Comparison of Injured Older Adults Included in vs Excluded From Trauma Registries With 1-Year Follow-up

Abstract

Importance: Trauma registries are the primary data mechanism in trauma systems to evaluate and improve the care of injured patients. Research has suggested that trauma registries may miss high-risk older adults, who commonly experience morbidity and mortality after injury.

Objective: To compare injured older adults who were included in with those excluded from trauma registries, with a focus on patients with serious injuries, requiring major surgery, or dying after injury.

Design, setting, and participants: This cohort study included all injured adults 65 years and older transported by 44 emergency medical services agencies to 51 trauma and nontrauma centers in 7 counties in Oregon and Washington from January 1, 2011, to December 31, 2011, with follow-up through December 31, 2012. Record linkage was used to match emergency medical services records with state trauma registries, state discharge databases, state death registries, and Medicare claims. Data were analyzed from August to November 2018.

Exposures: Inclusion in vs exclusion from a trauma registry.

Main outcomes and measures: Mortality up to 12 months, including time to death and causes of death.

Results: Of 8161 included patients, 5579 (68.4%) were women, and the mean (SE) age was 82.2 (0.10) years. A total of 1720 older adults (21.1%) were matched to a trauma registry record. Seriously injured patients not captured by trauma registries ranged from 18% (7 of 38 patients with abdominal-pelvic Abbreviated Injury Scale score of 3 or greater) to 80.0% (1792 of 2241 patients with extremity Abbreviated Injury Scale score of 3 or greater), while 68 of 186 patients requiring major nonorthopedic surgery (36.6%) and 1809 of 2325 patients requiring orthopedic surgery (77.8%) were not included in trauma registries. Of patients with serious injuries or undergoing major surgery missed by trauma registries (range by injury and procedure type, 36.0% to 57.1%), 36.4% (39.3% when excluding serious extremity injuries and orthopedic procedures) were treated at trauma centers, particularly level III through V hospitals. When registry and nonregistry groups were tracked over 12 months, 93 of 188 in-hospital deaths (49.5%) and 1531 of 1887 total deaths (81.1%) occurred in the nonregistry cohort.

Conclusions and relevance: In their current form, trauma registries are ineffective in capturing, tracking, and evaluating injured older adults, although mortality following injury is frequently due to noninjury causes. High-risk injured older adults are not included in registries because of care in nontrauma hospitals, restrictive registry inclusion criteria, and being missed by registries in trauma centers.

Figure 1.. Kaplan-Meier Survival Curves for Injured Older Adults With a Matched Trauma Registry Record vs Those With No Matched Trauma Registry Record

Of the 8161 included patients, 1720 were captured in a trauma registry and 6441 were not. The shaded regions represent 95% CIs. The log-rank test comparing Kaplan-Meier curves yielded a P value of .03. The curves are not adjusted for other covariates.

Figure 2.. Deaths Following Injury Among Older Adults

Injured older adults captured and not captured in a trauma registry were tracked for 1 year, with a total of 1887 deaths recorded. The percentage of deaths captured in the trauma registry cohort was calculated by day, using the cumulative number of deaths occurring in the trauma registry cohort divided by the cumulative total number of deaths at each time point. EMS indicates emergency medical services.

Figure 3.. Causes of Death Among Injured Older Adults Included vs Not Included in a Trauma Registry

Of the 1887 deaths occurring in the cohort over 12 months, death certificate data were available for 1800 (95.4%); only observed values are presented. Because a patient could have multiple contributing causes of death, the percentages do not add to 100%. A total of 517 deaths occurred 0 to 30 days after EMS contact (A), 398 deaths occurred 31 to 90 days after EMS contact (B), and 885 deaths occurred 91 to 365 days after EMS contact (C).

Figure 4.. High-Risk Patients Not Captured by Trauma Registries Among 8161 Injured Older Adults

A, Proportion of high-risk patients not captured by trauma registries and estimates after modifying registry inclusion criteria. The error bars indicate 95% CIs. B, High-risk older adults not captured by trauma registries by final hospital location. The brackets and percentages represent the proportion of patients not captured by trauma registries who were cared for in trauma centers. There were 27 patients with unknown type of final destination hospital.