Table 1.
Descriptive characteristics of study cohorts
| Characteristic | Derivation Cohort n = 8,700 | Validation Cohort n = 8,189 | ||
|---|---|---|---|---|
| Readmitted | n | % | n | % |
| 1230 | 14.1 | 1211 | 14.8 | |
| Demographic/social | ||||
| Male sex* | 3404 | 39.1 | 3204 | 39.1 |
| Age* | Mean | Stdev | Mean | Stddev |
| 60.6 | 20.6 | 65 | 20.7 | |
| Age group | n | % | n | % |
| • 18–35 Years | 1489 | 17.1 | 1392 | 17.0 |
| • 36–44 Years | 517 | 5.9 | 433 | 5.3 |
| • 45–64 Years | 2416 | 27.8 | 2200 | 26.9 |
| • 65 + Years | 4278 | 49.2 | 4164 | 50.8 |
| Married* | 4304 | 49.5 | 4057 | 49.5 |
| Lives alone* | 783 | 9.0 | 1487 | 18.1 |
| Race/Ethnicity | ||||
| • Asian | 10 | 0.1 | 14 | 0.2 |
| • African American | 519 | 6 | 507 | 6.2 |
| • Hispanic | 48 | 0.6 | 43 | 0.5 |
| • American Indian/Alaskan Native | 4 | 0.0 | 3 | 0.0 |
| • Unknown | 57 | 0.7 | 34 | 0.4 |
| • White | 8062 | 92.7 | 7588 | 92.7 |
| Healthcare utilization | ||||
| Payers | ||||
| • Commercial | 2241 | 25.8 | 2089 | 25.5 |
| • Medicare | 4811 | 55.3 | 4676 | 57.1 |
| • Medicaid | 952 | 10.9 | 922 | 11.3 |
| • Uninsured* | 644 | 7.4 | 467 | 5.7 |
| • Other | 52 | 0.6 | 35 | 0.4 |
| • Acute admission* | 5734 | 66.0 | 5466 | 66.8 |
| ED Visits within one year* | Median | IQR | Median | IQR |
| 0 | 0–2 | 0 | 0–1 | |
| n | % | n | % | |
| • 0 | 4476 | 51.5 | 5760 | 70.3 |
| • 1–2 | 2611 | 30.0 | 1877 | 22.9 |
| • 3–5 | 990 | 11.4 | 450 | 5.5 |
| • ≥6 | 623 | 7.2 | 102 | 1.2 |
| Inpatient Visits within one year* | Median | IQR | Median | IQR |
| 0 | 0–1 | 0 | 0–1 | |
| n | % | n | % | |
| • 0 | 4935 | 56.7 | 4598 | 56.1 |
| • 1–2 | 2641 | 30.4 | 2538 | 31.0 |
| • 3–5 | 886 | 10.2 | 832 | 10.2 |
| • ≥6 | 238 | 2.7 | 221 | 2.7 |
| Length of stay* | median | IQR | median | IQR |
| 3 | 2–4 | 3 | 2–4 | |
| n | % | n | % | |
| • 0–1 | 1641 | 18.9 | 1418 | 17.3 |
| • 2–3 | 4383 | 50.4 | 4178 | 51.0 |
| • 4–8 | 2236 | 25.7 | 2067 | 25.2 |
| • ≥9 | 440 | 5.1 | 526 | 6.4 |
| Inpatient medications* | Mean | Stdev | Mean | Stddev |
| 14.0 | 5.9 | 16.1 | 6.5 | |
| Ambulatory medications* | Median | IQR | Median | IQR |
| 2 | 0–6 | 2 | 0–5 | |
| n | % | n | % | |
| • 0–1 | 3661 | 42.1 | 3656 | 44.6 |
| • 2–5 | 2497 | 28.7 | 2543 | 31.1 |
| • 6–14 | 2035 | 23.4 | 1686 | 20.6 |
| • ≥15 | 507 | 5.8 | 304 | 3.7 |
| Modified Charlson Comorbidity Index* | Median | IQR | Median | IQR |
| 6 | 0–8 | 7 | 0–9 | |
| n | % | n | % | |
| • 0–3 | 3405 | 39.1 | 3064 | 37.4 |
| • 4–8 | 3453 | 39.7 | 2231 | 27.2 |
| • ≥9 | 1842 | 21.2 | 2894 | 35.3 |
*Candidate predictor variables. Variables that are not normally distributed, such as the mCCI, are grouped according to patterns and visual break-points observed in the distributions in our patient population.