Search

Background Procalcitonin (PCT) is often measured in patients with signs of bacterial infection. PCT is often elevated in bacterial pneumonia and septic shock and usually low in viral infections. Recent studies have found a correlation between PCT and disease severity in COVID-19, and most patients receive antibiotics despite bacterial co-infection being rare. We sought to characterize PCT trends in COVID-19, assess its relation to bacterial pneumonia, and assess its relation to clinical decision making around antimicrobial use. Methods We included patients ≥18 hospitalized at Michigan Medicine (3/1/20–10/31/21), positive for COVID-19, with ≥ 1 PCT measurement. Structured query was used to retrospectively extract data. Patients started on an antibiotic underwent retrospective chart review by 2 reviewers for presence of bacterial pneumonia (bPNA), and were classified as having proven, probable, possible, or no bPNA (Figure 1). Multivariable models controlling for time from start of the pandemic, demographics, and comorbidities were used to determine associations of PCT and bPNA with antimicrobial use. Figure 1:Flow diagram of patients included in analyses Results 793 patients met inclusion criteria, with 224 (28.2%) initiated on antibiotics. Of these 224, 33 (14.7%) had proven/probable bPNA, 125 (55.8%) had possible bPNA, and 66 (29.5%) had no bPNA. On average, patients had 2.6 ±3.7 (mean ±SD) PCT measurements, with 4.1 ±5.2 if on antibiotics vs. 2.0 ±2.6 if not. Initial PCT was higher in those on antibiotics and highest in those with proven/probable bPNA (Table 1). After adjustment for confounders, initial PCT was associated with antibiotic initiation (OR 1.68, 95% CI 1.47-1.91, p < 0.0001) (Table 2). Initial PCT (RR 1.11, 95% CI 1.03-1.20, p=0.008), change in PCT over time (RR 1.03, 95% CI 1.01-1.05, p=0.007), and bPNA category (RR 1.51, 95% CI 1.23-1.84, p < 0.0001) were associated with antibiotic duration (Table 3). Table 1:Number of PCT measures by antibiotic and bPNA statusTable 2:Multivariable logistic regression model results for covariate associations with antibiotic initiation (n=793)Table 3:Multivariable negative binomial model results for covariate associations with antibiotic duration (n=224) Conclusion PCT was elevated in patients with COVID-19, but more pronounced with bPNA. Antibiotics were started in > 25% of patients, regardless of bPNA. PCT trends associate with the decision to initiate antibiotics and treatment duration, independent of bPNA and comorbidities. Future prospective studies should determine if PCT can be used to safely make decisions around antibiotic treatment for bacterial infection during COVID-19. Disclosures Krishna Rao, MD, MS, Merck & Co., Inc.: Grant/Research Support|Seres Therapeutics: Advisor/Consultant|Summit Therapeutics: Advisor/Consultant.

BACKGROUND: With a unique influenza season occurring in the midst of a pandemic, there is interest in assessing the role of the influenza vaccine in COVID-19 susceptibility and severity. METHODS: In this retrospective cohort study, patients receiving a laboratory test for COVID-19 were identified. The primary outcome was comparison of positive COVID-19 testing in those who received the influenza vaccine versus those who did not. Secondary end points in patients testing positive for COVID-19 included mortality, need for hospitalization, length of stay, need for intensive care, and mechanical ventilation. RESULTS: A total of 27,201 patients received laboratory testing for COVID-19. The odds of testing positive for COVID-19 was reduced in patients who received an influenza vaccine compared to those who did not (odds ratio 0.76, 95% CI 0.68-0.86; P

Background Fluoroquinolones increase the risk of Clostridioides difficile infection and antibiotic resistance. Hospitals often use pre-prescription approval or prospective audit and feedback to target fluoroquinolone prescribing. Whether these strategies impact aggregate fluoroquinolone use is unknown. Methods This study is a 48-hospital, retrospective cohort of general-care, medical patients hospitalized with pneumonia or positive urine culture between December 2015–September 2017. Hospitals were surveyed on their use of pre-prescription approval and/or prospective audit and feedback to target fluoroquinolone prescribing during hospitalization (fluoroquinolone stewardship). After controlling for hospital clustering and patient factors, aggregate (inpatient and post-discharge) fluoroquinolone (ciprofloxacin, levofloxacin, moxifloxacin) exposure was compared between hospitals with and without fluoroquinolone stewardship. Results There were 11 748 patients (6820 pneumonia; 4928 positive urine culture) included at 48 hospitals. All hospitals responded to the survey: 29.2% (14/48) reported using pre-prescription approval and/or prospective audit and feedback to target fluoroquinolone prescribing. After adjustment, fluoroquinolone stewardship was associated with fewer patients receiving a fluoroquinolone (37.1% vs 48.2%; P = .01) and fewer fluoroquinolone treatment days per 1000 patients (2282 vs 3096 days/1000 patients; P = .01), driven by lower inpatient prescribing. However, most (66.6%) fluoroquinolone treatment days occurred after discharge, and hospitals with fluoroquinolone stewardship had twice as many new fluoroquinolone starts after discharge as hospitals without (15.6% vs 8.4%; P = .003). Conclusions Hospital-based stewardship interventions targeting fluoroquinolone prescribing were associated with less fluoroquinolone prescribing during hospitalization, but not at discharge. To limit aggregate fluoroquinolone exposure, stewardship programs should target both inpatient and discharge prescribing., In this 48-hospital cohort study of 11 748 patients hospitalized with pneumonia or positive urine cultures, hospital-based antibiotic stewardship interventions targeting fluoroquinolone prescribing were associated with less fluoroquinolone use during hospitalization, but appeared to partially increase fluoroquinolone prescribing at discharge.

Background The guidelines for peripherally inserted central catheters (PICCs) recommend avoiding insertion if the anticipated duration of use is =5 days. However, short-term PICC use is common in hospitals. We sought to identify patient, provider, and device characteristics and the clinical outcomes associated with short-term PICCs. Methods Between January 2014 and June 2016, trained abstractors at 52 Michigan Hospital Medicine Safety (HMS) Consortium sites collected data from medical records of adults that received PICCs during hospitalization. Patients were prospectively followed until PICC removal, death, or 70 days after insertion. Multivariable logistic regression models were fit to identify factors associated with short-term PICCs, defined as dwell time of =5 days. Complications associated with short-term use, including major (eg, venous thromboembolism [VTE] or central lineassociated bloodstream infection [CLABSI]) or minor (eg, catheter occlusion, tip migration) events were assessed. Results Of the 15,397 PICCs placed, 3902 (25.3%) had a dwell time of =5 days. Most (95.5%) short-term PICCs were removed during hospitalization. Compared to PICCs placed for >5 days, variables associated with short-term PICCs included difficult venous access (odds ratio [OR], 1.54; 95% confidence interval [CI], 1.40-1.69), multilumen devices (OR, 1.53; 95% CI, 1.39-1.69), and teaching hospitals (OR, 1.25; 95% CI, 1.04-1.52). Among those with short-term PICCs, 374 (9.6%) experienced a complication, including 99 (2.5%) experiencing VTE and 17 (0.4%) experiencing CLABSI events. The most common minor complications were catheter occlusion (4%) and tip migration (2.2%). Conclusion Short-term use of PICCs is common and associated with patient, provider, and device factors. As PICC placement, even for brief periods, is associated with complications, efforts targeted at factors underlying such use appear necessary.

Essentials How best to quantify thrombosis risk with peripherally inserted central catheters (PICC) is unknown. Data from a registry were used to develop the Michigan Risk Score (MRS) for PICC thrombosis. Five risk factors were associated with PICC thrombosis and used to develop a risk score. MRS was predictive of the risk of PICC thrombosis and can be useful in clinical practice. SummaryBackground Peripherally inserted central catheters (PICCs) are associated with upper extremity deep vein thrombosis (DVT). We developed a score to predict risk of PICC-related thrombosis. Methods Using data from the Michigan Hospital Medicine Safety Consortium, image-confirmed upper-extremity DVT cases were identified. A logistic, mixed-effects model with hospital-specific random intercepts was used to identify factors associated with PICC-DVT. Points were assigned to each predictor, stratifying patients into four classes of risk. Internal validation was performed by bootstrapping with assessment of calibration and discrimination of the model. Results Of 23 010 patients who received PICCs, 475 (2.1%) developed symptomatic PICC-DVT. Risk factors associated with PICC-DVT included: history of DVT; multi-lumen PICC; active cancer; presence of another CVC when the PICC was placed; and white blood cell count greater than 12 000. Four risk classes were created based on thrombosis risk. Thrombosis rates were 0.9% for class I, 1.6% for class II, 2.7% for class III and 4.7% for class IV, with marginal predicted probabilities of 0.9% (0.7, 1.2), 1.5% (1.2, 1.9), 2.6% (2.2, 3.0) and 4.5% (3.7, 5.4) for classes I, II, III, and IV, respectively. The risk classification rule was strongly associated with PICC-DVT, with odds ratios of 1.68 (95% CI, 1.19, 2.37), 2.90 (95% CI, 2.09, 4.01) and 5.20 (95% CI, 3.65, 7.42) for risk classes II, III and IV vs. risk class I, respectively. Conclusion The Michigan PICC-DVT Risk Score offers a novel way to estimate risk of DVT associated with PICCs and can help inform appropriateness of PICC insertion.

Two paradigms for the evaluation of surrogate markers in randomized clinical trials have been proposed: the causal effects paradigm and the causal association paradigm. Each of these paradigms rely on assumptions that must be made to proceed with estimation and to validate a candidate surrogate marker (S) for the true outcome of interest (T). We consider the setting in which S and T are Gaussian and are generated from structural models that include an unobserved confounder. Under the assumed structural models, we relate the quantities used to evaluate surrogacy within both the causal effects and causal association frameworks. We review some of the common assumptions made to aid in estimating these quantities and show that assumptions made within one framework can imply strong assumptions within the alternative framework. We demonstrate that there is a similarity, but not exact correspondence between the quantities used to evaluate surrogacy within each framework, and show that the conditions for identifiability of the surrogacy parameters are different from the conditions, which lead to a correspondence of these quantities.

Triggering events for acute aortic dissections are incompletely understood. We sought to investigate whether there is an association between admission for acute type A aortic dissection (ATAAD) to the University of Michigan Medical Center and the reported annual influenza activity by the Michigan Department of Health and Human Services. From 1996–2019 we had 758 patients admitted for ATAAD with 3.1 admissions per month during November-March and 2.5 admissions per month during April-October (p = 0.01). Influenza reporting data by the Michigan Department of Health and Human Services became available in 2009. ATAAD admissions for the period 2009–2019 (n = 455) were 4.8 cases/month during peak influenza months compared to 3.5 cases/month during non-peak influenza months (p = 0.001). ATAAD patients admitted during influenza season had increased in-hospital mortality (11.0% vs. 5.8%, p = 0.024) and increased 30-day mortality (9.7 vs. 5.4%, p = 0.048). The results point to higher admission rates for ATAAD during months with above average influenza rates. Future studies need to investigate whether influenza virus infection affects susceptibility for aortic dissection, and whether this risk can be attenuated with the annual influenza vaccine in this patient population.

In clinical trials, a surrogate outcome ( S) can be measured before the outcome of interest ( T) and may provide early information regarding the treatment ( Z) effect on T. Many methods of surrogacy validation rely on models for the conditional distribution of T given Z and S. However, S is a post-randomization variable, and unobserved, simultaneous predictors of S and T may exist, resulting in a non-causal interpretation. Frangakis and Rubin developed the concept of principal surrogacy, stratifying on the joint distribution of the surrogate marker under treatment and control to assess the association between the causal effects of treatment on the marker and the causal effects of treatment on the clinical outcome. Working within the principal surrogacy framework, we address the scenario of an ordinal categorical variable as a surrogate for a censored failure time true endpoint. A Gaussian copula model is used to model the joint distribution of the potential outcomes of T, given the potential outcomes of S. Because the proposed model cannot be fully identified from the data, we use a Bayesian estimation approach with prior distributions consistent with reasonable assumptions in the surrogacy assessment setting. The method is applied to data from a colorectal cancer clinical trial, previously analyzed by Burzykowski et al.

Background Most patients hospitalized with community-acquired pneumonia (CAP) can be safely treated with 5 days of antibiotics, but many are not. We determined whether a hospitalist-collaborative can reduce excess antibiotic duration in patients with CAP through partnership with antibiotic stewardship teams (AST), data feedback, pay-for-performance, and sharing best practices. Methods From April 2017 to October 2018, abstractors collected data (medical record, phone calls 30-days post-discharge) on adult, non-ICU patients hospitalized with CAP at 43 hospitals in Michigan. We used a guideline-based algorithm1 to determine appropriate antibiotic duration based on patient factors (e.g., clinical stability). All hospitals received a) quarterly reports on appropriate 5-day treatment rates (2016—current), b) best practice recommendations (2017—current) including toolkit and webinar (3/2018), and c) pay-for-performance based on 5-day CAP metric (2018—current). Generalized linear mixed models were used to evaluate change over time in a) proportion of patients with CAP eligible for 5-day treatment who received 5 ± 1 days and, after adjusting for patient factors and weighting by inverse probability of treatment, b) patient outcomes 30-days post-discharge. Results Of 6,229 patients hospitalized with CAP, 4,769 (76.6%) were eligible for 5-days of antibiotic treatment; 283 (5.9%) were excluded due to inability to determine antibiotic duration. Between April 2017 and October 2018, the proportion of patients eligible for a 5-day duration of antibiotic treatment who received 5 ± 1 days increased from 19.8% (181/914) to 30.9% (207/670; P = 0.01), a relative improvement of 56.1% (Figure 1). During this time period, there were no changes in 30-day post-discharge death, readmission, emergency room visit, Clostridioides difficile infection, or provider-documented antibiotic-associated adverse-events (Table 1). However, there was a decrease (3.3% to 1.7%, P = 0.03 for change over time; relative reduction: 48.5%) in patient-reported antibiotic-associated adverse events (Figure 2). Conclusion A hospitalist collaborative partnering with AST can safely reduce excess antibiotic duration and antibiotic-associated adverse-events in hospitalized patients with CAP. Disclosures All Authors: No reported Disclosures.

In clinical trials, an intermediate marker measured after randomization can often provide early information about the treatment effect on the final outcome of interest. We explore the use of recurrence time as an auxiliary variable for estimating the treatment effect on overall survival in phase three randomized trials of colon cancer. A multi-state model with an incorporated cured fraction for recurrence is used to jointly model time to recurrence and time to death. We explore different ways in which the information about recurrence time and the assumptions in the model can lead to improved efficiency. Estimates of overall survival and disease-free survival can be derived directly from the model with efficiency gains obtained as compared to Kaplan-Meier estimates. Alternatively, efficiency gains can be achieved by using the model in a weaker way in a multiple imputation procedure, which imputes death times for censored subjects. By using the joint model, recurrence is used as an auxiliary variable in predicting survival times. We demonstrate the potential use of the proposed methods in shortening the length of a trial and reducing sample sizes.

Catheter exchange over a guidewire is frequently performed for malfunctioning peripherally inserted central catheters (PICCs). Whether such exchanges are associated with venous thromboembolism is not known.We performed a retrospective cohort study to assess the association between PICC exchange and risk of thromboembolism. Adult hospitalized patients that received a PICC during clinical care at one of 51 hospitals participating in the Michigan Hospital Medicine Safety consortium were included. The primary outcome was hazard of symptomatic venous thromboembolism (radiographically confirmed upper-extremity deep vein thrombosis and pulmonary embolism) in those that underwent PICC exchange vs those that did not.Of 23,010 patients that underwent PICC insertion in the study, 589 patients (2.6%) experienced a PICC exchange. Almost half of all exchanges were performed for catheter dislodgement or occlusion. A total of 480 patients (2.1%) experienced PICC-associated deep vein thrombosis. The incidence of deep vein thrombosis was greater in those that underwent PICC exchange vs those that did not (3.6% vs 2.0%, P .001). Median time to thrombosis was shorter among those that underwent exchange vs those that did not (5 vs 11 days, P = .02). Following adjustment, PICC exchange was independently associated with twofold greater risk of thrombosis (hazard ratio [HR] 1.98; 95% confidence interval [CI], 1.37-2.85) vs no exchange. The effect size of PICC exchange on thrombosis was second in magnitude to device lumens (HR 2.06; 95% CI, 1.59-2.66 and HR 2.31; 95% CI, 1.6-3.33 for double- and triple-lumen devices, respectively).Guidewire exchange of PICCs may be associated with increased risk of thrombosis. As some exchanges may be preventable, consideration of risks and benefits of exchanges in clinical practice is needed.

BACKGROUNDPeripherally inserted central catheters (PICCs) are associated with central-line–associated bloodstream infections (CLABSIs). However, no tools to predict risk of PICC-CLABSI have been developed.OBJECTIVETo operationalize or prioritize CLABSI risk factors when making decisions regarding the use of PICCs using a risk model to estimate an individual’s risk of PICC-CLABSI prior to device placement.METHODSUsing data from the Michigan Hospital Medicine Safety consortium, patients that experienced PICC-CLABSI between January 2013 and October 2016 were identified. A Cox proportional hazards model with robust sandwich standard error estimates was then used to identify factors associated with PICC-CLABSI. Based on regression coefficients, points were assigned to each predictor and summed for each patient to create the Michigan PICC-CLABSI (MPC) score. The predictive performance of the score was assessed using time-dependent area-under-the-curve (AUC) values.RESULTSOf 23,088 patients that received PICCs during the study period, 249 patients (1.1%) developed a CLABSI. Significant risk factors associated with PICC-CLABSI included hematological cancer (3 points), CLABSI within 3 months of PICC insertion (2 points), multilumen PICC (2 points), solid cancers with ongoing chemotherapy (2 points), receipt of total parenteral nutrition (TPN) through the PICC (1 point), and presence of another central venous catheter (CVC) at the time of PICC placement (1 point). The MPC score was significantly associated with risk of CLABSI (PCONCLUSIONThe MPC score offers a novel way to inform decisions regarding PICC use, surveillance of high-risk cohorts, and utility of blood cultures when PICC-CLABSI is suspected. Future studies validating the score are necessary.Infect Control Hosp Epidemiol2017;38:1155–1166

Background The validation of intermediate markers as surrogate markers ( S) for the true outcome of interest ( T) in clinical trials offers the possibility for trials to be run more quickly and cheaply by using the surrogate endpoint in place of the true endpoint. Purpose Working within a principal stratification framework, we propose causal quantities to evaluate surrogacy using a Gaussian copula model for an ordinal surrogate and time-to-event final outcome. The methods are applied to data from four colorectal cancer clinical trials, where S is tumor response and T is overall survival. Methods For the Gaussian copula model, a Bayesian estimation strategy is used and, as some parameters are not identifiable from the data, we explore the use of informative priors that are consistent with reasonable assumptions in the surrogate marker setting to aid in estimation. Results While there is some bias in the estimation of the surrogacy quantities of interest, the estimation procedure does reasonably well at distinguishing between poor and good surrogate markers. Limitations Some of the parameters of the proposed model are not identifiable from the data, and therefore, assumptions must be made in order to aid in their estimation. Conclusions The proposed quantities can be used in combination to provide evidence about the validity of S as a surrogate marker for T.

Because of the time and expense required to obtain clinical outcomes of interest, such as functional limitations or death, clinical trials often focus the effects of treatment on earlier and more easily obtained surrogate markers. Preliminary work to define surrogates focused on the fraction of a treatment effect "explained" by a marker in a regression model, but as notions of causality have been formalized in the statistical setting, formal definitions of high-quality surrogate markers have been developed in the causal inference framework, using either the "causal effect" or "causal association" settings. In the causal effect setting, high-quality surrogate markers have a large fraction of the total treatment effect explained by the effect of the treatment on the marker net of the treatment on the outcome. In the causal association setting, high-quality surrogate markers have large treatment effects on the outcome when there are large treatment effects on the marker, and small effects on the outcome when there are small effects on the marker. A particularly important feature of a surrogate marker is that the direction of a treatment effect be the same for both the marker and the outcome. Settings in which the marker and outcome are positively associated but the marker and outcome have beneficial and harmful or harmful and beneficial treatment effects, respectively, have been referred to as "surrogate paradoxes". If this outcome always occurs, it is not problematic; however, as correlations among the outcome, marker, and their treatment effects weaken, it may occur for some trials and not for others, leading to potentially incorrect conclusions, and real-life examples that shortened thousands of lives are unfortunately available. We propose measures for assessing the risk of the surrogate paradox using the meta-analytic causal association framework, which allows us to focus on the probability that a given treatment will yield treatment effect in different directions between the marker and the outcome, and to determine the size of a beneficial effect of the treatment on the marker required to minimize the risk of a harmful effect of the treatment on the outcome. We provide simulations and consider two applications.

Background Broad-spectrum (BS) antibiotics directed against Pseudomonas aeruginosa and methicillin-resistant Staphylococcus aureus (MRSA) are commonly used for health-care associated pneumonia (HCAP) treatment. Many patients with HCAP do not have a microbiologically confirmed diagnosis. The goal of this study was to evaluate the impact of antibiotic de-escalation on clinical outcomes in patients with HCAP without a microbiological diagnosis. Methods This is a retrospective cohort study of adult, non-ICU, medical patients hospitalized with HCAP between January 2016 and February 2018 at 46 Michigan hospitals. Exclusions included extrapulmonary infection, severe immune suppression, or clinical instability on day 4. Included patients: (1) lacked any positive culture (blood/sputum); (2) started on empiric anti-P. aeruginosa and anti-MRSA therapy by hospital day 2; (3) switched to a narrow-spectrum (NS) regimen (no anti-P. aeruginosa or anti-MRSA coverage) or maintained on BS antibiotics (anti-P. aeruginosa ± anti-MRSA) by therapy day 4 (Figure 1). Mortality, readmission, Clostridium difficile infection, and adverse events from antibiotics were compared between the BS and NS groups. Data were analyzed using logistic generalized estimating equation models and inverse probability of treatment weighting. Results Of 363 patients with HCAP included, 73 (20%) were switched to an NS regimen. Of 290 patients maintained on anti-PSA BS regimens, 47.6% also continued anti-MRSA therapy. The median age was 72 (IQR, 61–81) and Charlson comorbidity index was 4 (IQR, 2–6) of the entire cohort. Baseline characteristics were similar between BS and NS groups, except more patients had chronic kidney disease in the BS group. On multivariable analysis, no other baseline factors were found to be associated with use of BS antibiotics on day 4. Both total and IV antibiotic duration were longer in the BS group (10 vs. 8 days, P = 0.002, and 4 vs. 3 days, P < 0.001, respectively). On adjusted analysis, there were no differences in patient outcomes (Figure 2). Conclusion Among patients with HCAP started on empiric MRSA and PSA coverage without microbiological diagnosis, clinical outcomes were similar in patients switched to an NS antibiotic and those maintained on BS antibiotics. Our findings suggest a potential role for antimicrobial stewardship in promoting antibiotic de-escalation in this population. Disclosures All authors: No reported disclosures.

Background Treatment of asymptomatic bacteriuria (ASB) is a major driver of inappropriate antibiotic use and an important target for antimicrobial stewardship. We identified patient- and hospital-level factors associated with treatment of ASB and its impact on patient outcomes. Methods In this retrospective cohort study, detailed data were abstracted from the medical record of adult non-ICU patients hospitalized with a positive urine culture (Ucx) between January 2016 and February 2018 at 46 Michigan hospitals. Exclusions included pregnancy, urologic surgery or abnormality, immune-compromise, or concomitant infection. ASB was defined as a positive Ucx without signs or symptoms attributable to a urinary tract infection (UTI). The treatment group received ≥1 antibiotic dose. Patient outcomes included mortality, readmissions, Clostridium difficile infection, and emergency room visits. Patient and hospital factors associated with ASB treatment were evaluated using logistic generalized estimating equation models; patient outcomes were inverse probability of treatment weighted. Results Of 2,733 included patients with ASB, 82.9% (n = 2,266) were treated with antibiotics for a median 7 days (IQR 4,9). Ceftriaxone (71.1%) was the most frequent initial therapy; fluoroquinolones (33.2%) were most common at discharge. In the multivariable model, patient variables associated with ASB treatment included: increased age, dementia, positive urinalysis, incontinence, indwelling urinary catheter, and nonambulatory status (Figure 1). Hospitals varied (Figure 2), but those that required a documented indication for antibiotics in the order or medical record had lower ASB treatment rates (OR = 0.5). There was no difference in patient outcomes for patients treated vs. not treated with antibiotics. Conclusion Antibiotic treatment of ASB, often broad-spectrum, is widespread. Certain patient characteristics (including advanced age, nonambulatory, dementia, and incontinence) and the misinterpretation of test results (including overemphasis of the urinalysis) drive clinicians to treat ASB. Requiring documentation of antibiotic indication may decrease inappropriate treatment. Future interventions may be more effective by incorporating these drivers of ASB treatment. Disclosures All authors: No reported Disclosures.

Importance 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. Objective 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. Exposures 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 difficileinfection (formerly known asClostridium difficile) at 30 days, and duration of hospitalization after urine testing. Results 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.

Background Randomized trials demonstrate no benefit from antibiotic treatment exceeding the shortest effective duration. Objective To examine predictors and outcomes associated with excess duration of antibiotic treatment. Design Retrospective cohort study. Setting 43 hospitals in the Michigan Hospital Medicine Safety Consortium. Patients 6481 general care medical patients with pneumonia. Measurements The primary outcome was the rate of excess antibiotic treatment duration (excess days per 30-day period). Excess days were calculated by subtracting each patient's shortest effective (expected) treatment duration (based on time to clinical stability, pathogen, and pneumonia classification [community-acquired vs. health care-associated]) from the actual duration. Negative binomial generalized estimating equations (GEEs) were used to calculate rate ratios to assess predictors of 30-day rates of excess duration. Patient outcomes, assessed at 30 days via the medical record and telephone calls, were evaluated using logit GEEs that adjusted for patient characteristics and probability of treatment. Results Two thirds (67.8% [4391 of 6481]) of patients received excess antibiotic therapy. Antibiotics prescribed at discharge accounted for 93.2% of excess duration. Patients who had respiratory cultures or nonculture diagnostic testing, had a longer stay, received a high-risk antibiotic in the prior 90 days, had community-acquired pneumonia, or did not have a total antibiotic treatment duration documented at discharge were more likely to receive excess treatment. Excess treatment was not associated with lower rates of any adverse outcomes, including death, readmission, emergency department visit, or Clostridioides difficile infection. Each excess day of treatment was associated with a 5% increase in the odds of antibiotic-associated adverse events reported by patients after discharge. Limitation Retrospective design; not all patients could be contacted to report 30-day outcomes. Conclusion Patients hospitalized with pneumonia often receive excess antibiotic therapy. Excess antibiotic treatment was associated with patient-reported adverse events. Future interventions should focus on whether reducing excess treatment and improving documentation at discharge improves outcomes. Primary funding source Blue Cross Blue Shield of Michigan (BCBSM) and Blue Care Network as part of the BCBSM Value Partnerships program.

Background Existing guidelines, including Choosing Wisely recommendations, endorse avoiding placement of peripherally inserted central catheters (PICCs) in patients with chronic kidney disease (CKD). Objective To describe the frequency of and characteristics associated with PICC use in hospitalized patients with stage 3b or greater CKD (glomerular filtration rate [GFR]

Purpose To evaluate rectal dose and post-treatment patient-reported bowel quality of life (QOL) following radiation therapy for prostate cancer. Methods Patient-reported QOL was measured at baseline and 2-years via the expanded prostate cancer index composite (EPIC) for 90 patients. Linear regression modeling was performed using the baseline score for the QUANTEC normal tissue complication probability model and dose volume histogram (DVH) parameters for the whole and segmented rectum (superior, middle, and inferior). Results At 2-years the mean summary score declined from a baseline of 96.0–91.8. The median volume of rectum treated to ⩾70 Gy (V70) was 11.7% for the whole rectum and 7.0%, 24.4%, and 1.3% for the inferior, middle, and superior rectum, respectively. Mean dose to the whole and inferior rectum correlated with declines in bowel QOL while dose to the mid and superior rectum did not. Low (V25–V40), intermediate (V50–V60) and high (V70–V80) doses to the inferior rectum influenced bleeding, incontinence, urgency, and overall bowel problems. Only the highest dose (V80) to the mid-rectum correlated with rectal bleeding and overall bowel problems. Conclusions Segmental DVH analysis of the rectum reveals associations between bowel QOL and inferior rectal dose that could significantly influence radiation planning and prognostic models.

In cancer clinical trials, patients often experience a recurrence of disease prior to the outcome of interest, overall survival. Additionally, for many cancers, there is a cured fraction of the population who will never experience a recurrence. There is often interest in how different covariates affect the probability of being cured of disease and the time to recurrence, time to death, and time to death after recurrence. We propose a multi-state Markov model with an incorporated cured fraction to jointly model recurrence and death in colon cancer. A Bayesian estimation strategy is used to obtain parameter estimates. The model can be used to assess how individual covariates affect the probability of being cured and each of the transition rates. Checks for the adequacy of the model fit and for the functional forms of covariates are explored. The methods are applied to data from 12 randomized trials in colon cancer, where we show common effects of specific covariates across the trials.

In clinical trials, a surrogate outcome variable (S) can be measured before the outcome of interest (T) and may provide early information regarding the treatment (Z) effect on T. Using the principal surrogacy framework introduced by Frangakis and Rubin (2002. Principal stratification in causal inference. Biometrics 58, 21–29), we consider an approach that has a causal interpretation and develop a Bayesian estimation strategy for surrogate validation when the joint distribution of potential surrogate and outcome measures is multivariate normal. From the joint conditional distribution of the potential outcomes of T, given the potential outcomes of S, we propose surrogacy validation measures from this model. As the model is not fully identifiable from the data, we propose some reasonable prior distributions and assumptions that can be placed on weakly identified parameters to aid in estimation. We explore the relationship between our surrogacy measures and the surrogacy measures proposed by Prentice (1989. Surrogate endpoints in clinical trials: definition and operational criteria. Statistics in Medicine 8, 431–440). The method is applied to data from a macular degeneration study and an ovarian cancer study.

Surrogates which allow one to predict the effect of the treatment on the outcome of interest from the effect of the treatment on the surrogate are of importance when it is difficult or expensive to measure the primary outcome. Unfortunately, the use of such surrogates can give rise to paradoxical situations in which the effect of the treatment on the surrogate is positive, the surrogate and outcome are strongly positively correlated, but the effect of the treatment on the outcome is negative, a phenomenon sometimes referred to as the "surrogate paradox." New results are given for consistent surrogates that extend the existing literature on sufficient conditions that ensure the surrogate paradox is not manifest. Specifically, it is shown that for the surrogate paradox to beman.est it must be the case that either there is (i) a direct effect of treatment on the outcome not through the surrogate and in the opposite direction as that through the surrogate or (ii) confounding for the effect of the surrogate on the outcome, or (iii) a lack of transitivity so that treatment does not positively affect the surrogate for all the same individuals for which the surrogate positively affects the outcome. The conditions for consistent surrogates and the results of the paper are important because they allow investigators to predict the direction of the effect of the treatment on the outcome simply from the direction of the effect of the treatment on the surrogate. These results on consistent surrogates are then related to the four approaches to surrogate outcomes described by Joffe and Greene (2009, Biometrics 65, 530–538) to assess whether the standard criterion used by these approaches to assess whether a surrogate is "good" suffices to avoid the surrogate paradox.

Background Analytic morphometrics provides objective data that may better stratify risk. We investigated morphometrics and outcome among colon cancer patients. Methods An IRB-approved review identified 302 patients undergoing colectomy who had CT scans. These were processed to measure psoas area (PA), density (PD), subcutaneous fat (SFD), visceral fat (VF), and total body fat (TBF). Correlation with complications, recurrence, and survival were obtained by t-tests and linear regression models after adjusting for age and Charlson index. Results The best predictor of surgical complications was PD. PMH, Charlson, BMI, and age were not significant when PD was considered. SF area was the single best predictor of a wound infection. While all measures of obesity correlated with outcome, TBF was most predictive. Final multivariate Cox models for survival included age, Charlson score, nodal positivity, and TBF. Conclusions Analytic morphometric analysis provided objective data that stratified complications and outcome better than age, BMI, or co-morbidities. J. Surg. Oncol. 2013; 108:169–175. © 2013 Wiley Periodicals, Inc.

Background Community-acquired pneumonia (CAP) guidelines recommend transition to an oral (PO) β-lactam (BL) regimen or fluoroquinolone (FQ) when patients are clinically stable. Due to collateral damage associated with FQs, stewardship efforts often focus on reducing initial FQ use for CAP therapy. We hypothesized that FQ use remains prevalent in CAP treatment despite initial intravenous (IV) BL therapy, and examined factors associated with switching to a PO BL vs. an FQ and the impact on outcomes. Methods In this retrospective cohort study, data were collected from January 2016 through February 2018 on non-ICU medical patients admitted with CAP at 46 Michigan hospitals. Patients were included if they received IV BL (ceftriaxone or ampicillin-sulbactam) plus macrolide/doxycycline by hospital day 2 and switched to a PO respiratory FQ or BL by therapy day 4. Exclusions included positive culture, concomitant infection, HCAP, unstable on day 4, or severe immune deficiency. Data were analyzed using logistic generalized estimating equation models, accounting for hospital-level clustering; outcomes were adjusted using the inverse probability of treatment weighting by propensity scores. Results. Of 555 included patients, 54.4% were switched to a PO BL vs. 45.6% to an FQ by day 4 of therapy. The groups had similar durations of therapy (8 days), time to clinical stability, prior antibiotics, and COPD, but the BL group was older, with a higher Pneumonia Severity Index, CURB-65, and more cardiovascular (CV) disease (Figure 1). In multivariable analysis, CV disease and higher CURB-65 were more common and diabetes (DM) less common in the PO BL group (Figure 2). Sharing antibiotic use data with providers were associated with less FQ use (83.1% vs. 70.8%, OR 0.51, 95% CI 0.27, 0.97, P = 0.04). On adjusted analysis there were no differences in patient outcomes (Figure 3). Conclusion. Among CAP patients started on an IV BL regimen, nearly half were switched to a PO FQ by therapy day 4, including more patients with DM. Although there were sicker patients in the BL group, there were no differences in outcomes between cohorts. To reduce FQ use, stewardship programs should share antibiotic use data and provide guidance for step-down therapy in clinically stable CAP patients. Disclosures All authors: No reported disclosures.

National guidelines1 recommend objective risk stratification for venous thromboembolism (VTE) in hospitalized medical patients. The Padua Prediction Score risk assessment model2 is recommended to categorize patients as high or low risk. The Michigan Hospital Medicine Safety Consortium (HMS), a statewide quality collaborative aimed at preventing adverse events in hospitalized medical patients, collects detailed data on VTE risk factors, prophylactic treatment, and outcomes. Using data from the HMS,3 we sought to determine whether patients in this cohort were receiving appropriate VTE prophylaxis.

Aromatase inhibitors (AIs), which are used to treat breast cancer, inhibit estrogen produc- tion in postmenopausal women. AI-associated musculoskeletal symptoms occur in approximately half of treated women and lead to treatment discontinuation in 20 to 30%. The etiology may be due in part to estrogen deprivation. In premenopausal women, lower estrogen levels have been associated with increased pain aswell as with impairment of descending pain inhibitory pathways, which may be a risk factor for developing chronic pain. We prospectively tested whether AI-induced estrogen deprivation alters pain sensitivity, thereby increasing the risk of developing AI-associated musculoskeletal symp- toms. Fifty postmenopausal breast cancer patients underwent pressure pain testing and conditioned pain modulation (CPM) assessment prior to AI initiation and after 3 and 6 months. At baseline, 26 of 40 (65%) assessed patients demonstrated impaired CPM, which was greater in those who had previously received chemotherapy (P = .006). No statistically significant change in pressure pain threshold or CPM was identified following estrogen deprivation. In addition, there was no association with either measure of pain sensitivity and change in patient-reported pain with AI therapy. AI-associated muscu- loskeletal symptoms are not likely due to decreased pain threshold or impaired CPM prior to treatment initiation, or to effects of estrogen depletion on pain sensitivity. Perspective: This article presents our findings of the effect of estrogen deprivation on objective measures of pain sensitivity. In postmenopausal women, medication-induced estrogen depletion did not result in an identifiable change in pressure pain threshold or CPM. Impaired CPM may be associated with chemotherapy.

Purpose Recently, much media attention has been given to the premature deaths in professional wrestlers. Since no formal studies exist that have statistically examined the probability of premature mortality in professional wrestlers, we determined survival estimates for active wresters over the past quarter century to establish the factors contributing to the premature mortality of these individuals. Methods Data including cause of death was obtained from public records and wrestling publications in wrestlers who were active between January 1, 1985 and December 31, 2011. 557 males were considered consistently active wrestlers during this time period. 2007 published mortality rates from the Center for Disease Control were used to compare the general population to the wrestlers by age, BMI, time period, and cause of death. Survival estimates and Cox hazard regression models were fit to determine incident premature deaths and factors associated with lower survival. Cumulative incidence function (CIF) estimates given years wrestled was obtained using a competing risks model for cause of death. Results The mortality for all wrestlers over the 26-year study period was.007 deaths/total person-years or 708 per 100,000 per year, and 16% of deaths occurred below age 50 years. Among wrestlers, the leading cause of deaths based on CIF was cardiovascular-related (38%). For cardiovascular-related deaths, drug overdose-related deaths and cancer deaths, wrestler mortality rates were respectively 15.1, 122.7 and 6.4 times greater than those of males in the general population. Survival estimates from hazard models indicated that BMI is significantly associated with the hazard of death from total time wrestling (p

Background Intermediate outcome variables can often be used as auxiliary variables for the true outcome of interest in randomized clinical trials. For many cancers, time to recurrence is an informative marker in predicting a patient's overall survival outcome and could provide auxiliary information for the analysis of survival times. Purpose To investigate whether models linking recurrence and death combined with a multiple imputation procedure for censored observations can result in efficiency gains in the estimation of treatment effects and be used to shorten trial lengths. Methods Recurrence and death times are modeled using data from 12 trials in colorectal cancer. Multiple imputation is used as a strategy for handling missing values arising from censoring. The imputation procedure uses a cure model for time to recurrence and a time-dependent Weibull proportional hazards model for time to death. Recurrence times are imputed, and then death times are imputed conditionally on recurrence times. To illustrate these methods, trials are artificially censored 2 years after the last accrual, the imputation procedure implemented, and a log-rank test and Cox model used to analyze and compare these new data with the original data. Results The results show modest, but consistent gains in efficiency in the analysis using the auxiliary information in recurrence times. Comparison of analyses show the treatment effect estimates and log-rank test results from the 2-year censored imputed data to be in between the estimates from the original data and the artificially censored data, indicating that the procedure was able to recover some of the lost information due to censoring. Limitations The models used are all fully parametric, requiring distributional assumptions of the data. Conclusions The proposed models may be useful in improving the efficiency in estimation of treatment effects in cancer trials and shortening trial length.

Dose to the Inferior Rectum Has the Strongest Association With Patient-Reported Bowel Quality of Life (QOL) After External Beam Radiation Therapy (EBRT) for Prostate Cancer M.H. Stenmark, S. Johnson, A.S. Conlon, S. Dainault, R. Marsh, T. Ritter, D. Litzenberg, H.M. Sandler, M.G. Sanda, and D.A. Hamstra; University of Michigan Medical Center, Ann Arbor, MI, Veterans Affairs Medical Center, Ann Arbor, MI, Cedars-Sinai Medical Center, Los Angeles, CA, Emory University, Atlanta, GA

9615 Background: AIAA affect up to half of AI-treated women with early stage breast cancer, and can lead to treatment discontinuation in 20-30%. The etiology is thought to be related to estrogen deprivation although the mechanism is unknown. In premenopausal women, lower estrogen levels have been associated with increased pain. Impaired descending pain inhibitory pathways, which may be a risk factor for developing chronic pain, have also been associated with lower estrogen levels. We prospectively tested whether AI-induced estrogen deprivation alters pain sensitivity, thereby increasing the risk of developing AIAA. Methods: Fifty postmenopausal women with early stage breast cancer initiating AI therapy were enrolled to the study. Subjects underwent experimental pressure pain testing and conditioned pain modulation (CPM) assessment and completed symptom questionnaires prior to AI initiation and after 3 months. Positive CPM values (>0) signify impaired descending pain inhibition. Serum estradiol concentrations were determined using an ultrasensitive assay. T-tests, Fisher’s exact test, and linear regression models were used to assess associations among baseline (BL) experimental pain measures, patient-reported pain, and clinical factors. P values

84 Background: Rectal dose-volume histogram (DVH) parameters and a consensus normal tissue complication probability (NTCP) model predict physician scored toxicity but have not been well studied for patient reported outcomes (PRO) following external beam RT for prostate cancer. Methods: PRO were prospectively collected at baseline and 12 month intervals by expanded prostate cancer index composite (EPIC) for 98 patients from 2004-2009. Linear regression modeling for bowel score at 1 and 2 years was performed using pre-treatment bowel score and DVH parameters [V80 (% of rectal volume >80 Gy), V75, V70, V50, V25, and NTCP]. Results: Median RT dose was 78 Gy with image-guidance in 81%, intensity-modulation in 72%, and pelvic RT in 11%. Rectal DVH parameters are presented in the Table. Pretreatment and 2 year PRO were available for 92% of patients. The mean pretreatment bowel summary score declined from 96.0 (sd 8.0) to 91.8 (13.5) at 2 years. Acute proctitis (grade 1-2) during RT was associated with a decline in 2 year bowel QOL, while acute diarrhea (grade 1-2) was not. At 2 years 63% of patients had a decline dose, NTCP and bowel QOL (Table). Akaike information criterion analysis followed by leave-one-out cross validation, indicated that V75 provided the best predictive model for 2 year bowel QOL followed by NTCP and V70. Conclusions: Most previous studies have highlighted V70 for toxicity (such as rectal bleeding). The association with V75 and rectal QOL has not been previously reported and supports potential changes in how prostate plans are designed to maintain QOL. This is also the first validation of a consensus NTCP model for bowel QOL. [Table: see text]

138 Background: Despite improved survival with adjuvant chemotherapy among older women, it is less often recommended for fear they cannot complete therapy or will have excessive toxicity. Analytic morphometrics uses objective imaging measurements such as muscle size, bone mineral density (BMD) and body composition to improve risk stratification. We examined whether morphometrics might help predict toxicity of chemotherapy in breast cancer patients. Methods: Our prospective IRB-approved breast cancer database was queried for all patients who underwent adjuvant or neoadjuvant chemotherapy and had CT scans of the chest, abdomen and pelvis prior to chemo. Complications of chemotherapy were graded according to the NCI-CTC. CT scans were processed using semi-automated algorithms (MATLAB v13.0) to measure psoas area (PA) and density (PD), BMD and subcutaneous fat (SF), visceral fat (VF), and total body area (TBA. Outcome (DFS and OS) was assessed by K-M, and logistic regression models and ANOVA were used for toxicity events. Results: We identified 129 patients, ranging in age from 24 to 83 (median 52). 105 (81%) received AC/T while 19% received alternate regimens, all full weight based. T-stage, nodal status and HER2 expression were significantly associated with measures of obesity, except BMI, but not sarcopenia. After adjusting for stage and Her-2, BMI remained a significant predictor of DFS (OR 1.07 (95%CI 1.01, 1.13)) and OS (OR 1.06 (95% CI 1.00, 1.12)).After controlling for age, PD and VF were predictors of chemotherapy completion. The OR for completing chemotherapy decreased 0.95 (95% CI: 0.90, 1.01, p=0.08) for every increase of 1000 in VF and increased 1.08 (95% CI: 1.01, 1.15, p=0.03) for every unit increase in PD. Sarcopenia was associated with an increased risk of pulmonary complications (p=0.01) and anemia (p=0.06). Conclusions: Although age is a significant predictor of increased toxicity and inability to complete chemotherapy, sarcopenia is an objective, independent predictor of chemotherapy completion. Psoas density, easily obtainable from a pre-treatment CT scan, can help the clinician older breast cancer patients for adjuvant chemotherapy.