A comparison of rural versus urban trauma care

INTRODUCTION

Injury continues to be a leading cause of death and disability throughout the United States. However, there are significant differences between injuries that occur in rural versus urban settings. Trauma patients in rural locations are typically older, less severely injured, and more likely to die at the scene than urban patients.[1] The majority of trauma in rural settings is blunt injury with fewer penetrating injuries, while the reverse is true for urban locations.[2] Similarly, the fatal crash rate is more than twice as high in rural than urban areas, with a rural rate of approximately 2.4 deaths/100 million vehicle miles traveled.[3] Despite this, the historical focus of trauma system development has centered only on the designation of urban-based trauma centers. Few non-trauma center alternative models of care existed, and very few trauma patient outcome studies have been conducted in the rural setting. Prior to 1999, California Trauma Regulations limited rural and urban areas to one, level I, II, or III, trauma center for every 350,000 residents (or 350 major trauma patients per year) and did not include a level IV trauma center option.

In rural communities, level III trauma centers have on-call trauma surgeons and in-house trauma teams, while level IV facilities usually do not. In a retrospective review of level III trauma hospitals in Oregon, Hedges et al. identified hospital interventions that were associated with improvement in survival.[4] They noted that the administration of blood products to patients with a presenting systolic blood pressure of less than 90 mmHg, and transferring patients both with an injury severity score (ISS) greater than 20 and a systolic blood pressure of greater than 90 mmHg to a higher level of care conferred increased survival in their patient population.

The northwest corner of California is a rural region with limited trauma care resources. North Coast Emergency Medical Services (NCEMS) operates across this remote tri-county region of approximately 6,838 square miles.[5] In the mid-1980s, NCEMS identified a high rate of preventable injury death in this region. Because of the long distances between hospitals, a central trauma center was not a practical solution. Instead, the State of California Emergency Medical Services Authority supported the development of a non-trauma center rural alternative, called the Model Rural Trauma Project (MRTP).[6] The rural alternative was an attempt to improve the management of major trauma in these remote communities. The MRTP included three key elements: A warning system for the arrival of major trauma patients using trauma triage criteria, early activation of a trauma team in the emergency department (ED), and periodic systems review and modification. In a three-year retrospective study, the MRTP was found to have similar patient survival in major trauma patients when comparing to the national Major Trauma Outcome Study (MTOS) expected norms.[5]

Given the experience with the MRTP, this study was conducted to compare the survival of moderately-to-severely injured trauma patients in urban versus this rural environments using Trauma and Injury Severity Score (TRISS) methodology to control for differences not related to location. With the implementation of the rural alternative to trauma care, we hypothesized that outcomes in urban versus rural trauma care would not be substantially different.

MATERIALS AND METHODS

We conducted an observational, cohort study using a consecutive series of all trauma patients brought to eight rural, northern California study hospitals and two urban trauma centers located in Los Angeles County (Harbor-UCLA Medical Center and Ronald Reagan UCLA Medical Center) between 09/01/95 and 8/31/96.

RESULTS

During the defined one-year study period, a total of 2201 trauma patients were transported to the two urban hospitals; of those, 786 (36%) were excluded because they had ISS scores less than 10, and an additional 222 (10%) were excluded because of missing ISS information. In the rural setting, 339 patients were entered in the registry as having an ISS ≥ 10, though 3 (1%) did not have an ISS score recorded and were thus excluded from the study.

In total, there were 1122 trauma patients with an ISS ≥ 10 entered into the analysis, with 336 (30.0%) of them presenting to one of the rural hospitals. Overall, 322 (28.7%) were female, 232 (20.7%) were 55-years-old or older, and 899 (80.1%) suffered blunt injuries. The median age was 32.5 (IQR: 21.5 −50.5) [Table 4].

The urban and rural trauma populations were different in multiple baseline characteristics. The rural victims tended to be older, had a higher percentage of females, and had a predominance of Caucasians. Blunt injuries were more frequent in the rural setting. While motor vehicle accidents represented approximately 31% of the mechanisms in both populations, falls were more common in the rural setting, whereas gunshot wounds and pedestrians struck by vehicles were relatively more frequent in the urban setting [Table 4].

No rural patients were sent home from the ED, whereas a small number of urban patients were discharged home. A considerably higher percentage of urban patients were admitted to a monitored setting or taken directly to the operating room. Hospitalizations were shorter on average for rural patients, and fewer rural patients died [Table 5].

The patients in the urban group tended to have modestly lower pre-hospital systolic blood pressure, a wider range of Glasgow Coma Scale (GCS) values (i.e., more patients with lower values), and a higher ISS [Table 5].

As noted above, GCS and Revised Trauma Score (RTS) values were calculated separately using pre-hospital and ED information, when available. We noted that there was a higher mortality rate among those patients for whom an ED-based RTSs could not be calculated because of missing ED information. In the urban setting, 43% of the patients who did not have complete information to calculate an ED-based RTS died as compared to the overall 16% of deaths among all urban patients. In the rural setting, 54% of those with missing ED-based RTS scores died versus 9% of all patients brought to rural EDs.

To address the missing values, the TRISS scores were calculated using multiple imputation.[10,11] The imputation included various demographic factors, mechanism of injury, individual components of the TRISS score (both pre-hospital and ED-based), transport times, resuscitation and imaging information, operative procedures performed, discharge status, ICU days, and survival. The physiologic and ISS components of the TRISS scores, after multiple imputations, are shown in Table 6.

After controlling for TRISS, the odds of survival in the rural population were not different than in the urban population, using either the pre-hospital or ED values for the TRISS calculation (Pre-hospital data: OR 0.73 (95% CI 0.41 – 1.31; ED data: OR 0.73 (95% CI 0.39-1.39).

DISCUSSION

Trauma patients in rural communities are known to have a higher rate of mortality than their peers in urban centers.[12] In a study evaluating the differences in times to trauma care, Esposito et al. found that the crude death rate in rural settings was three times that of urban areas.[12] In patients surviving long enough to reach the hospital, there is a threefold increase in risk of ED death among those injured in a region with limited access to a trauma care center, according to Gomez et al.[13] Theories to explain this discrepancy include differences in pre-hospital care, delays in transit time, prolonged discovery time, and severity of illness.[14,15]

In 1994, Karsteadt et al. examined this issue by comparing survival rates of rural trauma patients before and after implementing the MRTP.[5] They compared the actual survival rates of trauma patients to those predicted by TRISS methodology and found significant improvement in survival. Similarly, Zulick et al. examined survival at their rural level II trauma center and found that it was similar to the predictions of TRISS and comparable with care at urban level I trauma hospitals.[16] Norwood and Myers compared mortality rates at their rural-based level I trauma center in Urbana, IL with the MTOS.[17] Despite low patient volumes in their hospital, they observed a similar survival comparison to the MTOS population. Both Norwood and Zulick attributed their success to an organized trauma team with set protocols and a dedicated ED staff. Norwood noted that standardized orders and protocols were valuable as they helped maintain care standards across multiple providers. In Tulsa, OK where the city's population is 350,000, Thompson et al. examined the survival of injury patients to their level II trauma center.[18] When comparing their community hospital with predictions by MTOS, they found that outcomes were similar.

In this study, we compared the outcomes of two urban level I trauma centers to eight rural hospitals. Few studies have compared large hospital systems in this manner when evaluating trauma patient outcomes. Certain difficulties apply when comparing such a large and diverse patient cohort, such as differences in patient demographics and injury patterns in the two systems. In addition, there may be geographic and other uncontrollable factors that influence outcomes in ways not appreciated in our comparison. However, we found that overall survival was indistinguishable between rural and urban hospitals after correcting for differences in patient populations.

Our study reaffirms the assertion that rural trauma hospitals can have outcomes comparable to urban hospitals in survival, despite low patient volume. Several studies have noted that modest interventions such as set protocols for trauma, early activation of a trauma team, and standardized orders helped increase survival without the need for a formal level I trauma hospital. Our study supports this notion with a large patient cohort in rural northern California.

CONCLUSION

This study demonstrates that the overall mortality between the urban and rural trauma systems was not significantly different. A rural trauma system, with low volume and high blunt trauma rates, can effectively care for its population of trauma patients with an enhanced, committed trauma system, which allows for expeditious movement of patients toward definitive care, without hospital bypass of major trauma patients to designated trauma centers.