Table 1.
COHORT characteristics and unadjusted outcomesa
| All other hospitalizations | Excluded or not selected | Interviewed and enrolled | P | |
|---|---|---|---|---|
| Number of patientsb | 31,275 | 2,548 | 1,551 | –- |
| Age (Median, mean ± SD) | 67.0, 64.4 ± 18.1 | 72.0, 69.6 ± 16.9 | 67.0, 65.2 ± 15.1 | 0.0333 |
| Sex (% male) | 48.9 | 48.5 | 46.5 | 0.0588 |
| Race (%) | ||||
| White | 52.6 | 55.8 | 54.7 | 0.0968 |
| Black/African American | 15.8 | 16.6 | 21.0 | < .0001 |
| Hispanic | 14.0 | 11.4 | 9.1 | < .0001 |
| Asian | 14.9 | 13.9 | 12.4 | 0.0044 |
| Other/unknown race | 2.6 | 2.4 | 2.7 | 0.8563 |
| Charlson Comorbidity Index scorec (Median, mean ± SD) | 2.0, 2.7 ± 2.8 | 3.0, 4.0 ± 3.2 | 3.0, 3.4 ± 3.0 | < .0001 |
| COPS2 (Median, mean ± SD) | 21.0, 36.6 ± 38.3 | 45.0, 61.4 ± 56.0 | 31.0, 49.9 ± 47.2 | < .0001 |
| LAPS2 (Median, mean ± SD) | 45.0, 53.1 ± 38.9 | 68.0, 70.2 ± 42.3 | 57.0, 59.5 ± 38.0 | < .0001 |
| Admitted for observation (%) | 26.1 | 6.8 | 0.5 | < .0001 |
| Full code on admission (%) | 89.8 | 80.5 | 92.3 | 0.0003 |
| Ever admitted to ICU (%) | 12.5 | 18.2 | 14.9 | 0.0103 |
| Discharge diagnosesd (%) | ||||
| Sepsis | 12.1 | 18.8 | 19.5 | < .0001 |
| Community-acquired pneumonia | 1.4 | 1.7 | 1.1 | 0.2546 |
| Acute myocardial infarction | 3.1 | 2.7 | 3.4 | 0.5818 |
| Congestive heart fail | 0.5 | 0.9 | 0.6 | 0.7275 |
| Gastrointestinal bleeding | 1.3 | 1.0 | 1.4 | 0.8527 |
| All other | 81.6 | 74.9 | 74.1 | < .0001 |
| Length of stay (days (Median, mean ± SD) | 2.0, 3.4 ± 5.0 | 3.5, 5.9 ± 8.7 | 3.7, 5.0 ± 5.0 | < .0001 |
| Full code on discharge (%) | 86.1 | 73.7 | 90.3 | < .0001 |
| TSL scoree (Median, mean ± SD) | 9.0, 11.4 ± 7.9 | 12.0, 16.2 ± 11.7 | 9.0, 12.8 ± 9.5 | < .0001 |
| Died during initial hospitalization (%) | 2.2 | 2.3 | 0.3 | < .0001 |
| Non-elective hospitalization within 30 days of discharge (%) | 8.3 | 14.5 | 13.0 | < .0001 |
| Died within 30 days of discharge (%) | 2.6 | 6.6 | 2.1 | 0.1878 |
| Died or had non-elective hospitalization within 30 days of discharge (%) | 10.3 | 19.6 | 13.6 | 0.0002 |
aTable 1 provides information on interviewed patients, patients who were excluded or not selected, and all remaining patients (except 712 patients who refused and whose data could not be used). See text, Figure 1, and Appendix 2 for additional details regarding the recruitment process. SD = standard deviation ICU = intensive care unit. The P value shown compares interviewed and enrolled patients to all other hospitalizations; additional comparisons are provided in Appendix 4
bDuring the study period, a total of 36,086 adult patients were hospitalized in Oakland, San Leandro, and Walnut Creek hospitals. Of these, 1,551 patients agreed to be interviewed, 4 of whom died prior to discharge resulting in 1547 patients in the final analysis cohort; 712 patients refused to participate, and we could not use their data; lastly, 2,548 patients were excluded or not selected. For comparison purposes, we selected the first hospitalization experienced by patients who had multiple hospitalizations during the study period (N = 31,275, far left column)
cThe Charlson Comorbidity Index score (range, 0–40; higher scores indicate greater comorbidity burden) was calculated using the methodology of Deyo et al. [27]. COPS2 = COmorbidity Point Score, version 2 (COPS2, range, 0 to 1010, higher scores indicate increasing comorbidity burden) is assigned based on all diagnoses incurred by a patient in the 12 months preceding the index hospitalization. The univariate relationship of COPS2 with 30-day mortality is as follows: 0–39, 1.7%; 40–64, 5.2%; 65 + , 9.0%. LAPS2 = Laboratory-based Acute Physiology Score, version 2 (LAPS2, range, 0 to 414, higher scores indicating increasing physiologic derangement) is assigned based on a patient’s worst vital signs, pulse oximetry, neurological status, and 16 laboratory test results in the preceding 24 (hourly and discharge LAPS2) or 72 h (admission LAPS2). The univariate relationship of an admission LAPS2 with 30-day mortality is as follows: 0–59, 1.0%; 60–109, 5.0%; 110 + , 13.7%. See Escobar et al. [23]
dSee text and Escobar et al. [23] for description of how we grouped diagnosis codes into Primary Conditions
eTSL = Transition Support Level score. This score is assigned at 6 AM on the day of discharge to all adult hospitalized patients in Kaiser Permanente Northern California. The score, which is expressed as a percent, is calibrated against a composite outcome (non-elective hospitalization and/or death within 30 days of discharge). It is based on a patient’s LAPS2, COPS2, length of stay, recent hospital and emergency department utilization preceding the current hospitalization, and discharge care directive (full code or not); see Escobar et al. [17] for details. Patients with a TSL score of ≥ 25% receive additional assessments and follow-up calls and appointments