Original Contribution

Serial hematocrit testing does not identify Major injuries in trauma patients in an observation unit

Troy Madsen MD?, Matthew Dawson MD, Joseph Bledsoe MD, Philip Bossart MD

Division of Emergency Medicine, University of Utah, Salt Lake City, UT 84132, USA

Received 16 December 2008; revised 26 January 2009; accepted 27 January 2009

Abstract

Background: Routine serial hematocrit measurements are a component of the trauma evaluation for patients without serious injury identified on initial evaluation. We sought to determine whether serial hematocrit testing was useful in detecting significant Intra-abdominal injuries in trauma patients in our observation unit.

Methods: We performed a retrospective chart review of all trauma patients placed in our observation unit over a 14-month period. Patients had received trauma surgery evaluation before placement in the observation unit and routinely received serial hematocrit testing (>=2 Hematocrit levels) while in the observation unit. We compared trauma patients with a hematocrit drop of 5 points or more to those without a significant hematocrit drop.

Results: Three hundred sixty-five trauma patients were placed in the observation unit, and 310 patients (85%) had at least 2 hematocrits drawn during their stay. Of these patients, 20.6% had a hematocrit drop of 5 or more. Of patients with the hematocrit change, 18.8% were admitted to an Inpatient unit from the observation unit compared to 9.3% of patients without a significant hematocrit change (P = .034). In one of these patients who had a computed tomography scan before observation admission, which demonstrated free fluid, the hematocrit drop assisted in diagnosing significant intra-Abdominal injury. Conclusion: Although serial hematocrit testing may be useful in specific situations, routine use of serial hematocrit testing in trauma patients at a level I trauma center’s observation unit did not significantly aid in the detection of occult injury.

(C) 2010

Introduction

The use of serial hematocrit testing as a screening tool after the initial trauma evaluation has been suggested to aid in the diagnosis of intra-abdominal injury and may be the practice in some trauma centers [1,2]. One previous study has reported that a change in hematocrit of more than 5 points predicted serious injury. This study, however, evaluated the use of

* Corresponding author. Tel.: +1 801 581 2417; fax: +1 801 585 6699.

E-mail address: [email protected] (T. Madsen).

4-hour hematocrit change to predict injuries that had already been diagnosed in the initial trauma evaluation [1]. An additional study has demonstrated that a hematocrit change of 6.5 points at 15 minutes predicts significant injury in patients with penetrating trauma [2]. These studies, however, did not look specifically at patients who had received an initial trauma evaluation and then had serial hematocrit testing as a screening tool for undiagnosed injury.

The goal of our study was to determine whether routine serial hematocrit testing is useful in detecting occult serious injury in trauma patients after initial trauma evaluation. We

0735-6757/$ – see front matter (C) 2010 doi:10.1016/j.ajem.2009.01.034

specifically looked at the use of this testing in an observation unit, in which patients are routinely placed after a negative initial workup or a workup demonstrating minimal injury.

Materials and methods

University of Utah Medical Center is a level I trauma center in Salt Lake City, Utah. It is 1 of 2 level I trauma centers in the region that includes the state of Utah as well as parts of Idaho, Nevada, Wyoming, and Colorado. Trauma presenta- tions to the emergency department (ED) include all types, with most being blunt trauma from motor vehicle accidents.

In April 2006, the ED opened an observation unit. As part of the observation unit, we created a protocol for the evaluation of trauma patients. Admission to the observation unit was at the discretion of the trauma team and the attending emergency physician, with certain criteria established to exclude patients deemed too high risk for the observation unit. Before the creation of the observation unit, these were typically patients which the trauma team would admit to an inpatient unit. The unit is designed specifically for patients determined to be at low risk based on the results of the initial trauma evaluation and diagnostic testing. Exclusion criteria for observation admission included Abnormal vital signs, positive focused assessment with sonography for trauma examination, abnormal electrocardiogram, abnormal chest radiograph, abnormal head computed tomography scan, Glasgow Coma Scale score less than 14, or multiple system injury involvement. Patients were placed in the observation unit for serial examinations, pain control, additional testing, and to await specialty consultation. As part of the observation unit protocol, most patients had hematocrit levels drawn every 6 hours to aid in the detection of injuries that may not have been identified in the initial trauma evaluation. The use of serial hematocrit testing in the observation unit was per the discretion of the trauma service in discussion with the attending emergency physician.

This study was designed as a retrospective chart review of all patients admitted to the observation unit under the trauma protocol from April 1, 2006, to May 31, 2007. Fourth-year medical students, all of whom had worked in the ED, reviewed all charts of all trauma patients admitted to the ED observation unit during this period. They recorded patient characteristics, injuries, hematocrit levels, length of stay, additional injuries identified in the observation unit, and adverse outcomes, including loss of vital signs, intubation, death, and whether these patients were ultimately discharged from the hospital or admitted to an inpatient unit. Significant intra-abdominal injury was defined as any injury detected during observation that required operative repair or hospi- talization in an inpatient unit. For quality assurance purposes, one of the study’s lead investigators (JB) performed a review of 20% of the charts reviewed by the research associates (? = 0.98). These charts were randomly selected. An additional review was performed by the study’s primary investigator

(TM) of the charts of all patients admitted from the observation unit to an inpatient unit.

Patient hematocrit results and injuries identified in the observation unit were reviewed. Patients were considered to have a hematocrit drop of 5 points or more if any of the hematocrit levels after the initial hematocrit level (which was drawn on patient’s initial presentation to the ED) was 5 points or more lower. We used a hematocrit decrease of 5 points or more based on the previous study that established this threshold as a predictor of intra-abdominal injury [1]. Injuries identified during the observation unit stay were those which were not identified during the initial trauma workup and before placement in the observation unit. In cases of additional injuries identified in the observation unit, we reviewed the observation unit admission and progress notes, which were completed and timed at specific intervals in the observation stay. We did this to ensure these were not injuries that had been identified during the initial trauma evaluation before place- ment in the observation unit. In cases in which injuries were identified, the observation unit notes and dictated trauma surgery admission notes were reviewed to determine whether the hematocrit change had contributed to the decision-making process in the identification of these injuries.

In the case of hospital admission, injuries were reviewed and trauma surgery admission notes were reviewed to determine whether the change in hematocrit had influenced the decision to admit the patient to an inpatient unit from the observation unit. We also reviewed patient records to identify any return ED visits within 30 days of the observation unit stay to determine whether any additional injuries were missed during the observation unit stay.

The study was designed as a comparison between those patients with a hematocrit change of 5 points or more and those without this hematocrit change. Results are reported using ?², odds ratio, 95% confidence interval (CI), and descriptive statistics, with P b .05 considered statistically significant (SPSS v. 16.0; SPSS, Chicago, Ill).

Results

During the study period, 365 patients were placed in the observation unit, and 310 patients (85%) had at least 2

Table 1 injury mechanisms

Mechanism	Number	Percentage
Motor vehicle/Motorcycle accident	175	41.2
Fall	39	8.9
Ski/snowboard accident	24	5.5
Assault	19	4.4
Bicycle accident	18	4.1
Auto vs pedestrian	11	2.5
Stab wound	10	2.3
Horse accident	9	2.1
gunshot wound	5	1.1

Hematocrit change >=5		Hematocrit change b5	P
Age	33.4 y	35.2 y	.48
Sex (% male)	64%	67.5%	.61
Mechanism (% blunt trauma)	92.2%	91.5%	.85
CT abdomen performed	17.2%	15.8%	.80
before observation
Admitted to inpatient unit	18.8%	9.3%	.034
Returned to ED within 30 d	10.9%	5.3%	.10

hematocrits drawn between their initial evaluation and their observation unit stay. Initial hematocrit level was drawn at the time of the patient presentation to the ED. The observation unit trauma protocol called for hematocrit levels to be drawn every 6 hours. Ninety (29%) patients had 2 total hematocrit levels drawn, 170 (54.8%) had 3 levels drawn,

Table 2 Comparison of patients with hematocrit change of 5 or more vs those without change

and 48 (15.5%) had 4 levels drawn.

Of 310 with serial hematocrit testing performed, 282 (91%) had sustained blunt trauma, and 26 (8.4%) had sustained penetrating trauma. The most common specific mechanism was motor vehicle accident (175/310, 41.2%) (Table 1). Of the patients, 67.2% were male; and average age was 35.4 years (median age, 28 years). Average observation unit length of stay was 764 minutes (12 hours 44 minutes), with a median length of stay of 822 minutes (25th percentile

= 418 minutes, 75th percentile = 1039 minutes).

All patients had received a trauma evaluation, including FAST scan and CT scanning per the trauma team discretion, before placement in the observation unit. Fifty patients (16%) had a CT scan of the abdomen and pelvis performed before placement in the observation unit. The trauma protocol is to perform these scans with intravenous (IV) contrast only.

The average change in hematocrit was a decrease of 3 points (median, 3-point decrease). Of 310 patients, 64 (20.6%) had a hematocrit drop of 5 points or more during their observation unit stay. Comparison between patients with a significant hematocrit drop (>=5 points) and those without are listed in Table 2. Twelve (18.8%) of these patients with a hematocrit drop of 5 points or more were admitted from the observation unit to an inpatient unit vs 23

(9.3%) of 246 of patients without a significant hematocrit drop (P = .034; odds ratio, 2.0; 95% CI, 1.06-3.8). In only one of these cases did the admitting service cite the hematocrit change as the reason for inpatient admission, and this was a patient in whom the hematocrit change had led to a diagnosis of additional intra-abdominal injury.

Patient records were reviewed for 30 days after the observation unit visit. Nine patients (6.1%) returned to the ED. Return visit reasons included persistent pain, nausea, and urinary tract infection. In none of these patients were Significant injuries identified, which had not been diagnosed during the initial observation stay.

Three patients had additional significant intra-abdominal injuries detected during their stay in the observation unit. One patient who did not have a change in hematocrit had persistent abdominal pain and developed a fever while in the observation unit. This patient had a negative CT scan of the abdomen before placement in the observation unit. This patient was taken to the OR and discovered to have a bowel perforation. A second patient who did not have an abdominal CT scan before placement in the observation unit became tachycardic with increasing abdominal pain while in the observation unit. Computed tomography scan was ordered at this point (without a repeat hematocrit level available for use in the decision-making process for ordering this CT scan), and a second hematocrit level was sent. Computed tomography scan showed a grade 1 splenic laceration, for which the patient was admitted to an inpatient unit for continued observation, and the repeat hematocrit level had decreased 6.6 points from the first level drawn. The patient was observed in the inpatient unit and eventually discharged (Table 3).

A third patient with a significant additional intra- abdominal injury was the only patient in whom the serial hematocrit testing was cited as contributing to the identifica- tion of the injury. This patient had a positive FAST examination on presentation. The patient had a noncontrast CT scan due to a contrast allergy, which showed free fluid in the abdomen but no evidence of solid organ injury. The patient was placed in the observation unit, where the second hematocrit, drawn 2 hours after the first, decreased 1.3 points over the first, and the third hematocrit, drawn 6 hours after the first, decreased 5.8 points over the first. The patient was taken to the operating room and discovered to have a grade 4

Table 3

Patient

Patients with significant intra-abdominal injuries identified in observation unit

Significant injuries

1

Grade 4 Liver laceration, grade 1 to 2 splenic laceration

Small bowel perforation Grade 1 splenic laceration

Hematocrit decrease

5.8

CT finding before placement in observation

Moderate amount of free fluid in abdomen and pelvis

No acute intra-abdominal injury None performed

Outcome

Repaired in OR

2

3

2.8

6.6

Repaired in OR Observed in inpatient unit and discharged

OR indicates operating room.

liver laceration and grade 1 to 2 splenic laceration (Table 3). Of the 310 patients who had serial hematocrit testing performed, then, this testing modality identified 1 patient with an intra-abdominal injury requiring either operative repair or hospital admission (0.32%; 95% CI, 0%-0.95%).

Discussion

The goal of serial hematocrit testing is to detect occult, serious intra-abdominal injury. However, in our study, only 1 of 310 patients had an additional intra-abdominal injury detected while in observation due to serial hematocrit testing. This patient was one with extenuating circumstances, due to an IV contrast allergy, and had initial testing suggestive of intra-abdominal injury. In addition, this patient had a positive FAST examination on presentation to the ED, which would indicate that this patient was inappropriate for the observa- tion unit, based on the preestablished exclusion criteria. These findings suggest a low yield from serial hematocrit testing if used routinely in an observation unit for patients who have received a trauma evaluation and diagnostic testing before placement in the unit.

Several studies have shown the benefit of ED Observation Units in evaluating and risk-stratifying patients such as those with chest pain and transient ischemic attack, and treating dehydration and asthma [3-8]. Additional smaller studies have reported their use for the evaluation of trauma patients [9,10]. Because of the relative paucity of data regarding appropriate use of an observation unit for trauma patients, further research is needed to identify appropriate trauma patients for an observation unit and appropriate testing to be performed during this observation unit stay.

Based on our results, it would appear that it is only in certain cases where serial hematocrit testing may be helpful. This may aid in identifying injury in cases such as IV contrast allergy or large body habitus, which may prevent obtaining a CT scan. Trauma patients at a level I trauma center regularly have CT scanning performed, if indicated, as part of that evaluation, and studies to date on serial hematocrit testing have not evaluated the utility of this modality in detecting occult injury in patients who have already had a trauma surgery evaluation [1-2,11-13]. To place these trauma patients in an observation unit solely for the purpose of receiving serial hematocrit testing would not seem warranted. This does not imply that these patients do not warrant observation for pain control or serial examina- tions; however, the use of serial hematocrit testing in this setting does not appear to assist in diagnosing additional injuries. Still, the simple nature of performing serial blood draws, as well as the potential 1% incidence of injury detected (based on the 95% CI we report with the one patient with injury detected), may be enough to convince some practitioners of the utility of this testing modality.

There are several limitations to our study. This was a retrospective chart review, which has the potential for bias or

incorrect information based on the review of the patient records. As part of this study, we cite the decision-making process as outlined in the dictated admission notes, which also creates a potential for bias, as this does not include direct discussion with the admitting team and pointed questioning regarding the use of serial hematocrit in the decision-making process. Although the large majority of the observation unit patients had serial hematocrit testing performed, not all trauma patients underwent this testing. As this testing was per the discretion of the trauma team, this may create an additional potential for selection bias. This study also represents the results of only one institution and may not be applicable to other institutions where practice patterns differ. The varying use of CT scanning and quality of scanning at other institutions could change this dynamic, as well as the differences in trauma evaluations and protocols. There is also the issue of admissions based on hematocrit drop. Our study demonstrated that patients with a hematocrit drop of 5 points or more were admitted at a higher rate to an inpatient unit than those without the hematocrit drop. However, we reviewed the admission and discharge notes to see in which cases the admitting team cited the hematocrit drop as the reason for admission, and this was only cited in the case mentioned previously. In none of these additional patients were significant additional intra-abdominal injuries identified, whether in the observation unit, in the inpatient unit, or at 30-day follow-up. Although the admitting team did not cite the hematocrit change as the reason for admission, the possibility remains that this may have initially con- tributed to the admission, which would potentially increase resource expenditure based on the results of unnecessary

diagnostic testing.

Conclusion

At a level I trauma center, in which patients receive a trauma evaluation and diagnostic testing before placement in the observation unit, the use of routine serial hematocrit testing as a screening tool during observation rarely provides additional diagnostic information. These findings apply to patients who have undergone an initial trauma service evaluation and imaging and have been determined to be at low risk and appropriate for an observation unit. Serial hematocrit testing may be useful in specific situations and in patients with unique circumstances.

References

1. Zehtabchi S, Sinert R, Goldman M, Kaptianyan R, Ballas J. Diagnostic performance of serial haematocrit measurements in identifying Major injury in adult trauma patients. Injury 2006;37(1):46-52.
2. Paradis NA, Balter S, Davison CM, et al. Hematocrit as a predictor of significant injury after penetrating trauma. Am J Emerg Med 1997;15 (3):224-8.
3. Jagminas L, Partridge R. A comparison of emergency department versus inhospital chest pain observation units. Am J Emerg Med 2005; 23(2):111-3.
4. Goodacre S, Nicholl J, Dixon S, et al. Randomised controlled trial and economic evaluation of a chest pain observation unit compared with routine care. BMJ 2004;328(7434):254.
5. Lai C, Noeller TP, Schmidt K, King P, Emerman CL. Short-term risk after initial observation for chest pain. J Emerg Med 2003;25(4): 357-62.
6. Ross MA, Compton S, Medado P, Fitzgerald M, Kilanowski P, O’Neil BJ. An emergency department Diagnostic protocol for patients with transient ischemic attack: a randomized controlled trial. Ann Emerg Med 2007;50(2):109-19.
7. Mallory MD, Kadish H, Zebrack M, Nelson D. Use of a Pediatric observation unit for treatment of children with dehydration caused by gastroenteritis. Pediatr Emerg Care 2006;22(1):1-6.
8. Miescier MJ, Nelson DS, Firth SD, Kadish HA. Children with asthma admitted to a pediatric observation unit. Pediatr Emerg Care 2005;21 (10):645-9.
9. Henneman PL, Marx JA, Cantrill SC, Mitchell M. The use of an emergency department observation unit in the management of abdominal trauma. Ann Emerg Med 1989;18(6):647-50.
10. Ammons MA, Moore EE, Rosen P. Role of the observation unit in the management of Thoracic trauma. J Emerg Med 1986;4(4):279-82.
11. Winslow J, Hinshaw J, Hughes M, William R, Bozeman W. Quantitative assessment of diagnostic radiation doses in adult blunt trauma patients. Ann Emerg Med 2008;52(2):93-7.
12. Brenner DJ, Hall EJ. Computed tomography–an increasing source of radiation exposure. N Engl J Med 2007;357(22):2277-84.
13. Elton C, Riaz AA, Young N, Schamschula R, Papadopoulos B, Malka

V. Accuracy of computed tomography in the detection of blunt bowel and Mesenteric injuries. Br J Surg 2005;92(8):1024-8.