Bladder Pressure Measurements Are an Independent Predictor of Urinary Tract Infection in Trauma Patients

Patients and Methods

Epidemiologic data were collected prospectively from January 2003 through December 2005 on all trauma patients in a surgical trauma ICU (STICU) at our Level 1 trauma center. A dedicated epidemiology nurse evaluated each patient once daily, noting the presence or absence of a urinary drainage catheter, whether bladder pressure measurements (BPMs) were being taken, and the presence of a urinary tract infection (UTI) according to the definition of the U.S. Centers for Disease Control and Prevention (CDC) [14]. Patients with and without UTIs were compared using demographics, Injury Severity Score (ISS), and outcomes from data obtained from the trauma registry.

Blood pressure measurements were performed at the discretion of the trauma attending physician when there was a concern about intra-abdominal hypertension. An open system of measurement was used in which the catheter was disconnected from the collection bag to deliver 50 mL of saline into the bladder. An 18-gauge needle was then inserted into the catheter and connected to tubing and to a pressure monitor similar to an arterial line to follow pressure tracings.

Continuous variables were evaluated using the Wilcoxon rank test, and the nominal variables were evaluated using the Fisher exact test. Logistic regression was used to identify independent predictors of UTI and death. This study was approved by the Virginia Commonwealth University Institutional Review Board.

Results

A total of 938 patients were included in the trial. There were 50 patients without a catheter, and none of those patients had a UTI. There were 836 patients with catheters who did not have BPMs obtained, and 36 of these patients had a UTI (4.31%). In the remaining group of 52 patients who had catheters and BPMs, there were 12 patients who had a UTI (23.08%)(p < 0.0001). As shown in Table 1, patients with UTIs were more likely to be severely injured females who were, on average, older than patients without UTIs. These patients also had longer ventilator times (1.5 ± 4.9 days for those without UTI vs. 6.9 ± 12.2 days with UTI; p = 0.0069), intensive care unit stays (6.6 ± 10.4 days [no UTI] vs. 22.6 ± 31.8 days [UTI]; p < 0.0001), and hospital stays (13.4 ± 15.3 days [no UTI] vs. 48.9 ± 45.0 days [UTI]; p < 0.0001).

In the group of 52 patients who had catheters and BPMs, there was no difference in the number of days BPMs were obtained in those with UTIs (3 ± 2.26 days) and those without (4.075 ± 4.04 days) (p = 0.025). Using logistic regression, BPM was an independent predictor of UTI, with infection being seven times more likely to occur in these patients (Table 2). In addition, 19.2% of the patients with BPM died, a significantly higher rate than in those without catheters and those with catheters but without BPM (p = 0.0025) (Table 3). Along with age (odds ratio [OR] 1.039; 95% confidence interval [CI] 1.024-1.054) and ISS (OR 1.081; CI 1.056-1.106), having BPM was an independent predictor of death (OR 2.475; CI 1.191-6.328).

Discussion

Wallace et al. [15] noted more than a decade ago that trauma patients are at higher risk for nosocomial infections than are other surgical patients. Since that time, more emphasis has been placed on infection control measures, including minimization of device use. Miller et al. [16] significantly decreased the rates of BSI and UTI and reduced the occurrence of VAP with initiation of evidence-based plans, staff education, and commitment to protocol compliance. We showed similar results using such methods as a central line protocol and a checklist, decreasing both central line use and catheter-related BSI [13]. However, these standard infection control measures did not reduce our rate of UTIs. Hence, we investigated other factors that could have influenced the rate of UTI in our STICU population so we could improve these measures.

Many factors contribute to the development of UTIs, particularly in the critically ill patient. These usually are associated with manipulation of the genitourinary tract by catheters [17,18]. The catheter acts as a conduit for introduction of bacteria into the normally sterile bladder. This foreign body in the presence of a compromised host allows bacterial overgrowth and subsequent infection. Given the epidemiology of UTI, we hypothesized that greater manipulation of these catheters, such as by pressure measurements, would result in more infections.

As demonstrated in this study, those patients who had BPMs were much more likely to develop UTIs, and BPM was an independent predictor of infection and death. Such catheter manipulation has increased with greater awareness of the possibility of intra-abdominal hypertension and the need to perform BPMs to make the diagnosis [19]. Because treatment relies heavily on early diagnosis and aggressive decompression, the optimal method of obtaining the measurement must be determined. The specific complications of the technique have been poorly documented, however, which speaks to the value of this manuscript.

The approach at our institution during the time of this study was an open technique in which the catheter was compromised through insertion of an 18-gauge needle and disconnection of the catheter to allow instillation of 50 mL of saline into the bladder. Despite cleaning of the catheter, we show clearly that not only did this procedure increase the risk of UTI, it increased the risk of death. We are aware of only one other published study evaluating the relation between BPM for the diagnosis of intra-abdominal hypertension and UTIs, but these authors used a closed technique such that the drainage circuit was never broken [20]. Their study found no increase in the UTI rate with the use of BPM.

The limitations of these data include the small sample size (52 patients with BPMs) and the retrospective nature of the trauma registry data. Furthermore, no interventions were evaluated to identify ways to improve the UTI rate and lower the mortality rate. Consequently, in view of these data and the study by Cheatham et al. [20], we changed our approach to BPM to a closed technique. In addition, we instituted an aggressive checklist to question daily the need for the catheter; the indications for maintaining the catheter are listed in Table 4. Data from our institution, presented recently at the Society of Hospital Epidemiologists meeting (abstract unpublished), confirm that these interventions decreased the UTI rate.

As a Level 1 trauma center, we often are consulted by ICUs within our own institution and by outside institutions to educate the staff on the proper technique of obtaining BPMs. To provide appropriate guidance, it is important to look continually for ways to improve the approach, whether it is using a closed system, identifying the correct amount of saline to infuse into the bladder [21], or simplifying the process while maintaining accuracy [22,23]. We believe these data promote the goal of optimizing the approach so that we, our colleagues, and, most importantly, our patients benefit.