WTA 2014 PLENARY PAPER
Urinary tract infection in elderly trauma patients: Review of the Trauma Quality Improvement Program identifies the population at risk Stephanie F. Polites, MD, Elizabeth B. Habermann, PhD, Kristine M. Thomsen, Mahmoud A. Amr, MD, Donald H. Jenkins, MD, Scott P. Zietlow, MD, and Martin D. Zielinski, MD, Rochester, Minnesota
Elderly trauma patients are at high risk for urinary tract infection (UTI). Despite this, UTI has been deemed a potentially preventable problem and therefore not reimbursable by the Centers for Medicare and Medicaid Services. Early identification of UTI in these patients should lead to prompt treatment, improved outcomes, and cost savings. Risk factors for UTI development in this population must be elucidated to realize these goals. METHODS: The Trauma Quality Improvement Program (TQIP) database was used to analyze elderly patients (Q65 years) admitted as a result of injury during 2011. Patients with genitourinary injuries or undergoing dialysis before admission were excluded. Multivariable logistic regression analysis was conducted to identify UTI risk factors. Mean cost of UTI was calculated based on the assumption of $862 to $1,007 per UTI. RESULTS: In total, 33,257 patients were identified; 1,492 developed UTI (4.5%). Multiple significant risk factors were identified, including age greater than 75 years, female sex, ascites, moderate head injury, impaired sensorium, congestive heart failure, and duration of hospital stay (all p G 0.05). Assuming that UTIs diagnosed on hospital Day 1 were preexisting, the cost of UTI to TQIP hospitals ranged from $1,280,959 to $1,496,434 per year. CONCLUSION: Duration of stay has a profound impact on the development of UTIs in elderly trauma patients, but overall severity of injury does not. In addition, multiple nonmodifiable risk factors were identified, prompting the possibility for increased screening of occult UTIs. Reimbursement for care of UTI in this complicated patient population should be revisited. The TQIP database must improve urinary catheter data. (J Trauma Acute Care Surg. 2014;77: 952Y959. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level III. KEY WORDS: Urinary tract infection; trauma; elderly; injury; Medicare. BACKGROUND:
N
early 6% of hospitalized patients will develop a nosocomial infection, of which urinary tract infections (UTIs) are the most frequent.1Y3 Injured patients are particularly at risk for the development of a UTI because they often require prolonged hospitalization and have altered physiology that increases susceptibility to infection and urinary bacterial overgrowth.4 In addition to the known risk factors of age, female sex, and urinary catheterization in all hospitalized patients, UTI was found to be the most frequent nosocomial infection following spinal cord and head injury in trauma patients specifically.5,6 Since UTIs continue to plague injured patients, identification of additional risk for UTI development is needed for several reasons. Foremost among these reasons are the prolongation of Submitted: February 17, 2014, Revised: April 20, 2014, Accepted: April 21, 2014, Published online: September 22, 2014. From the Division of Trauma, Critical Care, and General Surgery (S.F.P., M.A.A, D.H.J., S.P.Z., M.D.Z.), Mayo Clinic, and The Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery (K.M.T., E.B.H.), Mayo Clinic, Rochester, Minnesota. This study was presented at the 44th Annual Meeting of the Western Trauma Association, March 2Y7, 2014, in Steamboat Springs, Colorado. The TQIP remains the full and exclusive copyrighted property of the ACS. The ACS is not responsible for any claims arising from works based on the original data, text, tables, or figures. Address for reprints: Martin D. Zielinski, MD, Division of Trauma, Critical Care, and General Surgery, Mary Brigh 2-810, St. Mary’s Hospital, Mayo Clinic, 1216 Second St. SW, Rochester, MN 55902; email: [email protected]. DOI: 10.1097/TA.0000000000000351
952
hospitalization and the increased risk of mortality associated with UTI.6Y8 Second, catheter-associated UTI (CAUTI) is considered a ‘‘reasonably preventable’’ and common in-hospital complication by the Centers for Medicare and Medicaid Services (CMS) and is no longer reimbursed. The cost of UTIs, estimated at $867 to $1,100 each, combined with costs associated with any delay in dismissal, results in a significant economic burden to hospitals.9 Since catheter use often cannot be avoided in trauma patients, resulting in lack of reimbursement of care for CAUTIs by CMS, identification of unmodifiable risk factors may provide leverage to modify policies that penalize institutions that care for high-risk injured patients. Due to the clinical implications of poor outcomes resulting from UTIs and the economic implications when they are catheter associated, identification of predictive features specific to this population is needed so that at-risk patients can be targeted for prevention efforts, earlier diagnosis, and ultimately, better outcomes. Therefore, the purpose of this study was to use the collective national experience of quality-seeking trauma centers through the Trauma Quality Improvement Program (TQIP) to determine risk factors of hospital-associated UTI in elderly trauma patients, regardless of catheter status.
PATIENTS AND METHODS We performed a retrospective study of injured patients 65 years or older. This age cutoff was chosen because it J Trauma Acute Care Surg Volume 77, Number 6
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
coincides with the start of Medicare eligibility. Within these patients, we identified those who developed a UTI during their hospitalization. Characteristics of patients with and without a UTI were compared, and independent predictors of UTI were determined.
Data Source Patients were identified from the 2011 TQIP Participant Use File (PUF). Built on the infrastructure of the American College of Surgeons’ (ACS) Committee on Trauma National Trauma Data Bank (NTDB), TQIP was designed to allow ACSdesignated Level I and II trauma centers to recognize and stratify their outcomes.10 Patient- and institution-level variables allow for risk adjustment and subsequent benchmarking with the goal of improving quality of care at participating trauma centers. In 2011, there were 160 centers participating in TQIP. Data are entered at each trauma center by trained, dedicated nurse abstractors and is validated externally on a quarterly basis. Standard definitions of comorbidities and in-hospital complications are used and can be determined through software registry programs, chart abstraction by registrar, and derivation from International Classification of DiseasesV9th Rev.VClinical Modification, (ICD-9-CM) codes.11 The definition of UTI in TQIP uses the NTDB definition and is based on a combination of clinical and microbiological criteria (Appendix 1).12 TQIP does not distinguish between CAUTI and UTI not associated with catheter use; therefore, this study examined risk factors for any UTI rather than CAUTI specifically. The ICD-9-CM procedure code for insertion of a urinary catheter (57.94) was used as a surrogate for catheter use. This procedure code includes urinary catheters inserted in any setting during the current hospitalization; thus, it would include catheters placed in the emergency department and operating room but not catheters placed en route or before transfer. Furthermore, because this code is only reflective of urinary catheter insertion, duration of use and sterility of placement could not be obtained. Rather than perform a separate analysis for CAUTI using this limited variable, this study analyzed all UTIs with insertion of a urinary catheter as a dependent variable.
Polites et al.
chemotherapy for cancer within 30 days, congenital anomalies, congestive heart failure (CHF), current smoker, cerebrovascular accident, diabetes mellitus (DM), disseminated cancer, functionally dependent health status, history of angina within 1 month, history of myocardial infarction within 6 months, history of revascularization or amputation for peripheral vascular disease, hypertension requiring medication, impaired sensorium, obesity, respiratory disease, steroid use, and cirrhosis. Length of stay (LOS) was categorized by 1, 2, 3 to 5, 6 to 8, 9 to 14, 15 to 21, and 22 or more days. Duration of stay in the intensive care unit and days of mechanical ventilation were also determined.
Statistical Analysis The cohort was described in aggregate using percentages, means (SDs), and medians (interquartile ranges [IQRs]). UTI rate by injury type was determined using ICD-9-CM diagnosis codes based on the methodology described by Fraser et al.6,11 Associations of demographic, clinical, and institutional factors with UTI were evaluated using W2 tests, and Fisher’s exact tests were used when low expected cell counts were observed. Associations of continuous factors with UTI were
Inclusion/Exclusion Criteria All patients 65 years or older in the 2011 TQIP PUF were identified for inclusion. Patients who required dialysis at the time of injury and those who underwent an operative procedure to repair a kidney, ureter, or bladder injury (ICD-9-CM codes 55.81Y55.99) were excluded. In addition, patients with missing data entries for age or UTI were excluded.
Variables Data collected included patient demographic and clinical information as well as institutional information. Institutional variables included ACS verification level (I vs. II), institutional tax status (for profit vs. nonprofit), teaching status (community, nonteaching, academic), and quartiles of hospital bed size. Patients were separated into age groups based on 5-year increments. Abbreviated Injury Scale (AIS) Yderived Injury Severity Score (ISS) and Glasgow Coma Scale (GCS) score were obtained directly from TQIP. Preexisting comorbidities included alcoholism, ascites within 30 days, bleeding disorder,
Figure 1. Identification of a cohort of elderly trauma patients from the 2011 TQIP PUF. Patients with missing age or UTI data and patients who required dialysis or underwent an operative procedure to repair a kidney, ureter, or bladder injury were excluded.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
953
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
TABLE 1. Rates of UTI Associated With Patient and Trauma Center Demographic Factors Patient demographics Age, y
Sex
Interhospital transfer
AIS-derived ISS
GCS score
Trauma center demographics Teaching status
ACS verification level
No. adult beds
All Patients (N = 33,257)
No UTI (n = 31,765)
UTI (n = 1,492)
p
65Y69 70Y74 75Y79 80Y84 85Y89* Not available Male Female Not available No Yes Not available e9 10Y16 17Y25 Q26 Not available 3Y8 9Y12 13Y15
6,212 5,744 6,513 7,814 6,974 6 15,068 18,183 11 21,009 12,237 630 14,710 9,148 6,313 2,456 1,374 2,278 961 28,644
5,989 (96.4) 5,511 (95.9) 6,196 (95.1) 7,446 (95.3) 6,623 (95.0) 6 14,523 (96.4) 17,236 (94.8) 11 20,083 (95.6) 11,671 (95.4) 607 14,109 (95.9) 8,776 (95.9) 5,992 (94.9) 2,281 (92.9) 1,304 2,179 (95.7) 858 (89.3) 27,424 (95.7)
223 (3.6) 233 (4.1) 317 (4.9) 368 (4.7) 351 (5.0) V 545 (3.6) 947 (5.2) V 926 (4.4) 566 (4.6) 23 601 (4.1) 372 (4.1) 321 (5.1) 175 (7.1) 70 99 (4.3) 103 (10.7) 1,220 (4.3)
G0.001
University Community Nonteaching Not available** I II Not available G350 350Y499 500Y699 Q700
18,107 12,774 2,376 7,650 17,466 8,141 62 7,419 7,753 8,808 9,215
17,158 (94.8) 12,315 (96.4) 2,292 (96.5) 7,255 16,620 (95.2) 7,890 (96.9) 55 7,192 (96.9) 7,430 (95.8) 8,365 (95.0) 8,723 (94.7)
949 (5.2) 459 (3.6) 84 (3.5) 395 846 (4.8) 251 (3.1) 7 227 (3.1) 323 (4.2) 443 (5.0) 492 (5.3)
G0.001
0.36
G0.001
G0.001
G0.001
G0.001
G0.001
*No patients 90 years or older were identified. **ACS verification level was not available for patients treated at trauma centers with state designation only. Values are presented as n (%). Percentages reported as row percents.
evaluated using t tests; if the factors were not sufficiently normally distributed, Wilcoxon rank-sum tests were used. Multivariable logistic regression was performed to identify independent predictors of UTI inclusive of urinary catheter insertion. Results of multivariable analysis are expressed as odds ratios (ORs) with 95% confidence intervals (CIs). Statistical significance was acknowledged when p G 0.05. All data analysis was completed using SAS version 9.3 (SAS Institute, Inc., Cary, NC).
RESULTS Of the 126,130 total patients in the 2011 TQIP PUF, 33,257 injured patients 65 years or older were included in this study (Fig. 1). The mean (SD) age was 77.5 (7.2) years, and the majority (54.7%) were female. No patients 90 years or older were identified. In-hospital mortality of the cohort was 7.5%. 954
TABLE 2. Rates of UTI by Injury Type Injury Type (ICD-9 Code)H
Rate of UTI, %
Colon (863.4Y5) Spinal cord (806, 952) Kidney, ureter, bladder, urethra (866, 867.0Y3) Thorax (lungs, bronchi, esophagus) (861.2, 861.3, 862) Small bowel (863.2Y3) Liver (864) Pelvic fractures (808) Spleen (865) Long bone fractures (812, 813, 820, 821, 823) Multiple rib fractures (807.0(2Y9), 807.1(2Y9)) Intracranial (800.1Y4, 800.6Y9, 801.1Y4, 801.6Y9, 851Y954)
11.1 8.1 7.2 7.0 6.8 6.6 6.4 5.7 5.1 4.9 4.5
*Based on methodology described by Fraser et al. for determining rates of infectious complications by injury type.6,11 Injury type categories include all subcodes of indicated ICD-9 codes when available.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
TABLE 3. Rates of UTI Associated With Patient ComorbiditiesH and LOS Hypertension on medication DM Bleeding disorder Impaired sensorium Respiratory disease CHF Cerebrovascular accident/residual neurologic deficit Current smoker Alcoholism Functionally dependent Myocardial infarction within 6 mo Obesity Disseminated cancer Angina within 1 mo Steroid use Chemotherapy within 30 d History of revascularization/amputation for peripheral vascular disease Cirrhosis Congenital anomalies Ascites within 30 d LOS,** d 1 2 3Y5 6Y8 9Y14 15Y21 Q22 Intensive care days 0 1Y3 4Y7 8Y14 Q15 Ventilator days 0 1Y2 3Y4 5Y9 10Y14 Q15
All Patients (N = 33,257)
No UTI (n = 31,765)
UTI (n = 1,492)
p
20,379 (61.9) 7,793 (23.7) 4,742 (14.4) 4,609 (14.0) 4,067 (12.4) 3,077 (9.3) 2,066 (6.3)
19,374 (61.6) 7,393 (23.5) 4,502 (14.3) 4,356 (13.9) 3,844 (12.2) 2,879 (9.2) 1,952 (6.2)
1,005 (67.9) 400 (27.0) 240 (16.2) 253 (17.1) 223 (15.1) 198 (13.4) 114 (7.7)
G0.001 0.002 0.044 G0.001 0.001 G0.001 0.021
1,967 (6.0) 1,334 (4.1) 1,308 (4.0) 1,138 (3.5) 1,044 (3.2) 659 (2.0) 617 (1.9) 335 (1.0) 171 (0.5) 165 (0.5)
1,879 (6.0) 1,264 (4.0) 1,244 (4.0) 1,085 (3.5) 979 (3.1) 623 (2.0) 579 (1.8) 316 (1.0) 161 (0.5) 152 (0.5)
138 (0.4) 98 (0.3) 26 (0.1) 2,171 (6.5) 2,976 (9.0) 12,259 (36.9) 7,632 (23.0) 4,910 (14.8) 1,800 (5.4) 1,501 (4.5) 18,037 (54.2) 9,012 (27.1) 3,243 (9.8) 1,687 (5.1) 1,278 (3.8) 28,236 (84.9) 1,993 (6.0) 779 (2.3) 944 (2.8) 556 (1.7) 749 (2.3)
128 (0.4) 94 (0.3) 22 (0.1) 2,167 (6.8) 2,956 (9.3) 12,045 (37.9) 7,318 (23.0) 4,517 (14.2) 1,568 (4.9) 1,186 (3.7) 17,439 (54.9) 8,764 (27.6) 3,017 (9.5) 1,477 (4.6) 1,068 (3.4) 27,194 (85.6) 1,922 (6.1) 711 (2.2) 844 (2.7) 480 (1.5) 614 (1.9)
88 (5.9) 70 (4.7) 64 (4.3) 53 (3.6) 65 (4.4) 36 (2.4) 38 (2.6) 19 (1.3) 10 (0.7) 13 (0.9) 10 (0.7) 4 (0.3) 4 (0.3) 4 (0.3) 20 (1.3) 214 (14.3) 314 (21.0) 393 (26.3) 232 (15.5) 315 (21.1) 598 (40.1) 248 (16.6) 226 (15.1) 210 (14.1) 210 (14.1) 1,042 (69.8) 71 (4.8) 68 (4.6) 100 (6.7) 76 (5.1) 135 (9.0)
0.95 0.18 0.48 0.79 0.006 0.23 0.045 0.30 0.39 0.036 0.12 90.99 0.028 G0.001
G0.001
G0.001
*Comorbidity data were missing for 345 patients. **LOS was missing for 8 patients. Values are presented as n (%). Percentages are reported as column percents.
Univariate Analysis of Patients With and Without UTI UTI was diagnosed in 1,492 patients (4.5%). The rate of UTI increased as patient age increased ( p G 0.001) and UTIs were more likely to occur in females than in males (5.2% vs. 3.6%, p G 0.001) (Table 1). The rate of UTI increased with greater ISS ( p G 0.001); however, UTI was more frequent in patients with a GCS score of 9 to 12 than those with 13 to 15 ( p G 0.001). There was no significant difference in UTI between patients treated at nonprofit and for-profit facilities (4.5% vs. 4.6%, p = 0.83), but UTIs were more frequent at university teaching facilities
( p G 0.001) and facilities with ACS Level I verification ( p G 0.001). The rate of UTI was highest in patients with colon injuries (11.1%) followed by spinal cord injuries (8.1%). The rate of UTI was lowest in patients with intracranial injuries (4.5%) (Table 2). The most common comorbidity was hypertension requiring medication, which was present in 61.9% of the patients, followed by DM (23.7%). Comorbidity information was missing in 345 patients (1.0%). Several comorbidities were associated with an increased frequency of UTI (Table 3). Insertion of an indwelling urinary catheter was documented in 11.8% of the patients (n = 3,913), and the frequency of UTI
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
955
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
(Table 4). Patients 75 years or older were at increased odds of UTI when compared with patients 69 years or younger. Females were at increased risk when compared with males ( p G 0.001; OR, 1.86; 95% CI, 1.65Y2.10). Although significant on univariable analysis, ISS was not predictive of UTI on multivariable analysis ( p 9 0.05). A moderate head injury resulting in GCS score of 9 to 12 was a risk factor when compared with GCS score of 13 to 15 ( p G 0.001; OR, 1.79; 95% CI, 1.41Y2.26) though this was not true for GCS score of 3 to 8 ( p 9 0.05). Institution-level risk factors were university teaching status and larger number of beds. ACS verification level was not included in the multivariable analysis because it was closely related to other institutional variables. Ascites within 30 days, CHF, and impaired sensorium were significant risk factors for UTI (all p G 0.05); however, many other comorbidities that were significant on univariable analysis were not significant after adjustment. Using a LOS of 3 days to 5 days as a reference point, decreased LOS was found to be protective against UTI, while increased LOS was predictive of UTI. Due to the close relationship of intensive care days and ventilator days with hospital LOS, these variables could not be included in the model. Patients with a documented indwelling urinary catheter were at an increased risk for UTI after adjusting for other variables (p = 0.001; OR, 1.27; 95% CI, 1.10Y1.47).
DISCUSSION
Figure 2. Duration of hospitalization, intensive care, and mechanical ventilation as well as rate of UTI. The rate of UTI increased exponentially with increased LOS and linearly with more intensive care and ventilator days.
was greater in these patients (7.1 vs. 4.1%, p G 0.001). Patients who required intensive care were more likely to develop UTI (5.9% vs. 3.3%, p G 0.001) as were patients who required mechanical ventilation (9.0% vs. 3.7%, p G 0.001). The rate of UTI increased linearly with the duration of intensive care and mechanical ventilation and exponentially with LOS (Fig. 2). The median (IQR) LOS was significantly longer for patients with a UTI (11 [7Y19] days vs. 5 [3Y8] days, p G 0.001).
Multivariable Analysis Multivariable analysis indicated that there are several patient-level risk factors for UTI in elderly trauma patients 956
UTIs are a significant burden of morbidity to trauma patients, who, because of their age, are susceptible to altered urine microbiology, transient immunosuppression, and frequent requirement of urinary catheters; however, predictive features specific to this population for UTI, outside of urinary catheterization, are unknown.4,13 These risk factors are needed to guide health care providers in prevention, early diagnosis, and effective treatment efforts. This study identified a 4.5% rate of UTI in trauma patients, age or 65 years or greater, treated at Level I and II trauma centers participating in the TQIP. Several patient demographic and clinical factors emerged as independent risk factors for UTI including age 75 years or greater, female sex, ascites, CHF, and impaired sensorium. Age and sex are consistent with existing trauma and critical care literature;14,15 however, this is the first description of specific comorbidity risk factors for UTI in elderly trauma patients. Although patients with DM have an increased incidence of bacteriuria, we did not find that diabetes was associated with UTI in this study.5 Moderate head injury and need for intensive care or mechanical ventilation were associated with UTI and should be used to risk stratify, rather than ISS. We suspect that patients with GCS score of 3 to 8 did not have an increased risk of UTI because of the high early mortality associated with severe head injuries. Similarly, this may have resulted in the relatively low rate of UTI among patients with intracranial injuries. The relatively high rate of UTI found in patients with spinal cord injuries is consistent with the limited existing literature on UTIs in injured patients.6 The risk of UTI was amplified as LOS, duration of intensive care, and duration of mechanical ventilation increased. * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
TABLE 4. Independent Risk Factors for UTI in Elderly Trauma Patients Using Logistic RegressionH Risk Factor Age, y (vs. 65Y69)
70Y74 75Y79 80Y84 85Y89
Female sex Teaching status (vs. community) No. adult beds (vs. G350 beds)
AIS derived ISS (vs. e9)
GCS score (vs. 13Y15)
Nonteaching University 350Y499 500Y699 Q700 10Y16 17Y25 Q26 3Y8 9Y12
Ascites within 30 d CHF Bleeding disorder Cerebrovascular accident/residual neurologic deficit DM History of angina within past 1 mo History of revascularization/amputation for peripheral vascular disease Impaired sensorium Obesity LOS, d (vs. 3Y5) 1 2 6Y8 9Y14 15Y21 Q22 Insertion of indwelling urinary catheter
p
OR
0.231 0.001 G0.001 G0.001 G0.001 0.59 G0.001 0.030 G0.001 G0.001 0.78 0.61 0.11 0.33 G0.001 0.046 0.028 0.41 0.32 0.08 0.09 0.16
1.13 1.40 1.41 1.57 1.86 1.08 1.33 1.23 1.47 1.43 0.98 0.96 1.18 0.89 1.79 3.36 1.21 1.07 1.11 1.13 1.35 1.56
0.92 1.16 1.17 1.30 1.65 0.82 1.16 1.02 1.22 1.20 0.85 0.82 0.96 0.70 1.41 1.02 1.02 0.91 0.90 0.99 0.95 0.84
1.38 1.69 1.70 1.91 2.10 1.41 1.52 1.49 1.76 1.72 1.13 1.12 1.44 1.12 2.26 11.06 1.44 1.25 1.38 1.28 1.93 2.90
0.004 0.76 G0.001 0.001 G0.001 G0.001 G0.001 G0.001 0.001
1.25 0.96 0.13 0.43 2.32 5.03 9.27 17.67 1.27
1.07 0.72 0.05 0.27 1.92 4.20 7.52 14.38 1.10
1.46 1.27 0.36 0.69 2.79 6.03 11.42 21.70 1.47
95% CI
*Missing data as outlined in Tables 1 and 2 were not included in the multivariable analysis.
Cause and effect relationships, however, cannot be inferred from our current study. Laupland et al.15 previously found that duration of intensive care unit stay predicted UTI. Others have concluded that nosocomial infections actually result in increased LOS and excess cost.7,16,17 Granular, prospective data including complications of UTI and timing of diagnosis and treatment are needed to elucidate this causative relationship in trauma patients. Urinary catheter use relied on patients’ having the appropriate ICD-9-CM diagnosis code, and we suspect the actual incidence was much greater. Furthermore, TQIP does not capture information regarding duration of urinary catheterization, the leading risk factor for UTI development and a requirement for CAUTI. Nonetheless, patients who experience UTIs, regardless of catheter use, continue to be at risk of poor outcomes. Catheter use is not the only risk factor in trauma patients, as demonstrated in the current analysis. Since not all catheterized patients develop UTI, there is room for further risk modification using these results. Furthermore, even if UTIs cannot be prevented, identification of those at highest risk may facilitate early diagnosis and treatment and, ultimately, avoidance of some of the possible downstream
negative effects including increased LOS, urinary sepsis, mortality, and lack of reimbursement to providers. Unfortunately, in the acute setting of trauma, most risk factors, particularly comorbidities, are nonmodifiable. The results of this study, however, provide several opportunities for prevention and early treatment efforts. Urinary catheters are often required in trauma patients because of the need for careful hemodynamic and fluid monitoring, surgical interventions, or mobility limitations. Nevertheless, it is paramount that all hospitals caring for injured patients develop processes that ensure removal of catheters at the earliest appropriate time. In addition, we recommend that providers have a high index of suspicion for UTI in elderly trauma patients who remain hospitalized for more than 2 days because of the increased risk associated with hospital stay and duration of intensive care and/or mechanical ventilation. It may be difficult for injured elderly patients to voice concerns about symptoms, resulting in delayed diagnosis and treatment. Although it is our institutional practice to use screening urinalysis in such patients, efficacy data are lacking. Prospective studies on effective screening practices for UTIs in elderly, injured patients are needed.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
957
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
Patients who received care at large or university teaching hospitals were at increased risk of UTI, suggesting that, despite inclusion of GCS and ISS in our analysis, the complexity of care at these trauma centers was not accounted for in our modeling or that there are processes inherent to these centers that result in more UTIs. Larger and higher-acuity centers may receive patients who are more likely to require intensive care, mechanical ventilation, or surgical care, which typically involves insertion of a urinary catheter. We also hypothesize that urinary catheters may not be as aggressively removed at university centers because of the presence of trainees. Providers at institutions with high-risk characteristics should be aware of these findings and initiate quality improvement efforts to minimize this risk. TQIP provides robust data on a large number of trauma patients from quality-seeking trauma centers; however, there are several more limitations of this study that must be acknowledged. Only ACS-verified Level I and II trauma centers can participate in TQIP with specific inclusion and exclusion parameters, resulting in a selection bias toward more severely injured patients. Although TQIP centers maintain their data prospectively, we used the TQIP data in a retrospective fashion, which introduces further bias. It is possible that we were not able to capture all potential patient- and institutional-level influences on the risk of UTI. It is also acknowledged that some UTIs may have been present on admission and, therefore, were not hospital acquired. Lastly, the rate of urinary catheter insertion was low and was likely underestimated in this study because of the reliance on ICD-9 coding. More specific data about urinary catheter placement and duration are required and should be implemented by the ACS Committee on Trauma for future TQIP data collection. Robust catheter data would allow for improved investigation of CAUTI specifically, the infections which result in the loss of reimbursement to TQIP hospitals. Since specific injuries are undoubtedly associated with varying degrees of catheter use, further investigation of injury types that are most susceptible to UTIs could not be conducted because of poor catheter data in the TQIP. Since hospital-acquired UTIs increase mortality and are not reimbursable by CMS if associated with a urinary catheter, it is necessary that trauma centers take action to reduce the occurrence of this nosocomial infection and to reduce the morbidity of patients who develop nonpreventable UTIs. While it is unlikely that UTIs will be eliminated entirely in elderly trauma patients, the results of this study will aid in risk stratification and earlier diagnosis. In addition to prompt urinary catheter removal, development of screening processes for UTIs in injured patients with prolonged duration of hospitalization and critical care may be of use. Lastly, results of this study should provide leverage to modify policies that penalize institutions that care for high-risk injured patients. AUTHORSHIP S.F.P. contributed to the literature search, study design, data interpretation, and manuscript writing. E.B.H. contributed to the study design, data analysis, and data interpretation. K.M.T. contributed to the data collection, data analysis, and critical revision of the manuscript. M.A.A. contributed to the study design and critical revision of the manuscript. D.H.J. contributed to the study design and critical revision of the manuscript. S.P.Z. contributed to critical review of the manuscript. M.D.Z.
958
contributed to the literature search, study design, data interpretation, and critical revision of the manuscript.
DISCLOSURE This publication was made possible by CTSA grant number UL1 TR000135 from the National Center for Advancing Translational Sciences (NCATS), a component of the National Institutes of Health (NIH).
APPENDIX 1. TQIP DEFINITION OF UTI* UTI: Defined as an infection anywhere along the urinary tract with clinical evidence of infection, which includes at least one of the following symptoms with no other recognized cause: 1. Fever of 38-C or greater 2. White blood cell (WBC) count greater than 100,000/HL or less than 3,000/HL 3. Urgency 4. Frequency 5. Dysuria 6. Suprapubic tenderness AND positive urine culture (Q100,000 microorganisms per microliter of urine with no more than two species of microorganisms) OR at least two of the following signs or symptoms with no other recognized cause: 1. Fever of 38-C or greater 2. WBC count greater than 100,000/HL or less than 3,000/HL 3. Urgency 4. Frequency 5. Dysuria 6. Suprapubic tenderness AND at least one of the following: 1. Positive dipstick for leukocyte esterase and/or nitrate 2. Pyuria (urine specimen with 910 WBC/HL or 93 WBC/ high-power field of unspun urine) 3. Organisms seen on Gram stain of unspun urine 4. At least two urine cultures with repeated isolation of the same uropathogen (gram-negative bacteria or Staphylococcus saprophyticus) with equal to or greater than 102 colonies per milliliter in nonvoided specimens 5. Equal to or less than 105 colonies per milliliter of a single uropathogen (gram-negative bacteria or S. saprophyticus) in a patient being treated with an effective antimicrobial agent for a UTI 6. Physician diagnosis of a UTI 7. Physician institutes appropriate therapy for a UTI Excludes asymptomatic bacteriuria and ‘‘other’’ UTIs that are more like deep space infections of the urinary tract. ICD-9 code range, 595.0 to 595.9 or 599.0 *From the ACS NTDB Data Dictionary.12 REFERENCES 1. Emmerson AM, Enstone JE, Griffin M, et al. The Second National Prevalence Survey of infection in hospitalsVoverview of the results. J Hosp Infect. 1996;32:175Y190.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
2. Chenoweth CE, Saint S. Urinary tract infections. Infect Dis Clin North Am. 2011;25:103Y115. 3. Haley RW, Culver DH, White JW, et al. The nationwide nosocomial infection rate. A new need for vital statistics. Am J Epidemiol. 1985;121: 159Y167. 4. Aubron C, Huet O, Ricome S, et al. Changes in urine composition after trauma facilitate bacterial growth. BMC Infect Dis. 2012;12:330. 5. Foxman B. Epidemiology of urinary tract infections: incidence, morbidity, and economic costs. Am J Med. 2002;113(Suppl 1A):5SY13S. 6. Fraser DR, Dombrovskiy VY, Vogel TR. Infectious complications after vehicular trauma in the United States. Surg Infect (Larchmt). 2011;12:291Y296. 7. Lim SC, Doshi V, Castasus B, et al. Factors causing delay in discharge of elderly patients in an acute care hospital. Ann Acad Med Singapore. 2006;35:27Y32. 8. Monaghan SF, Heffernan DS, Thakkar RK, et al. The development of a urinary tract infection is associated with increased mortality in trauma patients. J Trauma. 2011;71:1569Y1574. 9. Scott RD, Douglas R. The direct medical costs of healthcare-associated infections in US hospitals and the benefits of prevention. 2009. Available at: http://www.cdc.gov/hai/pdfs/hai/scott_costpaper.pdf. Atlanta, GA. Accessed February 14, 2014. 10. Hemmila MR, Nathens AB, Shafi S, et al. The Trauma Quality Improvement Program: pilot study and initial demonstration of feasibility. J Trauma. 2010;68:253Y262.
Polites et al.
11. National Center for Health Statistics. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). 2013. Available at: http://www.cdc.gov/nchs/icd/icd9cm.htm. Atlanta, GA. Accessed April 11, 2014. 12. National Trauma Data Bank. National trauma data standard data dictionary 2012 admissions. American College of Surgeons. 2012. Available at: http://www.ntdsdictionary.org/documents/NTDS2012_xsd_Final_102011_ 000.pdf. Salt Lake City, UT. Accessed February 14, 2014. 13. Nester TA, Rumsey DM, Howell CC, et al. Prevention of immunization to D+ red blood cells with red blood cell exchange and intravenous Rh immune globulin. Transfusion. 2004;44:1720Y1723. 14. Bochicchio GV, Joshi M, Shih D, et al. Reclassification of urinary tract infections in critically ill trauma patients: a time-dependent analysis. Surg Infect (Larchmt). 2003;4:379Y385. 15. Laupland KB, Zygun DA, Davies HD, et al. Incidence and risk factors for acquiring nosocomial urinary tract infection in the critically ill. J Crit Care. 2002;17:50Y57. 16. Kaye KS, Marchaim D, Chen TY, et al. Effect of nosocomial bloodstream infections on mortality, length of stay, and hospital costs in older adults. J Am Geriatr Soc. 2014;62:306Y311. 17. Pittet D, Tarara D, Wenzel RP. Nosocomial bloodstream infection in critically ill patients. Excess length of stay, extra costs, and attributable mortality. JAMA. 1994;271:1598Y1601.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
959
Urinary tract infection in elderly trauma patients: Review of the Trauma Quality Improvement Program identifies the population at risk Stephanie F. Polites, MD, Elizabeth B. Habermann, PhD, Kristine M. Thomsen, Mahmoud A. Amr, MD, Donald H. Jenkins, MD, Scott P. Zietlow, MD, and Martin D. Zielinski, MD, Rochester, Minnesota
Elderly trauma patients are at high risk for urinary tract infection (UTI). Despite this, UTI has been deemed a potentially preventable problem and therefore not reimbursable by the Centers for Medicare and Medicaid Services. Early identification of UTI in these patients should lead to prompt treatment, improved outcomes, and cost savings. Risk factors for UTI development in this population must be elucidated to realize these goals. METHODS: The Trauma Quality Improvement Program (TQIP) database was used to analyze elderly patients (Q65 years) admitted as a result of injury during 2011. Patients with genitourinary injuries or undergoing dialysis before admission were excluded. Multivariable logistic regression analysis was conducted to identify UTI risk factors. Mean cost of UTI was calculated based on the assumption of $862 to $1,007 per UTI. RESULTS: In total, 33,257 patients were identified; 1,492 developed UTI (4.5%). Multiple significant risk factors were identified, including age greater than 75 years, female sex, ascites, moderate head injury, impaired sensorium, congestive heart failure, and duration of hospital stay (all p G 0.05). Assuming that UTIs diagnosed on hospital Day 1 were preexisting, the cost of UTI to TQIP hospitals ranged from $1,280,959 to $1,496,434 per year. CONCLUSION: Duration of stay has a profound impact on the development of UTIs in elderly trauma patients, but overall severity of injury does not. In addition, multiple nonmodifiable risk factors were identified, prompting the possibility for increased screening of occult UTIs. Reimbursement for care of UTI in this complicated patient population should be revisited. The TQIP database must improve urinary catheter data. (J Trauma Acute Care Surg. 2014;77: 952Y959. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level III. KEY WORDS: Urinary tract infection; trauma; elderly; injury; Medicare. BACKGROUND:
N
early 6% of hospitalized patients will develop a nosocomial infection, of which urinary tract infections (UTIs) are the most frequent.1Y3 Injured patients are particularly at risk for the development of a UTI because they often require prolonged hospitalization and have altered physiology that increases susceptibility to infection and urinary bacterial overgrowth.4 In addition to the known risk factors of age, female sex, and urinary catheterization in all hospitalized patients, UTI was found to be the most frequent nosocomial infection following spinal cord and head injury in trauma patients specifically.5,6 Since UTIs continue to plague injured patients, identification of additional risk for UTI development is needed for several reasons. Foremost among these reasons are the prolongation of Submitted: February 17, 2014, Revised: April 20, 2014, Accepted: April 21, 2014, Published online: September 22, 2014. From the Division of Trauma, Critical Care, and General Surgery (S.F.P., M.A.A, D.H.J., S.P.Z., M.D.Z.), Mayo Clinic, and The Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery (K.M.T., E.B.H.), Mayo Clinic, Rochester, Minnesota. This study was presented at the 44th Annual Meeting of the Western Trauma Association, March 2Y7, 2014, in Steamboat Springs, Colorado. The TQIP remains the full and exclusive copyrighted property of the ACS. The ACS is not responsible for any claims arising from works based on the original data, text, tables, or figures. Address for reprints: Martin D. Zielinski, MD, Division of Trauma, Critical Care, and General Surgery, Mary Brigh 2-810, St. Mary’s Hospital, Mayo Clinic, 1216 Second St. SW, Rochester, MN 55902; email: [email protected]. DOI: 10.1097/TA.0000000000000351
952
hospitalization and the increased risk of mortality associated with UTI.6Y8 Second, catheter-associated UTI (CAUTI) is considered a ‘‘reasonably preventable’’ and common in-hospital complication by the Centers for Medicare and Medicaid Services (CMS) and is no longer reimbursed. The cost of UTIs, estimated at $867 to $1,100 each, combined with costs associated with any delay in dismissal, results in a significant economic burden to hospitals.9 Since catheter use often cannot be avoided in trauma patients, resulting in lack of reimbursement of care for CAUTIs by CMS, identification of unmodifiable risk factors may provide leverage to modify policies that penalize institutions that care for high-risk injured patients. Due to the clinical implications of poor outcomes resulting from UTIs and the economic implications when they are catheter associated, identification of predictive features specific to this population is needed so that at-risk patients can be targeted for prevention efforts, earlier diagnosis, and ultimately, better outcomes. Therefore, the purpose of this study was to use the collective national experience of quality-seeking trauma centers through the Trauma Quality Improvement Program (TQIP) to determine risk factors of hospital-associated UTI in elderly trauma patients, regardless of catheter status.
PATIENTS AND METHODS We performed a retrospective study of injured patients 65 years or older. This age cutoff was chosen because it J Trauma Acute Care Surg Volume 77, Number 6
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
coincides with the start of Medicare eligibility. Within these patients, we identified those who developed a UTI during their hospitalization. Characteristics of patients with and without a UTI were compared, and independent predictors of UTI were determined.
Data Source Patients were identified from the 2011 TQIP Participant Use File (PUF). Built on the infrastructure of the American College of Surgeons’ (ACS) Committee on Trauma National Trauma Data Bank (NTDB), TQIP was designed to allow ACSdesignated Level I and II trauma centers to recognize and stratify their outcomes.10 Patient- and institution-level variables allow for risk adjustment and subsequent benchmarking with the goal of improving quality of care at participating trauma centers. In 2011, there were 160 centers participating in TQIP. Data are entered at each trauma center by trained, dedicated nurse abstractors and is validated externally on a quarterly basis. Standard definitions of comorbidities and in-hospital complications are used and can be determined through software registry programs, chart abstraction by registrar, and derivation from International Classification of DiseasesV9th Rev.VClinical Modification, (ICD-9-CM) codes.11 The definition of UTI in TQIP uses the NTDB definition and is based on a combination of clinical and microbiological criteria (Appendix 1).12 TQIP does not distinguish between CAUTI and UTI not associated with catheter use; therefore, this study examined risk factors for any UTI rather than CAUTI specifically. The ICD-9-CM procedure code for insertion of a urinary catheter (57.94) was used as a surrogate for catheter use. This procedure code includes urinary catheters inserted in any setting during the current hospitalization; thus, it would include catheters placed in the emergency department and operating room but not catheters placed en route or before transfer. Furthermore, because this code is only reflective of urinary catheter insertion, duration of use and sterility of placement could not be obtained. Rather than perform a separate analysis for CAUTI using this limited variable, this study analyzed all UTIs with insertion of a urinary catheter as a dependent variable.
Polites et al.
chemotherapy for cancer within 30 days, congenital anomalies, congestive heart failure (CHF), current smoker, cerebrovascular accident, diabetes mellitus (DM), disseminated cancer, functionally dependent health status, history of angina within 1 month, history of myocardial infarction within 6 months, history of revascularization or amputation for peripheral vascular disease, hypertension requiring medication, impaired sensorium, obesity, respiratory disease, steroid use, and cirrhosis. Length of stay (LOS) was categorized by 1, 2, 3 to 5, 6 to 8, 9 to 14, 15 to 21, and 22 or more days. Duration of stay in the intensive care unit and days of mechanical ventilation were also determined.
Statistical Analysis The cohort was described in aggregate using percentages, means (SDs), and medians (interquartile ranges [IQRs]). UTI rate by injury type was determined using ICD-9-CM diagnosis codes based on the methodology described by Fraser et al.6,11 Associations of demographic, clinical, and institutional factors with UTI were evaluated using W2 tests, and Fisher’s exact tests were used when low expected cell counts were observed. Associations of continuous factors with UTI were
Inclusion/Exclusion Criteria All patients 65 years or older in the 2011 TQIP PUF were identified for inclusion. Patients who required dialysis at the time of injury and those who underwent an operative procedure to repair a kidney, ureter, or bladder injury (ICD-9-CM codes 55.81Y55.99) were excluded. In addition, patients with missing data entries for age or UTI were excluded.
Variables Data collected included patient demographic and clinical information as well as institutional information. Institutional variables included ACS verification level (I vs. II), institutional tax status (for profit vs. nonprofit), teaching status (community, nonteaching, academic), and quartiles of hospital bed size. Patients were separated into age groups based on 5-year increments. Abbreviated Injury Scale (AIS) Yderived Injury Severity Score (ISS) and Glasgow Coma Scale (GCS) score were obtained directly from TQIP. Preexisting comorbidities included alcoholism, ascites within 30 days, bleeding disorder,
Figure 1. Identification of a cohort of elderly trauma patients from the 2011 TQIP PUF. Patients with missing age or UTI data and patients who required dialysis or underwent an operative procedure to repair a kidney, ureter, or bladder injury were excluded.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
953
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
TABLE 1. Rates of UTI Associated With Patient and Trauma Center Demographic Factors Patient demographics Age, y
Sex
Interhospital transfer
AIS-derived ISS
GCS score
Trauma center demographics Teaching status
ACS verification level
No. adult beds
All Patients (N = 33,257)
No UTI (n = 31,765)
UTI (n = 1,492)
p
65Y69 70Y74 75Y79 80Y84 85Y89* Not available Male Female Not available No Yes Not available e9 10Y16 17Y25 Q26 Not available 3Y8 9Y12 13Y15
6,212 5,744 6,513 7,814 6,974 6 15,068 18,183 11 21,009 12,237 630 14,710 9,148 6,313 2,456 1,374 2,278 961 28,644
5,989 (96.4) 5,511 (95.9) 6,196 (95.1) 7,446 (95.3) 6,623 (95.0) 6 14,523 (96.4) 17,236 (94.8) 11 20,083 (95.6) 11,671 (95.4) 607 14,109 (95.9) 8,776 (95.9) 5,992 (94.9) 2,281 (92.9) 1,304 2,179 (95.7) 858 (89.3) 27,424 (95.7)
223 (3.6) 233 (4.1) 317 (4.9) 368 (4.7) 351 (5.0) V 545 (3.6) 947 (5.2) V 926 (4.4) 566 (4.6) 23 601 (4.1) 372 (4.1) 321 (5.1) 175 (7.1) 70 99 (4.3) 103 (10.7) 1,220 (4.3)
G0.001
University Community Nonteaching Not available** I II Not available G350 350Y499 500Y699 Q700
18,107 12,774 2,376 7,650 17,466 8,141 62 7,419 7,753 8,808 9,215
17,158 (94.8) 12,315 (96.4) 2,292 (96.5) 7,255 16,620 (95.2) 7,890 (96.9) 55 7,192 (96.9) 7,430 (95.8) 8,365 (95.0) 8,723 (94.7)
949 (5.2) 459 (3.6) 84 (3.5) 395 846 (4.8) 251 (3.1) 7 227 (3.1) 323 (4.2) 443 (5.0) 492 (5.3)
G0.001
0.36
G0.001
G0.001
G0.001
G0.001
G0.001
*No patients 90 years or older were identified. **ACS verification level was not available for patients treated at trauma centers with state designation only. Values are presented as n (%). Percentages reported as row percents.
evaluated using t tests; if the factors were not sufficiently normally distributed, Wilcoxon rank-sum tests were used. Multivariable logistic regression was performed to identify independent predictors of UTI inclusive of urinary catheter insertion. Results of multivariable analysis are expressed as odds ratios (ORs) with 95% confidence intervals (CIs). Statistical significance was acknowledged when p G 0.05. All data analysis was completed using SAS version 9.3 (SAS Institute, Inc., Cary, NC).
RESULTS Of the 126,130 total patients in the 2011 TQIP PUF, 33,257 injured patients 65 years or older were included in this study (Fig. 1). The mean (SD) age was 77.5 (7.2) years, and the majority (54.7%) were female. No patients 90 years or older were identified. In-hospital mortality of the cohort was 7.5%. 954
TABLE 2. Rates of UTI by Injury Type Injury Type (ICD-9 Code)H
Rate of UTI, %
Colon (863.4Y5) Spinal cord (806, 952) Kidney, ureter, bladder, urethra (866, 867.0Y3) Thorax (lungs, bronchi, esophagus) (861.2, 861.3, 862) Small bowel (863.2Y3) Liver (864) Pelvic fractures (808) Spleen (865) Long bone fractures (812, 813, 820, 821, 823) Multiple rib fractures (807.0(2Y9), 807.1(2Y9)) Intracranial (800.1Y4, 800.6Y9, 801.1Y4, 801.6Y9, 851Y954)
11.1 8.1 7.2 7.0 6.8 6.6 6.4 5.7 5.1 4.9 4.5
*Based on methodology described by Fraser et al. for determining rates of infectious complications by injury type.6,11 Injury type categories include all subcodes of indicated ICD-9 codes when available.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
TABLE 3. Rates of UTI Associated With Patient ComorbiditiesH and LOS Hypertension on medication DM Bleeding disorder Impaired sensorium Respiratory disease CHF Cerebrovascular accident/residual neurologic deficit Current smoker Alcoholism Functionally dependent Myocardial infarction within 6 mo Obesity Disseminated cancer Angina within 1 mo Steroid use Chemotherapy within 30 d History of revascularization/amputation for peripheral vascular disease Cirrhosis Congenital anomalies Ascites within 30 d LOS,** d 1 2 3Y5 6Y8 9Y14 15Y21 Q22 Intensive care days 0 1Y3 4Y7 8Y14 Q15 Ventilator days 0 1Y2 3Y4 5Y9 10Y14 Q15
All Patients (N = 33,257)
No UTI (n = 31,765)
UTI (n = 1,492)
p
20,379 (61.9) 7,793 (23.7) 4,742 (14.4) 4,609 (14.0) 4,067 (12.4) 3,077 (9.3) 2,066 (6.3)
19,374 (61.6) 7,393 (23.5) 4,502 (14.3) 4,356 (13.9) 3,844 (12.2) 2,879 (9.2) 1,952 (6.2)
1,005 (67.9) 400 (27.0) 240 (16.2) 253 (17.1) 223 (15.1) 198 (13.4) 114 (7.7)
G0.001 0.002 0.044 G0.001 0.001 G0.001 0.021
1,967 (6.0) 1,334 (4.1) 1,308 (4.0) 1,138 (3.5) 1,044 (3.2) 659 (2.0) 617 (1.9) 335 (1.0) 171 (0.5) 165 (0.5)
1,879 (6.0) 1,264 (4.0) 1,244 (4.0) 1,085 (3.5) 979 (3.1) 623 (2.0) 579 (1.8) 316 (1.0) 161 (0.5) 152 (0.5)
138 (0.4) 98 (0.3) 26 (0.1) 2,171 (6.5) 2,976 (9.0) 12,259 (36.9) 7,632 (23.0) 4,910 (14.8) 1,800 (5.4) 1,501 (4.5) 18,037 (54.2) 9,012 (27.1) 3,243 (9.8) 1,687 (5.1) 1,278 (3.8) 28,236 (84.9) 1,993 (6.0) 779 (2.3) 944 (2.8) 556 (1.7) 749 (2.3)
128 (0.4) 94 (0.3) 22 (0.1) 2,167 (6.8) 2,956 (9.3) 12,045 (37.9) 7,318 (23.0) 4,517 (14.2) 1,568 (4.9) 1,186 (3.7) 17,439 (54.9) 8,764 (27.6) 3,017 (9.5) 1,477 (4.6) 1,068 (3.4) 27,194 (85.6) 1,922 (6.1) 711 (2.2) 844 (2.7) 480 (1.5) 614 (1.9)
88 (5.9) 70 (4.7) 64 (4.3) 53 (3.6) 65 (4.4) 36 (2.4) 38 (2.6) 19 (1.3) 10 (0.7) 13 (0.9) 10 (0.7) 4 (0.3) 4 (0.3) 4 (0.3) 20 (1.3) 214 (14.3) 314 (21.0) 393 (26.3) 232 (15.5) 315 (21.1) 598 (40.1) 248 (16.6) 226 (15.1) 210 (14.1) 210 (14.1) 1,042 (69.8) 71 (4.8) 68 (4.6) 100 (6.7) 76 (5.1) 135 (9.0)
0.95 0.18 0.48 0.79 0.006 0.23 0.045 0.30 0.39 0.036 0.12 90.99 0.028 G0.001
G0.001
G0.001
*Comorbidity data were missing for 345 patients. **LOS was missing for 8 patients. Values are presented as n (%). Percentages are reported as column percents.
Univariate Analysis of Patients With and Without UTI UTI was diagnosed in 1,492 patients (4.5%). The rate of UTI increased as patient age increased ( p G 0.001) and UTIs were more likely to occur in females than in males (5.2% vs. 3.6%, p G 0.001) (Table 1). The rate of UTI increased with greater ISS ( p G 0.001); however, UTI was more frequent in patients with a GCS score of 9 to 12 than those with 13 to 15 ( p G 0.001). There was no significant difference in UTI between patients treated at nonprofit and for-profit facilities (4.5% vs. 4.6%, p = 0.83), but UTIs were more frequent at university teaching facilities
( p G 0.001) and facilities with ACS Level I verification ( p G 0.001). The rate of UTI was highest in patients with colon injuries (11.1%) followed by spinal cord injuries (8.1%). The rate of UTI was lowest in patients with intracranial injuries (4.5%) (Table 2). The most common comorbidity was hypertension requiring medication, which was present in 61.9% of the patients, followed by DM (23.7%). Comorbidity information was missing in 345 patients (1.0%). Several comorbidities were associated with an increased frequency of UTI (Table 3). Insertion of an indwelling urinary catheter was documented in 11.8% of the patients (n = 3,913), and the frequency of UTI
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
955
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
(Table 4). Patients 75 years or older were at increased odds of UTI when compared with patients 69 years or younger. Females were at increased risk when compared with males ( p G 0.001; OR, 1.86; 95% CI, 1.65Y2.10). Although significant on univariable analysis, ISS was not predictive of UTI on multivariable analysis ( p 9 0.05). A moderate head injury resulting in GCS score of 9 to 12 was a risk factor when compared with GCS score of 13 to 15 ( p G 0.001; OR, 1.79; 95% CI, 1.41Y2.26) though this was not true for GCS score of 3 to 8 ( p 9 0.05). Institution-level risk factors were university teaching status and larger number of beds. ACS verification level was not included in the multivariable analysis because it was closely related to other institutional variables. Ascites within 30 days, CHF, and impaired sensorium were significant risk factors for UTI (all p G 0.05); however, many other comorbidities that were significant on univariable analysis were not significant after adjustment. Using a LOS of 3 days to 5 days as a reference point, decreased LOS was found to be protective against UTI, while increased LOS was predictive of UTI. Due to the close relationship of intensive care days and ventilator days with hospital LOS, these variables could not be included in the model. Patients with a documented indwelling urinary catheter were at an increased risk for UTI after adjusting for other variables (p = 0.001; OR, 1.27; 95% CI, 1.10Y1.47).
DISCUSSION
Figure 2. Duration of hospitalization, intensive care, and mechanical ventilation as well as rate of UTI. The rate of UTI increased exponentially with increased LOS and linearly with more intensive care and ventilator days.
was greater in these patients (7.1 vs. 4.1%, p G 0.001). Patients who required intensive care were more likely to develop UTI (5.9% vs. 3.3%, p G 0.001) as were patients who required mechanical ventilation (9.0% vs. 3.7%, p G 0.001). The rate of UTI increased linearly with the duration of intensive care and mechanical ventilation and exponentially with LOS (Fig. 2). The median (IQR) LOS was significantly longer for patients with a UTI (11 [7Y19] days vs. 5 [3Y8] days, p G 0.001).
Multivariable Analysis Multivariable analysis indicated that there are several patient-level risk factors for UTI in elderly trauma patients 956
UTIs are a significant burden of morbidity to trauma patients, who, because of their age, are susceptible to altered urine microbiology, transient immunosuppression, and frequent requirement of urinary catheters; however, predictive features specific to this population for UTI, outside of urinary catheterization, are unknown.4,13 These risk factors are needed to guide health care providers in prevention, early diagnosis, and effective treatment efforts. This study identified a 4.5% rate of UTI in trauma patients, age or 65 years or greater, treated at Level I and II trauma centers participating in the TQIP. Several patient demographic and clinical factors emerged as independent risk factors for UTI including age 75 years or greater, female sex, ascites, CHF, and impaired sensorium. Age and sex are consistent with existing trauma and critical care literature;14,15 however, this is the first description of specific comorbidity risk factors for UTI in elderly trauma patients. Although patients with DM have an increased incidence of bacteriuria, we did not find that diabetes was associated with UTI in this study.5 Moderate head injury and need for intensive care or mechanical ventilation were associated with UTI and should be used to risk stratify, rather than ISS. We suspect that patients with GCS score of 3 to 8 did not have an increased risk of UTI because of the high early mortality associated with severe head injuries. Similarly, this may have resulted in the relatively low rate of UTI among patients with intracranial injuries. The relatively high rate of UTI found in patients with spinal cord injuries is consistent with the limited existing literature on UTIs in injured patients.6 The risk of UTI was amplified as LOS, duration of intensive care, and duration of mechanical ventilation increased. * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
TABLE 4. Independent Risk Factors for UTI in Elderly Trauma Patients Using Logistic RegressionH Risk Factor Age, y (vs. 65Y69)
70Y74 75Y79 80Y84 85Y89
Female sex Teaching status (vs. community) No. adult beds (vs. G350 beds)
AIS derived ISS (vs. e9)
GCS score (vs. 13Y15)
Nonteaching University 350Y499 500Y699 Q700 10Y16 17Y25 Q26 3Y8 9Y12
Ascites within 30 d CHF Bleeding disorder Cerebrovascular accident/residual neurologic deficit DM History of angina within past 1 mo History of revascularization/amputation for peripheral vascular disease Impaired sensorium Obesity LOS, d (vs. 3Y5) 1 2 6Y8 9Y14 15Y21 Q22 Insertion of indwelling urinary catheter
p
OR
0.231 0.001 G0.001 G0.001 G0.001 0.59 G0.001 0.030 G0.001 G0.001 0.78 0.61 0.11 0.33 G0.001 0.046 0.028 0.41 0.32 0.08 0.09 0.16
1.13 1.40 1.41 1.57 1.86 1.08 1.33 1.23 1.47 1.43 0.98 0.96 1.18 0.89 1.79 3.36 1.21 1.07 1.11 1.13 1.35 1.56
0.92 1.16 1.17 1.30 1.65 0.82 1.16 1.02 1.22 1.20 0.85 0.82 0.96 0.70 1.41 1.02 1.02 0.91 0.90 0.99 0.95 0.84
1.38 1.69 1.70 1.91 2.10 1.41 1.52 1.49 1.76 1.72 1.13 1.12 1.44 1.12 2.26 11.06 1.44 1.25 1.38 1.28 1.93 2.90
0.004 0.76 G0.001 0.001 G0.001 G0.001 G0.001 G0.001 0.001
1.25 0.96 0.13 0.43 2.32 5.03 9.27 17.67 1.27
1.07 0.72 0.05 0.27 1.92 4.20 7.52 14.38 1.10
1.46 1.27 0.36 0.69 2.79 6.03 11.42 21.70 1.47
95% CI
*Missing data as outlined in Tables 1 and 2 were not included in the multivariable analysis.
Cause and effect relationships, however, cannot be inferred from our current study. Laupland et al.15 previously found that duration of intensive care unit stay predicted UTI. Others have concluded that nosocomial infections actually result in increased LOS and excess cost.7,16,17 Granular, prospective data including complications of UTI and timing of diagnosis and treatment are needed to elucidate this causative relationship in trauma patients. Urinary catheter use relied on patients’ having the appropriate ICD-9-CM diagnosis code, and we suspect the actual incidence was much greater. Furthermore, TQIP does not capture information regarding duration of urinary catheterization, the leading risk factor for UTI development and a requirement for CAUTI. Nonetheless, patients who experience UTIs, regardless of catheter use, continue to be at risk of poor outcomes. Catheter use is not the only risk factor in trauma patients, as demonstrated in the current analysis. Since not all catheterized patients develop UTI, there is room for further risk modification using these results. Furthermore, even if UTIs cannot be prevented, identification of those at highest risk may facilitate early diagnosis and treatment and, ultimately, avoidance of some of the possible downstream
negative effects including increased LOS, urinary sepsis, mortality, and lack of reimbursement to providers. Unfortunately, in the acute setting of trauma, most risk factors, particularly comorbidities, are nonmodifiable. The results of this study, however, provide several opportunities for prevention and early treatment efforts. Urinary catheters are often required in trauma patients because of the need for careful hemodynamic and fluid monitoring, surgical interventions, or mobility limitations. Nevertheless, it is paramount that all hospitals caring for injured patients develop processes that ensure removal of catheters at the earliest appropriate time. In addition, we recommend that providers have a high index of suspicion for UTI in elderly trauma patients who remain hospitalized for more than 2 days because of the increased risk associated with hospital stay and duration of intensive care and/or mechanical ventilation. It may be difficult for injured elderly patients to voice concerns about symptoms, resulting in delayed diagnosis and treatment. Although it is our institutional practice to use screening urinalysis in such patients, efficacy data are lacking. Prospective studies on effective screening practices for UTIs in elderly, injured patients are needed.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
957
J Trauma Acute Care Surg Volume 77, Number 6
Polites et al.
Patients who received care at large or university teaching hospitals were at increased risk of UTI, suggesting that, despite inclusion of GCS and ISS in our analysis, the complexity of care at these trauma centers was not accounted for in our modeling or that there are processes inherent to these centers that result in more UTIs. Larger and higher-acuity centers may receive patients who are more likely to require intensive care, mechanical ventilation, or surgical care, which typically involves insertion of a urinary catheter. We also hypothesize that urinary catheters may not be as aggressively removed at university centers because of the presence of trainees. Providers at institutions with high-risk characteristics should be aware of these findings and initiate quality improvement efforts to minimize this risk. TQIP provides robust data on a large number of trauma patients from quality-seeking trauma centers; however, there are several more limitations of this study that must be acknowledged. Only ACS-verified Level I and II trauma centers can participate in TQIP with specific inclusion and exclusion parameters, resulting in a selection bias toward more severely injured patients. Although TQIP centers maintain their data prospectively, we used the TQIP data in a retrospective fashion, which introduces further bias. It is possible that we were not able to capture all potential patient- and institutional-level influences on the risk of UTI. It is also acknowledged that some UTIs may have been present on admission and, therefore, were not hospital acquired. Lastly, the rate of urinary catheter insertion was low and was likely underestimated in this study because of the reliance on ICD-9 coding. More specific data about urinary catheter placement and duration are required and should be implemented by the ACS Committee on Trauma for future TQIP data collection. Robust catheter data would allow for improved investigation of CAUTI specifically, the infections which result in the loss of reimbursement to TQIP hospitals. Since specific injuries are undoubtedly associated with varying degrees of catheter use, further investigation of injury types that are most susceptible to UTIs could not be conducted because of poor catheter data in the TQIP. Since hospital-acquired UTIs increase mortality and are not reimbursable by CMS if associated with a urinary catheter, it is necessary that trauma centers take action to reduce the occurrence of this nosocomial infection and to reduce the morbidity of patients who develop nonpreventable UTIs. While it is unlikely that UTIs will be eliminated entirely in elderly trauma patients, the results of this study will aid in risk stratification and earlier diagnosis. In addition to prompt urinary catheter removal, development of screening processes for UTIs in injured patients with prolonged duration of hospitalization and critical care may be of use. Lastly, results of this study should provide leverage to modify policies that penalize institutions that care for high-risk injured patients. AUTHORSHIP S.F.P. contributed to the literature search, study design, data interpretation, and manuscript writing. E.B.H. contributed to the study design, data analysis, and data interpretation. K.M.T. contributed to the data collection, data analysis, and critical revision of the manuscript. M.A.A. contributed to the study design and critical revision of the manuscript. D.H.J. contributed to the study design and critical revision of the manuscript. S.P.Z. contributed to critical review of the manuscript. M.D.Z.
958
contributed to the literature search, study design, data interpretation, and critical revision of the manuscript.
DISCLOSURE This publication was made possible by CTSA grant number UL1 TR000135 from the National Center for Advancing Translational Sciences (NCATS), a component of the National Institutes of Health (NIH).
APPENDIX 1. TQIP DEFINITION OF UTI* UTI: Defined as an infection anywhere along the urinary tract with clinical evidence of infection, which includes at least one of the following symptoms with no other recognized cause: 1. Fever of 38-C or greater 2. White blood cell (WBC) count greater than 100,000/HL or less than 3,000/HL 3. Urgency 4. Frequency 5. Dysuria 6. Suprapubic tenderness AND positive urine culture (Q100,000 microorganisms per microliter of urine with no more than two species of microorganisms) OR at least two of the following signs or symptoms with no other recognized cause: 1. Fever of 38-C or greater 2. WBC count greater than 100,000/HL or less than 3,000/HL 3. Urgency 4. Frequency 5. Dysuria 6. Suprapubic tenderness AND at least one of the following: 1. Positive dipstick for leukocyte esterase and/or nitrate 2. Pyuria (urine specimen with 910 WBC/HL or 93 WBC/ high-power field of unspun urine) 3. Organisms seen on Gram stain of unspun urine 4. At least two urine cultures with repeated isolation of the same uropathogen (gram-negative bacteria or Staphylococcus saprophyticus) with equal to or greater than 102 colonies per milliliter in nonvoided specimens 5. Equal to or less than 105 colonies per milliliter of a single uropathogen (gram-negative bacteria or S. saprophyticus) in a patient being treated with an effective antimicrobial agent for a UTI 6. Physician diagnosis of a UTI 7. Physician institutes appropriate therapy for a UTI Excludes asymptomatic bacteriuria and ‘‘other’’ UTIs that are more like deep space infections of the urinary tract. ICD-9 code range, 595.0 to 595.9 or 599.0 *From the ACS NTDB Data Dictionary.12 REFERENCES 1. Emmerson AM, Enstone JE, Griffin M, et al. The Second National Prevalence Survey of infection in hospitalsVoverview of the results. J Hosp Infect. 1996;32:175Y190.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
2. Chenoweth CE, Saint S. Urinary tract infections. Infect Dis Clin North Am. 2011;25:103Y115. 3. Haley RW, Culver DH, White JW, et al. The nationwide nosocomial infection rate. A new need for vital statistics. Am J Epidemiol. 1985;121: 159Y167. 4. Aubron C, Huet O, Ricome S, et al. Changes in urine composition after trauma facilitate bacterial growth. BMC Infect Dis. 2012;12:330. 5. Foxman B. Epidemiology of urinary tract infections: incidence, morbidity, and economic costs. Am J Med. 2002;113(Suppl 1A):5SY13S. 6. Fraser DR, Dombrovskiy VY, Vogel TR. Infectious complications after vehicular trauma in the United States. Surg Infect (Larchmt). 2011;12:291Y296. 7. Lim SC, Doshi V, Castasus B, et al. Factors causing delay in discharge of elderly patients in an acute care hospital. Ann Acad Med Singapore. 2006;35:27Y32. 8. Monaghan SF, Heffernan DS, Thakkar RK, et al. The development of a urinary tract infection is associated with increased mortality in trauma patients. J Trauma. 2011;71:1569Y1574. 9. Scott RD, Douglas R. The direct medical costs of healthcare-associated infections in US hospitals and the benefits of prevention. 2009. Available at: http://www.cdc.gov/hai/pdfs/hai/scott_costpaper.pdf. Atlanta, GA. Accessed February 14, 2014. 10. Hemmila MR, Nathens AB, Shafi S, et al. The Trauma Quality Improvement Program: pilot study and initial demonstration of feasibility. J Trauma. 2010;68:253Y262.
Polites et al.
11. National Center for Health Statistics. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). 2013. Available at: http://www.cdc.gov/nchs/icd/icd9cm.htm. Atlanta, GA. Accessed April 11, 2014. 12. National Trauma Data Bank. National trauma data standard data dictionary 2012 admissions. American College of Surgeons. 2012. Available at: http://www.ntdsdictionary.org/documents/NTDS2012_xsd_Final_102011_ 000.pdf. Salt Lake City, UT. Accessed February 14, 2014. 13. Nester TA, Rumsey DM, Howell CC, et al. Prevention of immunization to D+ red blood cells with red blood cell exchange and intravenous Rh immune globulin. Transfusion. 2004;44:1720Y1723. 14. Bochicchio GV, Joshi M, Shih D, et al. Reclassification of urinary tract infections in critically ill trauma patients: a time-dependent analysis. Surg Infect (Larchmt). 2003;4:379Y385. 15. Laupland KB, Zygun DA, Davies HD, et al. Incidence and risk factors for acquiring nosocomial urinary tract infection in the critically ill. J Crit Care. 2002;17:50Y57. 16. Kaye KS, Marchaim D, Chen TY, et al. Effect of nosocomial bloodstream infections on mortality, length of stay, and hospital costs in older adults. J Am Geriatr Soc. 2014;62:306Y311. 17. Pittet D, Tarara D, Wenzel RP. Nosocomial bloodstream infection in critically ill patients. Excess length of stay, extra costs, and attributable mortality. JAMA. 1994;271:1598Y1601.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
959
