Asymptomatic Urinary Tract Infection Treatment in Adults Admitted to Inpatient Psychiatry: A Retrospective COHORT study

Abstract

Introduction: Treatment of asymptomatic bacteriuria remains prevalent despite recommendations against treatment in most patient populations. Rates of asymptomatic treatment of urinary tract infection (UTI) has not been thoroughly evaluated within the inpatient psychiatry population. The objective of this study is to describe the rate of antibiotic use for the treatment of asymptomatic UTI in psychiatric inpatients and investigate factors contributing to overuse. Methods: This IRB approved retrospective cohort study evaluated adults admitted to inpatient psychiatry from May 1, 2021 to May 1, 2022 that received an antibiotic for UTI. The primary outcome assessed the rate of asymptomatic treatment, defined as treatment without urinary symptoms. Secondary outcomes evaluated most frequently prescribed antibiotics, determined the impact of altered mental status (AMS) on treatment, and correlated the incidence of UTI treatment with primary psychiatric disorder. Results: One hundred nine patients were identified and 61 were included for analysis. The rate of asymptomatic treatment for UTI was 84%. The most prescribed antibiotic was nitrofurantoin (48%). All patients with AMS (23%) were asymptomatic. Altered mental status did not significantly impact the rate of empiric treatment (P = .098). Primary psychiatric disorder did not significantly impact rate of empiric treatment for UTI (P = .696). Common disorders in this population were depression, schizophrenia, and bipolar disorder with rates of asymptomatic treatment of 79% (n = 19), 87% (n = 13), and 78% (n = 7), respectively. Discussion: Frequent asymptomatic treatment of UTI was identified in this inpatient psychiatry population. These results emphasize the need for antibiotic monitoring and stewardship in this setting.

Introduction

Treatment of asymptomatic bacteriuria (ASB) remains prevalent despite 2005 Infectious Disease Society of America (IDSA) recommendations against screening and treatment in most patient populations. ¹ A 2019 update to this guideline expanded recommendations against screening and treatment for various populations including older adults with delirium. ASB is defined as bacterial growth ≥ 10⁵ colony-forming units/mL in the absence of signs or symptoms attributable to urinary tract infection (UTI). ² A 2017 meta-analysis evaluating treatment of ASB across inpatient and outpatient settings reported inappropriate treatment at a rate of 45%. Factors that increase risk of inappropriate treatment include the presence of gram-negative isolates, female sex, positive nitrites, and pyuria. ³

Differentiation between ASB and UTI is dependent on presence of urinary symptoms. If patients cannot accurately or reliably report symptoms, the decision to treat can be challenging. This is a common dilemma in psychiatry as patients diagnosed with severe psychiatric disorders can frequently experience cognitive impairment. Rates of cognitive impairment may be up to 60% in older adults with severe psychiatric disorders. ⁴ Among a predominantly younger adult population with major depressive disorder, bipolar disorder, and schizophrenia, rates of cognitive impairment were between 58% and 84%. ⁵ Screening and treatment for ASB is not recommended in older adults with cognitive impairment and acute mental status change due to lack of evidence of benefit and potential adverse effects. ² With routine collection of urinalyses, overinterpretation of lab results, and lack of reliable symptom assessment, patients with psychiatric disorders are commonly treated for UTI despite absence of symptoms. ⁴ Inappropriate antibiotic prescribing contributes to increased healthcare cost, medication-related side effects, and antibiotic resistance.^6-8 More specific to this population, psychiatric side effects such as psychosis has been reported with antibiotics used for UTI including beta-lactams, fluoroquinolones, and sulfamethoxazole-trimethoprim. ⁹

The frequency of inappropriate treatment for UTI has not been thoroughly evaluated in the inpatient psychiatric population. A 2014 study reported that 61% of antibiotics prescribed for UTI were for asymptomatic patients at an inpatient psychiatric hospital. ¹⁰ A 2021 study reported 80% of geriatric psychiatry inpatients were over-diagnosed with UTI. ¹¹ This study aims to report the incidence of asymptomatic treatment initiation for UTI within inpatient psychiatry at a single institution and investigate factors that may contribute to antibiotic overuse.

Methods

This is an IRB approved retrospective cohort study of adults admitted to inpatient psychiatry from May 1, 2021 to May 1, 2022. Inpatient psychiatry at this institution includes 5 units which can be categorized as 2 acute, 2 general/mood disorder, and one geriatric. Patients were identified by having a resulted urinalysis and one of the following antibiotics administered within the first 5 days of admission: amoxicillin/clavulanate, ampicillin, ampicillin/sulbactam, cefepime, ceftriaxone, cephalexin, ciprofloxacin, fosfomycin, levofloxacin, meropenem, nitrofurantoin, piperacillin/tazobactam, and sulfamethoxazole/trimethoprim. If a patient was prescribed more than one antibiotic during the encounter, only the antibiotic prescribed first was used for data analysis. Patients were excluded if they met any of the following criteria: pregnancy, chronic urinary catheter use, plans to undergo a surgical urologic procedure, treatment for infection other than a UTI within first 5 days of admission, and those on study antibiotics prior to admission.

Pregnancy was defined by positive hCG. Planned urologic procedure was assessed by reviewing urology notes for procedures within 30 days after patient presentation. Treatment for infection other than UTI was evaluated by reviewing clinician notes at the time of prescribing. Use of study antibiotics prior to admission was determined by review of emergency department and psychiatry history and physical notes. Patients were excluded if antibiotic treatment was started on a medical unit before transfer to inpatient psychiatry. Duplicate antibiotic orders for the same patient during the same encounter were screened and removed as duplicates. Duplicate antibiotic orders in the medical record may result from changing the order such modifying the original order.

The primary outcome was to assess the rate of asymptomatic treatment of UTI for patients admitted to inpatient psychiatry. Asymptomatic treatment of UTI was defined as antibiotic treatment without presence of documented urinary symptoms. Secondary outcomes were to evaluate most frequently prescribed antibiotics, determine the impact of altered mental status (AMS) on treatment, and correlate the incidence of UTI treatment with primary psychiatric disorder based on DSM-V diagnostic category (Table 1).

Table 1.

Psychiatric disorder categories. ¹²

• Anxiety disorder • Bipolar and related disorder • Depressive disorder • Impulse control and conduct disorder • Neurocognitive disorder• Neurodevelopmental disorder • Obsessive-compulsive and related disorder• Other disorder • Personality disorder• Schizophrenia spectrum and other psychotic disorder• Sleep-wake disorder • Substance related and addictive disorder • Trauma and stressor related disorder

Presence of urinary symptoms was determined by clinician documentation suggestive of UTI such as urinary frequency, urinary urgency, dysuria, or costovertebral tenderness. ¹ Symptoms were collected from Emergency Medicine, Psychiatry, and Internal Medicine notes including but not limited to Review of Systems. For psychiatry notes, medical Review of Systems is only included in the history and physical notes. AMS was determined by inability to complete Review of Systems or not alert and oriented to person, place, and time by assessing ED and psychiatry history and physical notes. Psychiatric disorder was determined based on the primary diagnosis listed in the discharge summary.

Data was collected and managed using REDCap electronic data capture tools hosted by Cleveland Clinic.^13,14 The following data was collected: date of birth, sex assigned at birth, date of admission, antibiotic regimen prior to admission, presence of altered mental status upon admission, presence of urinary symptoms, hCG, antibiotic, unit assignment, presence of urinary catheter, initial prescribing service of antibiotic, urinalysis order, planned urologic procedure, infection of non-urinary source, and primary psychiatric diagnosis at discharge. Although temperature was not a collected data point, patients with fever are not candidates for psychiatric admission and were subsequently not a part of this population.

Descriptive and inferential statistics were performed. Nominal variables are presented as frequency with proportion. Continuous variables are presented as median with range. Inferential statistics were performed for secondary outcomes with Fisher’s exact tests. Analysis was performed using Stata Statistical Software. ¹⁵ Data cannot be shared to protect the privacy of the individuals included in the study.

Results

One hundred nine patients met inclusion criteria, 48 were excluded, and the remaining 61 unique patients were evaluated (Figure 1). Four patients were excluded for meeting 2 exclusion criteria each. Eighty-nine percent of patients were female, n = 54, and median age on admission was 51 years, with ages ranging from 19 to 95 years (IQR 31-66).

Figure 1.

Identification, exclusion criteria, and included records. Four records met 2 exclusion criteria.

The rate of asymptomatic treatment of UTI for patients admitted to inpatient psychiatry was 84%, n = 51. The most frequently prescribed antibiotics were nitrofurantoin, cephalexin, and sulfamethoxazole/trimethoprim, with prescription rate of 48%, 28%, and 16% respectively (Figure 2). Altered mental status did not have a significant impact on the rate of empiric treatment for UTI (P = .098). Primary psychiatric disorder did not have a significant impact on rate of empiric treatment for UTI (P = .696) (Table 2). The rate of asymptomatic treatment on acute units was 89%, n = 27, general/mood disorder units 65%, n = 17, and geriatric unit 94%, n = 17. Antibiotic therapy was initiated by Emergency Medicine in 43 encounters, Internal Medicine in 13 encounters, and Psychiatry in 5 encounters. The rate of asymptomatic treatment initiation by Emergency Medicine was 91%, n = 39, by Internal Medicine was 69%, n = 9, and by Psychiatry was 60%, n = 3.

Figure 2.

Antibiotics prescribed for urinary tract infection.

Table 2.

Treatment rates based on symptoms, diagnosis, and altered mental status.

	Asymptomatic n = 51 n (%)	Symptomatic n = 10 n (%)	P-value
Primary Diagnosis
Neurocognitive Disorder	7 (100)	0 (0)	P = .696
Substance Related Disorder	2 (100)	0 (0)
Anxiety Disorder	1 (100)	0 (0)
Other Disorder	1 (100)	0 (0)
Schizophrenia Spectrum	13 (87)	2 (13)
Depressive Disorder	19 (79)	5 (21)
Bipolar Disorder	7 (78)	2 (22)
Trauma and Stressor Disorder	1 (50)	1 (50)
Altered Mental Status
Yes	14 (100)	0 (0)	P = .098
No	37 (79)	10 (21)

Discussion

The rate of antibiotic treatment for UTI in this study in patients without documented urinary symptoms is 23% higher than previously reported by Cogdill et al. ¹⁰ Differences in patient selection, sample size, symptom definition, and study population may have contributed to this increase. The previous study selected patients whose treatment was initiated on a psychiatric unit. It also included AMS as a sign of UTI for patients greater than 65 years of age. Similar patients were classified as asymptomatic in our study based on the 2019 update on screening and treatment recommendations for this population. ² The proportion of patients with primary diagnosis of depressive disorder was also higher in our cohort than previously studied, 39% versus 20%, respectively. ¹⁰

At the study institution, urinalysis reflexes to culture if urine microscopic has >10 WBCs/hpf. Urinalysis is required for patients admitted to psychiatry for medical clearance and may increase the likelihood patients are treated based on potential contamination or pyuria rather than reported symptoms. Patients were intentionally identified based on urinalysis instead of urine culture to account for patients empirically treated for UTI whose urinalysis may not have reflexed to culture. Because urine culture data was not assessed, it may not be accurate to compare to incidence of ASB as defined by IDSA guidelines.

Cephalexin was prescribed more frequently than sulfamethoxazole/trimethoprim despite being considered alternative therapy for the treatment of UTI. ¹⁶ This is consistent with local antibiotic resistance patterns and hospital antibiogram demonstrating higher susceptibility of E.coli to cephalexin than sulfamethoxazole/trimethoprim. Fluoroquinolones were used in only 2 encounters (3%), which aligns with the FDA warning from 2016 to use these agents for uncomplicated UTI when no other treatment options are available. ¹⁷ The highest incidence of antibiotic initiation occurred in the emergency department and highlights an opportunity for targeted education. It also underscores the importance of symptom re-evaluation upon admission or psychiatric stabilization to assess for antibiotic de-escalation.

Primary psychiatric disorders in this cohort are consistent with that of the typical population at the study institution. Amongst the most common psychiatric disorder subgroups there may be a trend that patients with schizophrenia are more commonly treated for asymptomatic UTI than patients with depression. However, the increased rate of AMS in this subgroup (31% in schizophrenia group and 0% in depression group) may account for this trend. All patients in the neurocognitive disorder subgroup had AMS and were empirically treated for UTI. It is not unexpected that all the patients in the AMS group were asymptomatic as patients would likely not be able to reliably report urinary symptoms. The geriatric unit had the highest rate of asymptomatic treatment, followed by acute and general adult population. This suggests that patients with neurocognitive disorders or serious mental illness may be more likely to receive empiric treatment than those with mood disorders.

Initiatives at other institutions have made strides toward reducing unnecessary antibiotic prescribing in psychiatric patients with communication barriers. Rarrick and Hebbard developed an innovative UTI treatment algorithm for psychiatric inpatients with communication barriers that reduced inappropriate antibiotic prescribing from 12.5% to 0% in adults with dementia, delirium, autism spectrum disorder, or intellectual disability.^18,19 Based on the results of our study, these stewardship directives may also benefit patients with serious mental illness.

Outcomes in this study are dependent on consistent and accurate provider documentation during the patient encounter. Omitting documentation of urinary symptoms would have classified the patient as asymptomatic and contributed to an overestimation of asymptomatic patients treated for UTI. Documentation of urinary symptoms was only collected by clinician notes and therefore may not have captured if symptoms were documented in notes by other members of the healthcare team. In clinical practice, direct communication with prescribers and nurses can assist with determining if antibiotic therapy is appropriate. Data was manually collected by a single study member to minimize variability in assessment between reviewers. A second study member validated the data by conducting a secondary audit of 25% of the population. The criteria used to define AMS was chosen to be as clear and objective as possible but resulted in a definition that may not have captured all patients with altered cognition. Systemic signs of infection such as fever was not assessed, as patients with temperature >100.4 F do not meet medical clearance parameters at the study institution. Additional research is needed to validate the results of this study on a larger scale. The results of this study were presented and discussed at local antimicrobial stewardship and quality committees.

Conclusion

These results underscore overtreatment of asymptomatic patients and highlights unique challenges in UTI diagnosis within patients admitted to inpatient psychiatry. It emphasizes the need for diligent antibiotic monitoring and identifies an opportunity for targeted antimicrobial stewardship efforts in both the emergency department and inpatient setting for this patient population.