Is computerized tomography of trauma patients associated with a transfer delay to a regional trauma centre?

EM Advances

CJEM 2008;10(3):205-208

Abstract

Résumé

Introduction

The National Trauma Registry reported that in 2000 there were 198 040 injury admissions in Canada, of which 6560 people died in hospital.1 The pattern of mortality in trauma is trimodal. In the first peak, victims die at the scene and efforts to reduce deaths in this area are mainly societal (e.g., drunk driving laws, airbags, speed limit enforcement, gun control laws and trauma prevention education).2,3 In the second peak, death occurs within the first hour (the "golden hour") following the event.4 Trauma deaths in the third peak occur in the intensive care unit (ICU), often as a result of postoperative complications and delays to definitive care.5 Trauma care is thus extremely time-sensitive. The rapid transport of patients to the most appropriate facility and the timely identification and treatment of injuries are therefore fundamental components of modern trauma systems.6-11

Delays in the transfer of patients to regional trauma centres can occur for a variety of complex reasons, including the lack of available escort staff, the lack of available ambulances, poor weather and ongoing resuscitation procedures.11-13 Delays due to ongoing diagnostic imaging studies at the sending facility, however, are often avoidable and do not directly contribute to the patient's well-being. The objectives of this study were to determine what proportion of transfer trauma cases underwent CT scans before being transferred and if this imaging was associated with increased length-of-stay (LOS) at the referring hospital.

Methods

We performed a medical record review study of all consecutive trauma cases referred to a regional trauma centre during a 2-year period. The local institutional ethics review board approved the study as a quality assessment study exempt from formal review.

The primary outcome was the LOS, which was measured as the time from arrival at the primary hospital to the time of departure to the regional trauma centre. All data were abstracted from the trauma patient database and corroborated using the CritiCall (a regional critical care referral centre) database, ambulance transfer records and patients' charts. The regional trauma centre services 2.2 million people and 24 hospitals over an area of 13 400 km2. All patients over 18 years of age initially treated at another hospital and subsequently transferred to the regional trauma centre were selected for analysis. To maintain confidentiality, cases were assigned a unique study number during data abstraction.

The data abstractors were trained and reached an agreement on variable definitions at the outset. Following initial data abstraction, a randomly selected subset of 20% of the cases was reviewed for data extraction accuracy by a second abstractor who was blinded to the study hypothesis and to procedure codes. Comparisons were made with the main sample for consistency by a third researcher.

Mean Injury Severity Scores (ISS) were compared between groups using an unpaired t test. Because of the skew of the time interval (LOS) data, both means and medians are reported, and both parametric (t test) and nonparametric analyses (Mann-Whitney U Test) were used to compare groups. The impact of missing LOS data was determined by worst case sensitivity analysis (i.e., repeating the data analysis after replacing the missing time intervals LOS in each group with the mean time intervals from the comparison group).

The sample size was calculated based on a pilot study in which we found that the mean LOS for those patients receiving a CT scan at the primary hospital and those who did not was approximately 180 and 120 minutes, respectively, with a standard deviation of 120 minutes. With a desired α of 0.05, power of 0.8 and using a 2-sided test, we estimated that a minimum of 63 patients would be needed in each group to show a significant difference. The interrater reliability of the data abstraction process was measured using the intraclass correlation coefficient (ICC).

Results

During the study period, 249 trauma patients were transferred to the regional trauma centre from a primary hospital, of which 79 (31.7%) underwent CT scanning before transfer. Based on the CritiCall database, none of these patients had surgery at the sending facility, nor was the transfer initiated on the basis of the CT scan. The injury severity scores were similar between the 2 groups. On the other hand, both mean and median LOS intervals were substantially different between groups (p < 0.001). Patients who underwent CT scanning before transfer had a median LOS of 213 (interquartile range [IQR] 157, 278) minutes, compared with 118 (IQR 81, 175) minutes for those transferred without CT scans being done (Table 1).

Seven of the CT cases (11%) and 26 of the non-CT cases (15%) had missing time data that precluded calculation of the primary outcome (χ2 = 1.94, p = 0.163). Nevertheless, the substantial difference in LOS persisted in the worst case sensitivity analysis. The ICC of data abstractors for the times of arrival and departure from the primary hospital was 1.0 (perfect).

Discussion

It is widely accepted that trauma patients should be transferred to a trauma centre as soon as possible, provided that they meet the Maryland criteria.7-9,11 Despite this time imperative, approximately one-third of patients transferred to our trauma centre from another hospital underwent CT imaging before transfer. This substantial proportion reflects a shift toward increased diagnostic imaging in trauma patients in general, yet these CT scans did not appear to alter patient management at the primary hospital. We did not attempt to explore the reason(s) for performing the CT, whether or not it was repeated at the trauma centre, and whether the increased LOS associated with the CT group had any impact on clinical outcome.

CT scanning is relatively fast. While newer generations of CT scanners are able to acquire high-quality images in a matter of minutes, arriving at our trauma centre with a CT scan "in hand" was associated with a substantially longer stay at the referring hospital that far outweighed the actual scan time. We have previously shown that patients spend a median time of 1 hour out of the ED for a noncontrast CT of the abdomen for suspected ureteric colic.14 We suspect that the cumulative time needed for all steps in the CT process is often underestimated by physicians ordering the test.

Limitations

These findings are based on data from a single regional trauma centre, yet we are not aware of any reason why they should not be representative of other regional trauma systems in Canada. We used only the ISS to compare groups for severity. The ISS is the most commonly used measure of injury severity in the Canadian trauma system and it is the only injury severity measure recorded for each patient entered into Ontario regional databases.

As with many medical record reviews, we encountered missing data. Nevertheless, a "worst case scenario" sensitivity analysis demonstrated that the findings were robust. Other methods for managing missing data such as imputation are not amenable to small sample sizes.15

We did not seek to determine whether the CT findings at the local hospitals guided therapy, which studies were repeated at the trauma centre and whether the increased LOS had any impact on clinical outcome including mortality and subsequent ICU and receiving hospital LOS.

Finally, one of the data abstractors was aware of the study objectives and we did not monitor abstractor performance other than the assessing interrater reliability at the conclusion.15,16

Conclusion

One-third of trauma patients undergo CT scanning at the primary hospital before being transferred to a regional trauma centre. Obtaining a CT scan before transfer is associated with an increased LOS of 90 minutes before transfer, on average. This relatively common practice, which would not appear to contribute to definitive trauma management, contradicts the long-standing emphasis on rapid transfer to regional trauma centres in the interest of timely intervention.

References

1. Ladak N. National Trauma Registry 2003 Report. Canadian Insitute for Health Information. Toronto (ON); 2003.
2. Cornwell EE III, Jacobs D, Walker M, et al. National Medical Association Surgical Section: position paper on violence prevention: a resolution of trauma surgeons caring for victims of violence. JAMA 1995;273:1788-9.
3. Cornwell EE III, Berne TV, Blezberg H, et al. Health care crisis from a trauma center perspective: the L.A. story. JAMA 1996;276:940-4.
4. Meislin H, Criss EA, Judkins D, et al. Fatal rrauma: the modal distribution of time to death is a function of patient demographics and regional resources. J Trauma 1997;43:433-9.
5. Trunkey DD. Trauma. Accidental and intentional injuries account for more years of life lost in the U.S. than cancer and heart disease. Among the prescribed remedies are improved preventive efforts, speedier surgery and further research. Sci Am 1983;249:28-35.
6. Trunkey D. Initial management of patients with extensive trauma. N Engl J Med 1991;324:1259-63.
7. Cowley RA, Hudson F, Scanlan E, et al. An economically proved helicopter program for transporting the emergency critically ill and injured patient in Maryland. J Trauma 1973;13: 1029-38.
8. Campbell JE. BTLS. Englewood Cliffs (NJ): Prentice Hall; 1998.
9. Creel JH. Scene size up. In: Campbell JE, editor. Basic trauma life support: for paramedics and other advanced providers. 4th ed. Upper Saddle River (NJ): Brady; 2000. p. 2-3.
10. Jacobs LM, Sinclair A, Beiser A, et al. Prehospital advanced life support: benefits in trauma. J Trauma 1984;24:8-13.
11. Sampalis JS, Denis R, Lavoie A. Trauma care regionalization: a process-outcome evaluation. J Trauma 1999;46:565-79.
12. Hedges JR, Feero S, Moore B, et al. Factors contributing to paramedic onscene times during evaluation and management of blunt trauma. Am J Emerg Med 1988;6:443-8.
13. Colucciello SA. Blunt abdominal trauma. Emerg Med Clin North Am 1993;11:114-5.
14. Worster A, Haines T. Does replacing intravenous pyelography with noncontrast helical computed tomography benefit patients with suspected acute urolithiasis? Can Assoc Radiol J 2002;53: 144-8.
15. Worster A, Haines T. Advance statistics: medical record review (MRR) studies. Acad Emerg Med 2004;11:187-92.
16. Gilbert EH, Lowenstein SR, Koziol-McLain J, et al. Chart reviews in emergency medicine research: Where are the methods? Ann Emerg Med 1996;7:305-8.