Treatment outcomes of injured children at adult level 1 trauma centers: are there benefits from added specialized care?

Transcription

1 The American Journal of Surgery (2011) 201, Society of Black Academic Surgeons Treatment outcomes of injured children at adult level 1 trauma centers: are there benefits from added specialized care? Tolulope A. Oyetunji, M.D., M.P.H. a, *, Adil H. Haider, M.D., M.P.H., F.A.C.S. b, Stephanie R. Downing, M.D. a, Oluwaseyi B. Bolorunduro, M.D., M.P.H. a, David T. Efron, M.D., F.A.C.S. b, Elliott R. Haut, M.D., F.A.C.S. b, David C. Chang, M.P.H., M.B.A., Ph.D. d, Edward E. Cornwell III, M.D., F.A.C.S. a, Fizan Abdullah, M.D., Ph.D., F.A.C.S. c, Suryanarayana M. Siram, M.D., F.A.C.S. a a Department of Surgery, Howard University College of Medicine, Washington, DC, USA; b Department of Surgery and c Division of Pediatric Surgery, Johns Hopkins University School of Medicine, Baltimore, MD, USA; d Department of Surgery, University of California, San Diego School of Medicine, San Diego, CA, USA KEYWORDS: Pediatric; Trauma; Trauma center Abstract BACKGROUND: Accidental traumatic injury is the leading cause of morbidity and mortality in children. The authors hypothesized that no mortality difference should exist between children seen at ATC (adult trauma centers) versus ATC with added qualifications in pediatrics (ATC-AQ). METHODS: The National Trauma Data Bank, version 7.1, was analyzed for patients aged 18 years seen at level 1 trauma centers. Bivariate analysis compared patients by ATC versus ATC-AQ using demographic and injury characteristics. Multivariate analysis adjusting for injury and demographic factors was then performed. RESULTS: A total sample of 53,702 children was analyzed, with an overall mortality of 3.9%. The adjusted odds of mortality was 20% lower for children seen at ATC-AQ (odds ratio,.80; 95% confidence interval,.68.94). Children aged 3 to 12 years, those with injury severity scores 25, and those with Glasgow Coma Scale scores 8 all had significant reductions in the odds of death at ATC-AQ. CONCLUSIONS: Improved overall survival is associated with pediatric trauma patients treated at ATC-AQ Elsevier Inc. All rights reserved. Trauma remains the leading cause of death amongst children and adolescents aged 1 to 15 years, accounting for approximately 15,000 deaths each year and 50% of all Presented on April 30, 2010, at the 20th annual meeting of the Society of Black Academic Surgeons, Durham, NC. * Corresponding author. Tel.: ; fax: address: toyetunji@howard.edu Manuscript received July 17, 2010; revised manuscript October 20, 2010 pediatric deaths in the United States. 1 More than 20 million children sustain injuries that require treatment, resulting in 100,000 cases of permanent disability each year. 2 The leading mechanisms of death in this age group are motor vehicle collisions, drowning, burns (including those from house fires), homicides from nonaccidental traumas (ie, child abuse), firearms, and falls. In recent years however, there has been a markedly increased number of fatalities due to violence, including those secondary to firearms, especially in teenagers /$ - see front matter 2011 Elsevier Inc. All rights reserved. doi: /j.amjsurg

2 446 The American Journal of Surgery, Vol 201, No 4, April 2011 The trauma system in the United States has significantly helped mitigate the impact of trauma across all age groups since its inception and development over the past 40 years, 4 with the standardization of trauma care through trauma center verification. The American College of Surgeons (ACS), in addition to several individual states, designates some centers as level 1 or level 2 trauma centers, on the basis of the availability of manpower and resources, following a stringent review process. A level 1 trauma center is a regional resource hospital able to care for every form of injury, with around-the-clock coverage by a trained trauma surgeon and other trauma-related specialists available when needed. This represents the highest level of care available to an injured patient in the US trauma system. Each center must pass rigorous standards set by the ACS or state boards. For pediatric trauma patients, there exist some verified level 1 pediatric trauma centers (PTC). Unfortunately, these are very few in number and not easily accessible to all injured pediatric patients. Currently, there are more verified level 1 adult trauma centers (ATC) in existence than level 1 PTC. Some adult centers, in addition to being designated level 1 ATC, have added qualifications in pediatrics (ATC-AQ). This qualifies them to offer adequate and arguably better care to pediatric populations compared with ATC only. Improved survival has been demonstrated with children treated at PTC and compared with ATC-AQ and ATC only. 5 However, not all children in the country have access to pediatric trauma care, as demonstrated by Nance et al. 6 As a matter of fact, a higher proportion of pediatric traumas are typically seen at ATC-AQ or ATC because of unavailability of PTC in the region. 7 Thus, the role and impact of ATC in the treatment of pediatric trauma patients is fairly substantial. Potoka et al 5 first alluded to an improved outcome at ATC-AQ compared with ATC on the basis of their thorough and in-depth analysis of data from a statewide trauma system. Though limited by the nature of the analysis (stratified analysis vs multivariate regression) and the statebased nature of the data set, this difference has typically been an issue of controversy. Proponents of similar outcomes argue that there are no differences in care at level 1 ATC-AQ compared with level 1 ATC only, because these centers still represent the highest level of trauma care for severely injured adult patient. 8 A recent review of 60 published articles on pediatric trauma outcomes concluded that enough data did not exist to conclusively determine which type of trauma center was better in the delivery of pediatric care. 9 Unfortunately, most of the data included in this analysis were either descriptive in nature, single institutional, or unadjusted analysis at best. Although the pediatric trauma system remains embedded in its adult counterpart it is imperative to verify if there truly exists a difference in mortality among children treated ATC compared with ATC-AQ and, if so, which subset may benefit significantly from ATC-AQ. We therefore hypothesized in this study that ATC-AQ are more beneficial overall and for some select subsets of pediatric trauma patients. Methods The National Trauma Data Bank (NTDB) was used for this analysis. The NTDB is maintained by the ACS and contains data from 700 trauma centers across the United States and Puerto Rico. In its most recent multiyear version (version 7), the NTDB contains 1.8 million patient records over a 5-year period ( ). 10 The NTDB documents patient records, including demographics, injury characteristics, length of stay, discharge disposition, and trauma center characteristics such as verification level and number of intensive care unit beds, among other variables. It also captures data from prehospital care to emergency department care. It contains data across the age spectrum from 0 to 89 years. For this analysis, the NTDB was queried for patients aged 18 years to delineate the pediatric population. The age cutoff at 18 years is in keeping with many other studies of pediatric trauma. Because the purpose of the analysis was to compare care at level 1 ATC versus level 1 ATC-AQ, we excluded all patient records from non level 1 trauma centers. A level 1 ATC-AQ was defined in this analysis as any trauma center with ACS-designated and/or state-designated level 1 trauma status with added pediatric level 1 trauma status. The added qualification in pediatric trauma may be ACS designated or state designated. A level 1 ATC was defined as an ACS-designated and/or state-designated level 1 trauma center without any pediatric qualification. Burn patients were excluded from the analysis because of the difference in their course of care. All patients transferred in and out of facilities were also excluded. This was to circumvent the issue of severely injured patients being transferred to level 1 trauma centers who may have had poor prognoses on admission. This also enabled us to look at patients primarily received and cared for entirely at each facility. Also excluded from the analysis were mildly injured patients, defined as those with injury severity scores (ISS) 9. This was to ensure comparison of patients who truly require level 1 trauma care and not to unduly influence the outcomes of centers that attend to more severely injured patients. The outcome of interest in this study was in-hospital mortality. This excluded patients who were dead on arrival or dead before any treatment could be offered in the receiving facility. Because of limitations of the data set, other outcomes of interest (eg, short-term and long-term disabilities) could not be evaluated in this analysis, because patients were not followed beyond discharge. The primary comparison was ATC-AQ versus ATC, as defined above. Other covariates included in the model were ISS (further categorized as moderate [ISS 9 14], severe [ISS 15 24], or very severe [ISS 25]), mechanism of injury (including pedestrian struck, motor vehicle collision, gunshot wound, bicyclist, cyclist, and stab wound), insurance status (classified as commercially insured, government insured, uninsured, or other insurance), ethnicity (defined as white, African American, Hispanic, Asian, and others), presence of

3 T.A. Oyetunji et al. Pediatric outcomes at trauma centers 447 shock (with shock defined as systolic blood pressure 74, 78, 82, 86, and 90 mm Hg for patients aged 0 2, 3 4, 5 6, 7 8, and 8 years, respectively), Glasgow Coma Scale (GCS) motor component, head injury (defined as head Abbreviated Injury Scale score 3), and age ( 3, 3 12, and 12 years). The choice of age in these categories was based on physiology. Toddlers (aged years) are more likely to have similar causes and patterns of injury and were therefore grouped together. The choice of 12 years as a cutoff for the teen and adolescent group was based on a previously published work clearly demonstrating a different response in this age group after traumatic injury. 11 In addition, we wanted to separate the group with different physiology on the basis of the onset of puberty. Severe traumatic head injury (TBI) was defined as head maximum Abbreviated Injury Scale score 3. Univariate analysis was performed to describe the demographic and injury characteristic of the study population. This was followed by an unadjusted analysis to compare and statistically verify associations between in-hospital mortality and the different covariates on the basis of the primary comparison of ATC-AQ versus ATC. For the bivariate analysis, Student s t tests were used to compare all continuous variables, and 2 tests were used to compare categorical variables. All data analysis was carried out using Stata version 10 (StataCorp LP, College Station, TX). The adjusted (multiple logistic regression) analyses included all the covariates listed above to establish the independent effect of treatment at a given trauma center (ATC-AQ or ATC) on in-hospital mortality. Furthermore, subset analyses by GCS score (stratified as 8, 8 to 15, and 15), injury severity categories, and the defined age categories as stated above were performed to check if there was a difference in outcomes on the basis of which trauma center delivered care to the patient. Results A total of 53,702 children met the inclusion criteria and were included in the final analysis, with 55.5% seen at level 1 ATC and 44.5% at ATC-AQ. The median age was 11 years (interquartile range, 4 15 years), with a median ISS of 10 (interquartile range, 9 17). Female patients accounted for 34.1%. The majority of the children were white (53.4%), with African American, Hispanics, and other ethnicities accounting for 14.6%, 11.1%, and 12.9%, respectively. Overall mortality was 3.9%. Uninsured children constituted 9.3% of the study population, and commercially insured and government-insured children made up 54.4%. Other types of insurance accounted for the remaining. By injury severity, 13.2% of the children had very severe injuries (ISS 25), 19.8% with severe injuries (ISS 15 and 25). Children with severe TBIs constituted 17.8%. By age categories, children aged 3, 3 to 12, and 12 years accounted for 17.0%, 40.0%, and 43.0%, respectively. In unadjusted analysis comparing ATC with ATC-AQ, mortality was significantly lower in ATC-AQ versus ATC (3.2% vs 4.5%, P.001). By ethnicity, whites also accounted for the majority of children seen at both centers. The median ISS were comparable (10 and 9 at ATC and ATC-AQ, respectively), though significantly different statistically (P.001). However, ATC had a comparably higher proportion of severely injured children than ATC-AQ (17.6% vs 12.9%, respectively, P.001). Other demographic and injury characteristic comparisons are as shown in Table 1. The distribution by age categories also followed a similar pattern, with whites accounting for the majority of the population and a male predominance across all ages (Table 2). Additional injury and demographic factors by age categories are listed in Table 2. On multivariate analysis, the adjusted odds ratio (OR) of mortality for all patients seen at ATC-AQ compared with ATC was significantly lower by 20% (OR,.80; 95% confidence interval [CI],.68.94; P.05; Table 3). In the subset analysis by age group, children aged 3 to 12 years seen at ATC-AQ had a significant 43% reduction in the odds of death versus their counterparts seen at ATC irrespective of injury severity (OR,.57; 95% CI,.41.78; P.05). However, this difference was not seen in the other age groups. Children with very severe injuries demonstrated a 21% reduction in the odds of death (OR,.71; 95% CI, Table 1 Demographic and injury characteristics of the study population by type of trauma center Variable ATC ATC-AQ P Mortality Race White African American Hispanic Asian/Pacific Islander Native American/Alaskan Other category (y) Gender Male Female ISS 10 (9 17) 9 (9 16) ISS category Insurance status Commercial Government Uninsured Other Severe TBI Data are expressed as percentages or as median (interquartile range).

4 448 The American Journal of Surgery, Vol 201, No 4, April 2011 Table 2 Demographic and injury characteristics of the study population by age category.59.85; P.05). Children with severe TBIs (GCS score 8) also had a survival advantage, with a mortality OR of.81 (95% CI,.68.97; P.05; Table 3). Comments 3 y 3 12 y 12 y Mortality Race White African American Hispanic Asian Race Race Gender Male Female ISS category Insurance status Commercial Government Uninsured Others Severe TBI Data are expressed as percentages. The present study demonstrates that pediatric trauma patients have reduced mortality when treated at level 1 ATC-AQ compared with ATC only, in contrast to some articles on pediatric trauma care in the literature. 8 To the best of our knowledge, this study is the largest comparison of level 1 ATC-AQ versus ATC on a national level to date. Our study population included children aged 0 to 17 years. This is similar to the age inclusion criteria used in a study by Knudson et al. 8 Other investigators, however, have used different age cutoffs, as typified by Rhodes et al, 12 who chose 15 years as the cutoff. Potoka et al 5 demonstrated a difference in mortality between ATC and ATC-AQ in the Pennsylvania Trauma Outcome Study. In this statewide study, they compared PTC with all other levels of ATC, including level 1 ATC-AQ. The study concluded that children managed at PTC or ATC-AQ had significantly better outcomes compared with those treated at ATC only. However, this study failed to control for commonly known predictors of outcomes after trauma. Although our study results have numerous parallels compared with that study, in addition to some yet unidentified findings, we went a step further in this analysis to control for as many known independent predictors of P trauma of which we were aware. We adjusted for demographic factors (age, gender, insurance status, and ethnicity), injury severity characteristics (ISS, presence of shock, and GCS motor component), and mechanism of injury. This study still showed a 20% reduction in risk for mortality in children treated at ATC-AQ compared with ATC, similar to what was observed by Potoka et al. In addition, we demonstrated improved survival for children with severe TBIs, as typified by GCS scores 8. Although Potoka et al showed a difference in outcomes for severely injured children with ISS 15 treated at PTC, we demonstrated a similar difference in children with very severe injuries (ISS 25) treated at ATC-AQ. Because the cohort was a subset of children with ISS 15, it is possible that this significantly influenced the results of their study. Similar to Potoka et al, we demonstrated a significant difference in outcomes of children with severe TBIs treated at ATC-AQ. TBI is well known to be a leading cause of pediatric trauma death On the basis of the findings of this study, it is evident that children with TBIs will benefit from care at ATC-AQ, and this subset of patients should be considered a triage priority in geographic locations where this option exists. In addition to the survival benefit outlined above, a significant finding in this study is the age-dependent benefit of ATC compared with ATC-AQ, showing improved survival in children aged 3 to 12 years. A similar age subset of patients (0 10 years) has already been shown to have better outcomes at pediatric hospitals compared with adult-centered hospitals. 7 This indeed has a major impact on triage criteria for critically injured children. The reason for this survival benefit may be multifactorial. However, we postulate that this may be related to personnel availability and the setup of ATC-AQ. These facilities must meet additional stringent criteria to qualify as trauma centers for children. Although some investigators have concluded that no difference exists in outcomes irrespective of pediatric or adult designation, a closer review of the literature shows weaknesses in the methods used (simple descriptive articles, unadjusted analyses, small sample sizes), and such data are Table 3 ORs and 95% CIs after multivariate analysis on the overall population and by different subset categories Variable Reference OR (95% CI) P Overall ATC.80 (.68.94) category ATC 3 y.81 ( ) y.57 (.41.78) y.91 ( ).40 ISS category ATC ( ) ( ) (.59.85) GCS score ATC ( ) ( ) (.68.97).05

5 T.A. Oyetunji et al. Pediatric outcomes at trauma centers 449 typically from single institutions. 8,16 For the other age subsets, the reason for the lack of difference in infants and teenagers is unknown. Although speculations can be made regarding contributions from the improper use of safety equipment (infant seats and seatbelts), this issue warrants further investigation. Free-standing PTC have been demonstrated to have the best outcomes for injured children As mentioned previously, such centers are uncommon in every locality, making access to them an issue. In addition, construction of new PTC is an expensive venture, even with the best intentions. According to Nance et al, 6 approximately 17 million pediatric patients lack access to pediatric trauma care. Their definition of what constitutes a PTC was more liberal, including additional centers beyond PTC. This further underscores the role of ATC in the delivery of adequate pediatric trauma care. At present, the majority of pediatric trauma patients are typically seen at ATC or ATC-AQ. 7 Most states have at least one designated level 1 ATC with or without added pediatric qualifications. There is a need for policies that favor the triage of certain injured children to ATC-AQ rather than ATC. That said, the role of ATC alone should not be undermined by such triage criteria. We are of the opinion that a role currently exists for both centers in an evolving pediatric trauma system. Akin to the trauma level stratification that presently exist for trauma centers, PTC, ATC-AQ, and ATC can be seen as essential components of the pediatric trauma system with a two-way interhospital transfer potential, as determined by established triage criteria. This will further strengthen the quality of care and improve access in the pediatric population. In a recent review of pediatric trauma care, Nance et al 20 surmised, The lessons learned in trauma should be extended to include pediatric trauma care and generalized to inform the ongoing development of the emergency care system as a whole. No doubt the development of the pediatric trauma system is inherently intertwined with the advancement of the adult trauma system. However, it is important to appreciate the limitations of some ATC and acknowledge what makes for manageable pediatric trauma, with an overall goal of improving outcomes in all injured children. 21 This study is not without its limitations. The lack of data from PTC is a drawback, as this might have been another opportunity to demonstrate the role of PTC in pediatric trauma care. The NTDB is also a voluntary database and may be limited by who chooses to contribute data. We also acknowledge that the study may not necessarily reflect the standard of care at individual institutions. However, this is the largest national analysis to date comparing ATC with ATC-AQ. We are of the opinion that the power of this study will help mitigate some of these limitations. In conclusion, we have demonstrated that children treated at ATC-AQ had reduced mortality compared with those treated at ATC. Subsets of patients (those aged 3 12 years, very severely injured [ISS 15] patients, and those with severe TBIs [GCS score 8]) also benefited significantly when receiving care at ATC-AQ. Incorporating these findings into triage criteria may help improve the outcomes of pediatric trauma patients and may have policy implications for future pediatric trauma system development and growth. References 1. Ramenofsky ML. Pediatric abdominal trauma. Pediatr Ann 1987;16: Rescorla FJ, Grosfeld JL. Splenic and liver trauma in children. Indiana Med 1989;82: American College of Surgeons. First impression. In: Advanced trauma life support for doctors. 7th ed. Chicago, IL: American College of Surgeons; 2004: Mullins RJ. A historical perspective of trauma system development in the United States. J Trauma 1999;47(suppl):S Potoka DA, Schall LC, Gardner MJ, et al. Impact of pediatric trauma centers on mortality in a statewide system. J Trauma 2000;49: Nance ML, Carr BG, Branas CC. Access to pediatric trauma care in the United States. Arch Pediatr Adolesc Med 2009;163: Densmore JC, Lim HJ, Oldham KT, et al. Outcomes and delivery of care in pediatric injury. J Pediatr Surg 2006;41: Knudson MM, Shagoury C, Lewis FR. Can adult trauma surgeons care for injured children? J Trauma 1992;32: Petrosyan M, Guner YS, Emami CN, et al. Disparities in the delivery of pediatric trauma care. J Trauma 2009;67(suppl):S American College of Surgeons. National Trauma Data Bank research data set v.7.0 user manual. Available at: ntdb/ntdbmanual07.pdf. 11. Haider AH, Efron DT, Haut ER, et al. Mortality in adolescent girls vs boys following traumatic shock: an analysis of the national pediatric trauma registry. Arch Surg 2007;142: Rhodes M, Smith S, Boorse D. Pediatric trauma patients in an adult trauma center. J Trauma 1993;35: Tepas JJ III, DiScala C, Ramenofsky ML, et al. Mortality and head injury: the pediatric perspective. J Pediatr Surg 1990;25: Tepas JJ III, Mollitt DL, Talbert JL, et al. The pediatric trauma score as a predictor of injury severity in the injured child. J Pediatr Surg 1987;22: Tepas JJ III, Ramenofsky ML, Barlow B, et al. National Pediatric Trauma Registry. J Pediatr Surg 1989;24: Bensard DD, McIntyre RC Jr, Moore EE, et al. A critical analysis of acutely injured children managed in an adult level I trauma center. J Pediatr Surg 1994;29: Mooney DP, Rothstein DH, Forbes PW. Variation in the management of pediatric splenic injuries in the United States. J Trauma 2006;61: Potoka DA, Schall LC, Ford HR. Development of a novel age-specific pediatric trauma score. J Pediatr Surg 2001;36: Davis DH, Localio AR, Stafford PW, et al. Trends in operative management of pediatric splenic injury in a regional trauma system. Pediatrics 2005;115: Carr BG, Nance ML. Access to pediatric trauma care: alignment of providers and health systems. Curr Opin Pediatr 2010;22: Stone KP, Woodward GA. Pediatric patients in the adult trauma bay comfort level and challenges. Clin Pediatr Emerg Med 2010;11: