TABLE 3

Results From Subgroup Analysis of Children With and Without OSA

Dependent VariableCoefficientChildren With OSA (n = 60 535)Children Without OSA (n = 302 457)
Estimate (95% CI)aEstimate (95% CI)a
≥1 prescription fills for codeine after tonsillectomyIntercept39.6 (38.4 to 40.8)31.0 (30.3 to 31.8)
Monthb−0.4 (−0.4 to −0.3)−0.3 (−0.3 to −0.2)
Postb−16.2 (−18.2 to −14.3)−12.6 (−13.9 to −11.4)
Monthsafterb0.2 (0.1 to 0.3)0.1 (0.1 to 0.2)
≥1 prescription fills for an alternative opioid after tonsillectomyIntercept35.6 (34.4 to 36.7)35.0 (32.9 to 37.2)
Monthb0.3 (0.2 to 0.4)0.2 (0.1 to 0.3)
Postb−0.8 (−2.8 to 1.2)−1.3 (−3.8 to 1.1)
Monthsafterb−0.3 (−0.4 to −0.2)−0.2 (−0.4 to −0.1)
  • a Coefficients from regressions were multiplied by 100 and are expressed as percentage point changes.

  • b The coefficient for month represents the absolute percentage point change per month over the study period; the coefficient for post represents the immediate level change after the investigation ended; and the coefficient for monthsafter represents the slope change after the investigation ended.