Table. Change in Level and Weekly Growth Rate of Naloxone, Opioid, and All Prescriptions From Prepandemic to During Pandemica.
| Characteristic | Mean (95% CI) | ||||||
|---|---|---|---|---|---|---|---|
| Average No. of individuals (level) | % Changed | Weekly growth rate, % (trend) | |||||
| Prepandemic levelb | Pandemic levelb | Change in levelc | Prepandemic trende | Pandemic trende | Change in trendf | ||
| Individuals filling prescriptions | |||||||
| Naloxone | 1371.87 (1288.32 to 1455.42)g | 1010.78 (897.09 to 1124.47)g | −361.09 (−499.56 to −222.62)g | −26.32 | −0.07 (−0.31 to 0.19) | 0.29 (−0.20 to 0.78) | 0.36 (−0.17 to 0.88) |
| Opioids (in thousands) | 2651.46 (2590.49 to 2712.42)g | 2420.49 (2322.72 to 2518.26)g | −230.97 (−343.93 to −118.01)g | −8.71 | 0.09 (−0.01 to 0.18) | −0.15 (−0.33 to 0.04) | −0.23 (−0.43 to −0.03)g |
| All (in tens of thousands) | 5915.71 (5809.23 to 6022.19)g | 5042.76 (4895.40 to 5190.11)g | −872.95 (−1051.36 to −694.54)g | −14.79 | 0.46 (0.37 to 0.54)g | −0.11 (−0.25 to 0.02) | −0.57 (−0.72 to −0.42)g |
| Individuals filling naloxone by payer | |||||||
| Medicaid | 184.10 (169.46 to 198.74)g | 180.94 (163.49 to 198.38)g | −3.16 (−25.54 to 19.22) | −1.72 | −0.55 (−0.82 to −0.29)g | 0.05 (−0.40 to 0.50) | 0.60 (0.10 to 1.10)g |
| Medicare | 552.58 (510.28 to 596.88)g | 364.90 (317.34 to 412.46)g | −188.68 (−251.91 to −125.45)g | −34.15 | −0.01 (−0.33 to 0.31) | 0.51 (−0.05 to 1.07) | 0.52 (−0.10 to 1.15) |
| Commercial | 527.93 (497.98 to 557.88)g | 363.19 (318.79 to 407.59)g | −164.74 (−217.27 to −112.21)g | −31.20 | 0.04 (−0.20 to 0.27) | 0.44 (−0.07 to 0.96) | 0.41 (−0.14 to 0.95) |
| Cash | 106.26 (96.17 to 116.36)g | 101.75 (91.03 to 112.47)g | −4.51 (−18.99 to 9.97) | −4.24 | 0.11 (−0.30 to 0.51) | −0.73 (−1.26 to −0.21)g | −0.84 (−1.49 to −0.19)g |
The prepandemic period is defined as May 13, 2019, through the week of March 13, 2020. The pandemic period begins the following week and continues through December 20, 2020.
Predicted levels are calculated from a linear regression of weekly trends on the number of individuals filling prescriptions.
The change in level is calculated as the coefficient on the term that measures the breakpoint in the interrupted time series analysis.
Calculated as (change in level/prepandemic level × 100).
The trend is calculated as the coefficient on the weekly trend in a log-linear regression, where the dependent variable (number of individuals filling a prescription) is log transformed.
The change in trend is calculated as the coefficient on the interaction term in the interrupted time series analysis where the dependent variable (number of individuals filling a prescription) is log transformed.
Denotes P ≤ .05.