Patients readmitted to the intensive care unit during the same hospitalization: clinical features and outcomes.

OBJECTIVE

To determine the clinical features and outcomes of patients readmitted to the intensive care unit (ICU) during the same hospital stay and the causes for these readmissions.

DESIGN

Multicenter, cohort study.

SETTING

Three ICUs from two teaching hospitals and four ICUs from four community hospitals.

PATIENTS

All ICU admissions were collected prospectively for a registry database in the seven ICUs. We retrospectively analyzed ICU admissions between January 1, 1995 and February 29, 1996.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

During the study period, 236 (4.6%) of the patients discharged alive from the ICU were readmitted to the unit. Patients with gastrointestinal (GI) and neurologic diagnoses had the highest readmission rate. Of the readmissions, 45% had recurrence of the initial disease, 39% experienced new complications, and 14% required further planned operation. Among patients readmitted for the same illness, cardiovascular and respiratory problems were the most frequent diagnoses. Of patients readmitted with a new diagnosis, 30% initially had GI diseases, while respiratory diseases accounted for 58% of the new complications. Readmissions within 24 hrs occurred in 27% of all readmissions. Patients requiring readmission had a higher hospital mortality rate (31.4%) compared with those not requiring readmission (4.3%, p < .001), even after adjustment for disease severity score (odds ratio = 5.93, p < .001).

CONCLUSIONS

Patients with GI and neurologic diseases are at greatest risk of requiring ICU readmission. Respiratory diseases are the major reason for readmission due to new complications. Readmitted patients have a high risk of hospital death that may be underestimated by the usual physiologic indicators on either initial admission or readmission. Further studies are required to determine if patients at risk for readmission can be identified early to improve the outcome.