Error bars indicate 95% CIs.
aLow-volume hospital was defined as being in the 10th percentile or below for asymptomatic bacteriuria cases across the collaborative.
eAppendix. Outcome Adjustment Variables
eFigure. Flow Diagram of Study Population
eTable 1. Top Ten Primary Admitting and Discharge Diagnoses Among Treated vs Not Treated Asymptomatic Bacteriuria Patients, N = 2733
eTable 2. Sensitivity Analyses for Duration of Hospital Stay after Urine Testing for Treatment vs No Treatment of Asymptomatic Bacteriuria, N = 2733
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Petty LA, Vaughn VM, Flanders SA, et al. Risk Factors and Outcomes Associated With Treatment of Asymptomatic Bacteriuria in Hospitalized Patients. JAMA Intern Med. 2019;179(11):1519–1527. doi:10.1001/jamainternmed.2019.2871
What are the clinical characteristics and outcomes associated with antibiotic treatment of hospitalized patients with asymptomatic bacteriuria?
In this cohort study of 2733 hospitalized adults with asymptomatic bacteriuria, 82.7% received inappropriate antibiotic treatment; patients who were older, had altered mental status, or had abnormal urinalysis results were more likely to receive antibiotics. Antibiotic treatment was associated with a 37% longer duration of hospitalization after urine testing without improved clinical outcomes.
Antibiotic treatment of asymptomatic bacteriuria in hospitalized patients appears to be common, may not be associated with improved clinical outcomes, and may be associated with longer duration of hospitalization after urine testing.
Treatment of asymptomatic bacteriuria (ASB) with antibiotics is a common factor in inappropriate antibiotic use, but risk factors and outcomes associated with treatment of ASB in hospitalized patients are not well defined.
To evaluate factors associated with treatment of ASB among hospitalized patients and the possible association between treatment and clinical outcomes.
Design, Setting, and Participants
A retrospective cohort study was conducted from January 1, 2016, through February 1, 2018, at 46 hospitals participating in the Michigan Hospital Medicine Safety Consortium. A total of 2733 hospitalized medical patients with ASB, defined as a positive urine culture without any documented signs or symptoms attributable to urinary tract infection, were included in the analysis.
One or more antibiotic dose for treatment of ASB.
Main Outcomes and Measures
Estimators of antibiotic treatment of ASB. Secondary outcomes included 30-day mortality, 30-day hospital readmission, 30-day emergency department visit, discharge to post–acute care settings, Clostridioides difficile infection (formerly known as Clostridium difficile) at 30 days, and duration of hospitalization after urine testing.
Of 2733 patients with ASB, 2138 were women (78.2%); median age was 77 years (interquartile range [IQR], 66-86 years). A total of 2259 patients (82.7%) were treated with antibiotics for a median of 7 days (IQR, 4-9 days). Factors associated with ASB treatment included older age (odds ratio [OR], 1.10 per 10-year increase; 95% CI, 1.02-1.18), dementia (OR, 1.57; 95% CI, 1.15-2.13), acutely altered mental status (OR, 1.93; 95% CI, 1.23-3.04), urinary incontinence (OR, 1.81; 95% CI, 1.36-2.41), leukocytosis (white blood cell count >10 000/μL) (OR, 1.55; 95% CI, 1.21-2.00), positive urinalysis (presence of leukocyte esterase or nitrite, or >5 white blood cells per high-power field) (OR, 2.83; 95% CI, 2.05-3.93), and urine culture with a bacterial colony count greater than 100 000 colony-forming units per high-power field (OR, 2.30; 95% CI, 1.83-2.91). Treatment of ASB was associated with longer duration of hospitalization after urine testing (4 vs 3 days; relative risk, 1.37; 95% CI, 1.28-1.47). No other differences in secondary outcomes were identified after propensity weighting.
Conclusions and Relevance
Hospitalized patients with ASB commonly receive inappropriate antibiotic therapy. Antibiotic treatment did not appear to be associated with improved outcomes; rather, treatment may be associated with longer duration of hospitalization after urine testing. To possibly reduce inappropriate antibiotic use, stewardship efforts should focus on improving urine testing practices and management strategies for elderly patients with altered mental status.
Urinary tract infections (UTIs) are the second most common infection among hospitalized patients.1 Accurate diagnosis requires a combination of relevant signs and symptoms, generally supported by a positive urine culture (UC). Bacterial growth in a UC without accompanying symptoms is known as asymptomatic bacteriuria (ASB).2,3 Asymptomatic bacteriuria is commonly misdiagnosed as UTI.4 However, unlike UTI, treatment of ASB with antibiotics does not improve outcomes in outpatient and long-term care settings except in pregnancy or patients undergoing invasive urologic procedures.2,3,5 Despite national guidelines recommending against antibiotic therapy for ASB, high treatment rates persist.6-9
Inappropriate use of antibiotics can result in adverse events,10 increased antibiotic resistance, and Clostridioides difficile infection (CDI; formerly known as Clostridium difficile), prompting hospitals and antimicrobial stewardship programs to adopt strategies to reduce inappropriate antibiotic use, in particular, ASB treatment. However, data to guide interventions are limited; few studies have assessed factors and outcomes associated with antibiotic treatment for ASB in hospitalized populations.6-9,11,12 Surveys assessing clinician knowledge regarding ASB in inpatients found that deficits are prevalent. Even when clinicians are knowledgeable, guideline-discordant practices persist owing to concern for adverse outcomes if ASB is not treated.13-15 Therefore, a large, multihospital registry inclusive of a diverse patient population was used to identify factors possibly associated with inappropriate treatment of ASB and evaluate the clinical outcomes of antibiotic treatment.
The Michigan Hospital Medicine Safety (HMS) Consortium comprises 46 hospitals with shared goals of improving the quality of care and decreasing adverse events among hospitalized medical patients.16-18 Participation in the HMS Consortium is voluntary, representing half (46 of 92) of Michigan hospitals and including a mix of academic and community, small (<200 bed) and large (≥200 bed) hospitals.
Because the purpose of the HMS Consortium is to measure and improve the quality of existing care practices, this project received a status designation of not regulated by the University of Michigan Medical School’s Institutional Review Board, which waives the requirement of informed consent.
From January 1, 2016, through February 1, 2018 (10 hospitals in 2016, expanded to 46 in 2017), trained abstractors retrospectively collected data from patients with a positive UC collected during hospitalization (defined as a UC with any bacterial growth and identified as abnormal by the hospital’s microbiology policy). The HMS Consortium procedures for data collection and quality assurance have been previously described.16-18 Deidentified data were collected from 90 days before admission until follow-up was terminated by a major complication (eg, death) or 30 days after discharge. Data abstracted from medical records included demographics, signs and symptoms of UTI, laboratory and radiographic findings, vital signs, antibiotic type and duration, and outcomes. Symptoms were collected from clinician and nursing documentation 3 days before and after UC collection. At 30 days after discharge, patients were contacted for additional outcome data by scripted telephone follow-up (up to 3 attempts). A standardized data dictionary and random audits at each hospital by quality coordinators (including S.B.) ensured data integrity.
Abstractors at each hospital consecutively screened patients 30 days after discharge and included the first patient daily with a positive UC.18 Patients were not eligible for inclusion if they met any of the following criteria: (1) age younger than 18 years; (2) pregnant; (3) urinary stent, nephrostomy, altered urinary tract anatomy, or urologic surgery during hospitalization; (4) intensive care unit admission within 3 days before or after UC; (5) entered hospice during hospitalization; (6) left against medical advice; (7) concomitant infection (documentation by physician of an additional bacterial infection during hospitalization except CDI); (8) active treatment and/or prophylaxis for UTI on admission; (9) solid organ or bone marrow transplant recipient; (10) HIV with CD4 count less than 200 cells/mm3; (11) neutropenia (absolute neutrophil count <0.5 cells/μL [to convert to ×109 per liter, multiply by 0.001]) on hospital day 1 or 2; (12) isolated candiduria; or (13) within the 30 days after discharge from index hospitalization already abstracted for that patient.
Patients with bacteriuria were categorized as having ASB if there was no documentation of signs or symptoms meeting diagnostic criteria for UTI per Infectious Diseases Society of America Guidelines and National Healthcare Safety Network definitions.2,3,19 Specifically, patients could not have one of the following documented signs or symptoms: dysuria, urinary frequency/urgency, suprapubic pain, fever (temperature ≥38 °C), costovertebral pain/tenderness, hematuria, and autonomic dysreflexia or increased spasticity in patients with spinal cord injury (eFigure in the Supplement). Patients with acute alterations in mental status (AMS), who often cannot communicate symptoms, were categorized as having ASB if they had none of the aforementioned signs or symptoms and no systemic signs of possible infection (peripheral white blood cell count >10 000 cells/μL [to convert to ×109 per liter, multiply by 0.001], systolic blood pressure <90 mm Hg, or ≥2 criteria for systemic inflammatory response syndrome).3,20,21 Patients were excluded from analysis if relevant data were missing.
The primary outcome was antibiotic treatment, defined as receiving at least 1 dose of an oral or intravenous antibiotic (assessed 3 days before UC through discharge, including discharge prescriptions). Metronidazole and/or oral vancomycin alone was not considered ASB treatment.
Variables included (1) demographics, (2) receipt (in prior 90 days) of antibiotic, (3) urinary catheter presence, (4) nonspecific signs or symptoms not consistent with UTI definition (including fatigue, change in urine color/sediment), (5) severity of illness, and (6) laboratory results (including elevated peripheral white blood cell count, urinalysis, and UC results).
Secondary outcomes assessed included discharge to post–acute care facility, development of CDI within 30 days, duration of hospitalization after urine testing, and 30-day mortality, readmission, and emergency department visit. The CDI events included laboratory diagnosis (positive C difficile polymerase chain reaction and/or glutamate dehydrogenase level with toxin enzyme immunoassay testing)22 occurring 48 hours after UC or new CDI diagnosis documented within 30 days post discharge. Hospital duration was assessed from the day urine testing was performed (urinalysis or UC).
Sensitivity analyses were performed to account for potential unmeasured confounders of duration of hospitalization after urine testing. First assessed was the outcome of ASB treatment on hospital duration only for patients with urine testing performed on hospital day 1. This sample limit addressed potential differences among patients with a later change prompting testing, which would be a possible marker for a complicated hospitalization. Additional analysis assessed duration of hospitalization for patients who had testing and antibiotics ordered on day 1.
Factors in antibiotic treatment of ASB were summarized using descriptive analyses. Unadjusted associations of factors with antibiotic treatment were assessed using logistic generalized estimating equation models accounting for hospital level clustering, with results expressed as odds ratios (ORs) for dichotomous variables and relative risks (RRs) for continuous variables with 95% CIs. The final multivariable model was determined by a stepwise selection procedure based on covariate contributions to the model fit as measured by the Schwarz criterion.23 Associations of secondary outcomes with antibiotic treatment of ASB were assessed using logistic generalized estimating equation models, inverse probability of treatment24 weighted by baseline covariates identified to be significant in the bivariate and/or multivariate analysis, and other factors potentially associated with the outcome (eAppendix in the Supplement).25-33 To account for variables with missing data, a 10-fold multiple imputation was completed.34 A 2-sided P value <.05 was used to indicate significance. All analyses were performed in SAS, version 9.4 (SAS Institute Inc).
Between January 1, 2016, and February 1, 2018, 7252 patients with a positive UC were identified among 46 hospitals. Of these, 2772 patients (38.2%) lacked documented urinary symptoms and were classified as having ASB (eFigure in the Supplement). Among 2733 patients with ASB and complete treatment data, 2259 individuals (82.7%) received antibiotics for a median of 7 days (interquartile range [IQR], 4-9) (Table 1). Most individuals (1933/2259 [85.6%]) were treated with 3 or more days of antibiotics. The median age was 77 years (IQR, 66-86); 2138 of 2733 patients (78.2%) were women. At the time of UC, 375 of 2733 patients (13.7%) had an indwelling urinary catheter, 1138 patients (41.6%) had chronic kidney disease, 1076 patients (39.4%) had diabetes, and 560 patients (20.5%) had dementia. A total of 2308 patients (84.4%) were admitted to the hospital for reasons other than UTI. Discharge diagnoses more frequently included UTI for patients who received ASB treatment compared with those who did not receive treatment (485/2259 [21.5%] vs 0/474) (eTable 1 in the Supplement). Almost all patients (2499 [94.1%]) had pyuria. The most common urine pathogens isolated included Escherichia coli (50.2%) and Klebsiella species (15.3%). Ceftriaxone (61.6%) was the most common initial antibiotic treatment; fluoroquinolones (33.2%) were the antibiotics most frequently prescribed at discharge (Table 1). The indication for UC was documented in 1565 of the cases (57.3%) (Table 1). Of documented indications, abnormal urinalysis results was the most common (666/1565 [42.6%]). The proportion of patients with ASB treated with antibiotics varied by hospital, with 35 of the 46 facilities (76.1%) treating ASB in 80% or more of the patients (Figure).
Variables associated with ASB treatment included patient characteristics, such as older age and acute AMS, and laboratory results, such as positive urinalysis (presence of leukocyte esterase or nitrite, or >5 white blood cells per high-power field). Severity of illness, measured by qSOFA (quick sequential organ failure assessment)35 score (328/2733 [12.0%]) or 2 or more systemic inflammatory response syndrome criteria (909/2733 [33.3%]), was not associated with increased treatment rates (Table 2).
On multivariable analysis, patient characteristics of older age (OR, 1.10 per 10-year increase; 95% CI, 1.02-1.18), acute AMS (OR, 1.93; 95% CI, 1.23-3.04), dementia (OR, 1.57; 95% CI, 1.15-2.13), and urinary incontinence (OR, 1.81; 95% CI, 1.36-2.41), as well as laboratory test results including a positive urinalysis (OR, 2.83; 95% CI, 2.05-3.93), E coli bacteriuria (OR, 1.42; 95% CI, 1.12-1.79), high bacteriuria colony counts (>100 000 colony-forming units per high-power field) (OR, 2.30; 95% CI, 1.83-2.91), and peripheral leukocytosis (white blood cell count >10 000 cells/μL) (OR, 1.55; 95% CI, 1.21-2.00), were associated with ASB treatment (Table 3).
Within 30 days of discharge, 2375 of 2733 patients (86.9%) had follow-up terminated by a major complication or were assessed by record review and/or telephone follow-up. According to data presented in Table 4 after adjustment, no differences were found in mortality, readmission, emergency department visit, or CDI within 30 days post discharge or discharge to post–acute care facility between patients with ASB who received vs did not receive antibiotics. Most urine testing (2068/2733 [75.7%]) was performed on day 1 of hospitalization. Duration of hospitalization after urine testing was 37% longer (median, 4 vs 3 days; RR, 1.37; 95% CI, 1.28-1.47) in patients with ASB who received vs did not receive antibiotic treatment.
Treatment of ASB remained associated with duration of hospitalization on sensitivity analysis. When including only patients who had urine testing performed on hospital day 1 (n = 2068), treatment of ASB remained associated with longer hospitalization (median, 5 vs 3 days without treatment; RR, 1.33; 95% CI, 1.22-1.46) (eTable 2 in the Supplement). When only patients who had urine testing performed and antibiotics started on hospital day 1 were included (n = 1810), ASB treatment remained associated with longer hospitalization (median, 4 vs 3 days without treatment; RR, 1.27; 95% CI, 1.15-1.40).
In this study of 2733 hospitalized patients with ASB across 46 hospitals, 80% or more received antibiotic treatment. Both patient and laboratory characteristics were associated with ASB treatment, including older age, acute AMS, dementia, and abnormal urinalysis results. In addition, treatment of ASB was not associated with differences in most outcomes but was associated with longer duration of hospitalization after urine testing. These findings support guideline recommendations against ASB treatment among hospitalized patients and suggest potential harm associated with treatment.
We found high treatment rates of ASB in hospitalized patients and identified patient characteristics associated with antibiotic treatment of ASB. Similar to findings in prior studies, older age and AMS were risk factors for ASB treatment.9,36 In addition, our study identified new characteristics associated with treatment: dementia and urinary incontinence. These factors, although independently associated with treatment, are more common with older age, placing elderly patients—who have the highest morbidity and mortality due to antibiotic-associated adverse events—at the highest risk of receiving inappropriate antibiotic therapy. Acute AMS and dementia are particularly challenging because patients with these conditions are often not able to reliably communicate symptoms, causing clinician hesitation in withholding treatment.12,37,38 From reports of experience in the long-term care setting, patients with bacteriuria and with acute AMS and no localizing UTI symptoms or signs of systemic infection may be observed without antibiotics for 24 to 48 hours while other potential sources of AMS (eg, dehydration, medication adverse effects) are addressed.37-39 Prior studies describing lower treatment rates (47%-62%) in hospitalized patients with ASB have differing or unspecified ASB definitions.7,40,41 Higher treatment rates (72%-77%) of ASB were described among institutionalized patients with dementia or studies excluding AMS as a UTI criterion.8,9,42,43 Our inclusion of this population (ASB with AMS) likely contributed to our observed higher treatment rates and enabled us to identify an association of dementia and AMS with increased ASB treatment. Our approach in patients with AMS is supported by the recently updated Infectious Diseases Society of America ASB guidelines, which addressed the impaired elderly patient with bacteriuria and delirium, recommending assessment for other causes and careful observation instead of antibiotics.3 This population would likely benefit from targeted stewardship interventions.
We also found ASB treatment to be more common in patients with certain laboratory test results. Despite common misperceptions, positive urinalysis or UC results do not define a UTI or necessitate antibiotic therapy.2,44-46 A negative urinalysis result makes a UTI unlikely, but a positive urinalysis does not diagnose infection given its poor positive predictive value.2,46,47 However, abnormal urinalysis was the top indication documented for UC and the strongest factor associated with treatment. These findings are consistent with prior studies that identified pyuria, abnormal urinalysis, gram-negative bacteriuria, and higher bacterial colony counts to be associated with ASB treatment.6,9,11 Furthermore, the rate of pyuria in this cohort was 91.4%, which is similar to the rate in a prior study in an elderly institutionalized population.45 Given the high prevalence of both pyuria and ASB in this population, reflex urine culturing (UC performed after abnormal urinalysis) may potentially further contribute to inappropriate antibiotic use in hospitalized elderly patients, as clinicians may be more likely to misinterpret an abnormal urinalysis linked to a positive UC.47 To decrease inappropriate testing and treatment of ASB, diagnostic stewardship interventions and education should emphasize the poor positive predictive value of abnormal urinalysis results and emphasize symptoms as the correct prompt for urine testing and UTI diagnosis.
In addition, we found no difference in mortality, readmission, emergency department visits, or CDI within 30 days in patients with ASB treated with antibiotics compared with those who received no treatment. Multiple randomized clinical trials demonstrated no improvement in outcomes with ASB treatment for certain populations (patients who are elderly, diabetic, or institutionalized), and these data have been extrapolated to the hospitalized patient.2 Data on clinical outcomes in the hospitalized population with ASB are limited.12,48,49 In addition to our study finding no clinical benefit, antibiotic treatment of ASB was associated with a 37% increase in duration of hospitalization after urine testing, suggesting potential harm and additional cost. Similarly, a recent study found that hospitalized patients with a UC on admission, irrespective of their admitting diagnosis, had longer hospital stays compared with patients with no UC.50 The association of antibiotic treatment with longer duration of hospitalization was robust in sensitivity analyses attempting to account for differences in patients with testing performed later during hospitalization. It is possible that ASB treatment may be associated with longer hospitalization because clinicians may delay discharge in stable patients while awaiting antibiotic susceptibility test results. Given that UTI is one of the most commonly treated and misdiagnosed conditions in US hospitals, a 1-day difference in length of stay would have a substantial effect on hospitalization-related costs.
Our study has limitations. First, because this was an observational retrospective study, we cannot prove causation. Second, ascertaining urinary symptoms is limited by the accuracy of documentation, so ASB frequency may have been overestimated. Third, by excluding concomitant infections, we may not have captured the full scope of ASB treatment in hospitalized patients. Fourth, adverse outcomes were infrequent, limiting our power to detect differences. However, given the study size, if a difference in outcomes existed, it would be small. Fifth, although we adjusted for potential confounding variables and completed sensitivity analyses for duration of hospitalization, residual confounding may still exist. Sixth, one-third of the patients had 2 or more systemic inflammatory response syndrome criteria (by definition, none had AMS or fever), and 328 of 2733 patients (12.0%) met qSOFA criteria. Patients with sepsis syndrome may warrant empirical use of antibiotics, but systemic inflammatory response syndrome criteria are neither sensitive nor specific for infection, and indiscriminate early antibiotic use may cause harm without proven benefit in less sick populations.51-54 However, the qSOFA score should prompt consideration of possible infection in patients with nonlocalizing symptoms; thus, some proportion of early antibiotic use for this population may be appropriate.
Despite these limitations, our study has multiple strengths. Trained abstractors collected data uniformly for accuracy. We also included a large, heterogeneous cohort of hospitalized patients with ASB managed at diverse hospitals for generalizability. We found significant differences in ASB treatment between hospitals, suggesting that certain stewardship strategies may help to reduce ASB treatment and warrant future study. Furthermore, outcomes were assessed by both medical record review and telephone follow-up, increasing our ability to identify adverse events. Attempts were made to minimize potential bias by inverse probability of treatment weighting by propensity scores. Because a randomized clinical trial of hospitalized patients with ASB may be unethical given the preponderance of available outpatient data arguing against treatment, this multicenter retrospective cohort may provide the best data available to address this population.
This study has important implications for stewardship policy and intervention design. It suggests that inappropriate treatment of ASB is common in hospitals. Treatment was associated with both patient symptoms (AMS, incontinence) and test results (particularly urinalysis), likely representing a misinterpretation of the UTI diagnostic criteria. No benefit associated with ASB treatment in the hospitalized patient was identified, but potential harm includes an association with longer duration of hospitalization. Stewardship efforts should prioritize ASB, targeting older patients with dementia. Inappropriate antibiotic use has been associated with harm; these findings add to the growing body of literature supporting a less-is-more approach, especially in elderly patients who are at higher risk of experiencing adverse drug events. Furthermore, to reduce ASB treatment, programs should consider diagnostic stewardship interventions addressing decreasing urine testing in asymptomatic patients (eg, through computerized decision support and education).
Inappropriate antibiotic treatment of ASB is common among hospitalized patients. Treatment was associated with certain patient characteristics (elderly, dementia, and AMS) and laboratory test findings (particularly abnormal urinalysis results), which were possibly related to a misunderstanding of the diagnostic criteria for UTI. Treatment did not improve clinical outcomes; rather, it may have been associated with an increase in duration of hospitalization after urine testing. Our findings suggest that, to reduce inappropriate antibiotic use, stewardship efforts should focus on improving urine testing practices and management strategies for stable elderly patients with AMS.
Accepted for Publication: May 28, 2019.
Corresponding Author: Lindsay A. Petty, MD, Division of Infectious Diseases, Michigan Medicine, F4177 University Hospital S, 1500 E Medical Center Dr, SPC 5226, Ann Arbor, MI 48109-2800 (firstname.lastname@example.org).
Published Online: August 26, 2019. doi:10.1001/jamainternmed.2019.2871
Author Contributions: Drs Petty and Gandhi had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Petty, Vaughn, Flanders, Malani, Eschenauer, Gandhi.
Acquisition, analysis, or interpretation of data: Petty, Vaughn, Malani, Conlon, Kaye, Thyagarajan, Osterholzer, Nielsen, Eschenauer, Bloemers, McLaughlin, Gandhi.
Drafting of the manuscript: Petty, Vaughn, Malani, Nielsen, Bloemers, McLaughlin, Gandhi.
Critical revision of the manuscript for important intellectual content: Petty, Vaughn, Flanders, Malani, Conlon, Kaye, Thyagarajan, Osterholzer, Eschenauer, Gandhi.
Statistical analysis: Petty, Vaughn, Malani, Conlon, Kaye, Nielsen, Gandhi.
Administrative, technical, or material support: Bloemers, McLaughlin.
Supervision: Petty, Vaughn, Flanders, Malani, Kaye, Osterholzer, McLaughlin, Gandhi.
Conflict of Interest Disclosures: Dr Petty reported grants from Blue Cross Blue Shield of Michigan during the conduct of the study. Dr Vaughn reported grants from Blue Cross Blue Shield of Michigan during the conduct of the study. Dr Flanders reported personal fees from Expert Testimony and Wiley Publishing and grants from Blue Cross Blue Shield of Michigan and the Agency for Healthcare Research and Quality. Mr Nielsen reported other support from Blue Cross Blue Shield of Michigan during the conduct of the study. Ms McLaughlin reported other support from Blue Cross Blue Shield of Michigan during the conduct of the study. Dr Gandhi reported grants from Blue Cross Blue Shield of Michigan during the conduct of the study. No other disclosures were reported.
Funding/Support: This study was funded by a grant from Blue Cross Blue Shield of Michigan.
Role of the Funder/Sponsor: Blue Cross Blue Shield of Michigan had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: The authors thank David Ratz, MS (University of Michigan), for his statistical support and review of the manuscript. He received no additional compensation above his usual compensation for this role. Additionally, the authors thank Dr Preeti Malani, MD, MSI (University of Michigan), for her editorial support. She received no compensation for this role.