Do Patients with Pre-Existing Psychiatric Illness Have an Increased Risk of Infection after Injury?

Abstract

Background:

Trauma remains a leading cause of death and long term-morbidity. We have shown that patients who sustain traumatic injuries are at increased risk for the development of infectious complications. Psychiatric illnesses (PIs) are also noted to occur frequently among the general population. The presence of a PI has been shown to be a risk factor for the development of infections. Despite the prevalence of both traumatic injuries and psychiatric diseases, there are little data relating the impact of PI on the outcome of patients with trauma. We hypothesize that the presence of a PI will be associated with an increased risk of an infection developing after injury.

Patients and Methods:

This is a five year retrospective chart review of all admitted patients with trauma age 18 years and older. Patients with and without a major psychiatric illness were compared. Demographic data, mechanism of injury and Injury Severity Score (ISS) were reviewed. Co-morbidities included diabetes mellitus, obesity, pre-injury steroid use, and International Classification of Diseases, 9th edition, based psychiatric illness. All infections were diagnosed by microbiologic criteria (urinary tract infection [UTI], ventilator-associated pneumonia) or Centers for Disease Control and Prevention criteria for clinically evident infections (surgical site infection).

Results:

Of the 11,147 admitted trauma patients, 14.5% had a pre-injury PI diagnosis. The PI patients were older (61.5 ± 0.5 vs. 54.3; p < 0.001), more often female (56% vs. 39.1%; p < 0.001), and had no difference in blunt mechanism rates (88.4% vs. 89.9%; p = 0.06) or median ISS (9 vs. 9; p = 0.06). There was no difference between PI and non-PI patients in pre-injury diabetes mellitus (13.4% vs. 12.7%; p = 0.4), steroid use (2.5% vs. 1.9%; p = 0.1), but patients with PI were more likely to be obese (15.7% vs. 13.6%; p = 0.03). Patients with PI were more likely to have an infection develop (10.4% vs. 7.5%; p < 0.001). The most common infection in both groups was UTI (6.9% vs. 4.2%; p < 0.001). Compared with non-PI patients, adjusting for age, gender, ISS, diabetes mellitus, and obesity, patients with PI were more likely to have an infection develop (odds ratio 1.3, 95% confidence interval = 1.1–1.5)

Conclusions:

Patients with an underlying PI are at increased risk of having a UTI after traumatic injury. This study identifies a previously unknown independent risk factor for UTIs in patients with trauma. This stresses the need for increased awareness and attention to this vulnerable population.

Traumatic injuries are a leading cause of death and long-term morbidity. There are an estimated 2.3 million trauma-related admissions per year in the United States. Despite advances in trauma and intensive care unit (ICU) care with improvement in mortality rates [1], complications still occur and are often associated with increased long-term morbidity. Specifically, it is well recognized that traumatic injuries lead to immune dysfunction and immunoparalysis [2,3]. This immunosuppression is associated with the development of infectious complications in the trauma population. Further, pre-trauma medical co-morbidities, especially if poorly controlled, are also known to increase the risk of infection in patients with trauma. When infections do occur, they are associated with increased rates of secondary complications, increased length of stay (LOS), and ultimately increased risk of death [4 –6].

Psychiatric illnesses (PIs) are also prevalent with an estimated occurrence in 25% of the general U.S. adult population [7,8]. Patients with psychiatric illness have increased rates of co-morbidities including diabetes mellitus and cardiovascular disease [9 –12]. The presence of an associated PI often affects negatively the incidence, management, and outcomes of these medical co-morbidities. Despite controlling for these co-morbidities, it has been shown that patients with PI are at increased risk for the development of infections such as pneumonia, urinary tract infection (UTI), and appendiceal perforation [13]. Recent evidence has emerged that psychiatric co-morbidities act as a cofactor in infectious disease susceptibility and outcomes.

There has been little work that addresses the incidence and impact of PI on infectious complications among patients with trauma. We performed an analysis of the incidence of psychiatric disease among our trauma population and hypothesize that trauma patients with an associated psychiatric co-morbidity will have an increased risk of an infectious complication.

Patients and Methods

After Institutional Review Board approval, a retrospective chart review was conducted of prospectively collected data from our American College of Surgery verified level I trauma center registry. All admitted trauma patients 18 years of age or greater from January 2010 to January 2015 were reviewed. Exclusion criteria were those who were dead on arrival, were pronounced dead in the trauma bay, or who died in the operating room. The registry data were reviewed for patients and injury characteristics, co-morbidities, Injury Severity Score (ISS), and complications. Pre-trauma co-morbidities were recorded in the trauma registry based on information obtained at the time of presentation as well as any new information gathered during the hospital stay.

Functionally dependent is defined in our trauma registry as the following: Inability of the patient to complete age appropriate activities of daily living (ADL) including bathing, feeding, dressing, toileting, and walking because of cognitive or physical limitations relating to a pre-existing medical condition without the help of another person to complete some or all ADL. Outcomes reviewed included occurrence of infections, hospital course, discharge disposition, and death.

The primary outcome studied was the occurrence of any infection. All infectious workups were directed on the discretion of the treating trauma team. Infections were either clinically evident such as an abscess, wound infection, or cellulitis, or were based on culture data. A diagnosis of UTI required a urine culture with greater than 100,000 colony-forming units (CFU)/mL. A diagnosis of pneumonia was considered if a patient scored five or more points on the Clinical Pulmonary Infection Score (CPIS). This prompted direct bronchoalveolar lavage (BAL), and the diagnosis of pneumonia was confirmed if the BAL culture grew greater than 10,000 CFU/mL in intubated patients or via clinical diagnosis in non-intubated patients. All other infections including catheter related blood stream infections (CRBSI) and surgical site infections (SSI) were diagnosed by Centers for Disease Control and Prevention (CDC) criteria and cultures.

Patients were divided into two groups comprising patients with a pre-trauma diagnosis of PI and those who did not have a PI (non-PI). Major PI is defined as “documentation of the presence of pre-injury major depressive disorder, bipolar disorder, schizophrenia, anxiety/panic disorder, borderline or antisocial personality disorder, and/ or adjustment disorder.” Substance use disorders were not specifically included in the psychiatric illness group, and patients with substance use disorders existed in both groups.

Categoric variables were analyzed using the chi-square test, and continuous variables were analyzed using the Student unpaired samples t-test. A p value <0.05 was considered significant. To identify risk factors for infection after injury, simple and multivariable regression analysis was performed, and an odds ratio (OR) with 95% confidence intervals (CI) was calculated. Conditions in the multivariable analysis were those significant in the simple regression as well as factors deemed clinically significant. SPSS version 2.0 was used for the analysis.

Results

A total of 11,147 patients 18 years of age and older were admitted after traumatic injuries over the five-year period of review. There were 14.5% (1625 patients) who had a pre-injury diagnosis of PI and 9522 patients who were non-PI. Those with a PI were significantly older (61.5 ± 0.5 vs. 54.3 ± 0.2 y; p < 0.01), were less likely male (44% vs. 61%; p < 0.01) and had a slightly lower ISS (9.2 ± 0.2 vs. 9.6 ± 0.1; p = 0.049). The patients with PI had higher rates of penetrating injuries (12.3% vs. 9.5%; p < 0.01), self-inflicted injuries (10.3% vs. 1.2%; p < 0.01), and falls (58.1% vs. 47.2%; p < 0.01) but lower rates of motor vehicle collisions (17.9% vs. 29.2%; p < 0.01) and assaults (7.5% vs. 10.5%; p < 0.01) (Table 1).

Table 1.

Comparisons of Characteristics between Patients with and without Pre-Trauma Psychiatric Illness

	Psychiatric	Non-psychiatric
	N = 1625	N = 9522	p
Demographics
Age (years)	61.5 ± 0.6	54.3 ± 0.3	<0.01
Male gender (%)	44%	61%	<0.01
ISS (0–75)	9.2 ± 0.2	9.6 ± 0.1	0.049
Mechanism
Blunt	86.7%	87.6%	NS
Penetrating	11%	9.5%	NS
Burn	2.3%	2.9%	NS

ISS = injury severity score; MVC = motor vehicle crash.

Patients with PI were noted to have higher rates of obesity (15.7% vs. 13.6%; p = 0.02), respiratory disease (15% vs. 8.4%; p < 0.01), and congestive heart failure (8.3% vs. 5.6%; p < 0.01). Patients with PI were noted to be more often functionally dependent (21% vs. 9.1%; p < 0.01). Notably, there was no difference between PI and non-PI with respect to diabetes mellitus, steroid use, or smoking status (Table 2).

Table 2.

Rates of Medical Co-morbidities

	Psychiatric	Non-psychiatric
	n = 1625	n = 9522	p
Co-morbidity
Diabetes mellitus	13.5%	12.7%	N/S
Alcoholism and drug use disorder	18%	8.6%	<0.01
Obesity	15.7%	13.6%	0.02
Respiratory disease	15%	8.4%	<0.01
Functionally dependent	21%	9.1%	<0.01
Congestive heart failure	8.3%	5.6%	<0.01
Steroid use	2.5%	1.9%	N/S
Cirrhosis	1.6%	1.3%	N/S
Current smoker	25.3%	24.2%	N/S

The overall infection rate was significantly higher in patients with PI (10.3% vs. 7.5%; p < 0.01). Among all patients, both PI and non-PI, in the study, UTIs occurred most frequently, followed by pneumonia. In comparison, UTIs occurred more frequently in patients with PI (6.9% vs. 4.2%; p < 0.01) (Table 3).

Table 3.

Infectious Outcomes and Hospital Course

	Psychiatric	Non-psychiatric
	n = 1625	n = 9522	p
Infection
Catheter related BSI	0.1%	0.2%	N/S
SSI	1.5%	1.3%	N/S
PNA	2.6%	2.9%	N/S
UTI	6.9%	4.2%	<0.01
Sepsis	0.5%	0.6%	N/S
Infection rate	10.3%	7.5%	<0.01
Hospital course
Required operative intervention	36.9%	43.8%	<0.001
Hospital LOS	8 ± 0.4	7.5 ± 0.1	N/S
ICU stay	27%	29%	N/S
ICU LOS	6.4 ± 0.3	6.2 ± 0.2	N/S
Drug and/or alcohol withdrawal	7%	4%	<0.01
Death	4.3%	4.7%	N/S
Disposition
Home	45.7%	63%	<0.01
Facility	50%	32.3%	<0.01

BSI = blood stream infection; SSI = surgical site infection; PNA = pneumonia; UTI = urinary tract infection; LOS = length of stay; ICU = intensive care unit.

There were no differences between the groups with respect to hospital LOS, need for the ICU, or ICU LOS. Patients with PI were more likely to sustain substance withdrawal during the hospital stay (7% vs. 4%; p < 0.01). Patients with PI were less likely to be discharged to home (45.7% vs. 63%; p < 0.01) and more likely to be discharged to a care facility (50%; vs. 32%; p < 0.01). Overall, there was no difference in deaths between groups (4.3% vs. 4.7%; p = 0.52) (Table 3).

Across all patients with and without PI, simple logistic regression analyses identified age, male gender, ISS, and diabetes mellitus as being associated with development of an infection after admission. Obesity was not associated with increased odds of infection. Multiple logistic regression analysis adjusting for age, gender, ISS, diabetes mellitus, and obesity demonstrated that the presence of a pre-trauma PI was significantly associated with an increased risk for the development of an infection (OR 1.3; 95% CI = 1.1–1.5) (Table 4).

Table 4.

Univariable and Multivariable Analysis for Infection

Univariable analysis	Odds ratio	95% CI
Age (y)	1.02	1.01–1.02
Male gender	1.3	1.1–1.5
ISS	1.08	1.07–1.09
Diabetes mellitus	1.6	1.3–1.9
Obesity	1.1	0.93–1.4
Psychiatric illness	1.45	1.2–1.7
Multivariable analysis
Psychiatric illness	1.3	1.1–1.5

CI = confidence interval; ISS = Injury Severity Score.

Adjusting for age, gender, ISS, diabetes mellitus, and obesity, psychiatric illness was independently associated with development of an infection.

Discussion

Despite the significant prevalence of PI in the general population, there has been little investigation into the impact of PI on persons who sustain traumatic injuries. It has been shown that PI increases the risk of being hospitalized for an infection [13]. In this review of more than 11,000 adult patients with trauma, we have demonstrated that despite controlling for factors known to increase the risk of development of an infection, the presence of a pre-trauma PI is an independent risk factor for the development of a UTI among patients with trauma. Patients with PI were not shown to be at increased risk for development of other infections. Hospital-acquired infections, specifically UTIs, have been deemed a marker of the quality of acute adult trauma care [4], and these infections have severe implications with regard to healthcare costs [5,6]. More emphasis is being placed on preventing health-care–acquired infections. It is then important to recognize potential risk factors to guide practice changes such as early extubation, central line removal, and urinary catheter removal to reduce their incidence.

Urinary tract infections were the most common infection in our study and occurred significantly more frequently in those with a pre-admission PI. This offers a target for improvement in care. Zielinski et al. [14] recommends a screening urinalysis and urine culture at admission for elderly patients with trauma as a cost-effective measure based on their prospective observational clinical trial. The UTIs have been associated with changes in mental status [15] and have been found in up to 20% of geri-psychiatric admissions [16,17]. As one of the most common unrecognized medical conditions on admission to a geriatric psychiatry unit, routine screening of elderly patients with trauma for UTI seems prudent.

This increased rate of UTIs in our PI population affirms the importance of reviewing indications for initial urethral catheter placement and strategies for early catheter removal and decannulation of other tubes, lines, and drains. With UTI noted to be the most common hospital-acquired infection, increased attention has been directed to reduce the incidence especially over the last decade. In 2008, Medicare selected catheter-associated UTI as a condition for which treatment would not be reimbursed. Since then, some authors have even described catheter-associated UTI to be a “never event.” The UTI risk is markedly reduced if a catheter is never placed, so it should be uniform practice to review guidelines such as the CDC's 2009 list of indications for indwelling urethral catheter use. Numerous articles have also shown decreased incidence of UTI when urinary retention protocols are used. This decreased incidence is likely related to limiting the number of catheters that are placed [28]. When a catheter is required, it is then important to have daily “stop reminders” or signals in the electronic record to prompt physicians and nurses to review the necessity of the catheter and remove it if possible.

Recently, urethral catheter duration and UTI rates have been shown to be decreased by nursing empowered protocols to review catheter necessity daily and remove the catheter without requiring physician orders [29]. Patients with PI who may be unable to advocate for certain elements of their care such as catheter and device placement and removal, it is especially important to have protocols in place.

Pneumonia is the infection most commonly associated with death in patients with trauma. We, however, did not show a difference in rates of pneumonia between the PI and non-PI groups. Patients with dementia are at an increased risk of both pneumonia development [18] as well as dying from pneumonia [19]. In a large cohort of surgical patients in a 2012 study, the presence of PI was associated with a 2.4-fold increase risk of post-operative pneumonia developing [20]. Ahmedani et al. [21] noted that patients with PI are at increased risk of needing re-admission within 30 days after inpatient management of illnesses such as pneumonia. The authors noted, however, that aggressive screening for PI as well as having a PI-targeted component to the discharge plan decreased the risk of re-admission for pneumonia.

Unlike these authors, we did not show a significant difference in pneumonia rates in patients with PI. In patients with trauma at our institution and more broadly, pneumonia most commonly occurs in mechanically ventilated patients in the ICU. Pneumonia is also more likely to result in death in critically ill patients. To address this, we have had a set of guidelines in the ICU for the last several years to reduce the incidence of pneumonia and specifically ventilator-associated pneumonia: Head of bed elevation, selective oral decontamination, daily awakening trials, spontaneous breathing trials. The common goals of measures such as awakening and spontaneous breathing trials are to minimize duration of mechanical ventilation through early extubation. It is likely that patients in the ICU, whether they have a PI diagnosis or not, now receive more uniform care based on these protocols that promote early extubation, oral care, and head of bed elevation. In reviewing these older studies that showed differences in pneumonia rates, it is unclear whether such protocols were in place.

Ramifications of hospital-acquired pneumonia—namely death and increased ICU LOS—are likely more often considered than those of UTI where death is less likely. This may affect protocol adherence and lead to greater priority being given to early extubation compared with urethral catheter removal and partly explain why we have found a difference in UTI rates but not pneumonia rates in patients with PI. We, however, did not show an increase in ICU or hospital LOS in patients with PI, which may be because of the differences in infection rates being limited to UTIs.

Similar to previous reviews, we found that patients with a PI have higher rates of co-morbidities than the general population [9 –12,22]. Despite the higher incidence of co-morbidities, even when controlling for age, obesity, and diabetes mellitus, having a PI was still associated with increased odds of infection. Our data, however, are limited by a lack of a definable measure of how well these medical co-morbidities were controlled. It has been speculated that persons with PI tend to display overall worse control and long-term management of other medical co-morbidities.

Lustman et al. [23], in a meta-analysis, showed depression to be associated with higher HbA1c levels in patients with diabetes mellitus. Further, patients with PI and diabetes mellitus also have been shown to have less access to care, poorer compliance to treatment, and worse overall glycemic control [24]. McIntyre et al. [25], in a review of medical co-morbidities in patients with bipolar disorder, noted an increased cause-specific death from cardiovascular and cerebrovascular disease in the bipolar population when compared with the general population. The authors contended that potential worse control of medical co-morbidities pre-admission in patients with PI may lead to worse hospital outcomes including the noted increased re-admission rates. Others have speculated that increased risk of infections has been attributed to increased smoking prevalence and other negative health habits among those with a PI [26]. In our patient population, however, we did not find a difference in the number of current smokers between the two groups. This may be, in part, because of the generally high prevalence of smoking in the trauma population [27].

Our study is limited by its retrospective nature. We were limited by the trauma registry's information about the psychiatric history of our patients and lacked specific psychiatric diagnoses, duration of the PI, information on treatment history, and whether patients with PI received frequent psychiatric follow-up. We also were unable to reliably determine psychiatric medications that were being used by patients at the time of injury. Many PIs resolve or remit, with or without treatment, and at this point it is difficult to know the impact or severity of the diagnosed illness at the time of the injury presentation. It is likely that PI that is poorly controlled—for example, untreated severe depression—may place patients in a greater state of physiologic stress at the time of injury and affect infection risk. Future work will be aimed at determining severity of PI on presentation and whether this relates to risk of infection after injury. This would lead potentially to possible early psychiatric and pharmacologic interventions aimed at reducing this risk.

Conclusion

Psychiatric illness is a common pre-injury co-morbidity and is associated with more frequent UTIs after admission for injury. The impact of pre-existing PI in patients with trauma warrants future, prospective study. Future work will seek to further identify the role that PI has in the multifactorial association that has been shown between PI and occurrence of infections. These data add to the growing body of literature implicating complex relationships between mental health and outcomes with medical and surgical illness.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

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.

Marik

, Flemmer

. The immune response to surgery and trauma: Implications for treatment. J Trauma Acute Care Surg, 2012; 73:801–808.

Ottinger

, Monaghan

, Gravenstein

, et al. The geriatric cytokine response to trauma: Time to consider a new threshold. Surg Infect (Larchmt), 2014; 15:800–805.

Moore

, Lauzier

, Stelfox

, et al. Complications to evaluate adult trauma care: An expert consensus study. J Trauma Acute Care Surg, 2014; 77:322–329.

Rahmqvist

, Samuelsson

, Bastami

, Rutberg

. Direct health care costs and length of hospital stay related to health care-acquired infections in adult patients based on point prevalence measurements. Am J Infect Control, 2016; 44:500–506.

Roberts

, Scott

2nd , Hota

, et al. Costs attributable to healthcare-acquired infection in hospitalized adults and a comparison of economic methods. Med Care. 2010; 48:1026–1035.

Simeone

, Ward

, Rotella

, et al. An evaluation of variation in published estimates of schizophrenia prevalence from 1990–2013: A systematic literature review. BMC Psychiatry, 2015; 15:193.

Rund

, Demler

, et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry, 2005; 62:593–602.

Chouinard

, Pingali

, Chouinard

, et al. Factors associated with overweight and obesity in schizophrenia, schizoaffective and bipolar disorders. Psychiatry Res, 2016; 237:304–310.

10.

Suvisaari

, Keinanen

, Eskelinen

, Mantere

. Diabetes and schizophrenia. Curr Diab Rep, 2016; 16:16.

11.

Hsu

, Chien

, Lin

, et al. Increased risk of chronic obstructive pulmonary disease in patients with schizophrenia: A population based study. Psychosomatics, 2013; 54:345–351.

12.

Olfson

, Gerhard

, Huang

, et al. Premature mortality among adults with schizophrenia in the United States. JAMA Psychiatry, 2015; 72:1172–1181.

13.

Davydow

, Ribe

, Pedersen

, et al. Serious mental illness and risk for hospitalizations and rehospitalizations for ambulatory care-sensitive conditions in Denmark: A national population-based cohort study. Med Care, 2016; 54:90–97.

14.

Zielinski

, Kuntz

, Polites

, et al. A prospective analysis of urinary tract infections among elderly trauma patients. J Trauma Acute Care Surg, 2015; 79:638–642.

15.

Juthani-Mehta

, Quagliarello

, Perrelli

, et al. Clinical features to identify urinary tract infection in nursing home residents: A cohort study. J Am Geriatr Soc, 2009; 57:963–970.

16.

Manepalli

, Grossberg

, Mueller

. Prevalence of delirium and urinary tract infection in a psychogeriatric unit. J Geriatr Psychiatry Neurol, 1990; 3:198–202.

17.

Woo

, Daly

, Allen

, et al. Unrecognized medical disorders in older psychiatric inpatients in a senior behavioral health unit in a univeristy hospital. J Geriatric Psychiatry Neurol, 2003; 16:121–125.

18.

Brunnstrom

, Englund

. Cause of death in patients with dementia disorders. Eur J Neurol, 2009; 16:488–492.

19.

Keene

, Hope

, Fairburn

, Jacoby

. Death and dementia. Int J Geriatr Psychiatry, 2001; 16:969–974.

20.

, Liao

, Chang

, et al. Postoperative adverse outcomes in surgical patients with dementia: A retrospective cohort study. World J Surg, 2012; 36:2051–2058.

21.

Ahmedani

, Solberg

, Copeland

, et al. Psychiatric comorbidity and 30-day readmissions after hospitalization for heart failure, AMI, and pneumonia. Psychiatr Serv, 2015; 66:134–140.

22.

Kim

, Woo

, Jung

, et al. Positive association between serious psychiatric outcomes and complications of diabetes mellitus in patients with depressive disorders. Int J Psychiatry Med, 2015; 50:131–146.

23.

Lustman

, Anderson

, Freedland

, et al. Depression and poor glycemic control: A meta-analysis review of the literature. Diabetes Care, 2000; 23:934–942.

24.

Gonzalez

, Safren

, Cagliero

, et al. Depression, self-care, and medication adherence in type 2 diabetes: Relationships across the full range of symptom severity. Diabetes Care, 2007; 30:2222–2227.

25.

McIntyre

, Soczynska

, Beyer

, et al. Medical comorbidity in bipolar disorder: Re-prioritizing unmet needs. Curr Opin Psychiatry, 2007; 20:406–416.

26.

Lasser

, Boyd

, Woolhandler

, et al. Smoking and mental illness: A population-based prevalence study. JAMA, 2000; 284:2606–2610.

27.

Ferro

, Goslar

, Romanovsky

, Petersen

. Smoking in trauma patients: The effects on the incidence of sepsis, respiratory failure, organ failure, and mortality. J Trauma, 2010; 69:308–312.

28.

Meddings

, Rogers

, Krein

, et al. Reducing unnecessary urinary catheter use and other strategies to prevent catheter-associated urinary tract infection: An integrative review. BMJ Qual Saf, 2014; 23:277–289.

29.

Leis

, Corpus

, Rahmani

, et al. Medical directive for urinary catheter removal by nurses on general medical wards. JAMA Intern Med, 2016; 176:113–115.