Baseline Demographic, Clinical, and Utilization Characteristics of Intervention Subjects and Matched Controls
| Characteristic | Intervention (N=146) | Control (N=437) | P Value |
|---|---|---|---|
| Demographic characteristics | |||
| Age, y, mean (SD) | 78.6 (8.2) | 77.3 (7.9) | .09 |
| Female, n (%) | 96 (66) | 287 (66) | .99 |
| Census-linked socioeconomic characteristics* | |||
| Median household income, mean | 54,580 | 54,370 | .90 |
| Proportion of adults who graduated from high school, mean (SD) | 0.88 (0.09) | 0.88 (0.08) | .55 |
| Proportion of adults of non-white race, mean (SD) | 0.17 (0.15) | 0.22 (0.19) | .01 |
| Clinical characteristics | |||
| Diabetes, n (%) | 36 (25) | 107 (25) | .97 |
| Heart disease, n (%) | 58 (40) | 140 (32) | .09 |
| Cancer diagnosis, n (%) | |||
| None | 118 (81) | 362 (83) | |
| Within 5 years | 13 (9) | 34 (8) | .94 |
| 5–10 years from index date | 9 (6) | 23 (5) | |
| >10 years from index date | 6 (4) | 18 (4) | |
| Dementia or Parkinson’s Disease, n (%) | 16 (11) | 27 (6) | .06 |
| Recent depression, n (%) | 4 (3) | 10 (2) | .76 |
| RxRisk score, mean (SD) | 3210 (2990) | 3470 (3700) | .44 |
| Charlson Comorbidity Index, n (%) | |||
| 0 | 118 (81) | 342 (78) | |
| 1 | 11 (8) | 44 (10) | .84 |
| 2 | 8 (5) | 23 (5) | |
| ≥3 | 9 (6) | 28 (6) | |
| Propensity score, mean (SD | 0.29 (0.15) | 0.29 (0.15) | .84 |
| Healthcare utilization | |||
| Hospitalizations, mean (SD)† | 0.31 (0.68) | 0.26 (0.63) | .44 |
| One or more hospitalizations, n (%) | 32 (22) | 83 (19) | .44 |
| Medical/surgical specialist visits, mean (SD) | 4.8 (7.5) | 4.8 (6.3) | .97 |
| Outpatient visits, mean (SD) | 18.7 (14.0) | 18.3 (13.2) | .74 |
| Any nursing home admission, n (%) | 11 (7.5) | 29 (6.6) | .71 |
| Number of 14-day equivalent high-risk prescriptions, mean (SD)‡ | 4.2 (11.2) | 4.9 (15.6) | .58 |
| Total healthcare costs, median $§ | 5068 | 4977 | .72 |
* Block-level 2000 United States Census data as identified by geocoding techniques using subjects’ 2005 street addresses and ZIP codes. Education and non-white race status pertain to adults over age 25 years. Income and education data were missing for 89 subjects (15.3%), and non-white race data were missing for 105 subjects (18.0%).
† During year prior to index date.
‡ Includes selected sedative-hypnotics, antihistamines, muscle relaxants, and tricyclic antidepressants.
§ Pre-index date costs were missing/incomplete for 6 control subjects. P value derived from nonparametric χ2 test for difference in medians.