Who has life-sustaining therapy withdrawn after injury?

Trauma scoring systems have been developed to help surgeons predict who will die after injury. However, some patients may not actually die of their injuries but may undergo withdrawal of life-sustaining therapy (WLST). The goal of this study was to determine which factors were associated with WLST among older patients who died. We hypothesized that patients with comorbid illnesses, higher injury severity scores (ISS), complications, and existing advanced directives (AD) would be more likely to have WLST and that patients having WLST would receive more medication for symptom relief in the 24 hours before death.

Data were collected via a retrospective chart review of patients age 55 years and older admitted to the intensive care unit after injury who subsequently died. In addition to demographic and injury information, documentation of family discussions regarding care wishes and formal ADs were evaluated. Patients dying despite curative attempts were compared with those who died after WLST by Student's t test and chi test where appropriate.

In a 3-year period, of 330 patients age 55 and older admitted to the intensive care unit, 66 (20%) died. Complete records were available for 64 patients. More than half of those who died (n = 35, 54.7%) had WLST. ADs were available for 15 patients (23.4%), and 11 (17.2%) patients had expressed to their families desires to not undergo aggressive curative care. Family discussions were documented for 50 (78%) cases. Comorbid illnesses were present in 46 (71.9%) patients and 35 (54.7%) developed at least one complication. Among people with ADs, 73% had WLST versus 49% of people without ADs (p = 0.09). WLST was independent of comorbid illnesses (p = 0.3), complications (p = 0.8), age (p = 0.5), and ISS (p = 0.2). Patients for whom there was documentation of a family discussion were more likely to have WLST than those without (91.4% versus 62.1%, p = 0.005). Morphine and benzodiazepine dosing in the 24 hours preceding death were greater in the WLST group than the curative therapy group (p = 0.02 and p = 0.05, respectively).

Expected associations with WLST such as age, ISS, comorbidities, and complications were not present in this population. Although trends may exist regarding patient wishes and ADs, larger studies are needed to corroborate these findings. Given the percentage of patients having supportive care withdrawn, trauma registries and scoring systems should include WLST.