Abstract
Objective:
To identify predictors of under-triage among older injured Medicare beneficiaries, identify any regions where under-triage is more likely to occur, and examine additional factors associated with under-triage at a national level.
Methods:
Using 2008-2014 Medicare claims data, we identified older adults (age ≥ 65) diagnosed with traumatic injury, and linked claims with trauma center designation records, from the American Trauma Society. Under-triage was defined as NTC treatment with an injury severity score (ISS) ≥16, consistent with the ACS-COT benchmark. We used multivariable logistic regression to estimate odds of under-triage by census region, adjusting for sex, race, age, ISS, TC proximity and mode of transportation.
Results:
Of the 7,920,026 eligible trauma patients identified, 26% were treated at TC (n=2,095,491) and 74% at a NTC (n=5,824,535). Among eligible patients, 3.4% (n=272,858) were considered severely injured. Forty-six percent of severely injured patients (n=126,081) were treated at a NTC. Compared to patients in the Northeast, adjusted odds of under-triage were 72% higher for patients in Southern states (OR 1.72, 95% CI 1.67-1.73) and 51% higher in Western states (OR 1.51, 95% CI 1.47-1.55). Patients in the Midwest had lower odds of being under-triaged (OR 0.85, 95% CI 0.83-0.87).
Conclusion:
Nearly half of older adult trauma patients are under-triaged; it increases with age, distance to care, and is most common in Southern and Western states. Improvements to field triage and trauma center access for older patients are urgently needed.
Introduction
Older adults (≥ 65 years) are the fastest growing age group in the United States (US). This population is projected to increase by 105% from 2015 (47.8 million people) to 2060 (98.2 million people), which will represent nearly one quarter of the US population.1 Traumatic injury and mortality in older adults are following the same trend, with falls representing more than half (55%) of unintentional death by injury in older adults in the US.2 Treating traumatic injuries in this population has intrinsic challenges, due to concomitant preexisting medical conditions and polypharmacy, which have important implications for field-triage, in-patient care, and long-term functional outcomes after trauma.3 Trauma care systems need to evolve accordingly to meet the growing burden of older adult trauma, yet limited resources are currently employed to better understand and improve trauma care outcomes for older adults.
The most recent Centers for Disease Control and Prevention (CDC) Guidelines for Field Triage of Injured Patients,4 updated in 2011, call for the “highest level care” within the trauma system for patients meeting criteria. Level I and Level II TCs provide the highest level of care and offer definitive care for all injured patients. Moreover, the American College of Surgeons Committee on Trauma (ACS-COT), benchmarked the use of ISS ≥16 nationally to define the patient population who would benefit from treatment at a Level I or Level II TC.5 Table 1 outlines key differences amongst TC level of designation.6 These guidelines acknowledged under-triage as an issue, and include special considerations for older adults due to elevated risk of poor injury outcomes. Such considerations include different thresholds for systolic blood pressure (SBP) indicative of traumatic shock (SBP ≤ 90 mmHg for adults younger than 65, and SBP ≤ 110 for older adults) and increased priority for low-impact mechanisms (e.g. ground level falls). TC care has been shown to have increased probability of survival in older adult trauma patients.7 However, several studies suggest that older adults may not fully benefit from advanced trauma care systems in the US as numerous older adult patients are under-triaged to NTC.8-11 Reported under-triage rates are variable and range from 33% to 49.9% in regional,8 and multiregional10,12 studies with mixed findings on mortality, and complications.9,10,13 There are two national studies of under-triage to our knowledge,11,14 one of them focusing on older adults.11 Both used encounter-level data from the Nationwide Emergency Department Sample,15 which suffers from several key limitations related to triage decisions, including lack of information on mode of prehospital transportation, and the inability to determine the source of definitive care for transfer patients.
Table 1.
Trauma Center (TC) designation and patient care criteria, by level of designation: Level I and II TCs have far more resources than lower level TCs and Non-Trauma Centers (NTC).
| Trauma Center Level |
Minimum Patient Threshold |
24/7 Trauma Surgeon |
24/7 Anesthesia Services |
24/7 Neuro-surgeon |
24/7 Orthopedic Surgeon |
24/7 Operating Room |
PT*, OT**, and Rehab Available |
Transfer Complex Patients |
|---|---|---|---|---|---|---|---|---|
| I | Yes | In-house | In-house | On-call | On-call | Yes | Yes | No |
| II | No | On-Call*** | In-house | On-call | On-call | Yes | Yes | Possibly |
| III | No | On-Call | On-Call | No | On-call | Yes | PT Only | Yes |
| IV/V/NTC | No | No | No | No | No | No | No | Yes |
PT = physical therapists
OT = occupational therapists
Senior Resident in house with a Trauma Surgeon available in <15 mins
Medicare covers around 46 million Americans aged 65 years or older,16 regardless of income or health status across the US, and therefore provides a unique analytic vantage point to better understand traumatic injury in a national sample of older adults.17 Studying under-triage among Medicare beneficiaries allows us to validate prior estimates of under-triage, and identify potential determinants of under-triage, including mode of transportation and transfer status. These factors may be sensitive to policy interventions designed to ensure quality and efficiency of care as trauma health care systems grow and adapt to meet the increasing demand for services.
Our objective was to identify predictors of under-triage for older injured patients at a national level, identify any regions where under-triage is more likely to occur, and examine additional factors associated with under-triage. We hypothesize that patient, injury severity, and geographic characteristics independently predict treatment by TC level.
Methods
Data source
Using Medicare claims data for inpatient and emergency department encounters from 2008-2014, we identified trauma patients aged ≥ 65 years based on International Classification of Disease, 9th Revision, Clinical Modification (ICD-9-CM) diagnosis codes for traumatic injury (800.0-959.9, excluding 905-909, 910-924, and 930-939). Medicare data was obtained from the Centers for Medicare and Medicaid Services (CMS) through virtual access to the CMS Virtual Research Data Center (VRDC), managed by the Research Data Assistance Center (ResDAC).18 We linked Medicare claims with hospital and TC identification information from the American Trauma Society (ATS) Trauma Information Exchange Program (TIEP)19 and the American Hospital Association (AHA) Annual Survey.20The TIEP maintains a database of hospital characteristics for TC in the US, and shares data with institutional members of the ATS. The AHA conducts an annual survey of hospitals in the US, including characteristics of NTC not included in the TIEP data. This allowed us to determine the facilities where older patients (≥ 65y) with traumatic injury received care in the US from 2008-2014 and identify factors associated with under-triage. This study was approved and exempt by our Institutional Review Board.
Variables
We included the following variables, derived from Medicare patient characteristics and claims data in our analyses: TC characteristics, demographic characteristics, census region, distance to a trauma center (miles), mode of transportation (ambulance vs. private), injury severity, and fall as a mechanism of injury.TC care was defined as definitive treatment at a hospital designated as Level I or II, based on verification from the ACS-COT, or from a state agency responsible for trauma system oversight. NTC care was defined as definitive treatment at a designated Level III/IV/V TC or at a hospital without TC verification. Demographic variables included age (65-69, 70-75, 76-80, > 80), sex (male, female), race (White, Black, Asian, North American Native, Other), census region (Midwest, Northeast, South and West), and mode of transport (ambulance vs. private vehicle). Trauma center proximity was measured based on estimated driving distance from the population-weighted centroid of each patient’s residential ZIP code to the nearest trauma center. Driving distances were estimated using road network data from OpenStreetMap (Open StreetMap Foundation)21 and a proprietary algorithm developed by (Redivis, Inc)22, and were categorized as ≤ 15 miles, 15-30 miles, or ≥ 30 miles. Driving distances based on population-weighted centroids have been validated as a measure of access to tertiary medical care,23 and specifically examined as a proxy for ambulance driving distance in preliminary work associated with this analysis.24 ISS was calculated based on ICD-9-CM codes, using an adapted version of the validated25 ICD Programs for Injury Categorization (ICDPIC)26 for SAS, and categorized as ≤ 9, 10-15, 16-25, or ≥ 25. Falls were identified using ICD-9-CM codes E880.0-E899 (unintentional).
Outcomes
The primary outcome of interest was under-triage, defined as NTC care for patients with an ISS ≥16, aligned with the Field Triage Guidelines4 of injured patients, and the ACS-COT5 benchmark. Patients who were initially treated at a NTC and subsequently transferred to a TC were considered properly triaged in our analyses. To account for patients’ dead on arrival we excluded those who died in the emergency department prior to inpatient admission. Secondary outcomes included census region, distance to trauma center care, and mode of transportation.
Analysis
We used descriptive statistics (chi-square for categorical variables and t-tests, ANOVA or Kruskal-Wallis for continuous variables) to examine demographic, geographic, transport and injury characteristics overall, and comparing TCs and NTCs. We used multivariable logistic regression to estimate odds of under-triage, adjusting for sex, race, age, ISS, TC proximity, and mode of transportation. We also conducted subgroup analysis to examine patterns of under-triage by trauma center proximity and mode of transportation. Statistical analyses were performed in SAS, version 9.4 (SAS Institute, Cary, NC).
Results
Among 7.9 million (n=7,920,026) trauma patients identified, 26% were treated at TC (n=2,095,491) and the rest at NTC (74%, n=5,824,535). Approximately one quarter of older adults treated at TC and three quarters of older adults treated at NTC, consistently with slight variations across age groups, gender, race, ambulance transport, and injury mechanism (falls) for all Medicare Trauma patients (Table 2). TC treatment was most common in the Midwest (38%, n=683,359) and least common in the South (19%, n=639,119). Eighty percent (n=11,863) of patients with an ISS >25 were treated at a TC, and 75%(n=5,569,452) of those with an ISS ≤9 at an NTC. Falls were the mechanism of injury for half of our patient population (n=4,005,682).
Table 2.
Trauma Center Demographics and Characteristics for Medicare Patients (≥ 65 y) Overall and Severely Injured (ISS ≥16), 2008-2014.
| All Medicare Trauma Patients | Subset of Medicare Trauma Patients with an ISS ≥16 (3.4%) | |||||
|---|---|---|---|---|---|---|
| Total N=7,920,026 |
Trauma Center N=2,095,491 (26%) |
Non-Trauma Center N= 5,824,535 (74%) |
Total ISS ≥16 N= 272,858 |
Trauma Center ISS ≥16 N=146,777 (54%) |
Non-Trauma Center ISS ≥16 N= 126,081 (46%) |
|
| Age Mean (SD) | 80.1 | 80.2 (8.4) | 79.9 (8.5) | 80.9 | 80.9(8.0) | 82.2 (7.9) |
| Age Group | ||||||
| 65-69 | 1,117,637 | 25.4% | 74.6% | 24,121 | 61.9% | 38.1% |
| 70-75 | 1,558,318 | 25.6% | 74.4% | 44,048 | 58.1% | 41.9% |
| 76-80 | 1,381,928 | 26.3% | 73.7% | 49,343 | 55.2% | 44.8% |
| >80 | 3,862,143 | 27.2% | 72.8% | 155,346 | 50.9% | 49.1% |
| Sex | ||||||
| FEMALE | 5,223,471 | 26% | 74% | 146,413 | 52.3% | 47.7% |
| MALE | 2,696,555 | 27.4% | 72.6% | 126,445 | 55.6% | 44.4% |
| Race | ||||||
| WHITE | 7,183,171 | 26.2% | 73.8% | 245,073 | 54.1% | 45.9% |
| BLACK | 437,414 | 29.9% | 70.1% | 13,349 | 54.3% | 45.7% |
| HISPANIC | 120,268 | 25.4% | 74.6% | 4,719 | 44.6% | 55.4% |
| ASIAN | 78,198 | 29.1% | 70.9% | 4,983 | 47.3% | 52.7% |
| NORTH AMERICAN NATIVE | 27,789 | 23.3% | 76.7% | 1,150 | 53% | 47% |
| OTHER/NA | 14,075 | 28.5% | 71.7% | 3,584 | 53.3% | 46.7% |
| Region | ||||||
| Midwest | 1,796,634 | 38% | 62% | 63,239 | 63.5% | 36.5% |
| Northeast | 1,549,660 | 29.6% | 70.4% | 54,990 | 59.7% | 40.3% |
| South | 3,342,024 | 19.1% | 80.9% | 110,689 | 46.9% | 53.1% |
| West | 1,231,708 | 25.6% | 74.4% | 43,940 | 49.8% | 50.2% |
| Distance to Trauma Center (miles) | ||||||
| 0 -15 | 5,386,841 | 32.1% | 67.9% | 186,600 | 55.5% | 44.5% |
| 15- 30 | 1,398,691 | 16.7% | 83.3% | 47,626 | 51.5% | 48.5% |
| > 30 | 1,134,494 | 11.7% | 88.3% | 38,632 | 48.2% | 51.8% |
| Transported by Ambulance | ||||||
| No | 5,120,956 | 25.2% | 74.8% | 144,556 | 55.9% | 44.1% |
| Yes | 2,799,070 | 28.8% | 71.2% | 128,302 | 51.4% | 48.6% |
| Transported by Private vehicle | ||||||
| No | 7,758,805 | 26.1% | 73.9% | 252,050 | 51.5% | 48.5% |
| Yes | 161,221 | 44.5% | 55.5% | 20,808 | 81.3% | 18.7% |
| Injury Severity Score (ISS) Median (IQR) | 4.0 (8.00) | 4.0 (4.00) | 17 (4.00) | 16.0 (1.00) | ||
| <=9 | 7,422,387 | 25% | 75% | / | / | / |
| 10-15 | 224,781 | 41.1% | 58.9% | / | / | / |
| 16-25 | 258,054 | 52.3% | 47.7% | 258,054 | 52.3% | 47.7% |
| >25 | 14,804 | 80.1% | 19.9% | 14,804 | 80.1% | 19.9% |
| ISS ≥16 | ||||||
| No | 7,647,168 | 25.5% | 74.5% | / | / | / |
| Yes | 272,858 | 53.8% | 46.2% | 272,858 | 53.8% | 46.2% |
| FALL | ||||||
| No | 3,914,344 | 25% | 75% | 95,433 | 57% | 43% |
| Yes | 4,005,682 | 27.9% | 72.1% | 177,425 | 52.1% | 47.9% |
Under-triage
Incidence of under-triage varied by census region, trauma center proximity, and by ISS category. Of the injury patients identified in Medicare claims, 3.4% (n=272,858) had an ISS ≥ 16 (Table 2). Fifty four percent of these patients (n=146,777/272,858) were treated at a TC and 46% (n=126,081/272,858) at a NTC (Table 2). Approximately 50% older adults were treated in both TC and NTC, across most categories. Differences were observed by age group, census region, private transportation, and ISS as follows. Incidence of under-triage was lowest among patients age 65-69 (38%, n=9,188), and highest among those age >80 (49%, n=76,314). Similarly, under-triage was lowest in the Midwest (37%, n=23,072), and highest in the South (53%, n=58,800). Nineteen percent (n=3,901) of older adults traveled by private vehicle to a NTC. Twenty percent (n= 2,941) of older adults with an ISS >25 were treated at a NTC. Compared to patients in the Northeast, adjusted odds of under-triage were 70% higher for older patients in Southern states (OR 1.70, 95% CI 1.66-1.73; Table 3) and 51% higher in Western states (OR 1.51, 95% CI 1.47-1.55). Older patients in the Midwest, had lower odds of being under-triaged (OR 0.85, 95% CI 0.83-0.87). Odds of under-triage were higher for female patients, compared to male patients (OR 1.09, 95% Cl 1.07-1.11). Hispanic (OR 1.33, 95% Cl 1.25-1.41) and for Asian (OR 1.28, 95% Cl 1.21-1.35) patients also had higher odds of under-triage compared to White patients. Compared to patients age 65-69, odds of under-triage increased by 53% for patients age > 80 (OR 1.53, 95% CI 1.48, 1.57). The odds of under-triage increased by 270% (OR 3.70, 95% Cl 3.55-3.85) for very severely injured patients (ISS >25).
Table 3.
Multivariable logistic regression for under-triage in Medicare Patients with and ISS ≥16.
| Parameter | Odds ratio |
95% confidence interval | p value | |
|---|---|---|---|---|
| Region | Northeast | Reference | ||
| Midwest | 0.85 | (0.83-0.87) | <.0001 | |
| South | 1.70 | (1.66-1.73) | <.0001 | |
| West | 1.51 | (1.47-1.55) | <.0001 | |
| Sex | Male | Reference | ||
| Female | 1.09 | (1.07-1.11) | <.0001 | |
| Race | White | Reference | ||
| Black | 0.99 | (0.96-1.03) | 0.6365 | |
| Hispanic | 1.33 | (1.25-1.41) | <.0001 | |
| Asian | 1.28 | (1.21-1.35) | <.0001 | |
| North American Native | 0.99 | (0.84-1.07) | 0.4110 | |
| Other | 1.10 | (1.02-1.18) | 0.0109 | |
| Age Group | 65-69 | Reference | ||
| 70-75 | 1.15 | (1.12-1.19) | <.0001 | |
| 76-80 | 1.23 | (1.25-1.33) | <.0001 | |
| >80 | 1.53 | (1.48-1.57) | <.0001 | |
| ISS | 16-25 | Reference | ||
| >25 | 3.70 | (3.55-3.85) | <.0001 | |
| Distance to Trauma Center (miles) | 0 -15 | Reference | ||
| 15-30 | 1.18 | (1.15-1.20) | <.0001 | |
| > 30 | 1.38 | (1.35-1.41) | <.0001 | |
| Transported by Ambulance | Yes | Reference | ||
| No | 1.19 | (1.17-1.21) | <.0001 | |
Distance to TC
Among patients who lived within 0-15 miles of a TC, 32% (n=1,729,788) were treated at a TC, while 17% (n=233,217) of patients who lived 15-30 miles from a TC, and 12% (n=132,486) of patients who lived more than 30 miles from a TC received TC treatment (Table 2). Distance to a TC was associated with increasing odds of under-triage with 38% (OR 1.38, 95% Cl 1.35-1.41) higher odds for older adults living more than 30 miles from a TC compared to patients within 15 miles (Table 2).
In multivariable subgroup analysis of older adults with an ISS ≥16, stratified by distance to trauma center care, the association between patient age and under-triage was more pronounced for patients who lived more than 30 miles from a TC, with odds of under-triage increasing by 59% for patients older than 80 years (OR 1.59, 95% CI 1.48-1.70; Table 4). Similarly, the odds of treatment at a NTC increased to 69% (OR 1.69, 95% CI 1.62-1.76) for patients who were transported by private vehicle and lived more than 30 miles from a TC. Conversely, the association between census region and under-triage was attenuated for patients who lived within 30 miles from a TC, with 81% higher odds of under-triage in Southern states (OR 1.81, 95% CI 1.77-1.85), and 52% higher odds of under-triage in Western states (OR 1.52, 95% CI 1.48-1.56). Midwest states had 17% lower odds of under-triage (OR 0.83, 95% CI 0.81-0.85) when compared to the Northeast for patients living within 30 miles of TC care. Very severely injured patients had increased odds of under-triage in both distance to TC groups, 267% (OR 3.67, 95% CI 3.50-3.84) within 30 miles, and 271% (OR 3.71, 95% CI 3.37-4.09) more than 30 miles, when compared to older adults with ISS 16-25.
Table 4.
Multivariable logistic regression for under-triage in Medicare Patients with and ISS ≥16, and distance to Trauma Center care.
| Parameter | Trauma Center within 30 Miles | Trauma Center more than 30 Miles | |||||
|---|---|---|---|---|---|---|---|
| Odds ratio | 95% confidence interval | p value | Odds ratio | 95% confidence interval | p value | ||
| Region | Northeast | Reference | Reference | ||||
| Midwest | 0.83 | 0.81-0.85 | <.0001 | 0.90 | 0.83-0.97 | 0.0056 | |
| South | 1.81 | 1.77-1.85 | <.0001 | 1.16 | 1.08-1.25 | <.0001 | |
| West | 1.52 | 1.48-1.56 | <.0001 | 1.34 | 1.23-1.46 | <.0001 | |
| Sex | Male | Reference | Reference | ||||
| Female | 1.08 | 1.06-1.10 | <.0001 | 1.15 | 1.11-1.20 | <.0001 | |
| Race | White | Reference | |||||
| Black | 0.97 | 0.93-1.01 | 0.1074 | 1.13 | 1.01-1.26 | 0.0333 | |
| Hispanic | 1.30 | 1.23-1.39 | <.0001 | 1.15 | 0.93-1.43 | 0.1983 | |
| Asian | 1.24 | 1.17-1.32 | <.0001 | 1.64 | 1.18-2.27 | 0.0033 | |
| North American Native | 1.11 | 0.96-1.29 | 0.143 | 0.71 | 0.57-0.88 | 0.0016 | |
| Other | 1.06 | 0.99-1.15 | 0.117 | 1.38 | 1.05-1.81 | 0.0209 | |
| Age Group | 65-69 | Reference | Reference | ||||
| 70-75 | 1.16 | 1.12-1.20 | <.0001 | 1.13 | 1.04-1.22 | 0.0028 | |
| 76-80 | 1.28 | 1.24-1.33 | <.0001 | 1.29 | 1.19-1.39 | <.0001 | |
| >80 | 1.51 | 1.46-1.55 | <.0001 | 1.59 | 1.48-1.70 | <.0001 | |
| ISS (≥16) | 16-25 | Reference | Reference | ||||
| >25 | 3.67 | 3.50-3.84 | <.0001 | 3.71 | 3.37-4.09 | <.0001 | |
| Transported by Ambulance | Yes | Reference | Reference | ||||
| No | 1.12 | 1.10-1.14 | <.0001 | 1.69 | 1.62-1.76 | <.0001 | |
Mode of transportation
Fifty-one percent and forty-nine percent of severely injured older adults were transported by ambulance at TC and NTC respectively. Eighty-one percent and nineteen percent of severely injured older adults were transported by private vehicle at TC and NTC respectively (Table 2). Compared to older adults transported by ambulance, the odds of receiving treatment at a NTC were 19% higher for patients using private transportation (OR 1.19, 95% Cl 1.17-1.21, Table 2). In multivariable subgroup analysis of patients with an ISS ≥16, stratified by mode of transportation (ambulance vs. private vehicle), the estimated effects of ISS and trauma center proximity were more pronounced for patients transported by private vehicle (Table 5). Among private vehicle transports, ISS > 25 was associated with a four-fold increase in odds of treatment at a NTC (OR 4.20, 95% CI 3.95-4.46). The association between trauma center proximity and under-triage was more pronounced among patients transported by ambulance with a 39% increase in odds of under-triage for patients who lived within 15-30 miles of a TC (OR 1.39, 95% CI 1.34-1.43), and 80% increase in odds of treatment at a NTC for patients who lived more than 30 miles from a TC (OR 1.80, 95% CI 1.74-1.86). Conversely, the association between trauma center proximity and treatment at a NTC was attenuated for patients transported by private vehicle. The associations between census region, sex, race, and age group were similar for both modes of transportation.
Table 5.
Multivariable logistic regression for undertriage in Medicare Patients with and ISS ≥16, and means of transportation to a Trauma Center.
| Parameter | Transported by Ambulance | Private vehicle | |||||
|---|---|---|---|---|---|---|---|
| Odds ratio | 95% confidence interval | p value | Odds ratio | 95% confidence interval | p value | ||
| Region | Northeast | Reference | Reference | ||||
| Midwest | 0.84 | 0.81-0.87 | <.0001 | 0.87 | 0.84-0.90 | <.0001 | |
| South | 1.71 | 1.65-1.76 | <.0001 | 1.70 | 1.65-1.75 | <.0001 | |
| West | 1.48 | 1.43-1.54 | <.0001 | 1.55 | 1.50-1.61 | <.0001 | |
| Sex | Male | Reference | Reference | ||||
| Female | 1.11 | 1.09-1.13 | <.0001 | 1.07 | 1.05-1.10 | <.0001 | |
| Race | White | Reference | Reference | ||||
| Black | 0.99 | 0.93-1.04 | 0.2732 | 0.99 | 0.95-1.04 | 0.7808 | |
| Hispanic | 1.32 | 1.20-1.44 | <.0001 | 1.32 | 1.22-1.43 | <.0001 | |
| Asian | 1.19 | 1.09-1.30 | 0.0001 | 1.34 | 1.24-1.45 | <.0001 | |
| North American Native | 1.17 | 0.95-1.43 | 0.5793 | 0.87 | 0.75-1.01 | 0.0720 | |
| Other | 1.05 | 0.94-1.17 | 0.2456 | 1.13 | 1.03-1.25 | 0.0105 | |
| Age Group | 65-69 | Reference | Reference | ||||
| 70-75 | 1.15 | 1.09-1.21 | <.0001 | 1.16 | 1.11-1.21 | <.0001 | |
| 76-80 | 1.31 | 1.24-1.38 | <.0001 | 1.27 | 1.22-1.33 | <.0001 | |
| >80 | 1.57 | 1.50-1.65 | <.0001 | 1.49 | 1.43-1.54 | <.0001 | |
| ISS (≥16) | 16-25 | Reference | Reference | ||||
| >25 | 3.27 | 3.09-3.47 | <.0001 | 4.20 | 3.95-4.46 | <.0001 | |
| Distance to Trauma Center (miles) | 0 -15 | Reference | Reference | ||||
| 15- 30 | 1.39 | 1.34-1.43 | 0.0477 | 1.02 | 0.99-1.05 | 0.2374 | |
| > 30 | 1.80 | 1.74-1.86 | <.0001 | 1.12 | 1.08-1.15 | <.0001 | |
Limitations
There are several limitations that need to be acknowledged in our study. First, this study is based on a retrospective review of Medicare claims data. The most relevant to our study include, confounding, missing data, the fact that there are no timestamps, no vital signs or ancillary test results reported, and that data is obtained from billing data; a more extensive description on limitations associated with this study design has been described elsewhere.17,27,28 Second, we are determining if an older adult was “adequately triaged” based on Medicare claims post-factually. In other words, we do not have access to pre-hospital level data which could provide further insights towards the rational of EMS personnel or family members to seek care. Unfortunately, no reliable pre-hospital databases that would allow national analyses of this magnitude exist at this time. Standardized collection of pre-hospital trauma care is urgently needed to improve quality of care and outcomes for injured patients. Third, we rely on ISS, CDC guidelines, and ACS-COT benchmark to determine under-triage of older adults, but these criteria have not been validated in this population and could inappropriately categorize need for TC care. Older adults present with well documented differences in physiologic response to injury,29,30 preexisting medical conditions, polypharmacy,3,31 and mechanisms of injury. These known differences have led to an Ad Hoc Geriatric Trauma Committee,32 which calls for reassessment and standardization of triage criteria for geriatric trauma patients.33 Finally, although we excluded patients who died in the emergency room, and considered older adults transferred from NTC to TC to be appropriately triaged, it is possible that some patients were not stable enough to be transferred to TC care or died as inpatients at NTC. Despite these limitations, this study used a comprehensive dataset that enabled us to examine national patterns of under-triage, including examination of several factors that were not assessed in previous studies. Medicare data used in this study is the best available data to examine under-triage of older adults in the absence of nationally standardized pre-hospital data collection.
Discussion
This is the first national analysis to quantify the magnitude and identify predictors of under-triage among injured Medicare beneficiaries. Our results echo findings from previous national, regional, and state8-12,14 analyses. This study offers unique sub-analyses on distance traveled to receive care and mode of transportation used (ambulance vs. private vehicle). Almost half (46%) of severely injured patients in our study received care at a NTC despite the CDC recommendation that older injured patients be triaged to the highest level of trauma care available.4 We also found higher odds of under-triage in Southern and Western states, with increased age, higher injury severity, and increased distance to TC care.
Regional and Distance Findings
Our results suggest that incidence of under-triage is highest in Southern and Western states, a pattern that persists across injury severity categories. One potential explanation for this pattern could be the increased older population living in the South34 vs. the Midwest, and the geographic distribution of the population3 relative to TC and NTC locations.35 “Under-triage” or treatment at NTC could reflect lack of access to a Level I or Level II. In fact, 69.2% of the US population lived within 45 minutes’ access to a TC.36 In our study, 14.3% of older adults (lived more than 30-miles from a TC, indicating that a considerable number of older adults, likely from rural areas, lack access to TC care. Living more than 50 miles away form a TC has been described as a predictor of not receiving TC care, though our results suggest that TC distances as short as 15 miles’ impact likelihood of TC care.37 Further research is needed to better understand driving factors of under-triage at local and state levels, as well as distance and access to TC care in rural regions.
Patient and Family Preference Findings
Other potential determinants of under-triage include the high incidence of fall-related injuries in our older adults and the influence of patients or family preferences in hospital destination. Prehospital providers and caregivers making triage decisions might underestimate the severity of fall-related injuries, a mechanism that is highly prevalent and especially worrisome for older adults due to increased risk of death associated with aspirin use and intracranial bleed for older patients falling from their own height.31 Patient or family choice has also been shown to influence hospital selection by emergency medical services (EMS) providers after accounting for field triage protocols, particularly as age increases, up to 75.8% among patients older than 90 years old.38 There are important implications of these findings. First, although the revised Field Triage Guidelines4 have incorporated special considerations for older adults, evidence suggest that “Geriatric-Specific” EMS triage guidelines might be granted in this population.33 Having specific geriatric field triage guidelines might help reduce under-triage driven by patient or family choice. Some of the above-mentioned factors, such as patient/family choice-previous treatment at a NTC, or initial ISS, might be contributing to the increasing under-triaged rates by distance for older adults transported by ambulance. Moreover, analyses of outcomes by mechanism of injury and within TC levels are beyond the scope of this paper. Ciesla and colleagues,39 showed that severely injured patients do not necessarily require TC treatment for optimal care. Further research is needed to better understand outcomes of different injury mechanisms in these patients. Second, our study shows that 80% of severely injured adults driven by private vehicles were treated at a TC vs 20% at NTC. This might be attributable to proximity to TC, further research is granted to determine the reasons behind these decisions in lay persons. Nevertheless, a national public health intervention to raise awareness about the implications of ground level falls in older adults, could potentially lead to an informed adequate level of care selection, and thus reducing morbidity and mortality in severely injured older adults. The relationship between mode of transportation, triage decisions, and patient preference warrants further investigation to support improved triage protocols.
Under-triage in older adults is a complex multilayered issue that cannot be explained by one factor alone, but rather an interconnection of factors at a patient level (comorbid conditions, polypharmacy, frailty, mechanism of injury) and at a system level (pre-hospital: access, in-hospital: capacity, and post-hospital: rehabilitation) to name a few. Assessing pre-hospital outcomes, in-hospital outcomes and long-term outcomes for severely injured older adults is greatly needed to better understand the factors behind these results.
This analysis confirms the previously noted patterns of under-triage for older injury patients in a national sample. Nationally, nearly half of severely injured older adult trauma patients ≥ 65 are under triaged to a NTC. This problem appears to be particularly worse in the Southern US, despite the fact that Field Triage Guidelines4 currently call for expedited transport of older trauma patients to TC. Severely injured older adults are less likely to be under-triaged if they live within 30 miles from TC care and are transported by ambulance. Tools that improve field triage and enhance access to trauma centers for older patients are urgently needed.
To sum up, reducing fatal injuries continues to be a growing public health concern, as it was outlined by the Healthy People 2020.40 Although progress has been made to advance the areas of specific research established in the Field Triage Guidelines4, our results highlight the need for ongoing assessment and policy interventions. Our findings offer important information for stake holders with health policy implications to improve trauma systems. Limited resources can be targeted to improving access to TC care in the Southern and Western states and improving field-triage and access to quality care, particularly for patients with falls.
Acknowledgments
This work is funded by the National Institute on Aging (NIA) grant #1R56AG048452-01A1: “Comparative Effectiveness of Trauma Center Care for Older Americans”.
Footnotes
The authors declare that we have no conflicts of interest relevant to this work to report.
References
- 1.U.S. Census Bureau, Population Division. Table 3. Projections of the Population by Sex and Selected Age Groups for the United States: 2015 to 2060 (NP2014-T3). December 2014. [Google Scholar]
- 2.Kramarow E, Chen LH, Hedegaard H, Warner M. Deaths from unintentional injury among adults aged 65 and over: United States, 2000-2013. NCHS data brief. 2015(199):199. [PubMed] [Google Scholar]
- 3.Bonne S, Schuerer DJ. Trauma in the older adult: epidemiology and evolving geriatric trauma principles. Clinics in geriatric medicine. 2013;29(1):137–150. [DOI] [PubMed] [Google Scholar]
- 4.Sasser SM, Hunt RC, Faul M, et al. Guidelines for field triage of injured patients: recommendations of the National Expert Panel on Field Triage, 2011. MMWR Recommendations and reports : Morbidity and mortality weekly report Recommendations and reports. 2012;61(Rr-1):1–20. [PubMed] [Google Scholar]
- 5.Resources for Optimal Care of the Injured Patient. 6th. ed: American College of Surgeons.; 2014. Available at: https://www.facs.org/~/media/files/quality%20programs/trauma/vrc%20resources/resources%20for%20optimal%20care.ashx Accessed on June 6, 2018. [Google Scholar]
- 6.American College of Surgeons. Resources for Optimal Care of the Injured Patient 2014/Resources Repository. .
- 7.Pracht EE, Langland-Orban B, Flint L. Survival advantage for elderly trauma patients treated in a designated trauma center. The Journal of trauma. 2011;71(1):69–77. [DOI] [PubMed] [Google Scholar]
- 8.Chang DC, Bass RR, Cornwell EE, Mackenzie EJ. Undertriage of elderly trauma patients to state-designated trauma centers. Archives of surgery (Chicago, Ill : 1960). 2008;143(8):776–781; discussion 782. [DOI] [PubMed] [Google Scholar]
- 9.Lehmann R, Beekley A, Casey L, Salim A, Martin M. The impact of advanced age on trauma triage decisions and outcomes: a statewide analysis. American journal of surgery. 2009;197(5):571–574; discussion 574-575. [DOI] [PubMed] [Google Scholar]
- 10.Staudenmayer KL, Hsia RY, Mann NC, Spain DA, Newgard CD. Triage of elderly trauma patients: a population-based perspective. Journal of the American College of Surgeons. 2013;217(4):569–576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kodadek LM, Selvarajah S, Velopulos CG, Haut ER, Haider AH. Undertriage of older trauma patients: is this a national phenomenon? The Journal of surgical research. 2015;199(1):220–229. [DOI] [PubMed] [Google Scholar]
- 12.Newgard CD, Fu R, Zive D, et al. Prospective Validation of the National Field Triage Guidelines for Identifying Seriously Injured Persons. Journal of the American College of Surgeons. 2016;222(2):146–158.e142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Haas B, Gomez D, Zagorski B, Stukel TA, Rubenfeld GD, Nathens AB. Survival of the fittest: the hidden cost of undertriage of major trauma. Journal of the American College of Surgeons. 2010;211(6):804–811. [DOI] [PubMed] [Google Scholar]
- 14.Xiang H, Wheeler KK, Groner JI, Shi J, Haley KJ. Undertriage of major trauma patients in the US emergency departments. The American journal of emergency medicine. 2014;32(9):997–1004. [DOI] [PubMed] [Google Scholar]
- 15.Healthcare Cost and Utilization Project (HCUP). Overview of the Nationwide Emergency Department Sample (NEDS). Available at: https://www.hcup-us.ahrq.gov/nedsoverview.jsp Accessed on June 26, 2018. [Google Scholar]
- 16.Altman D, Frist WH. Medicare and Medicaid at 50 Years: Perspectives of Beneficiaries, Health Care Professionals and Institutions, and Policy Makers. Jama. 2015;314(4):384–395. [DOI] [PubMed] [Google Scholar]
- 17.Ghaferi AA, Dimick JB. Practical Guide to Surgical Data Sets: Medicare Claims Data. JAMA surgery. 2018. [DOI] [PubMed] [Google Scholar]
- 18.Research Data Assistance Center (ResDAC), CMS Virtual Research Data Center (VRDC). https://www.resdac.org/cms-data/request/cms-virtual-research-data-center Accessed May 23, 2018. [Google Scholar]
- 19.American Trauma Society. https://www.amtrauma.org/default.aspx Accessed May 23, 2018. [Google Scholar]
- 20.American Hospital Association (AHA). © 2016. Health Forum LLC. [Google Scholar]
- 21.OpenStreetMap.United States. 2018. https://www.openstreetmap.org/#map=2/4.6/-103.0 Accessed May 23, 2018.
- 22.REDIVIS.data.(re)imagined. https://redivis.com/ Accessed May 23, 2018. [Google Scholar]
- 23.Berke EM, Shi X. Computing travel time when the exact address is unknown: a comparison of point and polygon ZIP code approximation methods. International journal of health geographics. 2009;8:23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Jarman MP, Sturgeon DJ, Mathews I, Uribe-Leitz T, Haider AH. GIS estimates for ambulance driving distance: estimated the prehospital interval for Medicare claims data. Poster accepted for presentation at the American Public Health Association annual meeting; November 12, 2018; San Diego, CA. [Google Scholar]
- 25.Fleischman RJ, Mann NC, Dai M, et al. Validating the Use of ICD-9 Code Mapping to Generate Injury Severity Scores. Journal of trauma nursing : the official journal of the Society of Trauma Nurses. 2017;24(1):4–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Clark David E. & Osler Turner M. & Hahn David R., 2009. "ICDPIC: Stata module to provide methods for translating International Classification of Diseases (Ninth Revision) diagnosis codes into standard injury categories and/or scores," Statistical Software Components S457028, Boston College Department of Economics, revised 29 Oct 2010. [Google Scholar]
- 27.Sarrazin MS, Rosenthal GE. Finding pure and simple truths with administrative data. Jama. 2012;307(13):1433–1435. [DOI] [PubMed] [Google Scholar]
- 28.Haut ER, Pronovost PJ, Schneider EB. Limitations of administrative databases. Jama. 2012;307(24):2589; author reply 2589-2590. [DOI] [PubMed] [Google Scholar]
- 29.Heffernan DS, Thakkar RK, Monaghan SF, et al. Normal presenting vital signs are unreliable in geriatric blunt trauma victims. The Journal of trauma. 2010;69(4):813–820. [DOI] [PubMed] [Google Scholar]
- 30.Martin JT, Alkhoury F, O'Connor JA, Kyriakides TC, Bonadies JA. 'Normal' vital signs belie occult hypoperfusion in geriatric trauma patients. The American surgeon. 2010;76(1):65–69. [PubMed] [Google Scholar]
- 31.Bhattacharya B, Maung A, Schuster K, Davis KA. The older they are the harder they fall: Injury patterns and outcomes by age after ground level falls. Injury. 2016;47(9):1955–1959. [DOI] [PubMed] [Google Scholar]
- 32.Kozar RA, Arbabi S, Stein DM, et al. Injury in the aged: Geriatric trauma care at the crossroads. The journal of trauma and acute care surgery. 2015;78(6):1197–1209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Ichwan B, Darbha S, Shah MN, et al. Geriatric-specific triage criteria are more sensitive than standard adult criteria in identifying need for trauma center care in injured older adults. Annals of emergency medicine. 2015;65(1):92–100.e103. [DOI] [PubMed] [Google Scholar]
- 34.Older Americans Month: May 2017. Profile America Facts for Features: CB17-FF.08 March 27, 2017. https://www.census.gov/content/dam/Census/newsroom/facts-for-features/2017/cb17-ff08.pdf Accessed May 25, 2018. [Google Scholar]
- 35.MacKenzie EJ, Hoyt DB, Sacra JC, et al. National inventory of hospital trauma centers. Jama. 2003;289(12):1515–1522. [DOI] [PubMed] [Google Scholar]
- 36.Branas CC, MacKenzie EJ, Williams JC, et al. Access to trauma centers in the United States. Jama. 2005;293(21):2626–2633. [DOI] [PubMed] [Google Scholar]
- 37.Hsia RY, Wang E, Saynina O, Wise P, Perez-Stable EJ, Auerbach A. Factors associated with trauma center use for elderly patients with trauma: a statewide analysis, 1999-2008. Archives of surgery (Chicago, Ill : 1960). 2011;146(5):585–592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Newgard CD, Mann NC, Hsia RY, et al. Patient choice in the selection of hospitals by 9-1-1 emergency medical services providers in trauma systems. Academic emergency medicine : official journal of the Society for Academic Emergency Medicine. 2013;20(9):911–919. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Ciesla DJ, Pracht EE, Tepas JJ 3rd, et al. Measuring trauma system performance: Right patient, right place-Mission accomplished? The journal of trauma and acute care surgery. 2015;79(2):263–268. [DOI] [PubMed] [Google Scholar]
- 40.HHS. Office of Disease Prevention and Health Promotion. Healthy People; 2020. Washington,DC. Available from: http://www.healthypeople.gov/2020 Accessed May 21, 2018. [Google Scholar]