Logo of annsurgLink to Publisher's site
Ann Surg. Jul 1999; 230(1): 87.
PMCID: PMC1420849

Trauma Center Maturation

Quantification of Process and Outcome


Background and Objective

The regional trauma system with the trauma center as its center is a model for health care networks. However, trauma center maturation has not been defined in the literature. The authors’ hypothesis was that maturation of the trauma center would affect quantitatively both process and patient outcome.

Materials and Methods

A total of 15,303 trauma patients were admitted from 1987 to 1995. Annual admissions increased from 813 to 2669. Resources were generated as patient volume increased. Time to the operating room, length of stay, and complications were determined. TRISS methodology was used to calculate z scores and w values to compare actual with predicted mortality rates.


Time to the operating room for laparotomy decreased from 62 ± 73 to 35 ± 47 minutes, from 32 ± 32 to 20 ± 17 minutes in hypotensive patients, and for craniotomy decreased from 88 ± 54 to 67 ± 49 minutes. The incidence of infectious, airway, neurologic, orthopedic, respiratory, gastrointestinal, and procedure-related complications declined significantly. Z scores and w values increased for penetrating and blunt injuries. Deaths for patients with ISS >15 declined significantly. Hospital length of stay decreased for all ranges of injury severity.


As the trauma center matured, the process of delivering patient care became more efficient. The result was improved survival, fewer complications, and a shorter length of stay.

The regional trauma system with the trauma center (TC) as its center serves as a model for health care networks. The standards for TC accreditation by the American College of Surgeons Committee on Trauma provide the template for TC development. 1 More than any other program, the TC requires commitment from multiple surgical and medical specialities, nursing services, and administration. Trauma care is expensive and labor-intensive.

The “volume–outcome relationship” has been emphasized in the recent health care literature. 2–12 Luft et al 10,11 studied the relationship between surgical volume and surgical procedures, including coronary bypass grafting, abdominal aortic aneurysm repair, and complex gastrointestinal and biliary procedures. They reported a strong inverse relationship between the volume of a particular surgical procedure and the postoperative mortality rate. Other authors have made similar observations in the outcome of ruptured abdominal aortic aneurysms 13 and Whipple procedures. 4,5 An evaluation of the relationship between volume and outcome in seriously injured trauma patients in the Chicago trauma system found that low-volume TCs (<140 patients/2 years) had higher mortality rates than higher-volume TCs (>200 patients/2 years). 14 A seriously injured patient’s chance of dying was 30% greater at a low-volume TC.

The Pennsylvania Trauma Outcome Study 9 attempted to correlate measures of patient volume per TC and per surgeon with normative survival (w = 0). The authors found no correlation between outcome and volume per TC or volume per trauma surgeon for penetrating or pediatric trauma. They did suggest that a trauma surgeon should treat ≥35 seriously injured patients per year (28 adult patients with blunt injury) to achieve normative survival for blunt injuries in adults. However, the volume of admissions per surgeon was an estimate derived by dividing the number of admissions to a given hospital by the number of surgeons on call, rather than the actual value.

Richardson et al 15 recently evaluated the relationship between outcome and annual trauma volume per surgeon and years of experience. They reported no difference in outcome in either morbidity rate or mortality rate that correlated with annual volume of patients treated or years of experience.

In the pediatric population, Tepas et al 16 reported a trend of increasing mortality rate with increasing patient volume at pediatric TCs. They suggested that the increasing mortality rate with severe injuries as patient volume increased may reflect overdemand on resources.

Thus, the impact of patient volume and experience in the delivery of care to injured patients remains poorly defined. 2,3,10

Several recent papers have used “normative survival” (w = 0) as an end point for adequate trauma care. 6,9,17,18 Our expectations should be beyond achieving “normative survival.” We should be able to quantify improved outcome in the TC by decreased length of stay (LOS), decreased numbers of preventable complications and deaths, 19 and improved z scores and w values. 17,18,20–28

The TC undergoes a maturation process that has not been defined in the literature. The purpose of this paper was to describe the process of TC development and maturation. Our hypothesis was that maturation of the TC would affect both process and patient outcome in parameters that could be quantified.


The University of Pittsburgh Medical Center committed to development as a level I TC in 1984. Pennsylvania began a formal TC designation process in 1987 and the university’s TC was designated within the year. This study includes 15,303 adult trauma patients admitted from 1987 through 1995. The hospital trauma registry was the source of information for the study. Demographic data, injury patterns, revised trauma score (which includes the Glasgow Coma Scale [GCS], systolic blood pressure, and respiratory rate), injury severity score (ISS), 20 admission blood pressure, length of stay in the hospital and in the intensive care unit (ICU), complications, survival, and condition at disposition were reviewed.

The ISS, an anatomic indicator of overall severity of injury, 18,20 is calculated by adding the squares of the highest AIS (Abbreviated Injury Scale) in the three most severely injured body regions and ranges from 0 (no injury) to 75 (injury incompatible with life). Major injuries are defined as ISS >15. 23,29–32

Preventable deaths were determined by a multidisciplinary committee that met monthly.

The TRISS methodology, described by Boyd et al and others, 18,21,23 combines age, physiologic criteria on admission to the hospital, and anatomic sites and degree of injury to predict a patient’s probability of survival compared with a national norm (Major Trauma Outcome Study [MTOS]): (Equation 1) where b = b0 + b1 (Revised Trauma Score) + b2 (ISS) + b3 (age); age = 0 for age younger than 55 years and age = 1 for age 55 years or older. The “b’s” are coefficients that differ for blunt and penetrating injury. The Revised Trauma Score is a physiologic index of the severity of injury derived from a weighted sum of coded values of the GCS, systolic blood pressure, and unassisted respiratory rate: 21,23 (Equation 2) The TRISS methodology allows calculation of the z score and w statistic, which indicate unexpected outcomes. 21,23,30 Deaths among patients with a probability of survival >0.50 and survival among patients with a probability of survival <0.50 are unexpected outcomes. The z score compares the actual number of survivors (A) in the study with those expected (E) from the MTOS outcome norms with the equation: (Equation 3) where S is a scale factor that accounts for statistical variation. The z score indicates the statistical significance of the difference between the number of actual versus expected survivors. An absolute z score >1.96 indicates a statistically significant difference (p < 0.05). When the difference between actual and expected survivors is statistically significant, the clinical significance of this difference is determined with the w statistic: (Equation 4) where A is the actual number of survivors, E is the expected number of survivors based on MTOS, and n is the sample size. A positive w statistic indicates more survivors than predicted per 100 patients; a negative w statistic indicates fewer survivors than predicted per 100 patients.

equation image
equation image
equation image
equation image

Continuous variables were analyzed using a weighted linear regression with both quadratic and linear terms. Discrete variables were analyzed using a Cochran–Armitage test (two-sided). A probability value <0.05 was considered significant (*). Data are expressed as mean ± standard deviation.


Many components were involved in the development of the TC. Important organizational details are provided for 1984 to 1995 (Fig. 1). The hospital was accredited in 1987–1988 as a level I TC by the state. Complete patient data are available from 1987, the time of the upgrading of the trauma registry, to the present.

figure 13FF1
Figure 1. Timetable of major changes and additions in the trauma center. Key components in the organization of the trauma center were the hiring of the trauma director, trauma administrator, and nurse coordinator, the organization of the trauma team, ...

Figure 1 is a timeline showing major changes and additions to the TC over the years 1984 to 1995. Organizationally, the key components included the hiring of the trauma director, trauma administrator, and nurse coordinator and the formation of a dedicated trauma team with defined roles for each member and a dedicated trauma service. In the early years of the TC, general surgery, neurosurgery, and orthopedic call was provided by available staff. As the demands and expectations of the TC changed, attrition of noncommitted faculty and the addition of faculty whose academic careers were dedicated to trauma resulted. This concentration of expertise was furthered by building the trauma ICU and medical/surgical unit dedicated solely to the care of trauma patients.

Key advances in education included the weekly trauma conference and critique of videotaped trauma resuscitations, which are both educational and quality improvement tools. These multidisciplinary sessions provided a forum to critique, analyze, and learn from the management of a complex trauma patient as a group. The development of patient care protocols and a comprehensive trauma handbook 33 based on available literature and consensus agreement standardized our care and provided important quality improvement instruments.

The essential changes in the physical plant to provide quality care to the increasing patient volume included a dedicated trauma ICU, new trauma resuscitation areas in the emergency department, a dedicated trauma operating room (OR), and the moving of a computed tomography unit to the trauma resuscitation area.

The number of patients admitted to the TC increased progressively (Fig. 2). The proportion of patients injured by blunt (82%) versus penetrating (18%) mechanisms was constant. Table 1 shows the distribution of patients based on injury severity using ISS. During the last 7 years, the number of significantly injured patients (ISS >15) increased steadily, but the proportion of patients with ISS >15 was stable (20%).

figure 13FF2
Figure 2. Although the number of patients with penetrating and blunt mechanisms of injury increased over the study period, the proportion of patients injured by blunt (82%) and penetrating (18%) mechanisms was consistent.
Table thumbnail
Table 1. ISS Distribution

Table 2 lists the mode of arrival of the trauma patients. The growth of the air medical service was a major component of the growth of the TC.

Table thumbnail
Table 2. Mode of Arrival

As the total volume of patients increased, the need for immediate surgical intervention increased (Table 3). Although the number of patients requiring craniotomy was relatively constant, the time to OR for craniotomy decreased (p = 0.06) The number of patients requiring laparotomy, thoracotomy, and open reduction and internal fixation (ORIF) of extremity fractures increased. The time to OR for laparotomy declined from 62 ± 73 minutes in 1989 to 35 ± 47 minutes in 1995 (p < 0.05). The time to OR for laparotomy in hypotensive trauma patients (systolic blood pressure <90 mmHg) fell from 32 ± 32 minutes in 1989 to 20 ± 17 minutes in 1995 (p < 0.05). The time to OR for hypotensive patients who required simultaneous laparotomy and craniotomy also decreased over the time of the study (p < 0.05). Patients were processed through the system progressively more expediently.

Table thumbnail
Table 3. Surgical Procedures

The average LOS declined from 1987 through 1995 (p < 0.05), whereas the ICU LOS did not change significantly (Fig. 3). This decrease in hospital LOS was seen within the ISS <9, ISS 9 to 15, and ISS >15 subgroups (Table 4) (p < 0.05).

figure 13FF3
Figure 3. The number of trauma patients admitted to the trauma center increased progressively. Length of stay declined significantly over the same period.
Table thumbnail
Table 4. Length of Stay (Days)

Outcome using TRISS analysis is shown in Table 5. As the volume of admissions increased over time, z scores and concomitant w scores increased for both blunt and penetrating mechanisms of injury (Fig. 4). The number of unexpected survivors increased, and preventable deaths decreased to zero to two patients per year. Similarly, the mortality rate for patients with ISS >15 declined from 20.0% to 16.2% (p < 0.05).

figure 13FF4
Figure 4. As the number of patients increased over the study period, z scores and w scores for blunt and penetrating injury increased progressively.
Table thumbnail
Table 5. TRISS Outcomes

Table 6 shows that the proportion of patients discharged to home remained constant. Complications are listed in Table 7. The definitions of complications were redefined in 1988, and thus data are not shown for 1987 and 1988. Over the years 1989 to 1995, the majority of complications declined significantly (p < 0.05). Notably, procedure-related complications fell from 10.9% to 2.2%. Over the time of the study, patient volume increased, LOS declined, and preventable morbidity and mortality rates declined significantly.

Table thumbnail
Table 6. Posthospital Disposition (%)
Table thumbnail
Table 7. Complications (%)


Optimal care of the patient with multiple-system injury requires coordination of numerous surgical specialities, consumes extensive resources, and is disruptive to the elective schedules of the hospital. Coordination of this complex level of care is the goal of the TC. 6,34–41 The data in this paper support that this process of maturation of the TC has an important impact on patient outcome. The major observation from these data is that as the TC matured and the delivery of patient care improved (the process), patients were provided care with a shorter LOS, lower rates of complications, a decreased incidence of preventable deaths, and a significant increase in unexpected survivors (the outcome).

The improvement in trauma care delivery could not be attributed to a single change in the trauma system; rather, it resulted from multiple changes in all facets of trauma care. The general trend in the organizational structure of the TC was consolidation of the trauma patients on dedicated units and provision of care by physician and nonphysician staff who were committed primarily to the care of trauma patients. This was accomplished by limiting the general surgery, orthopedic surgery, critical care medicine, and neurosurgical attendings on the trauma call schedule to those with trauma as their academic focus.

The volume–outcome relationship has been well documented for many complex general surgical and cardiothoracic procedures. 2–5,9–12 The relationship between the mortality rate and volume seems to be curvilinear and can be represented by a logarithmic function. 10 However, patient volume at a TC would be expected to increase over time, as the TC becomes established within a region. 34–41 Improvement in the quality of care over time must be considered as a contributory factor in outcome. Both are part of the maturation of the TC. Intense experience gained in a short period of time with a large number of surgical procedures is as important in outcome as experience over several years. 5,10 Further, the experience of an individual trauma surgeon or volume of patients seen in a year by an individual surgeon did not affect the outcome variables in our study (data not shown). Differences in outcome seem to be more closely associated with the features of the hospital where the care is delivered than the characteristics of the individual surgeons, thus emphasizing the value of the TC. 2,3,9,10,11,15

The resident staff were assigned to cover only the trauma service during the rotation (daylight and night coverage). They also had the opportunity to rotate on the service as interns and junior and senior residents, with repeated exposure to trauma patients and trauma attendings. Similarly, the development of the trauma ICU and the dedicated trauma medical/surgical floor provided expertise in the nonphysician care of multiply injured patients as well. This directed focus resulted in improved patient outcome.

The educational components of the program were provided for all trauma care providers using a multidisciplinary approach. The goal of the weekly trauma conferences, the Advanced Trauma Life Support Course, and the trauma protocols was standardization of care. This process has progressed to structured trauma care protocols that provide tools for quality improvement monitoring. The changes in physical plant were accomplished as needed based on the clinical volume of the program (the volume of patient admissions increased severalfold). Failure to provide sufficient resources to optimize patient care increases rather than decreases the mortality rate as patient volume increases. 16

The recognition of patients who required early surgical intervention improved significantly. Patients who required laparotomy, especially those who were hypotensive and required prompt control of hemorrhage, arrived in the OR significantly more rapidly.

As the process of patient care improved quantitatively, outcomes in trauma patients also improved. The LOS declined significantly. It is not apparent why ICU LOS was relatively stable.

The incidence of complications fell significantly. Because the occurrence of complications has been demonstrated to prolong LOS significantly, 19 this decrease in preventable complications is essential. Despite the earlier discharge times, the rate of readmission for complications was unchanged. In addition, early discharge was appropriate as indicated by patient function at discharge: FIM (Functional Independence Measure) scores averaged 18 throughout the study (data not shown), and the discharge destinations remained consistent. Thus, a significant increase in unexpected survivors was achieved with good functional outcome.

TRISS analysis allows quantification of the outcome of a TC compared with a national norm. TRISS combines age, physiologic criteria, and anatomic sites and magnitude of injuries to estimate the probability of a patient’s survival. The w score measures the clinical significance and the z score the statistical significance of outcome. The w score represents the difference between the number of patients actually surviving and the number of survivors expected per 100 patients treated. Thus, a w value of +4 indicates that 4 more patients survived per 100 patients than would have been predicted. If z is more negative than −1.96, significantly more patients died than were predicted. If z exceeds +1.96, significantly more patients survived than predicted, and w indicates the number of unexpected survivors per 100 admissions. Thus, the progressive rise in z scores and w values for adults with penetrating or blunt injuries suggests that as the TC matured and the quality of care improved, a measurable improvement in outcome was demonstrated as lives saved. This conclusion is further supported by the significant decline in the mortality rate for patients with ISS >15 and the decrease in the number of preventable deaths each year.

Particularly in this era of managed care, expedient provision of patient care as measured by process and outcome data becomes critical. The concentration of patients in centers of excellence, as epitomized by the TC, may be a model for other programs. 42 The maturation of the TC provided more efficient care, fewer complications, and improved survival.


Correspondence: Andrew B. Peitzman, MD, Room A1010, Presbyterian University Hospital, Pittsburgh, PA 15213.

Supported in part by National Institutes of Health grant P50–GM53789.

Accepted for publication December 2, 1998.


1. American College of Surgeons Committee on Trauma. Resources for optimal care of the injured patient. Chicago: American College of Surgeons; 1993.
2. Flood AB, Scott WR, Ewy W. Does practice make perfect? Medical Care 1984; 22: 98–124. [PubMed]
3. Flood AB, Scoot WR, Ewy W, et al. Effectiveness in professional organizations. The impact of surgeons and surgical staff organizations on the quality of care in hospitals. Health Services Res 1982; 17: 341–363. [PMC free article] [PubMed]
4. Gordon TA, Bowman HM, Tielsch JM, et al. Statewide regionalization of pancreaticoduodenectomy and its effect on in-hospital mortality. Ann Surg 1998; 228: 71–78. [PMC free article] [PubMed]
5. Gordon TA, Burleyson GP, Tielsch JM, Cameron JL. The effect of regionalization on cost and outcome for one general high-risk surgical procedure. Ann Surg 1995; 221: 43–49. [PMC free article] [PubMed]
6. Hamilton SM, Letourneau S, Pekeles E, Voaklander D, Johnston WC. The impact of regionalization on a surgery program in the Canadian Health Care System. Arch Surg 1997; 132: 605–610. [PubMed]
7. Hannan EL, O’Donnell JF, Kilburn H Jr, Bernard HR, Yazic A. Investigation of the relationship between volume and mortality for surgical procedures performed in New York State hospitals. JAMA 1989; 262: 503–510. [PubMed]
8. Hughes RG, Hunt SS, Lutt HS. Effects of surgeon volume and hospital volume on quality of care in hospitals. Medical Care 1987; 25: 489–503. [PubMed]
9. Konvolinka CW, Copes WS, Sacco WJ. Institution and per-surgeon volume versus survival outcome in Pennsylvania’s trauma centers. Am J Surg 1995; 170: 333–340. [PubMed]
10. Luft HS. The relation between surgical volume and mortality: an exploration of causal factors and alternative models. Med Care 1980; 18: 940–959. [PubMed]
11. Luft HS, Hunt SS, Maerki SC. The volume/outcome relationship: practice makes perfect on selective referral patterns? Health Services Research 1987; 22: 153–182. [PMC free article] [PubMed]
12. Pasquale MD, Peitzman AB, Wasser TE. Outcome analysis of Level I and Level II trauma centers. J Trauma (in press).
13. Rutledge R, Oller DW, Meyer AA, et al. A statewide, population-based, time-series analysis of the outcome of ruptured abdominal aortic aneurysm. Ann Surg 1996; 223: 492–505. [PMC free article] [PubMed]
14. Smith RF, Frateschi L, Sloan EP, et al. The impact of volume on outcome in seriously injured trauma patients: two years’ experience of the Chicago trauma system. J Trauma 1990; 30: 1066–1076. [PubMed]
15. Richardson JD, Schmieg R, Boaz P, et al. Impact of trauma attending surgeon case volume on outcome: is more better? J Trauma 1998; 44: 266–272. [PubMed]
16. Tepas JJ III, Patel JC, DiScala C, et al. Relationship of trauma patient volume to outcome experience: can a relationship be defined? J Trauma 1998; 44: 827–831. [PubMed]
17. Norwood S, Myers MB. Outcomes following injury in a predominantly rural populated-based trauma center. Arch Surg 1994; 129: 800–805. [PubMed]
18. Sacco WJ, Copes WS, Forrester Staz C, Smith JS Jr, Buckman RF Jr. Status of trauma patient management is measured by survival/death outcomes: looking toward the 21st century. J Trauma 1994; 36: 297–298. [PubMed]
19. McAleese P, Odling-Smee W. The effect of complications on length of stay. Ann Surg 1994; 220: 740–744. [PMC free article] [PubMed]
20. Baker SP, O’Neill B, Hadden W Jr, Lang WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974; 14: 187–196. [PubMed]
21. Boyd CR, Tolson MA, Copes WS. Evaluating trauma care: the TRISS method. J Trauma 1987; 27: 370–378. [PubMed]
22. Brenneman FD, Boulanger BR, McLellan BA, Culhane JP, Redelmeier DA. Acute and long-term outcomes of extremely injured blunt trauma victims. J Trauma 1995; 39: 320–324. [PubMed]
23. Champion HR, Copes WS, Sacco WJ, et al. The Major Trauma Outcome Study: establishing national norms for trauma care. J Trauma 1990; 30: 1356–1365. [PubMed]
24. Champion HR, Sacco WJ, Copes WS. Improvement in outcome from trauma center care. Arch Surg 1992; 127: 333–338. [PubMed]
25. Copes WS, Forester Staz C, Konvolinka CW, Sacco WJ. American College of Surgeons audit filters: associations with patient outcome and resource utilization. J Trauma 1995; 38: 432–438. [PubMed]
26. Karmy-Jones R, Copes WS, Champion HR, et al. Results of a multi-institutional outcome assessment: results of a structured peer review of TRISS-designated unexpected outcomes. J Trauma 1992; 32: 196–203. [PubMed]
27. Sampalis JS, Lavoie A, Boukas S, et al. Trauma center designators: initial impact on trauma-related mortality. J Trauma 1995; 39: 232–239. [PubMed]
28. Stewart TC, Lane PL, Stefanits T. An evaluation of patient outcomes before and after trauma center designators using trauma and injury severity score analysis. J Trauma 1995; 39: 1036–1040. [PubMed]
29. Demetriades D, Berne TV, Belzberg H, et al. The impact of a dedicated trauma program on outcome in severely injured patients. Arch Surg 1995; 130: 216–220. [PubMed]
30. Gillot AR, Copes WS, Langan E, Najarian M, Wiseman J, Sacco WJ. TRISS unexpected survivors—a statistical phenomenon or a clinical reality? J Trauma 1992; 33: 743–748. [PubMed]
31. Rogers FB, Simons R, Hoyt DB, Shackford SR, Holbrook T, Fortlage D. In-house board-certified surgeons improve outcome for severely injured patients: a comparison of two university centers. J Trauma 1993; 34: 871–877. [PubMed]
32. Shackford SR, Mackersie RC, Hoyt DB, et al. Impact of a trauma system on outcome of severely injured patients. Arch Surg 1987; 122: 523–527. [PubMed]
33. Peitzman AB, Rhodes M, Schwab CW, Yealy DM. The trauma manual. Philadelphia: Lippincott-Raven; 1998.
34. Hill DA, West RH, Duflou J. Value of the prospective before and after study as a methodology to evaluate outcome in a trauma center. Aust NZ J Surg 1993; 63: 940–945. [PubMed]
35. Hunt J, Hill D, Besser M, West R, Roncal S. Outcome of patients with neurotrauma: the effect of a regionalized trauma system. Aust NZ J Surg 1995; 65: 83–86. [PubMed]
36. Kane G, Wheeler NC, Cook S, et al. Impact of the Los Angeles County trauma system on the survival of seriously injured patients. J Trauma 1992; 32: 576–583. [PubMed]
37. Mullins RJ, Veum-Stone J, Hedges JR, et al. Influence of a statewide trauma system on location of hospitalization and outcome of injured patients. J Trauma 1996; 40: 536–541. [PubMed]
38. Waddell TK, Kolman PG, Goodman SJL, Girotti MJ. Is outcome worse in a small volume Canadian trauma center? J Trauma 1991; 31: 958–961. [PubMed]
39. West JG, Trunkey DD, Lim RC. Systems of trauma care. A study of two countries. Arch Surg 1979; 114: 455–460. [PubMed]
40. Shackford SR, Hollingsworth-Fridlund P, Cooper GF, Eastman AB. The effect of regionalization upon the quality of trauma care as assessed by concurrent audit before and after institution of a trauma system: a preliminary report. J Trauma 1986; 26: 812–820. [PubMed]
41. Mullins RJ, Veum-Stone J, Helfand M, et al. Outcome of hospitalized injured patients after institution of a trauma system in an urban area. JAMA 1994; 271: 1919–1924. [PubMed]
42. Moore EE. Trauma systems, trauma centers, and trauma surgeons: opportunity in managed competition. J Trauma 1995; 39: 1–11. [PubMed]

Articles from Annals of Surgery are provided here courtesy of Lippincott, Williams, and Wilkins
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...