Impact of Device-Associated Infection on Trauma Patient Outcomes at a Major Trauma Center

Abstract

Background:

Catheter-associated urinary tract infection (CAUTI) and ventilator-associated pneumonia (VAP) are considered performance measures. We analyzed the incidence, prevalence, and risk of CAUTI and VAP in trauma patients, as well as the demographic and injury factors related to these infections and their relative risks of negative outcomes (prolonged length of stay [LOS], sepsis, and death).

Methods:

Trauma registry data were analyzed (age >18 y; LOS >24 h) from January 1, 2007, to December 31, 2011. Demographics and injury location, severity, and type were analyzed relative to outcomes along with device-associated infection, as defined by the U.S. Centers for Disease Control and Prevention. The outcomes analyzed were intensive care unit (ICU) and hospital LOS, sepsis, and in-hospital death. Multivariable logistic regression was then used to identify the factors contributing to sepsis, including device-associated infections.

Results:

The included population (n=10,755) was 66.6% male and had a mean age of 45.1 y, with blunt trauma in 91.8% and a median Injury Severity Score (ISS) of 10 points. Patients developing CAUTI (n=324; 3.0%; p<0.005) were more likely to be female (59.4%), had a higher median ISS (20.5), and were older (56.7 years). Patients with VAP (n=161; 1.5%; p<0.005) had a higher median ISS (27). Patients with sepsis (n=149; 1.4%; p<0.005) had a higher median ISS (24.0) and were older (52.3 y). Sepsis was associated with prolonged LOS and death, as expected (p<0.005). In multivariable analysis, independent predictors of sepsis were CAUTI (odds ratio [OR] 16.15; p<0.001), VAP (OR 6.95; p<0.001), ISS (OR 1.05 per unit; p<0.001), age (OR 1.02 per year; p<0.001), and penetrating, abdominal, pelvic, or chest injury.

Conclusion:

Development of CAUTI and VAP are significantly associated with a higher risk of sepsis in trauma patients after adjustment for age and injury type, location, and severity. This study suggests the importance of device-associated infections as vectors for sepsis in trauma and highlights the importance of prevention initiatives.

Sepsis accounts for a substantial proportion of the deaths in patients who survive injury initially [1]. Nosocomial infection and sepsis increase healthcare cost [2,3]. Catheter-associated urinary tract infection (CAUTI) [4] and ventilator-associated pneumonia (VAP) [5] are known causes of sepsis. Trauma victims inherently suffer from limited mobility (e.g., secondary to specific injuries such as severe pelvic fractures, spinal injury, and traumatic brain injury) that may lead to prolonged catheterization and a higher risk of urinary tract infection (UTI) and VAP [4,6]. Thoracic injury may independently increase the rate of endotracheal intubation, thus raising the risk of VAP. Some VAP and CAUTI cases result in sepsis. Sepsis associated with CAUTI carries a higher mortality rate than community-acquired bacteremia/sepsis [7].

The risks and prevalence of CAUTI, VAP, and associated sepsis in trauma patients are not well known, nor is it known how risk factors for CAUTI, VAP, and sepsis combine to change the odds of a negative outcome (prolonged length of stay [LOS], sepsis, and death). This project analyzed the incidence, prevalence, and risk of CAUTI and VAP in trauma patients and the relative risk of negative outcomes to identify the factors contributing to the incidence of CAUTI, VAP, and sepsis and to determine whether device-associated infections are an independent risk factor for sepsis. We hypothesized that CAUTI and VAP have identifiable risk factors and increase the risk of prolonged LOS, sepsis, and death. We did not have sufficient cases (n<5) of central line-associated blood stream infection (CLABSI) to make a statistical analysis.

Patients and Methods

Inclusion criteria were admission to the University of Kentucky (LOS >24 h) between January 1, 2007 and December 31, 2011 with a diagnosis of blunt or penetrating trauma and age 18 y or greater. The U.S. Centers for Disease Control and Prevention (CDC) limits the definition of CAUTI to a UTI (diagnosed clinically or by culture) that occurs at least two d after the placement of an indwelling catheter or within 1 d after discontinuation of a catheter [8]. Similarly, pneumonia (diagnosed clinically or by culture) must occur at least 2 d after initiation of mechanical ventilation or within one d after discontinuation of mechanical ventilation [9]. The diagnoses of CAUTI and VAP must have been made during the hospital stay. The exclusion criteria were burns or age less than 18 y. At discharge, trauma nursing staff entered all patient data in the University of Kentucky Trauma Registry and determined whether infections noted in the charts met the CDC criteria for VAP or CAUTI [8,9]. Data were collected retrospectively from this registry. The Institutional Review Board approved the study protocol and found it exempt from full review on the basis of the low risk to the research subjects of using data from the Trauma Registry.

The patients were sorted into the following groups, which are not mutually exclusive: CAUTI (sepsis/no sepsis), VAP (sepsis/no sepsis), and no device-associated infection (sepsis/no sepsis). The patients' medical records were evaluated for the following data points: Demographics (age, gender), injury type (blunt or penetrating) Injury Severity Score (ISS), Abbreviated Injury Score (AIS), duration of device use, intensive care unit (ICU) and hospital LOS, and disposition (discharge to home or a rehabilitation center, or death). Each patient had a unique identifier once data were exported from the Research Electronic Data Capture (REDCap). This number, along with the study data, was collected and managed using REDCap tools hosted at the University of Kentucky. REDCap is a secure, Web-based application designed to support data capture for research studies, providing validated data entry, audit trails, data downloads, and data import from external sources [10].

The outcomes analyzed were ICU and hospital LOS, sepsis, and in-hospital death. We defined prolonged LOS as a stay in the top quartile (more than 8 d). We set the significance threshold at p<0.005 to allow multiple comparisons. Multivariable logistic regression was used to identify factors contributing to sepsis, including device-associated infections.

Results

The included population, defined in Table 1 (n=10,755), was 66.6% male, had a mean age of 45.1 y, was mostly blunt-injured (91.8%), and had a median ISS of 10 points. Overall, 3% had CAUTI, 1.5% had VAP, 1.4% had sepsis, 4.1% died, and 24.6% had an LOS >8 d (Table 2). The patients who developed CAUTI (n=324; 3.0%; p<0.005) were more likely to be female (59.4%) and older (56.7 y) and to have a higher overall median ISS (20.5 points). These patients had a higher (p<0.001) prevalence of VAP (13.9%), LOS >8 d (83.0%), sepsis (16%), and death (8.6%). The patients who developed a VAP (n=161; 1.5%; p<0.005) had a higher overall median ISS (27 points). These patients had a higher (p<0.001) prevalence of CAUTI (28.0%), LOS >8 d (100.0%), sepsis (14.9%), and death (5.6%).

Table 1.

Patient and Trauma Characteristics in Outcome Groups

		Adverse Outcome Subgroups
	All Pts. (n=10,755)	CAUTI (n=324)	VAP+(n=161)	Sepsis (n=149)	Death (n=446)	LOS >8 d (n=2,644)
% Male	66.6	40.6^d	73.8	71.6	69.1	64.9
% Blunt trauma	91.8	95.7	96.9	88.6	89.2	93.7
Median ISS, points	10.0	20.5^d	27.0^d	24.0^d	25.0^d	17.0^d
Mean age, y	45.1	56.7^d	47.9	52.3^d	57.6^d	49.4^d
H, N, & F inj (%)
Moderate^a	22.9	25.3^d	18.0^d	22.8^d	11.0^d	24.8^d
Serious^b	11.6	18.5	24.2	18.1	13.2	14.9
Severe+^c	10.7	24.7	37.9	22.8	51.1	18.5
Chest inj (%)
Moderate	12.6	13.6^d	15.5^d	12.8^d	10.5^d	13.7^d
Serious	18.3	23.8	32.9	28.9	20.9	26.2
Severe+	6.0	17.6	30.4	25.5	17.7	12.5
Abdominal inj (%)
Moderate	17.6	27.2^d	21.1^d	20.8^d	20.2^d	23.4^d
Serious	6.4	8.6	14.9	16.1	7.0	9.6
Severe+	2.6	5.6	9.3	13.4	8.7	5.4
Leg/pelvis inj (%)
Moderate	18.0	17.6^d	16.1^d	12.1^d	9.6^d	19.3^d
Serious	19.8	29.0	18.6	30.2	16.1	30.6
Severe+	0.9	4.0	5.6	3.4	2.7	2.5

Moderate=abbreviated injury score (AIS) 2.

Serious=AIS 3.

Severe plus=AIS ≥4.

p<0.005 for differences in groups with and without the outcome.

H, N, & F=head, neck, and face.

Table 2.

Outcome Rates and Relations Between Outcomes

Variable (% of pts.)	All Trauma Pts. (n=10,755)	CAUTI (n=324)	VAP (n=161)	Sepsis (n=149)	Death (n=446)	LOS >8 d (n=2,644)
CAUTI	3.0		28.0^a	34.9^a	6.3^a	10.2^a
VAP	1.5	13.9^a		16.1^a	2.0	6.1^a
Sepsis	1.4	16.0^a	14.9^a		7.6^a	5.4^a
Death	4.1	8.6^a	5.6^a	22.8^a		4.0
LOS >8 d	24.6	83.0^a	100^a	95.3^a	23.6

p<0.001 (Fisher exact test).

Patients who developed sepsis, regardless of etiology (n=149; 1.4%; p<0.005), had a higher median ISS (24.0 points) and were older (52.3 y). These factors were associated with a higher (p<0.001) prevalence of CAUTI (34.9%), VAP (16.1%), LOS >8 d (95.3%), and death (22.9%). Sepsis was associated with prolonged LOS and death, as expected (p<0.005). Patients with an LOS >8 d had a significantly higher (p<0.001) prevalence of CAUTI (10.2%), VAP (6.1%), and death (5.4%). Patients who died had a higher (p<0.001) prevalence of CAUTI (6.3%) and sepsis (7.6%).

By multivariable analysis, shown in Table 3 and Fig. 1, independent predictors of sepsis were CAUTI (odds ratios [OR] 16.15; p<0.001), VAP (OR 6.95; p<0.001), severe (AIS=3) and serious (AIS=4) abdominal injury (OR 2.60; p<0.001 and OR 3.00; p<0.001, respectively), penetrating injury (OR 2.54; p<0.005), severe chest injury (OR 2.00; p<0.05), serious pelvic/leg injury (OR 1.51; p<0.05), ISS (OR 1.05 per unit over 10 points; p<0.001), age (OR 1.02 per year over 45.1; p<0.001), CAUTI and VAP together (OR 0.10; p<0.001), and female gender (OR 0.51; p<0.001).

FIG. 1.

Odds ratios for risk factors for sepsis in patients with trauma adjusted for age, gender, and total Injury Severity Score. p<0.05 for all.

Table 3.

Multivariable Odds Ratio for Sepsis

	Odds Ratio (95% CI; p value)
ISS, per unit	1.05 (1.03–1.06, <0.001)
Age, per year	1.02 (1.01–1.03, <0.001)
Female	0.51 (0.34–76, 0.001)
Abdominal Injury Severity vs. None-mild
Moderate	1.02 (0.64–1.64, 0.934)
Serious	2.60 (1.54–4.40, <0.001)
Chest Injury Severity vs. None-mild
Moderate	1.51 (0.85–2.69, 0.157)
Serious	1.59 (0.99–2.55, 0.056)
Leg/Pelv. Injury Severity vs. None-mild
Moderate	0.78 (0.44–1.38)
Serious	1.51 (1.00–2.28, 0.049)
Severe+	0.61 (0.20–1.85)
Device-associated infection
CAUTI and VAP	0.10 (0.03–0.29, <0.001)

CAUTI=catheter-associated urinary tract infection; CI=confidence interval; ISS=Injury Severity Score; VAP=ventilator – associated pneumonia.

Discussion

The chances of a typical trauma patient developing a CAUTI or VAP are 3% and 1.5%, respectively. This confers a baseline 1.4% chance of developing sepsis and a 4.1% chance of death for the typical trauma patient. Factors that increase a person's chance of developing a CAUTI include being female, having a higher ISS, and being of greater age. Injuries with a strong associated risk for CAUTI are moderate head, neck, and face injury and abdominal injury. Traumatic brain injuries and spinal cord injuries have been reported to increase the odds of CAUTI [11]. A higher ISS (particularly with an abdominal injury) increases a person's risk of VAP. Pulmonary complications in trauma patients are reported to increase with advancing age, traumatic brain injury, and chest injury that requires a thoracic intervention such as thoracotomy, chest tube placement, or mechanical ventilation [12].

Some types of injuries appear to be associated with higher risks of sepsis, and these injury groups might therefore be targets of focused prevention efforts. Injuries with a strong associated risk for sepsis were severe chest injury (OR 2.0), severe abdominal injury (OR 3.0), and penetrating injury (OR 2.5). A higher ISS increases the odds of developing sepsis by 1.05 per point over 10 points. Certain injuries increase the odds of developing sepsis, such as serious pelvic/leg injury (1.51-fold), severe chest injury (2.00-fold), and abdominal injury (2.60-fold). Abdominal injury is linked to a higher risk of sepsis [13]. If the injury is penetrating, it increases the odds of sepsis 2.54-fold.

Both CAUTI and VAP are more potent than any other risk factor for sepsis. The presence of VAP increases the odds of sepsis 6.95-fold, and CAUTI increases the odds by 16.10. Increasing age over 45.1 y raises the odds of developing sepsis 1.02-fold per year. One study found that the odds of sepsis increase above 46 rather than 45 years of age [14]. Female gender reduces risk of sepsis in trauma victims by one-half. Published research on female gender and the risk of sepsis support this finding. The possibility of a protective effect of female hormones [15] vs. the harmful effect of male hormones [16] is being investigated.

Developing CAUTI and VAP together is not associated with a higher risk of sepsis above the baseline risks and may in fact be statistically protective. The reason for this paradox is unclear. A possible explanation is that having multiple antibiotics treating both infections is synergistic and thus more protective.

This study did not explore the timing and causal relations of events, merely associations. Thus, the exact causality of a device-associated infection and sepsis is unclear, and the sepsis may be from another source but coincident with device-associated infection. For instance, a patient who develops sepsis secondary to a penetrating chest injury may need to be ventilated mechanically and may subsequently develop VAP. However, the profound magnitudes of the odds ratios for CAUTI and VAP are not likely to be explained by this alone. Although it would be ideal to have culture data to draw conclusive connections between events, this was not feasible in this retrospective study. First, many CAUTI and VAP episodes were diagnosed clinically without positive cultures. However, the CDC does not require a positive culture for a diagnosis of CAUTI or VAP. Therefore, these cases were included in this study. Second, for those with a positive culture, the diversity of organisms reported makes it statistically difficult to draw any meaningful conclusions. Third, the majority of cases of sepsis were diagnosed clinically without documented bacteremia. Furthermore, even if different pathogens were identified as causing a device-associated infection and the episode of sepsis, it is possible that the treatment of the primary infection led to a super-infection [17,18].

Culture data in one study showed that pathogens causing VAP in trauma patients are more likely to be non-resistant strains of bacteria, leading to a lower mortality rate among trauma patients [19]. This may explain why VAP has a lower sepsis risk than CAUTI. The most common pathogens causing CAUTIs are Providencia stuartti and Proteus mirabilis [20] instead of the Escherichia coli commonly associated with community-acquired UTI [21]. Proteus mirabilis [22] and P. stuartti [23] are particularly virulent and may increase the mortality rate when associated with sepsis. More focused prospective research may confirm this hypothesis.

Conclusion

Device-associated infections increase the risk of sepsis and death after trauma. However, the incidence of device-associated infections is associated with greater injury severity and age. Thoracic injury, in particular, increases the risk of VAP, CAUTI, prolonged LOS, and sepsis. Once a patient has one device-associated infection, getting a second infection does not greatly increase the risk of sepsis. This study demonstrates the devastating effects of device-associated infections in trauma and highlights the importance of prevention initiatives, particularly given that key risk factors often are present on arrival and may not be modifiable.

Author Disclosure Statement

No competing financial interests exist.

References

Consunji

, Serrato Marinas

, Aspuria Maddumba

, Dela Paz

Jr.

A profile of deaths among trauma patients in a university hospital: The Philippine experience. J Inj Violence Res, 2011; 3:85–89.

Andersen

, Sharkey

, Schwartz

, McLellan

. Injury severity score, head injury, and patient wait days: Contributions to extended trauma patient length of stay. J Trauma, 1992; 33:219–220.

Niven

, Fick

, Kirkpatrick

, et al. Cost and outcomes of nosocomial bloodstream infections complicating major traumatic injury. J Hosp Infect, 2010; 76:296–299.

Mandell

, Bennett

, Dolin

(eds). Mandell, Douglas, and Bennett's Principles and Practice of Infectious Disease, 7th ed. Philadelphia: Churchill Livingstone, 2009.

Michetti

, Fakhry

, Ferguson

, et al. Ventilator-associated pneumonia rates at major trauma centers compared to a national benchmark: A multi-institutional study of the AAST. Presented at the 70th Annual Meeting of the American Association of the Surgery of Trauma. Chicago. September 14, 2011.

Davoodain

, Nematee

, Sheikhvatan

. Inappropriate use of urinary catheters and its common complications in different hospital wards. Saudi J Kidney Dis Transpl, 2012; 23:63–67.

Melzer

, Welch

. Outcomes in UK patients with hospital-acquired bacteraemia and the risk of catheter-associated urinary tract infections. J Postgrad Med, 2013; 89:329–334.

Centers for Disease Control and Prevention, Device Associated Module–NHSN. Catheter Associated Urinary Tract Infection Event: Guidelines and Procedures for Monitoring CAUTI. www.cdc.gov/nhsn/pdfs/7psccauticurrent.pdf

Centers for Disease Control and Prevention, Device Associated Module–NHSN. Ventilator Associated Pneumonia Event: Guidelines and Procedures for Monitoring VAP. www.cdc.gov/nhsn/pdfs/pscmanual/6pscvapcurrent.html Accessed March 23, 2015.

10.

Harris

, Taylor

, Thielke

, et al. Research Electronic Data Capture (REDCap): A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform, 2009; 42:377–381.

11.

D'Hondt

, Everaert

. Urinary tract infections in patients with spinal cord injuries. Curr Infect Dis Rep, 2011; 13:544–551.

12.

Hoyt

, Simons

, Winchell

, et al. A risk analysis of pulmonary complications following major trauma. J Trauma, 1993; 35:524–531.

13.

Morales

, Villegas

, Villavicencio

, et al. Intra-abdominal infection in patients with abdominal trauma. Arch Surg, 2004; 139:1278–1285.

14.

Deschinger

, Siegel

, Ho

, Kufera

. Effect in change of velocity on the development of medical complications in patients with multisystem trauma sustained in vehicular crashes. Accid Anal Prevent, 1998; 30:831–837.

15.

Choudhry

, Bland

, Chaundry

. Gender and susceptibility to sepsis following trauma. Endocr Metab Immune Disord Drug Targets, 2006; 6:127–135.

16.

Zeckey

, Andruszkow

, Neunaber

, et al. Protective effects of finasteride on the pulmonary immune response in a combined model of trauma–hemorrhage and polymicrobial sepsis in mice. Cytokine, 2011; 56:305–311.

17.

Tillotson

, Finland

. Bacterial colonization and clinical superinfection of the respiratory tract complicating antibiotic treatment of pneumonia. J Infect Dis, 1969; 119:597–624.

18.

Stoutenbeek

, van Saene

, Miranda

, et al. The prevention of superinfection in multiple trauma patients. J Antimicrob Chemother, 1984; 14:203–211.

19.

Magret

, Amaya-Villar

, Garnacho

; EU-VAP/CAP Study Group. Ventilator-associated pneumonia in trauma patients is associated with lower mortality: Results from EU–VAP study. J Trauma, 2010; 69:849–854.

20.

Dedeic-Ljubovic

, Hukic

. Catheter-related urinary tract infection in patients suffering from spinal cord injuries. Bosn J Basic Med Sci, 2009; 9:2–9.

21.

Bosch

, van Vuuren

, Joubert

. Antimicrobial resistance patterns in outpatient urinary tract infections: The constant need to revise prescribing habits. S Afr Med J, 2011; 101:328–331.

22.

Chen

, Chen

, Lu

, et al. Proteus mirabilis urinary tract infection and bacteremia: Risk factors, clinical presentation, and outcomes. J Microbiol Immunol Infect, 2012; 45:228–236.

23.

Woods

, Watanakunakorn

. Bacteremia due to Providencia stuartii: Review of 49 episodes. South Med J, 1996; 89:221–224.