Impact of Multi-Drug–Resistant Pneumonia on Outcomes of Critically Ill Trauma Patients

Abstract

Background:

Rates of infections with multi-drug–resistant organisms (MDROs) are increasing among critically ill patients. Among non-surgical patients, MDROs increase directly the risk of adverse secondary events including death. However, similar effects do not appear to occur among surgical patients. Specifically, among critically injured trauma patients, it is unknown whether degree of injury versus the presence of an MDRO increases the risk of death.

Methods:

This is a retrospective chart review of admitted adult trauma patients. Data included demographics, medical comorbidities, injury severity score, infections, occurrence of pneumonia including microbiology sensitivity profile, hospital course, and outcomes.

Results:

Patients requiring adminission to the intensive care unit (ICU) were more severely injured with greater degree of thoracic and head trauma and had a greater burden of pre-trauma medical comorbidities. Among those admitted to the ICU, 93 patients developed pneumonia. Patients who developed pneumonia were younger and more severely injured, with higher rates of thoracic and head injuries and higher rates of smoking. Development of pneumonia was associated with worse outcomes. However, among patients with pneumonia, comparing MDRO to pan-sensitive (PanSens) infections, PanSens infection occurred earlier and were more likely associated with pre-trauma smoking status. There was no difference in injury patterns, medical comorbidities, or outcomes.

Conclusion:

The development of pneumonia among trauma patients reflects degree of injury and underlying medical status. However, development of MDRO versus PanSens pneumonia did not affect trauma-related outcomes further. This information will guide family discussions and critical care decisions better among vulnerable trauma patients.

Infectious complications and multi-organ failure remain leading contributors to delayed deaths after trauma. Traumatic injuries impose a marked degree of immune dysfunction and immunodepression [1], leading to increased risk of secondary infections. The occurrence of a secondary infection after traumatic injury is often a reflection of the degree of injury and physiologic insult suffered by the patient. Among trauma patients, ventilator-associated pneumonia (VAP) remains the leading cause of trauma-related complications. Despite the fact that direct trauma-related mortality may have decreased, the development of a VAP continues to contribute substantial added morbidity in trauma patients. Furthermore, trauma-related pneumonia, especially among geriatric trauma patients, is associated with increased ventilator days, an increased risk for needing tracheostomy, and a greater need for long-term post-discharge care.

However, not all infections are the same [2,3]. Multi-drug–resistant organisms (MDROs) have gained considerable attention across both developed and developing nations, being associated with greater need for broader spectrum antimicrobial agents, increased need for admission to the intensive care unit (ICU), as well as adding to overall healthcare expenditure [3,4] with patient isolation and the use of single occupancy hospital rooms. Reports have also associated MDROs with increased risk of adverse long-term events. However, the impact of these MDROs is debated between direct attributable mortality versus a reflection of underlying severity of illness. Among non-surgical patients, several investigators have noted that MDROs compared with pan-sensitive (PanSens) organisms have been associated with increased rates of complications and ultimately increased risk of death [5 –7]. These have been attributed to potential virulence attributes of the MDRO. However, most of these observations were noted among non-surgical patients presenting with coronary artery events, diabetes, or chronic lung disease.

Multi-drug–resistant organisms-related infections as primary events appear to be considerably different from secondary MDRO infectious events. Although secondary infectious events do contribute to an increased risk of further adverse events, MDRO infections among surgical patients may not add to the already high rates of morbidity and mortality. Several small studies, among predominantly orthopedic [8] or burn patients, have previously demonstrated no additive effect comparing MDRO with PanSens infections. In a relatively large analysis of critically ill surgical patients, Rosenberg et al. [9] demonstrated that the occurrence of methicillin-resistant Staphylococcus areus (MRSA) versus PanSens organisms did not further impact mortality. However, this study was undertaken in a mixed ICU population. To date, there is no study comparing outcomes of trauma patients with MDRO versus PanSens pneumonia. Given the major immunologic and physiologic impact of traumatic injuries, we hypothesize that when an infection occurs among critically injured trauma patients, the type of organism (MDRO vs. PanSens) will not impact trauma-related outcomes.

Patients and Methods

This is a retrospective chart review of all admitted blunt trauma patients, aged 18 years and older, from a single level I trauma center over a three-year period. Patients were excluded if they were dead upon arrival or if they died in the trauma bay or immediately in the operating room. Charts were reviewed for patient demographics (age and gender), injuries sustained, and Injury Severity Score (ISS) focusing on thoracic trauma, rib fractures and head injuries, all interventions, and hospital course including all complications, outcomes, disposition, and mortality. The number of rib fractures sustained was also recorded given that number of rib fractures is a central component of rib fracture protocol dictating need for ICU admission among patients who sustain thoracic trauma [10]. Specifically, we focused upon ICU patients, noting the need for mechanical ventilation development of pneumonia, need for tracheostomy, hospital length of stay, mortality, and, among survivors, post-hospital disposition.

Pre-trauma medical comorbidities had been recorded into the trauma registry from information gathered at the time of presentation as well as any new information collected during the course of the hospital stay. Comorbidities were counted both individually to calculate actual number of comorbidities per patient as well as being divided into groups by systems: pulmonary, cardiac, endocrine, renal, etc. Given that both chronic obstructive pulmonary disease (COPD) and smoking are both known contributors to risk of pulmonary infections, we also reviewed whether patients were active smokers at the time of the trauma.

The diagnosis of pneumonia and all treatment decisions had been made by the primary treating team. Clinical Pulmonary Infection Score (CPIS) was used as the trigger to consider the diagnosis of PNA. Following clinical consideration, if the CPIS was greater than 5, bronchoscopy and lavage was performed at the time of CPIS calculation without delay. Pneumonia was diagnosed by bronchoalveolar lavage (BAL) with >10,000 colony forming units per milliliter (CFU/mL). At the time of obtaining the BAL samples, our protocol is to commence antimicrobial coverage consisting of vancomycin and piperacillin-tazobactam immediately, unless an allergy dictates an alternative choice. Antimicrobial agents are then tailored to the microbiology results. Final microbiology data was reviewed to determine the infecting organism(s) and also to determine whether the organisms were PanSens versus MDRO. Micro-organisms known to exhibit intrinsic antimicrobial resistance were considered to be resistant for the purpose of this study. Among patients with polymicrobial infections, if any organisms displayed drug resistance, they were then considered among the MDRO cohort. The time to diagnosis of pneumonia was calculated as the time from hospital admission to the time of performing BAL. Analysis initially consisted of reviewing ICU patients who did versus did not develop a VAP. Next, among VAP patients, we reviewed the impact of drug-resistant organisms by comparing characteristics and outcomes between VAP patients with PanSens versus MDRO infecting organism.

Categorical data was analyzed using χ² analysis. Continuous data were analyzed using either Student t-test (and expressed as mean ± standard error of the mean) or Mann-Whitney U test (and expressed as median (interquartile range [IQR]). SigmaPlot 12.5 (Systat Software, Inc., San Jose, CA) was used for data analysis. Statistical significance was set at p < 0.05.

Results

Over the study period, 10,985 patients were admitted after traumatic injuries. The overall average age was 58.9 years; falls were the most predominant blunt mechanism. Patients who required ICU level of care (n = 2,389) were older and had higher median ISS. Intensive care unit patients had greater rates of rib fractures and head injury and were more likely to have a pre-trauma medical diagnosis of heart failure, COPD, and diabetes, but were less likely to be active smokers (16.6% vs. 27.2%; p < 0.001; Table 1).

Table 1.

Patients Admitted to the Trauma Intensive Care Unit Were Older, More Severely Injured with Higher Rates of Head and Thoracic Trauma and Had Higher Rates of Pre-Trauma Medical Comorbidities

	Non-ICU	ICU	P
Number of patients	8,596	2,389
Age	58.7 ± 0.2	62.4 ± 0.9	<0.001
Male gender	58.5%	53.9%	<0.001
ISS (median [IQR])	9 (4–10)	13 (9–25)	<0.001
Injuries
Fall	52%	67%	<0.001
Rib fractures	17.5%	27.6%	<0.001
Head injury	17.1%	41.2%	<0.001
Spleen injury	2.1%	4.1%	<0.001
Medical comorbidities
Heart failure	4.4%	10.5%	<0.001
Chronic lung disease	6.9%	9.4%	<0.001
Smoker	27.2%	16.7%	<0.001
Diabetes	12.9%	14.8%	0.02

ICU = intensive care unit; ISS = Injury Severity Score; IQR = interquartile range.

Among patients admitted to the ICU, 93 patients developed pneumonia. Patients who developed pneumonia compared with those who did not develop pneumonia were younger (50 [IQR = 34–61] vs. 80 years [IQR = 56–91]; p < 0.01) and more likely male (78.5% vs. 52.9%; p < 0.001). Among patients with pre-trauma medical comorbidities, there was no difference in median number of medical comorbidities between those who did versus did not develop pneumonia (2 [1 –3] vs. 3 [1 –4]; p = 0.2). Specifically, there was no difference in rates of diabetes, COPD, or steroid use between patients with and without pneumonia. However, patients who developed pneumonia were markedly more likely to be active smokers at the time of the trauma (36.5% vs. 15.9%; p < 0.001). Regarding injury profile, patients with compared with patients without pneumonia had higher median ISS (26 vs. 13; p < 0.001). Regarding specific injuries, patients with pneumonia were noted to have higher rates of rib fractures (48.4% vs. 26.7%; p < 0.001), higher median number of rib fractures, higher rates of head injuries (70.9% vs. 40.6%; p < 0.001), and higher rates of pelvic fractures (17.2% vs. 8.9%; p = 0.01). Patients with pneumonia had markedly higher rates of needing a tracheostomy (69.9% vs. 4.8%; p < 0.001) and longer length of stay (5 vs. 30 days; p < 0.001; Table 2).

Table 2.

Patients Who Developed Pneumonia Had Higher Rates of Rib Fractures and Head Injuries and Were More Likely to Be Active Smokers

	No pneumonia	Pneumonia	P
Age	80 (56–91)	50 (34–61)	<0.001
Male gender	52.9%	78.5%	<0.001
Injuries
ISS (median [IQR])	13 (9–25)	26 (16–34)	<0.001
Rib fractures	26.7%	48%	<0.001
Median number of rib fractures	4 (2–4)	4 (4–4)	0.01
Head injury	40.4%	60.2%	<0.001
Pelvic fracture	8.9%	17.2%	0.01
Medical comorbidities
Number of comorbidities	2 (1–3)	3 (1–4)	0.2
(median [IQR])	2 (1–3)	3 (1–4)	0.2
Diabetes	14.9%	12.9%	0.7
COPD	9.6%	10.7%	0.6
Steroid use	2.3%	5.3%	0.07
Smoker	15.9%	36.5%	<0.001
Outcomes
Need for tracheostomy	4.8%	69.9%	<0.001
Length of stay(median [IQR])	5 (3–9)	30 (19.75–51.25)	<0.001

ISS = Injury Severity Score; IQR = interquartile range; COPD = chronic obstructive pulmonary disease.

Microbiology reports were reviewed noting the organisms involved as well as the sensitivity versus resistance profile. Twelve patients were noted to have polymicrobial infections. Among the 93 patients with pneumonia, 62 were noted to have PanSens organisms, whereas 31 patients (33.3%) had drug-resistant organisms. The most common organism among PanSens patients was methicillin-sensitive Staphylococcus aureus (MSSA) occurring in 27 of 62 patients, with Haemophilus influenzae and Streptococcus pneumoniae being next most common. Among patients with MDRO pneumonia the most common organism was MRSA, which occurred in 14 of 31 patients with Pseudomonas aeruginosa and Escherichia coli being the next most common (Table 3). Among the 93 patients who developed pneumonia, five patients were noted to develop a second episode of pneumonia during the hospital stay, two of whom were in the MDRO group and three in the PanSens group. None of the five patients crossed over in groups, such that if the first pneumonia was MDRO then the repeat infection was also MDRO, and if the first infection was PanSens then the second PNA was also PanSens. There were no characteristic differences between patients who developed one versus multiple episodes of pneumonia.

Table 3.

The Top Three Most Common Isolated Organisms in Each Category

PanSens		MDRO
n = 62		n = 31
MSSA	27	MRSA	14
Haemophilus influenzae	24	Pseudomonas aeruginosa	5
Streptococcus pneumoniae	4	Escherichia coli ESBL	3

Polymicrobial	5	Polymicrobial	7

Among PanSens MSSA was most common occurring in 27 patients and among MDRO infections MRSA was the most common occurring in 14 patients.

PanSens = pan-sensitive; MDRO = multi-drug–resistant organisms; MSSA = methicillin-sensitive Staphylococcus aureus; MRSA = methicillin-resistant Staphylococcus aureus; ESBL = extended spectrum β-lactamase.

Comparing PanSens patients with MDRO PNA patients, there was no difference with respect to age (47.7 ± 2.3 vs. 53 ± 3.5 years; p = 0.22) or male gender (76.1% vs. 84.6%; p = 0.54). Compared with PanSens pneumonia, MDRO pneumonia occurred slightly later in the hospital course (7 vs. 6 days; p = 0.04). There was no difference in rates of COPD or diabetes between the groups, however, patients who developed MDRO pneumonia were less likely to have been smokers at the time of the trauma (16.1% vs. 46.8%; p = 0.006). Regarding injury patterns, comparing PanSens with MDRO there was no difference with respect to ISS (26.2 ± 1.8 vs. 27.3 ± 2.4; p = 0.54), rates of rib fractures (49.3% vs. 46.2%; p = 0.97), or rates of head injury (71.6% vs. 69.1%; p = 0.98; Table 4). Furthermore, there was no association between abdominal or specifically intestinal injury and development of either gram-negative infection or of an MDRO organism. Among the eight patients who sustained an intestinal injury, three had a gram-negative infection, all three of which were due to Haemophilus influenzae. Regarding outcomes after development of pneumonia, there was no difference in rates of tracheostomy (69.2% vs. 70.1%; p = 0.87), need for long-term rehabilitation (70.1% vs. 73.1%; p = 0.98) or overall hospital length of stay (35.4 ± 4.3 vs. 39.2 ± 3.9 days; p = 0.62). There was no difference in mortality between PanSens versus MDRO patients (14.5% vs. 22.5%; p = 0.4; Table 5).

Table 4.

Characteristics between Patients with Pan-Sensitive versus Multi-Drug–Resistant Organism Pneumonia

	PanSens	MDRO	P
	n = 62	n = 31	P
Age	47.7 ± 2.3	53 ± 3.5	0.22
Male gender	76.1%	84.6%	0.54
Time to pneumonia	6 (4–8)	7 (5–11)	0.04
(median days [IQR])	6 (4–8)	7 (5–11)	0.04

ISS	26.2 ± 1.8	27.3 ± 2.4	0.36
Rib fractures	49.3%	46.2%	0.97
Head injury	59.7%	61.3%	0.98

COPD	11.3%	9.6%	0.9
Diabetes	11.3%	16.1%	0.5
Smoker	46.8%	16.1%	0.006

PanSens = pan-sensitive; MDRO = multi-drug–resistant organisms; IQR = interquartile range; ISS = Injury Severity Score; COPD = chronic obstructive pulmonary disease.

Table 5.

Outcomes Differences between Pan-Sensitive versus Multi-Drug–Resistant Infections

	PanSens	MDRO	P
	n = 62	n = 31	P
Outcomes
Need for tracheostomy	69.2%	70.1%	0.87
Mortality	14.5%	22.5%	0.4
Need for long-term rehabilitation	70.1%	73.1%	0.98
Length of stay (d)	35.4 ± 4.3	39.2 ± 3.9	0.62

PanSens = pan-sensitive; MDRO = multi-drug–resistant organisms.

Discussion

Trauma remains a leading cause of death and of long-term morbidity in survivors. Trauma-related outcomes are related directly to the degree of injury. Furthermore, specific injury patterns such as head or thoracic trauma impose profound long-term burden. This is the first report addressing the impact of drug-resistant versus PanSens micro-organisms among trauma patients. We demonstrated that trauma patients requiring ICU level of care were older, more injured, and had a greater burden of pre-trauma medical comorbidities. Among ICU patients, the development of pneumonia was associated with greater injury burden, higher ISS, higher rates of rib fractures, and a greater median number of ribs that were fractured. Our rib fracture protocol focuses on number of rib fractures as a central criterion for admission to the ICU in an effort to decrease the risk of trauma mortality [10]. Within the geriatric population, this protocol focused mostly on prevention strategies to minimize the risk of complications such as pneumonia, rather than utilizing the ICU resources as a rescue after the occurrence of pneumonia. However, the development of pneumonia was not associated with differences in pre-trauma medical comorbidities, except for higher rates of smoking among patients who developed pneumonia. Contrary to the effect of injury profile, the nature of the infecting micro-organism—MDRO versus PanSens—was neither associated with injury pattern nor affected trauma-related outcomes. Interestingly, the rate of smoking was higher among patients with PanSens pneumonia.

Despite a blunting of the trimodal distribution of trauma-related deaths [11], delayed deaths and adverse outcomes are driven by infectious complications and end organ failure. These events continue to impose considerable short- and long-term burdens on trauma patients. Specifically, the occurrence of pneumonia is associated with increased need for ICU levels of care, mechanical ventilation, increased mortality, and among survivors, long-term care post-discharge. The impact of pneumonia among trauma patients has changed over the past three decades [12,13]. Magnotti et al. [13] reported that patients die with rather than die from pneumonia, describing trauma-related pneumonia as an epiphenomenon. In keeping with our data, they also noted that it was the degree of severity of the traumatic injuries contributed to death and that the development of an infection is more of a marker of this severity of the injury. Cook et al. [12] also noted that the occurrence of pneumonia had a markedly lower effect on mortality compared with pneumonia occurring in non-trauma patients. They identified that trauma was a substantial for the development of pneumonia, with increasing level of severity of injuries being the greatest risk factor. Interestingly, Cook et al. [12] also noted that trauma patients were more likely to have undergone BAL and microbiology-directed pneumonia care and that gram-negative organisms were more likely to occur in trauma versus non-trauma patients.

Resistance to antimicrobial agents is an increasing public health concern, occurring within both gram-positive and gram-negative organisms. Drug-resistant infections were initially considered a major concern among hospitalized patients, however, over the past 20 years rates of drug resistance have risen dramatically among outpatients. Since MRSA was first described in 1960, MRSA-related nosocomial infections have increased steadily among critically ill patients and within intensive care units [3,7,14,15]. The increased rates of MDRO infections have been most noted to occur within, and have greatest impact upon, critically ill patients [5]. In keeping with our observations, Staphylococcus aureus is noted to be one of the most commonly isolated bacterial pathogens [16]. Consequently, MRSA, both community- and hospital-acquired, is a major resistant pathogen, especially among organisms encountered by the surgeon [15].

Many of the initial reports implied that MDRO infections occurred among either critically ill or chronically ill patients as a direct result of exposure to multiple antimicrobial agents. This included the dramatic increase of drug-resistant Pseudomonas aeruginosa among patients with cystic fibrosis [17], recurrent MRSA infection among patients with COPD [18], or skin and soft tissue or patient with diabetic foot ulcers [19] as well as a spectrum of MDROs from within intra-abdominal infections [20]. Mera et al. [21] noted that rates of resistance among Staphylococcus aureus had risen from approximately 30% to more than 50% of Staphylococcus aureus isolates. Mechanisms potentially include selection pressure induced by antibiotic administration leading to the induction of genes allowing microbes to evade or directly counter antimicrobial agents [22].

Most reports implicating MDRO compared with PanSens organisms as a driver of worse outcomes have focused predominantly on non-surgical patients. In a large meta-analysis of the effect of infection with MRSA versus MSSA bacteremia on mortality among non-surgical patients, Cosgrove et al. [6] noted a twofold increase in mortality among patients with MRSA infection. Although the increased mortality with MRSA remained after the authors attempted to adjust for variations in severity of illness and source of primary infection, considerable heterogeneity remained between MRSA and MSSA patients. Within a large meta-analysis of 16 studies reviewing the effect of extended spectrum β-lactamase (ESBL)-producing Enterobacteriaceae only one study adjusted for cofounders such as severity of illness [23]. Lautenbach et al. 24 noted that cases of EBSL-producing Escherichia coli and Klebsiella pneumoniae had greater need for central venous access, longer hospital length of stay and higher APACHE II scores, but in their analysis only adjusted for type of organism and site of infection and not level of illness. Gupta et al. [25] reported on infections among patients with cirrhosis. The most common infecting organisms was Escherichia coli. Among all infections, the authors noted that the presence of MDROs was an independent predictor of mortality. However, it was also associated with increased level of illness. Overall, a common finding notes that MDRO infections occur within patients with higher severity of illness, and a common belief reflects that microbial resistance is a reflection of host severity of illness and lack of physiologic or immunologic reserve [9,26].

Among surgical patients, several authors have suggested that the occurrence of resistant pathogens is an indication of severity of illness [9,12, 26]. In a study of gram-negative infections in surgical patients, Raymond et al. [26] described an association between occurrence of resistant gram-negative infection and degree of illness as denoted by APACHE II score. Although on univariable analysis gram-negative infection was associated with increased risk of mortality. When adjusted for age, site of infection, and medical comorbidities it was noted that resistant gram-negative infection was no longer associated with an increased risk of mortality [26]. Our observations are consistent with those of Raymond et al. [26], wherein we noted that older, sicker, more injured patients were at highest risk of developing infections. Furthermore, the occurrence of an infection imposed added risk of trauma-related adverse events including increased length of stay, however, once infection occurred the type of organism (MDRO or PanSens) did not affect outcomes. Barrasa-Villar et al. [2] noted that the presence of MDROs among non-surgical patients was associated with increased risk of mortality. However, no added effect was noted among patients with surgical site infections. Their study was limited by the differing sites of infection as well as specific organism, multi-drug–resistant Escherichia coli versus MRSA, between the mixed surgical and non-surgical populations. Rosenberg et al. [9] reviewed a large surgical ICU population with respect to the impact of MDROs on surgical outcomes. The two most frequently isolated resistant gram-negative organism was Staphylococcus aureus and most common resistant gram-negative was Pseudomonas aeruginosa. Despite increasing total numbers of resistant organisms over the period of the study, there was a marked decline in mortality rates. The authors ascribe this improved mortality to improved overall ICU care including time of appropriate culture directed antimicrobial therapy that may offset the impact of resistant organisms.

The greatest limitation of this study was the inability to document how well or how poorly controlled the medical comorbidities had been prior to the trauma. For example, we did not find any difference in the presence of diabetes among MDRO versus PanSens, however, we were unable to determine efficacy of pre-trauma diabetes management. Furthermore, although there was no difference in rates of COPD between groups, we were unable to assess degree of severity of COPD including prior hospital admissions for COPD exacerbations of prior episodes of pneumonia among patients with COPD. Another limitation is the single institution nature of this study. It is unclear how these data may translate across differing surgical populations. Furthermore, we are unable to address the issue of the impact of operative versus non-operative traumatic injuries. A future larger multi-center study may help address these issues.

Conclusion

Despite considerable advances in trauma care, many patients suffer long-term effects, many of which are driven by the development of secondary events or infectious complications. We have demonstrated that the degree of injury burden affected trauma-related outcomes. However, once pneumonia occurs, there is no additive effect of whether the infection was due to MSRO versus PanSens organisms.

Footnotes

Acknowledgment

All authors contributed to concept generation, data collection/analysis, manuscript generation, and final approval.

Funding Information

This work was supported, in part, by the National Institutes of Health [K08-GM110495] (to D.S.H.).

Author Disclosure Statement

No competing financial interests exist.

References

Heffernan

, Monaghan

, Thakkar

, et al. Failure to normalize lymphopenia following trauma is associated with increased mortality, independent of the leukocytosis pattern. Crit Care, 2012; 16:R12.

Barrasa-Villar

, Aibar-Remon

, Prieto-Andres

, et al. Impact on morbidity, mortality, and length of stay of hospital-acquired infections by resistant microorganisms. Clin Infect Dis, 2017; 65:644–652.

Fry

. The continued challenge of Staphylococcus aureus in the surgical patient. Am Surg, 2013; 79:1–10.

Evans

, Lefrak

, Lyman

, et al. Cost of gram-negative resistance. Crit Care Med, 2007; 35:89–95.

Cosgrove

, Carmeli

. The impact of antimicobial resistance on health and economic outcomes. Clin Infect Dis, 2003; 36:1433–1437.

Cosgrove

, Sakoulas

, Perencevich

, et al. Comparison of mortality assoicated with methicillin-resistant and methicillin-susceptible Staphylococcus aureus bacteremia: A meta-analysis. Clin Infect Dis, 2003; 36:53–59.

Friedman

, Temkin

, Carmeli

. The negative impact of antibiotic resistance. Clin Microbiol Infect, 2016; 22:416–422.

Tyas

, Marsh

, Oswald

, et al. Antibiotic resistance profiles of deep surgical site infections in hip hemiarthroplasty: Comparing low dose single antibiotic versus high dosedual antibiotic impregnated cement. J Bone Joint Infect, 2018; 3:123–129.

Rosenberger

, LaPar

, Sawyer

. Infections caused by multidrug resistant organisms are not associated with overall, all cause mortality in the surgical intensive care unit: The 20,000 foot view. J Am Coll Surg, 2012; 214:747–755.

10.

Connolly

, Monaghan

, Heffernan

, et al. Practice makes perfect: The evolution of blunt chest trauma. R I Med J, 2019; 102:27–29.

11.

Demetriades

, Kimbrell

, Salim

, et al. Trauma deaths in a mature urban trauma system: Is “trimodal” distribution a valid concept?. J Am Coll Surg, 2005; 201:343–348.

12.

Cook

, Norwood

, Berne

. Ventilator-associated pneumonia is more common and of less consequence in trauma patients compared with other critically ill patients. J Trauma, 2010; 69:1083–1091.

13.

Magnotti

, Croce

, FAbian

. Is ventilator-associated pneumonia in trauma patients an epiphenomenon or a cause of death?. Surg Infect, 2004; 5:237–242.

14.

Boucher

, Corey

. Epidemiology of methicillin-resistant Staphylococcus aureus. Clin Infect Dis, 2008; 46(Suppl 5):S344–S349.

15.

Piper

, Kaplan

. Antibiotic heterogeneity optimizes antimicrobial prescription and enables resistant pathogen control in the intensive care unit. Surg Infect, 2012; 13:194–202.

16.

David

, Daum

. Community associated methicillin-resistant Staphylococcus aureus: Epidemiology and clinical consequences of an emerging epidemic. Clin Microbiol Rev, 2010; 23:616–687.

17.

Ciofu

, Tolker-Nielsen

. Tolerance and resistance of pseudomonas aeruginosa biofilms to antimicrobial agents: How P. aeruginosa can escape antibiotics. Front Microbiol, 2019; 10:913.

18.

Miravitlles

, Anzueto

. Chronic respiratory infection in patients with chronic obstructive pulmonary disease: What is the role of antibiotics?. Int J Molec Sci, 2017; 18:E1344.

19.

Guillamet

, Kollef

. How to stratify patients at risk for resistant bugs in skin and soft tissue infections?. Curr Opin Infect Dis, 2016; 29:116–123.

20.

Labricciosa

, Sartelli

, Abbo

, et al. Epidemiology and risk factors for isolation of multi-drug–resistant organisms inpatients with complicated intra-abdominal infections. Surg Infect, 2018; 19:264–272.

21.

Mera

, Suaya

, Amrine-Madsen

, et al. Increasing role of staphylococcus aureus and community-acquired methicillin-resistant Staphylococcus aureus infections in the United States: A 10-year trend of replacement and expansion. Microb Drug Resist, 2011; 17:321–328.

22.

Waglechner

, Wright

. Antibiotic resistance: It's bad, but why isn't it worse?. BMC Biol, 2017; 15:84.

23.

Schwaber

, Carmeli

. Mortality and delay in effective therapy associated with extended spectrum beta-lactamase production in enterobacteriaceae bacteremia: A systematic review and meta-analsysis. J Antimicrob Chemother, 2007; 60:913–920.

24.

Lautenbach

, Patel

, Bilker

, et al. Extended-spectrum beta-lactamase producing Escherichia coli and Klebsiella pneumoniae: risk factors for infection and impact of resistance on outcomes. Clin Infect Dis, 2001; 32:1162–1171.

25.

Gupta

, Lochan

, Verma

, et al. Presence of multi-drug resistant infections at admission predict 28 day mortality in acute decompensation of cirrhosis. J Gastroenterol Hepatol, 2019; 10.

26.

Raymond

, Pelletier

, Crabtree

, et al. Impact of antibiotic-resistant Gram-negative bacilli infections on outcome in hospitalized patients. Crit Care Med, 2003; 31:1035–1041.