Table 3.
Impact of overall and brand cap implementation on prescription utilization
| Outcome | Impact of cap implementation (% change) | 95 % Confidence interval | ||
|---|---|---|---|---|
| OVERALL CAP IMPLEMENTATION | ||||
| Essential medications | ||||
| Prescriptions | Change in level | −0.28 | −0.75 | 0.19 |
| Change in slope | −0.13** | −0.21 | −0.05 | |
| Expenditures | Change in level | −0.28 | −1.52 | 0.97 |
| Change in slope | −0.17 | −0.66 | 0.32 | |
| Preventive essential medications | ||||
| Prescriptions | Change in level | −0.47 | −1.05 | 0.1 |
| Change in slope | −0.28** | −0.46 | −0.11 | |
| Expenditures | Change in level | −0.44 | −1.35 | 0.46 |
| Change in slope | −0.30** | −0.43 | −0.17 | |
| Symptomatic essential medications | ||||
| Prescriptions | Change in level | 0.19** | 0.07 | 0.31 |
| Change in slope | 0.14 | −0.05 | 0.34 | |
| Expenditures | Change in level | 0.15 | −0.31 | 0.61 |
| Change in slope | 0.11 | −0.49 | 0.71 | |
| BRAND CAP IMPLEMENTATION | ||||
| All brand medications | ||||
| Prescriptions | Change in level | −2.29* | −4.16 | −0.42 |
| Change in slope | −0.02 | −0.23 | 0.18 | |
| Expenditures | Change in level | −1.26* | −2.36 | −0.16 |
| Change in slope | 0.08 | −0.19 | 0.35 | |
| Brand medications in classes with generic replacements a | ||||
| Prescriptions | Change in level | −0.74** | −1.23 | −0.25 |
| Change in slope | 0.02 | −0.22 | 0.26 | |
| Expenditures | Change in level | −1.27** | −2.01 | −0.53 |
| Change in slope | 0.05 | −0.29 | 0.40 | |
| Generic medication in classes with generic replacements a | ||||
| Prescriptions | Change in level | 0.79** | 0.20 | 1.38 |
| Change in slope | −0.08 | −0.19 | 0.03 | |
| Expenditures | Change in level | 0.60 | −0.12 | 1.31 |
| Change in slope | −0.03 | −0.13 | 0.08 | |
Interrupted time-series analyses modeling the impact of cap policy implementation on the proportion of prescriptions for selected groups of medications, controlling for Medicare Part D implementation and season. Medication use in states without prescription caps was used as a control. “Level” refers to the immediate impact of cap policy implementation on proportion of medication use. “Slope” refers to the subsequent rate of change per calendar quarter in proportion of use resulting from the cap policy. Complete model parameters can be found in Additional file 1. *p < 0.05; **p < 0.01
a Selected classes include: ACE-inhibitors, ARBs, CCBs, statins, NSAIDs, PPIs, SSRIs, and SNRIs