The optimum follow-up period for assessing mortality outcomes in injured older adults

Abstract

Objectives: To compare mortality rates of hospitalized injured aged 67 and older across commonly used follow-up periods (e.g., in-hospital, 30-day, 1-year) and to determine the postinjury time after which mortality rates stabilize.

Design: Retrospective analysis of Medicare claims.

Setting: Oregon and Washington Medicare patients.

Participants: Patients admitted to 171 Oregon and Washington facilities during 2001/02 with injuries identified according to International Classification of Diseases, Ninth Revision, code and followed for 1 year.

Measurements: The primary outcome was in-hospital mortality and mortality at 30, 60, 90, 180, and 365 days. Kaplan-Meier survival curves and daily postadmission mortality rates were also evaluated. The rate of change (slope) of the survival curves and daily mortality rates were analyzed to select the point after which mortality rates were no longer decreasing.

Results: There were 32,135 injured older adults hospitalized over the 2-year period, with a median age of 82 (interquartile range 77-88). Cumulative in-hospital mortality and at 30, 60, 90, 180, and 365 days was 4.1%, 9.7%, 13.6%, 16.1%, 21.3%, and 28.4%, respectively. Mortality rates stabilized by 6 months after injury, with 89% of the change occurring within 60 days. Although serious injuries, medical comorbidities, and preinjury nursing facility residence were all associated with higher mortality, they did not affect the pattern of mortality after injury.

Conclusion: In-hospital mortality is much lower than postdischarge mortality in injured older adults, with a substantial portion of persons dying shortly after discharge from the hospital. Mortality appears to stabilize by 6 months after injury, although 60-day postadmission follow-up captures most of the excess daily mortality rate.

Figure 1. Kaplan-Meier Curves For Post-Injury Survival Among Hospitalized Medicare Patients (n = 32,135)

*Seriously injured subgroup, defined by ICD-9 based Injury Severity Score (ICISS) ≤ 0.9 (n=1,607). †High co-morbidity subgroup, defined by a Charlson co-morbidity score ≥3 (n=4,414). ‡Skilled nursing facility (SNF) residents prior to injury (n=1,421). §Hip fracture (n=11,738).

Figure 2. Daily Mortality Rates Beginning With Admission For Hospitalized Injured Medicare Patients (n=32,135)

The daily mortality rate was calculated as the number dying on a given day divided by the number alive at the beginning of that day, divided by 1000 for clarity. These rates could also be described as the smoothed hazard estimates × 1000. *Seriously injured subgroup, defined by ICD-9 based Injury Severity Score (ICISS) ≤ 0.9 (n=1,607). †High co-morbidity subgroup, defined by a Charlson co-morbidity score ≥3 (n=4,414). ‡Skilled nursing facility (SNF) residents prior to injury (n=1,421). §Hip fracture (n=11,738).