence history; everyone in group 3 had exposure to other psychotropic medications in the third trimester but no history of opioid misuse/dependence or alcohol/non-opioid drug misuse; everyone in group 4 had history of smoking but no history of opioid misuse/dependence or alcohol/non-opioid drug misuse or exposure to other psychotropic medications in the third trimester; and everyone in group 5 had none of the four identified risk factors.
§Other psychotropic medication included were tricyclic antidepressants, selective serotonin re-uptake inhibitors, selective nor epinephrine re-uptake inhibitor, benzodiazepines, antipsychotics, antiepileptics, and non-benzodiazepine sedative hypnotic.