Characteristics of all pneumonia discharges with a readmission and abstracted sample*
| All pneumonia discharges with a readmission (n=11 278) | PPR status—fully abstracted cases (n=100) | |||
|---|---|---|---|---|
| Variable | Not fully abstracted (n=11 178) | Fully abstracted (n=100) | Yes (n=77) | No (n=23) |
| Age, mean (SD)† | 71.0 (12.2) | 71.1 (13.1) | 73.0 (12.2) | 67.0 (15.2) |
| Sex, male, n (%) | 10 950 (98.0) | 96 (96.0) | 75 (97.4) | 21 (91.3) |
| Race, n (%)‡ | ||||
| White | – | – | 56 (72.7) | 14 (60.9) |
| Black | – | – | 10 (13.0) | 6 (26.1) |
| Hispanic | – | – | 5 (6.5) | 2 (8.7) |
| Other | – | – | 6 (7.8) | 1 (4.3) |
| Length of stay, days, median (25th, 75th centile) | 5 (3, 8) | 5 (3, 9.5) | 5 (3, 10) | 5 (4, 8) |
| Time to readmission, days, median (25th, 75th centile) | 12 (5, 20) | 9.5 (4.5, 19) | 9 (4, 19) | 12 (6, 16) |
| Severity of illness, n (%)§ | ||||
| 1—Minor | 611 (5.5) | 2 (2) | 2 (2.6) | 0 (0) |
| 2—Moderate | 4665 (41.7) | 47 (47.0) | 36 (46.8) | 11 (47.8) |
| 3—Major | 4867 (43.5) | 41 (41.0) | 30 (39.0) | 11 (47.8) |
| 4—Extreme | 1034 (9.3) | 10 (10) | 9 (11.7) | 1 (4.4) |
| Number of Elixhauser comorbidities, mean (SD)¶ | 5.9 (2.5) | 5.8 (2.4) | 6.0 (2.5) | 5.1 (2.1) |
| Selected Elixhauser comorbidities, n, %¶ | ||||
| Heart failure | 3901 (34.9) | 34 (34.0) | – | – |
| Chronic pulmonary disease | 6874 (61.5) | 69 (69.0) | – | – |
| Metastatic cancer | 858 (7.7) | 3 (3.0) | – | – |
| Solid tumour without metastasis† | 3022 (27.0) | 19 (19.0) | – | – |
| Liver disease | 828 (7.4) | 8 (8.0) | – | – |
| Renal failure | 3105 (27.8) | 31 (31.0) | – | – |
| Psychoses | 2152 (19.3) | 23 (23.0) | – | – |
| Depression | 2929 (26.2) | 24 (24.0) | – | – |
| Selected EMR-abstracted comorbidities, n (%)** | ||||
| Chronic lung disease | – | – | 53 (68.8) | 13 (56.5) |
| COPD†† | – | – | 46 (59.7) | 11 (47.8) |
| Cancer | – | – | 6 (7.8) | 3 (13.0) |
| Liver disease | – | – | 5 (21.7) | 4 (17.4) |
| Heart failure | – | – | 23 (29.9) | 3 (13.0) |
| Stroke | – | – | 10 (13.0) | 2 (8.7) |
| Chronic kidney disease | – | – | 19 (24.7) | 7 (30.4) |
| Receiving home oxygen† | – | – | 20 (26.0) | 2 (8.7) |
| Intensive care unit admission, n (%) | – | – | 15 (19.5) | 2 (8.7) |
| Nursing home patient, n (%) | – | – | 14 (18.2) | 2 (8.7) |
*No significant differences (ie, no p<0.05) between groups (not fully abstracted pneumonia readmissions vs abstracted cases and PPR–yes vs PPR–no cases).
†Indicates differences with p values between 0.05 and 0.10. For age, this only applies to the PPR–yes vs PPR–no comparison. All other p values were >0.10.
‡Our administrative dataset did not contain race.
§This is derived from the APR-DRGs which classify patients according to their reason for admission and severity of illness. Severity of illness level is APR-DRG-specific and takes into account the patient's age, principal diagnosis, secondary diagnoses and procedures from the index admission.29
¶Consists of 29 Elixhauser comorbidities obtained from administrative data (both inpatient and outpatient) from year before admission up to and including the index admission.27
**Comorbidities obtained from EMR.
††COPD; subset of chronic lung disease.
APR-DRGs, all patient refined-diagnosis related groups; COPD, chronic obstructive pulmonary; EMR, electronic medical record; PPR, Potentially Preventable Readmission.