Incidence and Risk Factors Associated with Multi-Drug–Resistant Pathogens in a Critically Ill Trauma Population: A Retrospective Cohort Study

Abstract

Background:

Multi-drug resistance is considered a serious health threat particularly in the intensive care unit (ICU) setting. Studies evaluating multi-drug–resistant (MDR) pathogens in critically ill trauma patients are limited. The objectives were to describe the incidence of MDR, extensive-drug–resistant (XDR), and pan-drug–resistant (PDR) organism growth in ICU patients admitted with traumatic injuries and to identify any risk factors associated with MDR growth.

Patients and Methods:

This was a retrospective single-center cohort study of all ICU adult patients identified via the institution's trauma registry from January 1, 2016 to August 31, 2017. Patients were included if they had positive culture growth with susceptibility data taken during the index hospitalization. Patients were excluded if their cultures were drawn within 48 hours of emergency department triage. Study groups were defined based on the presence of at least one MDR pathogen during the index hospitalization.

Results:

A total of 2,578 charts were reviewed and 95 patients (mean age, 60 years; 66 males [69%]) with 201 total cultures were included. The majority of positive cultures were from respiratory (69%) and urinary (16%) sources. Of the 201 positive cultures, the majority of species identified was Enterobacteriaceae (47%), Staphylococcus (32%), Enterococcus (7%), Acinetobacter (5%), and Pseudomonas (3%). Of the 95 patients with positive cultures, the incidence of MDR, XDR, and PDR organisms was found to be 31%, 17%, and 0%, respectively. Augmented renal clearance (ARC) was the only risk factor associated with an increased risk for MDR organism growth (adjusted odds ratio 9.78, 95% confidence interval [CI] 2.56–37.41; p = 0.001).

Conclusions:

In this cohort of critically ill trauma patients, the incidence of an MDR pathogen occurred in 31% of patients. This is the first study to find an association of ARC and multi-drug resistance, which should be further validated as a potential cause for MDR organisms.

According to the U.S. Centers for Disease Control and Prevention, multi-drug resistance is considered a serious health threat and is a particular concern in intensive care units (ICU). It accounts for up to two million illnesses per year and is responsible for at least 23,000 deaths annually [1]. Patients who acquire multi-drug–resistant (MDR) pathogens are more likely to demonstrate poor outcomes, including increased mortality and longer ICU length of stay [2,3]. Surveillance data in the ICU setting reported increased resistance rates, predominantly with Enterobacteriaceae species, including Klebsiella spp., Escherichia coli, and Pseudomonas aeruginosa [4,5]. However, studies identifying patients at risk for MDR pathogens are limited, especially for ICU patients presenting to the hospital after sustaining traumatic injuries.

Studies evaluating MDR pathogens specifically in the civilian trauma population are minimal. The incidence of MDR pathogens causing ventilator-associated or hospital-acquired pneumonia in 72 trauma ICU patients was found to be 30.6% [6]. A 2018 cohort of 397 trauma patients over a 10-year period with ventilator-associated pneumonia caused by either Pseudomonas aeruginosa or Acinetobacter baumannii were evaluated for risk factors for MDR growth. The only risk factor identified for MDR growth was prophylactic antibiotic days [7]. Both of these studies found that patients growing MDR pathogens were more likely to have inadequate empiric antimicrobial therapy.

Further research is needed to evaluate the incidence of MDR pathogens in this population, including both gram-negative and gram-positive pathogens, extensive-drug–resistant (XDR) and pan-drug–resistant (PDR) pathogens, and other sites of infection. It is unknown if there are other risk factors for MDR growth that exist in this population.

The primary objective of this study was to describe the incidence of MDR organism growth. Secondary objectives were to describe the incidence of subcategories of multi-drug resistance, including XDR and PDR organism growth, and to identify risk factors associated with MDR organisms in a critically ill trauma population.

Patients and Methods

This was a retrospective, single-center, cohort study at a level I trauma center evaluating the incidence of and risk factors associated with MDR organisms for ICU patients experiencing a traumatic injury. The source population included patients admitted to any ICU (surgical/medical, neurosurgical, or cardiovascular) from January 1, 2016 through August 31, 2017 who were identified via the institution's trauma registry (TraumaBase version 6.1, Clinical Data Management). The medical/surgical ICU is a 24-bed mixed unit for surgical, trauma, or medical patients and patients may be placed in other units, including neurosurgical or cardiovascular, depending on the availability of beds. Patients eligible for inclusion were those 18 years of age and older who were admitted to an ICU for at least 24 hours after sustaining a traumatic injury. Patients had to have positive culture growth with documented susceptibility data from blood, body fluid, sputum, urine, and/or wound cultures drawn at least 48 hours after emergency department triage. Patients were excluded if their positive cultures were drawn less than 48 hours from emergency department triage in order to only include patients with hospital-acquired pathogens. This study was approved by the Institutional Review Board and the requirement for informed consent was waived.

Group assignment was based on patients whose cultures grew at least one MDR organism (MDR group) and patients with cultures that only grew non-MDR organisms (non-MDR group) on all cultures during the index hospitalization. Patients with multiple positive cultures were only included once in the final analysis. An MDR organism was defined as a pathogen that was non-susceptible to at least one agent in three or more antimicrobial categories tested for susceptibility, excluding antimicrobials to which the pathogen is inherently non-susceptible [8 –10]. All methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococci (VRE), extended-spectrum β-lactamase (ESBL), and carbapenem-resistant Enterobacteriaceae (CRE) organisms were also considered to be MDR [11]. An ESBL organism was defined as an organism that was reported to be ESBL-producing based on modified Hodge test result or was reported to be resistant to first-, second-, and third-generation cephalosporins [12]. An XDR organism was defined as a pathogen that was non-susceptible to at least one agent in all but two antimicrobial categories tested for susceptibility, excluding antimicrobials to which the pathogen is inherently non-susceptible [8,9]. A PDR organism was defined as a pathogen that was non-susceptible to all agents in all antimicrobial categories tested for susceptibility [8-9].

Culture information was extracted from the microbiology data in the medical records. Respiratory cultures included bronchoalveolar lavage, tracheal aspirates, or sputum samples. Tracheal aspirate samples were obtained from either endotracheal or tracheostomy tubes; however, it was not possible to distinguish between these. Urine cultures encompassed those taken from indwelling catheters or clean-catch/non-catheter samples. Blood cultures were reported to be from central or peripheral access sites. Body fluids and wound cultures were deemed to be from surgical or non-surgical sites. However, if the exact site of the sample was unclear, it was noted to be unspecified.

All study data were collected from the institution's medical records and the data was managed using REDCap electronic data system [13]. Baseline demographic and clinical data (age, gender, creatinine clearance, body mass index [BMI], race as reported by the patient, comorbidities, primary service, injury severity scores [ISS], including the abbreviated injury scale [AIS] scores, need for mechanical ventilation, time to positive culture, in-hospital mortality, and ICU and hospital length of stay) were collected. Comorbidities included diabetes mellitus, chronic obstructive pulmonary disease (COPD), end-stage renal disease (ESRD) undergoing dialysis, and immunocompromised state, which included presence of any of the following: immunosuppressant medications, severe neutropenia, transplant recipients, and human immunodeficiency virus. Prior to admission (PTA) antimicrobial agents were obtained from the admission medication list verified by either the nursing or pharmacy staff. The creatinine clearance was estimated by the Cockcroft-Gault equation using the patient's admission serum creatinine and adjusting for body size [14,15].

The following potential predictor variables were collected in order to include in the regression analysis: augmented renal clearance (ARC; defined as a creatinine clearance >130 mL/min), renal impairment (defined as a creatinine clearance <60 mL/min), any hospitalization within the previous 90 days, any MDR culture within the previous year, blunt or penetrating injuries, any surgical intervention within the index hospitalization, one or more take back surgery for the same system or injury, early antibiotic exposure (i.e., within the first 48 hours) or late antibiotic exposure (i.e., after 48 hours) from emergency department triage, any blood product administration, and vasopressor utilization [16 –18].

The primary outcome was the incidence of MDR organism growth in critically ill trauma patients and the predictive value of risk factors for MDR organism growth. Secondary outcomes were the incidence of XDR and PDR organism growth. Incidence of certain MDR organisms, including MRSA, VRE, ESBL, and CRE organisms, were also described. Predictor parameters were determined a priori and tested in the univariable analysis. Parameters that were included were designated into four broad categories, including medical history and comorbidities, trauma characteristics, surgical interventions, and hospital course.

Categorical data were presented as frequency and proportions and analyzed by χ² or Fisher exact test as appropriate. Continuous parameters were analyzed by Student t-test or Mann-Whitney U test based on normality of data distribution and presented as mean with standard deviation or median with interquartile range [Q1–Q3]. Odds ratios and 95% CI were reported to examine the relation between presence of MDR organisms and predictor parameters. Predictor parameters with a p value less than 0.15 in the univariable analysis were selected for inclusion into a logistic regression model using backward elimination. Significance of parameters in the model was determined by the Wald test (p < 0.05). Adjusted odds ratios (AOR) and 95% CI for parameters in the model are presented. Statistical analysis was performed using IBM SPSS statistics version 24.0 (IBM Corp., Armonk, NY) and the level of significance was set at p < 0.05 (two-sided). A biostatistician aided in the analysis of the data.

Results

Overall, 2,567 ICU patients with traumatic injuries were identified from the trauma registry during the study period. Patients were excluded because of an ICU stay shorter than 24 hours (n = 2,003), absence of cultures taken during the hospitalization (n = 325), cultures drawn less than 48 hours after emergency department triage (n = 65), cultures without positive growth (n = 63), undocumented susceptibility (n = 14), and age less than 18 years (n = 2). A total of 95 patients were included in the study and assigned to the non-MDR (n = 65) or MDR (n = 30) groups based on identification of organisms during the index hospitalization (Fig. 1).

FIG. 1.

Flow diagram depicting inclusion and exclusion criteria for study cohort.

Baseline demographic data is presented in Table 1. The overall cohort included 66 (69%) males with a mean (± standard deviation) age of 60 (±22.5) years. Patient comorbidities included diabetes mellitus and COPD in 22 (23%) and 10 (10%) of patients, respectively. Only small proportion of patients (n = 3, 3%) received antimicrobial agents prior to hospitalization according to the PTA medication list. The majority of patients (84%) was located in the medical/surgical ICU. The patients in this cohort had a high severity of illness represented by a median ISS of 22 [IQR 10–30], median AIS of 3 [IQR 3–5], and 68 (71%) of patients requiring mechanical ventilation for at least 24 hours. The mean (± standard deviation) length of stay for the ICU and hospital was 13.0 (±8) days and 16.9 (±7) days, respectively. There was no significant difference in the time to positive culture between the non-MDR and MDR groups (mean [± standard deviation], 6.3 [3.8] vs. 6.0 (2.9), p = 0.733). Baseline characteristics were similar between the MDR and non-MDR groups, with the exception of creatinine clearance (median 111 [IQR 78–176] vs. 89 [IQR 74–108] mL/min, p = 0.013) and AIS extremity score (median 2 [IQR 0–3] vs. 0 [IQR 0–2], p = 0.029) being higher in the MDR group.

Table 1.

Demographic and Clinical Data for All Patients and by Study Groups

	Total	Non-MDR	MDR	p
	n = 95	n = 65	n = 30	p
Age (y)^a	60.0 (22.5)	62.4 (22.2)	54.7 (22.7)	0.127
Gender (male)^b	66 (69)	48 (74)	18 (60)	0.173
CrCl (mL/min)^c	93 [75–120]	89 [74–108]	111 [78–176]	0.013
BMI (kg/m²)^c	27 [24–32]	26 [24–32]	28 [24–33]	0.337
Race^b				0.564
Caucasian	78 (82)	55 (85)	23 (77)
African American	14 (15)	8 (12)	6 (20)
Other/not reported	3 (3)	2 (3)	1 (3)
Comorbidities^b
Diabetes mellitus	22 (23)	17 (26)	5 (17)	0.308
COPD	10 (10)	6 (9)	4 (13)	0.720
ESRD on dialysis	2 (2)	0 (0)	2 (7)	0.097
Immunocompromised	4 (4)	3 (5)	1 (3)	>0.99
PTA antibiotic AGENTS^b	3 (3)	2 (3)	1 (3)	>0.99
Primary service^b				0.128
Trauma	85 (89)	59 (91)	26 (87)
Neurosurgery	2 (2)	1 (1)	1 (3)
Orthopedic	4 (4)	1 (1)	3 (10)
Medicine	4 (4)	4 (6)	0 (0)
ICU location^b				0.239
MICU/SICU	80 (84)	54 (83)	26 (87)
NSICU	14 (15)	11 (17)	3 (10)
CVICU	1 (1)	0 (0)	1 (3)
ISS^c	22 [10–30]	18 [10–29]	24 [10–33]	0.458
AIS^c	3 [3–5]	3 [3–5]	3.5 [3–5]	0.711
AIS head/neck	3 [0–5]	3 [0–5]	3 [0–5]	0.881
AIS face	0 [0–0]	0 [0–0]	0 [0–0]	0.601
AIS abdomen	0 [0–0]	0 [0–0]	0 [0–0]	0.647
AIS spine	0 [0–2]	0 [0–2]	0 [0–2]	0.459
AIS extremity	0 [0–2]	0 [0–2]	2 [0–3]	0.029
AIS chest	0 [0–3]	1 [0–3]	0 [0–3]	0.150
Mechanical ventilation^b	68 (71)	48 (74)	20 (67)	0.471
In-hospital mortality^b	10 (10)	5 (8)	5 (17)	0.279
ICU LOS (days)^a	13.0 (8.0)	12.8 (8.0)	13.5 (8.2)	0.707
Hospital LOS (days)^a	16.9 (7.0)	16.5 (7.0)	17.7 (7.1)	0.475
Time to positive culture (days)^a	6.2 (3.5)	6.3 (3.8)	6.0 (2.9)	0.733

Presented as ^amean (SD) ^bn (%) ^cmedian [IQR].

BMI = body mass index; CrCl = creatinine clearance; COPD = chronic obstructive pulmonary disease; ESRD = end-stage renal disease; PTA = prior to admission; MICU = medical intensive care unit; SICU = surgical intensive care unit; SD = standard deviation; NSICU = neurosurgical intensive care unit; CVICU = cardiovascular intensive care unit; ISS = injury severity score; AIS = abbreviated injury scale; ICU = intensive care unit; LOS = length of stay.

Outcomes data for the total positive cultures and for patients with positive cultures is presented in Table 2. Of the 201 positive cultures assessed from 95 patient encounters, 57 (28%) grew at least one MDR pathogen with 8 (14%) and 0 (0%) being an XDR and PDR organism, respectively. Additionally, MRSA 18 (9%), ESBL 10 (5%), CRE 5 (2%), and VRE 1 (0.5%) were identified from all positive cultures. Of the 95 patients with positive cultures, the incidence of at least one MDR pathogen was found in 30 (31%) patients. Of those 30 patients, there were 5 (17%) and 0 (0%) of patients growing XDR and PDR pathogens, respectively. Of the patients with positive cultures, MRSA 12 (13%), ESBL 9 (9%), CRE 4 (4%), and VRE 1 (1%) pathogens were present.

Table 2.

Outcomes Data: Incidence of Drug-Resistant Organism Growth

	Total positive cultures	Positive culture patients
	n = 201	n = 95
Non-MDR	144 (72)	65 (68)
MDR	57 (28)	30 (31)
XDR	8/57 (14)	5/30 (17)
PDR	0 (0)	0 (0)
MRSA	18 (9)	12 (13)
ESBL	10 (5)	9 (9)
CRE	5 (2)	4 (4)
VRE	1 (0.5)	1(1)

Presented as n (%).

MDR = multi-drug–resistant; XDR = extensive-drug–resistant; PDR = pan-drug–resistant; MRSA = methicillin-resistant Staphylococcus aureus; ESBL = extended-spectrum beta-lactamase; VRE = vancomycin-resistant Enterococci; CRE = carbapenem-resistant Enterobacteriaceae.

A summary of organism growth in all the cultures and the culture source is presented in Table 3. The majority of positive cultures were from respiratory 139 (69%), including tracheal aspirates (57%), sputum (10%) and bronchoalveolar lavage (3%), and urinary 32 (16%) sources with the majority obtained from unspecified (9%) and clean-catch/non-catheter (5%) sites. Of the 201 positive cultures, the majority were Enterobacteriaceae species 94 (47%), followed by Staphylococcus 64 (32%), Enterococcus 14 (7%), Acinetobacter 11 (5%), and Pseudomonas 6 (3%). There were no substantial differences regarding the pathogens grown on culture or the culture source between the MDR and non-MDR groups.

Table 3.

Culture Source and Organism Growth

Culture source and organism growth	Total positive cultures	Non-MDR cultures	MDR cultures	p
Culture source and organism growth	n = 201	n = 144	n = 57	p
Respiratory	139 (69)	102 (71)	37 (65)	0.413
Tracheal aspirate	114 (57)	85 (59)	29 (51)
Sputum	19 (10)	14 (10)	5 (9)
BAL	6 (3)	3 (2)	3 (5)
Urine	32 (16)	25 (17)	7 (12)	0.374
Unspecified	18 (9)	15 (10)	3 (5)
Clean-catch/non-catheter	10 (5)	7 (5)	3 (5)
Indwelling catheter	4 (2)	3 (2)	1 (2)
Wound	13 (6)	6 (4)	7 (12)	0.053
Non-surgical	5 (3)	2 (1)	3 (5)
Surgical	8 (4)	4 (3)	4 (7)
Body fluid	12 (6)	8 (6)	4 (7)	0.744
Non-surgical	5 (3)	3 (2)	2 (4)
Unspecified	7 (4)	5 (4)	2 (4)
Blood	5 (2)	3 (2)	2 (3)	0.623
Peripheral	4 (2)	2 (1)	2 (3)
Central	1 (0.5)	1 (1)	0 (0)
Organism growth
Enterobacteriaceae	94 (47)	63 (44)	31 (54)	0.210
Citrobacter spp.	3 (1)	2 (1)	1 (2)
Enterobacter spp.	23 (11)	20 (14)	3 (5)
Escherichia coli	27 (13)	17 (12)	10 (17)
Klebsiella spp.	25 (12)	15 (10)	10 (17)
Morganella morganii	1 (0.5)	1 (0.7)	0 (0)
Proteus spp.	7 (3)	7 (5)	0 (0)
Providencia spp.	1 (0.5)	0 (0)	1 (2)
Serratia spp.	5 (2)	1 (0.7)	4 (7)
Other	2 (1)	0 (0)	2 (3)
Staphylococcus spp.	64 (32)	50 (35)	14 (24)	0.182
Staphylococcus aureus	59 (29)	47 (33)	12 (21)
Staphylococcus non-aureus	5 (2)	3 (2)	2 (3)
Enterococcus spp.	14 (7)	12 (8)	2 (3)	0.357
Enterococcus faecalis	11 (5)	11 (8)	0 (0)
Enterococcus faecium	1 (0.5)	0 (0)	1 (2)
Other	2 (1)	1 (0.7)	1 (2)
Acinetobacter spp.	11 (5)	6 (4)	5 (9)	0.299
Pseudomonas spp.	6 (3)	2 (1)	4 (7)	0.055
Stenotrophomonas maltophilia	5 (2)	4 (3)	1 (2)	>0.99
Other non-Enterobacteriaceae	4 (2)	4 (3)	0 (0)	0.330
Streptococcus pneumoniae	3 (1)	3 (2)	0 (0)	0.560

Presented as n (%).

BAL = bronchoalveolar lavage; NA = not applicable.

The univariable analysis and multivariable logistic regression analysis for covariates associated with an MDR pathogen is presented in Table 4. Nine covariates were significant in the prediction for the MDR group and put into the multivariable logistic regression model using the pre-defined p value of 0.15. Male gender, age greater than 64, age less than 51, ARC, ESRD on dialysis, two or more surgeries during the index hospitalization, at least one take-back surgery, transfusion of any blood product, vasopressor use for at least 24 hours were all found to be noteworthy in the univariable analysis (p < 0.15). In the multivariable analysis, the only covariate found to be associated with an increased risk of MDR organism growth was ARC (AOR 9.78, 95% CI 2.56–37.41, p = 0.001). An age greater than 64 years old was found to decrease the risk for MDR growth (AOR 0.25, 95% 0.08-0.77, p = 0.016). This model was statistically significant χ²(4) = 24.1, p < 0.001.

Table 4

Univariable Analysis and Multivariable Logistic Regression Analysis for Covariates Associated with Multi-Drug–Resistant Growth

Univariable analysis for predictor parameters associated with MDR growth
Predictor parameters	Non-MDR	MDR	OR (95% CI)	p
Predictor parameters	n = 65	n = 30	OR (95% CI)	p
Medical history
Male gender	48 (74)	18 (60)	0.53 (0.21, 1.33)	0.127
Age >64	38 (58)	12 (40)	0.47 (0.20, 1.14)	0.094
Age ≤50	18 (28)	13 (43)	2.00 (0.81, 4.93)	0.131
BMI ≥30 (kg/m²)	20 (31)	10 (33)	1.12 (0.45, 2.83)	0.803
ARC	7 (11)	12 (40)	5.52 (1.89, 16.13)	0.001
Renal impairment	6 (9)	1 (3)	0.34 (0.04, 2.95)	0.426
Hospitalization within previous 90 d	4 (6)	3 (10)	1.69 (0.35, 8.10)	0.675
MDR cultures within previous 1 y	3 (5)	0 (0)	—	0.549
Diabetes mellitus	17 (26)	5 (17)	0.56 (0.19, 1.71)	0.308
COPD	6 (9)	4 (13)	1.51 (0.39, 5.82)	0.720
ESRD on dialysis	0 (0)	2 (7)	—	0.097
Immunocompromised	3 (5)	1 (3)	0.71 (0.07, 7.15)	>0.99
Trauma characteristics
Blunt injury	62 (95)	27 (90)	2.30 (0.43, 12.11)	0.376
ISS ≥16	40 (61)	19 (63)	1.308 (0.44, 2.64)	0.867
AIS ≥4	30 (46)	15 (50)	1.17 (0.49, 2.77)	0.727
Surgical interventions
Any surgery during admission	41 (63)	21 (70)	1.37 (0.54, 3.46)	0.510
Surgeries during admission ≥2	22 (34)	16 (53)	2.23 (0.92, 5.40)	0.072
Take-back surgeries for same system/injury ≥1	6 (9)	8 (27)	3.58 (1.11, 11.48)	0.034
Hospital characteristics
ICU length of stay ≥5 d	52 (80)	24 (80)	1.00 (0.34, 2.95)	>0.99
Early antibiotic exposure	30 (46)	18 (60)	1.75 (0.73, 4.21)	0.210
Late antibiotic exposure	37 (57)	18 (60)	1.13 (0.47, 2.74)	0.778
Invasive mechanical ventilation ≥ 24 h	48 (74)	20 (67)	0.71 (0.28, 1.81)	0.471
Received any transfusion	27 (41)	21 (70)	3.28 (1.30, 8.27)	0.010
Vasopressor use ≥24 h	8 (12)	9 (30)	3.05 (1.04, 8.95)	0.037

Multivariable logistic regression for covariates associated with MDR growth
Predictor parameters	Adjusted OR	95% CI	P
Age >65	0.25	(0.08, 0.77)	0.016
ARC	9.78	(2.56, 37.41)	0.001

Risk factors with p < 0.15 selected for inclusion for the multivariable logistic regression analysis: male gender, age ≥65, age ≤50, ARC, ESRD on dialysis, number of surgeries in OR ≥2, ≥1 take-back surgeries, received any transfusion, and vasopressor use ≥24 h.

BMI = body mass index; ARC = augmented renal clearance; MDR = multi-drug resistant; COPD = chronic obstructive pulmonary disease; ESRD = end-stage renal disease; ISS = injury severity score; AIS = abbreviated injury scale; ICU = intensive care unit.

Discussion

This retrospective cohort study evaluated the incidence of MDR organisms and risk factors associated with MDR organism growth in critically ill trauma patients. The incidence of an MDR organism was approximately one-third of patients with positive cultures with a relatively low rate of MRSA, VRE, ESBL and CRE organisms. The incidence of XDR pathogens occurred in 17% of patients and no PDR organisms were identified. According to the multivariable analysis, the only parameter found to be associated with increased risk of an MDR pathogen was ARC, which was represented by an estimated creatinine clearance above 130 mL/min.

There were notable similarities and differences from this study compared with previous evaluations. The incidence of MDR pathogens was similar to previous reports with ICU trauma populations, although the previous reports were solely limited to respiratory infections and specific pathogens, including Pseudomonas aeruginosa or Acinetobacter baumannii [6,7]. In contrast, we included all cultures with susceptibility data from any site. To our knowledge, this is the first study to evaluate the incidence of XDR and PDR pathogens.

There were considerable differences in the risk factors associated with MDR organism growth in this study as compared to studies that included medical and surgical ICU patients [17,18]. Independent risk factors found to be important in those studies included presence of any gram-negative bacteria within six months, any surgery during the index hospitalization, ESRD on dialysis, prior use of a carbapenem within six months, ICU length of stay more than five days, hospitalization within the previous year, baseline oxygen dependence, and respiratory failure requiring endotracheal intubation [17,18]. Whereas several risk factors have been found to be associated with MDR pathogens in these studies, they did not include an ICU trauma population. Overall, variables identified as independent risk factors in these medical and surgical ICU patients were not found to be significant in this cohort of patients.

There were also distinct differences in our findings compared with the previous evaluations that included a critical ill trauma population [6,7]. Early (i.e., within 48 hours) or late exposure to antimicrobial agents, whether these were given for prophylaxis or treatment, were not found to be noteworthy for MDR growth in the current study. Lewis et al. [7] found that prolonged prophylactic antibiotic exposure was associated with MDR growth. However, the current study did not evaluate for total duration of therapy or overall exposure, and courses for prophylaxis or treatment could not be determined or delineated.

Overall, the only risk factor associated with an increased risk for an MDR pathogen was the presence of ARC, which has not been evaluated in previous studies. Augmented renal clearance has been previously shown to affect a large proportion of trauma patients with one analysis showing an incidence up to 67% [19]. The major concern for ARC is the potential of under-dosing antibiotics [20], which then may lead to drug resistance. Several studies have shown that ARC is related to under-dosing by low or undetectable levels of antimicrobials, including levofloxacin [21], meropenem [22], piperacillin/tazobactam [23 –25], and vancomycin [26,27]. However, ARC may affect any medication that undergoes renal elimination. While several studies have evaluated ARC in relation to under-dosing of medications, this is the first evaluation to find an association with ARC and multi-drug resistance.

Limitations

The primary limitation was the single-center design limiting data collection, therefore, collecting data for previous culture growth, antibiotic agents, or hospitalizations from outside institutions could not be assessed. It was difficult to assess antibiotic utilization in the previous 90 days based on institutional data alone. However, current antimicrobial use was documented on the PTA medication list, which was found to be a small number of patients (3% overall). The results may not be representative of other institutions with different susceptibility data and may limit the generalizability of the results. Additionally, presence of an active infection was not assessed and could not determined to be pathogenic or as colonization because of the retrospective design. The retrospective design also limited the granularity of culture information for some samples. However, this was minimal with 13% of samples (urine and body fluid) being noted as unspecified. There were potentially additional risk factors that exist for MDR organism growth in critically ill trauma patients that were not assessed in this study, although the variables chosen were based on previous research in this field [17,18] and expert opinion of the authors.

Another potential limitation of this study was the lack of a measured creatinine clearance to verify that ARC was present. Although the Cockcroft-Gault equation has been found to be inaccurate for estimating ARC, the elevated creatinine clearance is likely related to ARC because the calculated equation has been found to be substantially lower than the measured [16 –20]. Another potential limitation is the relatively small sample size, which was limited by the data integrity within the trauma surveillance system prior to the study dates.

Conclusions

According to this evaluation of critically ill patients sustaining traumatic injuries, growth of an MDR culture occurred with a similar incidence compared with previously reported literature. This was the first study to also report the incidence of XDR and PDR growth and describe other sites growing infectious pathogens. Risk factors identified in the medical and surgical ICU patients in previous studies were not found to be significant in this trauma ICU population. The only risk factor found to be associated with increased risk of an MDR pathogen was ARC, which was represented by an estimated creatinine clearance >130 mL/min. This is the first study find an association of ARC and multi-drug resistance, which should be further validated as a potential cause for MDR organisms. Further prospective studies are needed to evaluate risk factors for MDR pathogens in this population.

Footnotes

Acknowledgments

The authors would like to thank the Cleveland Clinic Akron General Research Department and Trauma Registry for their assistance.

Author Disclosure Statement

All the authors of this article have nothing to disclose regarding possible financial or personal relationships with commercial entities that may have a direct or indirect interest in the subject matter of this study.

References

Centers for Disease Control and Prevention. Antibiotic/Antimicrobial Resistance. www.cdc.gov/drugresistance/biggest_threats.html. (Last accessed December 3, 2018).

Martin-Loeches

, Torres

, Rinaudo

, et al. Resistance patterns and outcomes in intensive care unit (ICU)-acquired pneumonia. Validation of European Centre for Disease Prevention and Control (ECDC) and the Centers for Disease Control and Prevention (CDC) classification of multidrug resistant organisms. J Infect, 2015; 70:213–222.

Tedja

, Nowacki

, Fraser

, et al. The impact of multidrug resistance on outcomes in ventilator-associated pneumonia. Am J Infect Control, 2014; 42:542–545.

Sader

, Farrell

, Flamm

, et al. Antimicrobial susceptibility of gram-negative organisms isolated from patients hospitalized in intensive care units in United States and European hospitals (2009–2011). Diagn Microbiol Infect Dis, 2014; 78:443–448.

Sader

, Farrell

, Flamm

, et al. Antimicrobial suspeptibility of Gram-negative organisms isolated from patients hospitalized with pneumonia in US and European hospitals: Results from the SENTRY Antimicrobial Surveillance Program, 2009–2012. Int J Antimicorb Agents, 2014; 43:328–334.

Becher

, Hoth

, Neff

, et al. Multidrug-resistant pathogens and pneumonia: Comparing the trauma and surgical intensive care units. Surg Infect, 2011; 12:267–272.

Lewis

, Sharpe

, Swanson

, et al. Reinventing the wheel: Impact of prolonged antibiotic exposure on multidrug-resistant ventilator-associated pneumonia in trauma patient. J Trauma Acute Care Surg, 2018; 85:256–262.

Magiorakos

, Srinivasan

, Carey

, et al. Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: An international expert proposal for interim definitions for acquired resistance. Clin Microbiol Infec, 2012; 18:268–281.

Mandell

, Bennet

, Dolin

(eds): Mandell, Douglas, and Bennett's: Principles and Practice of Infectious Disease. Volume 2, 8th edition. Philadelphia, PA: Elsevier, 2015.

10.

Gilbert

, Chambers

, Eliopoulos

, et al. (eds): The Sanford Guide to Antimicrobial Therapy. Sperryville, VA: Antimicrobial Therapy, Inc., 2017.

11.

Siegel

, Rhinehart

, Jackson

, et al. Management of multidrug-resistant organisms in healthcare settings, 2006. Am J Infect Control, 2007; 35:S165–S193.

12.

Rawat

, Nair

. Extended-spectrum β-lactamases in gram negative bacteria. J Glob Infect Dis, 2010; 2:263–274.

13.

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.

14.

Cockcroft

, Gault

. Prediction of creatinine clearance from serum creatinine. Nephron, 1976; 16:31–41.

15.

Winter

, Guhr

, Berg

. Impact of various body weights and serum creatinine concentrations on the bias and accuracy of the Cockcroft-Gault equation. Pharmacotherapy, 2012; 32:604–612.

16.

Baptista

, Udy

, Sousa

, et al. A comparison of estimates of glomerular filtration in critically ill patients with augmented renal clearance. Crit Care, 2011; 15:R139.

17.

Vasudevan

, Mukhopadhyay

, Li

, et al. A prediction tool for nosocomial multi-drug resistant gram-negative bacilli infections in critically ill patients: Prospective observational study. BMC Infect Dis, 2014; 14:615.

18.

Sonti

, Conroy

, Welt

, et al. Modeling risk for developing drug resistant bacterial infections in an MDR-naïve critically ill population. Ther Adv Infect Dis, 2017; 4:95–103.

19.

Barletta

, Mangram

, Byrne

, et al. Identifying augmented renal clearance in trauma patients: Validation of the Augmented Renal Clearance in Trauma Intensive Care scoring system. J Trauma Acute Care Surg, 2017; 82:665–671.

20.

Barletta

, Mangram

, Bryne

, et al. The importance of empiric antibiotic dosing in critically ill trauma patients: Are we under-dosing based on augmented renal clearance and inaccurate renal clearance estimates?. J Trauma Acute Care Surg, 2016; 81:1115–1121.

21.

Roberts

, Cotta

, Cojutti

, et al. Does critical illness change levofloxacin pharmacokinetics?. Antimicrob Agents Chemother, 2016; 60:1459–63.

22.

Carlier

, Carrette

, Roberts

, et al. Meropenem and piperacillin/tazobactam prescribing in critically ill patients: Does augmented renal clearance affect pharmacokinetic/pharmacodynamics target attainment when extended infusions are used?. Crit Care, 2013; 17:R84.

23.

Akers

, Niece

, Chung

, et al. Modified augmented renal clearance score predicts rapid piperacillin and tazobactam clearance in critically ill surgery and trauma patients. J Trauma Acute Care Surg, 2014; 77:S163–70.

24.

Huttner

, Von Dach

, Renzoni

, et al. Augmented renal clearance, low β-lactam concentrations and clinical outcomes in the critically ill: An observational prospective cohort study. Int J Antimicrob Agents, 2015; 45:385–392.

25.

Udy

, Dulhunty

, Roberts

, et al. Association between augmented renal clearance and clinical outcomes in patients receiving β-lactam antibiotic therapy by continuous or intermittent infusion: A nested cohort study of the BLING-II randomized, placebo-controlled, clinical trial. Int J Antimicrob Agents, 2017; 49:624–360.

26.

Chu

, Luo

, Qu

, et al. Application of vancomycin in patients with varying renal function, especially those with augmented renal clearance. Pharm Bio, 2016; 54:2802–2806.

27.

Baptista

, Sousa

, Martins

, et al. Augmented renal clearance in septic patients and implications for vancomycin optimisation. Int J Antimicrob Agents, 2012; 39:420–423.