EXHIBIT 3.
Estimates of the effect of PCIP on quality for 6, 12, 18, and 24 months of exposure to PCIP across levels of technical assistance
| Quality measures | ||||||||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| All quality measures | EHR sensitive quality measures | |||||||
|
| ||||||||
| Cumulative technical assistance visits | Cumulative technical assistance visits | |||||||
|
| ||||||||
| Exposure period | Overall | none | 3 visits | 8 visits | Overall | none | 3 visits | 8 visits |
| 6 months | −0.24 (0.53) |
−0.30 (0.77) |
−0.33 (0.54) |
0.04 (0.77) |
0.73 (0.66) |
0.00 (0.90) |
0.74 (0.71) |
1.85 (1.05) |
| 12 months | −0.25 (0.72) |
−0.65 (1.01) |
−0.40 (0.73) |
0.25 (0.91) |
1.24 (0.88) |
−0.09 (1.16) |
1.01 (0.90) |
2.70* (1.22) |
| 18 months | −0.02 (0.81) |
−0.86 (1.05) |
−0.23 (0.82) |
0.61 (0.90) |
1.71 (0.95) |
0.07 (1.22) |
1.13 (0.94) |
2.90* (1.15) |
| 24 months | 0.42 (0.91) |
−0.74 (1.15) |
0.14 (0.94) |
1.10 (1.07) |
2.31* (1.05) |
0.81 (1.33) |
1.41 (1.06) |
2.79* (1.29) |
|
| ||||||||
| Observations | 7,622 | 3,589 | ||||||
| Number of total physicians |
720 | 516 | ||||||
| Number of PCIP physicians |
360 | 258 | ||||||
SOURCE: Authors’ analysis
Note 1: p<.05
Note 2: 3 cumulative technical assistance visits is the median number of visits and 8 cumulative visits is the 90th percentile among practices that went live during the study period.
Note 3: Effects are interpreted as the incremental percentage point change in quality of care for a given exposure to PCIP
Note 4: Standard errors robust to groupwise heterskedasticity at the level of the match shown in ()
Note 5: EHR sensitive measures include breast cancer screening for women, retinal exam for diabetics, urine testing for diabetics, chlamydia screening for women, and colorectal screening