TABLE 1

Patient and Provider Characteristics Associated with Codeine Prescription During Pediatric Emergency Department Visits

CharacteristicNo. of Unweighted ED Visit Records With Codeine Rx (N = 56 375)Weighted Proportion of Visits With Codeine Rx, %P2)Adjusted OR (95% CI) for Codeine RxaP (Adjusted OR)
All visits with codeine Rx20593.8
Year20593.8.0090.97 (0.95–0.99).026
Age group (y)
 3 to 76573.3Reference
 8 to 126994.5<.0011.42 (1.21–1.67)<.001
 13 to 176853.41.04 (0.87–1.24).69
Gender
 Male9803.7.11Reference
 Female10793.61.07 (0.95–1.21).28
Race/ethnicity
 White, non-Hispanic12564.3Reference
 Black, non-Hispanic3312.6<.0010.67 (0.56–0.80)<.001
 Hispanic3283.20.83 (0.68–1.01).07
 Other633.00.88 (0.59–1.30).52
Payment source
 Private insurance9914.2Reference
 Medicaid7513.4<.0010.84 (0.71–0.98).03
 Other534.20.86 (0.72–1.04).11
US census region
 Northeast2672.0Reference
 Midwest5654.4<.0011.86 (1.50–2.29)<.001
 South7753.91.85 (1.49–2.31)<.001
 West4524.62.09 (1.63–2.68)<.001
Type of provider
 Physician18823.8Reference
 Resident/trainee physician282.6<.0010.69 (0.41–1.15).15
 Nurse practitioner/physician assistant1193.91.06 (0.86–1.32).56
 Otherb301.30.30 (0.19-0.48)<.001
Metropolitan statistical areac
 Metropolitan17253.6.38d
 Nonmetropolitan3344.2d
  • CI, 95% confidence interval; ED, emergency department; Rx, prescription.

  • a Logistic regression analysis using survey weights with the following variables included in the model: year, age, gender, race/ethnicity, payment source, US census region, and type of provider.

  • b “Other” providers included unlicensed and missing or unknown provider types.

  • c Metropolitan statistical area reflects whether a provider was practicing in an urban or non-urban setting.

  • d Metropolitan statistical area was not included in the multivariable model because this variable was not nominally associated with codeine prescription in χ2 testing.