Table 3:
Performance of Quality Indicators Pre- and Post-Intervention among Unselected Patients based on Administrative Data and Intention-to-treat Analysis: the VA MI-Plus Study, 2002–2008
| Indicator (No. patients with quality indicator/ No. patients eligible for indicator) | Intervention Group-Pre-Intervention | Intervention Group-Post-Intervention | Control Group-Pre-Intervention | Control Group-Post-Intervention | Unadjusted Difference in Improvement† | Adjusted Difference in Improvement (95% CI)‡ | P- value |
|---|---|---|---|---|---|---|---|
| Monitoring | |||||||
| LDL cholesterol measurement in previous 18 months, % | 85.2 (3,429/4,024) |
88.0 (2,711/3,080) | 87.0 (3,241/3,727) | 89.1 (2,594/2,911) | 0.7% | −0.1% (−8.9 to 4.7) |
>0.2 |
| HbA1c measurement in previous 18 months for adults with diabetes, % | 75.3 (1,603/2,129) |
86.8 (1,357/1,563) |
79.1 (1,515/1,916) |
88.6 (1,291/1,457) |
2.0% | 0.4% (−1.4 to 1.9) |
>0.2 |
| Treatment | |||||||
| Prescription for beta- blocker, % | 70.0 (2,818/4,024) |
85.5 (2,633/3,080) | 71.9 (2,678/3,727) | 84.0 (2,446/2,911) | 3.4% | 2.6% (0.1 to 4.1) |
0.04 |
| Prescription for statin, % | 87.4 (2,994/3,424) |
95.9 (2,708/2,825) |
87.8 (2,773/3,157) |
95.3 (2,506/2,630) |
1.0% | 0.7% (−1.3 to 2.1) |
>0.2 |
| Prescription for ACE or ARB, % | 69.7 (2,806/4,024) |
74.6 (2,299/3,080) | 70.6 (2,631/3,727) | 73.6 (2,141/2,911) | 1.9% | 1.8% (−1.4 to 4.8) |
>0.2 |
| Achievement of Target Levels | |||||||
| Last LDL cholesterol level <100 mg/dL, % | 52.3 (1,738/3,322) |
72.2 (1,922/2,663) | 55.2 (1,699/3,077) | 70.4 (1,766/2,509) | 4.7% | 4.0% (−0.3 to 7.9) |
0.06 |
| Last HbA1c level <8% in adults with diabetes, % | 71.3 (1,141/1,600) |
77.3 (1,048/1,356) | 73.1 (1,107/1,514) | 73.8 (953/1,291) | 5.3% | 4.6% (−1.5 to 9.8) |
0.13 |
| Composite Clinical Indicator Score | |||||||
| All patients, % | 69.2 | 79.7 | 71.1 | 79.0 | 2.6 | 2.3 (−0.6- to 5.1) |
0.12 |
| Patients with diabetes and hyperlipidemia, % | 75.0 | 83.8 | 77.5 | 83.4 | 2.9 | 2.6 (−0.4 to 5.7) |
0.09 |
| Physiologic Levels | |||||||
| Last LDL cholesterol level, mg/dL, mean reduction ± SD | 102.4 ± 36.1 | 87.0 ± 33.8 | 100.5 ± 36.2 | 88.8 ± 33.2 | 3.7 | 3.5 (−0.8 to 7.7) |
0.10 |
| Last HbA1c level, %, mean reduction ± SD | 7.4 ± 1.8 | 7.2 ± 1.5 | 7.4 ± 1.7 | 7.3 ± 1.6 | 0.1 | 0.2 (−0.1 to 0.5) |
0.12 |
ACE = angiotensin-converting enzyme; ARB = angiotensin-receptor blocker; HbA1c = hemoglobin A1c; LDL = low-density lipoprotein.
Difference in improvement was the change (post-intervention minus pre-intervention, except for the physiologic differences, where the signs are reversed because decrease signifies improvement) among intervention clinics minus the change among control clinics. A positive difference reflects greater improvement among intervention clinics.
With use of generalized mixed-effects regression models, these differences and the associated 95% CIs and P-values reflect adjustment for patient clustering within clinics, patient-level covariates (age, sex, race, and co-morbidity score) and random clinic effects (to account for correlation of patient measures within individual clinics). Co-morbidity score was calculated using a composite score based on 12 major health conditions: hypertension, hyperlipidemia, diabetes mellitus, heart failure, emphysema, depression, chronic kidney disease, stroke, cancer (excluding skin, prostate or thyroid), peripheral vascular disease, dementia, and asthma. LDL cholesterol and HgbA1c levels for eligible patients who had measurement within 18 months of the index visit. Laboratory level or prescription measured at the time of the index visit. The patient-level composite quality indicator score summed the 7 clinical indicators for which the patient was eligible. All members of the cohort were eligible for three contributors to the composite score (beta-blocker, ACE/ARB, and LDL measurement), while diabetes and hyperlipidemia each added two additional condition-specific indicators.