A Pharmacoepidemiologic Approach to Evaluate Real-world Effectiveness of Hormonal Contraceptives in the Presence of Drug-drug Interactions.

BACKGROUND: Accurate estimation of conception is critical in the assessment of the effects of drugs used during pregnancy or to prevent pregnancy. In a novel application, we studied the effectiveness of oral contraceptives (OCs), where misclassification of conception relative to OC exposure may obscure effect estimates. METHODS: We studied OC failure, in a large claims database, among women who used antiepileptic drugs with metabolizing enzyme-inducing properties (carbamazepine or oxcarbazepine), which reduce OC's effectiveness or enzyme-neutral properties (lamotrigine or levetiracetam), with no expected impact on OC effectiveness. We compared conception rates in women 12-48 years of age concomitantly using OCs and enzyme-inducing drugs with rates in concomitant users of OCs and enzyme-neutral drugs. We measured conception with a validated algorithm that estimates gestational age based on pregnancy endpoints. We estimated relative and attributable risk using generalized estimating equation models after standardized mortality ratio weighting. RESULTS: We identified 89,777 concomitant use episodes with adjusted contraceptive failure rates of 1.6 (95% confidence interval (CI) = 1.4, 1.8) per 100 person-years among users of enzyme-neutral drugs and 18,964 episodes with a rate of 2.3 (1.9, 2.8) among users of enzyme-inducing drugs. The relative risk of conception for enzyme-inducing group was 1.4 (1.1, 1.8), and the rate difference was 0.7 (0.2, 1.2). CONCLUSIONS: OCs in combination with antiepileptic drugs that interact with metabolic enzymes were associated with increased contraceptive failure rates. Measurement of conception in claims data had adequate accuracy to uncover a strong drug-drug interaction, offering promise for broader application in comparative effectiveness studies on hormonal contraceptives to inform clinical and regulatory decisionmaking.

Usability of encounter data for Medicaid comprehensive managed care vs traditional Medicaid fee-for-service claims among pregnant women.

BACKGROUND: The completeness of medical encounters capture among Medicaid enrollees in comprehensive managed care (CMC) has been shown to vary across states and years. CMC penetration has grown, and CMC encounter capture specific to pregnancy care is understudied. OBJECTIVES: To compare the completeness of encounter data for pregnant beneficiaries in CMC versus traditional fee-for-service (FFS) in Texas and Florida between 2007 and 2010. METHODS: Using Medicaid Analytic eXtract (MAX) data linked to Florida and Texas birth certificate records, for each state and study year, we compared proportions using seven themes: (a) delivery; (b) prenatal visits; (c) dispensed prescriptions during pregnancy; (d) gestational diabetes and blood glucose testing; (e) antidiabetics and diagnosis of diabetes mellitus; (f) antibiotics for urinary tract infection and outpatient encounter; and (g) bacterial vaginosis and dispensing for metronidazole or clindamycin. We considered CMC data to be acceptable if proportions were no less than 10% below the corresponding (2007 to 2010) FFS control values. RESULTS: Pregnancy-related characteristics of FFS vs CMC denominators were comparable. Proportions for the seven measures among FFS controls ranged from 26% to 98%. In Texas, CMC encounter data met the thresholds for all measures between 2007 and 2010. Florida had usable CMC encounter data starting from 2009 with incomplete medical and pharmacy records in 2007 and 2008. CONCLUSIONS: The quality of CMC encounter data in MAX files for pregnant women varied in Florida and Texas and improved over time. Use of pregnancy-specific measures can aid researchers in selecting states and years with acceptable encounter data quality.

Estimating Minimally Important Differences for the Bladder Cancer Index Using Distribution and Anchor Based Approaches.

PURPOSE: The BCI (Bladder Cancer Index) is a validated, condition specific health questionnaire assessing urinary, bowel and sexual function and quality of life among patients with bladder cancer. We aimed to establish minimally important difference score thresholds that signal clinical importance. MATERIALS AND METHODS: For 1 year after surgery we followed a prospective cohort of 150 patients who underwent radical cystectomy between 2013 and 2016. Usable data on 138 patients were analyzed. The BCI and the Medical Outcomes Study SF-36 (36-Item Short Form Health Survey) questionnaires were completed prior to cystectomy, and 3, 6 and 12 months postoperatively. Distribution based, minimally important differences were estimated at ⅓ and ½ SD for each index domain across time points. Changes in index domain scores anchored to changes in a SF-36 overall health assessment question were used to estimate anchor based, minimally important differences. Pooled averages are reported between time points and methods. RESULTS: The distribution based, minimally important difference of ⅓ SDs for urinary, bowel and sexual domains ranged between 5.3 and 7.3, 4.6 and 5.6, and 6.0 and 8.9 points, respectively. Ranges of ½ SDs were 8.8 and 10.9, 6.8 and 8.4, and 8.9 and 13.5 points, respectively. The anchor based approach resulted in minimally important difference estimates of 6.2, 7.3 and 6.8 points, respectively. Aggregated results across the 2 approaches resulted in minimally important differences of 6 to 9, 5 to 8 and 7 to 11 points for urinary, bowel and sexual domains, respectively. CONCLUSIONS: Using 2 independent approaches to our knowledge we established the first minimally important difference estimates for the BCI. Defining patient reported outcome thresholds is important to interpret changes or differences in BCI scores.

Internal validation of Medicaid Analytic eXtract (MAX) data capture for comprehensive managed care plan enrollees from 2007 to 2010.

PURPOSE: Because of concerns over incomplete medical encounter capture in Medicaid capitated comprehensive managed care (CMC) plans, researchers have traditionally confined analyses to fee-for-service (FFS) enrollees. We aimed to evaluate the usability of data for CMC enrollees in Medicaid Analytic eXtract (MAX) files for 29 states from 2007 to 2010. METHODS: We applied 7 measures to MAX inpatient, other therapy, and prescription drug files for each state and study year. Four measures were based on "connectivity" criteria where we expected use of a select essential service to be closely connected to another, resulting in "service pairs." Three measures were based on "continuity" criteria where we expected patients to continue chronically used services or treatments when they switched enrollment from FFS to CMC plans. High proportions of continuity and comparable proportions of patients with complete service pairs relative to FFS enrollees may suggest complete data capture for CMC enrollees. Data of states that met preset criteria were considered usable for research and policy analyses. RESULTS: The completeness of CMC enrollees' data in MAX varied by states. Among 22 states having at least 5% CMC plan enrollment, data of 12 states met our quality standard and were considered usable starting in 2007. Four states had usable data starting in 2008 and one in 2009. CONCLUSIONS: The completeness of CMC enrollees' data in MAX improved over the study period. In 17 out of 29 states, CMC enrollees' data in selected years were comparable with FFS enrollees and can be considered for use in analysis.

Preoperative Patient Reported Mental Health is Associated with High Grade Complications after Radical Cystectomy.

PURPOSE: Psychological distress has been associated with an impaired immune response and poor wound healing. We hypothesized that preoperative patient reported mental health would be associated with high grade 30-day complications after radical cystectomy. MATERIALS AND METHODS: We retrospectively identified patients who underwent radical cystectomy for bladder cancer who completed Short Form 12 (SF-12) surveys for self-assessment of health status less than 6 months before surgery. Median physical and mental composite scores were calculated. An expert model including known predictors of postoperative high grade complications was developed, and SF-12 physical composite score and mental composite score were added to determine their association with this end point. RESULTS: From January 2010 to August 2014, 472 patients underwent radical cystectomy for bladder cancer, of whom 274 (58.1%) completed preoperative SF-12 questionnaires. Responders were more likely to be white (p=0.024), have higher preoperative albumin (p=0.037), receive neoadjuvant chemotherapy (p=0.002), have pT3/T4 disease (p=0.044) and have positive soft tissue surgical margins (p=0.006). Median SF-12 physical composite score was 43.1 (IQR 33.0-51.5) and mental composite score was 48.5 (IQR 39.5-54.7) in responders. Overall 46 (16.8%) responders experienced a high grade 30-day complication. Patients with a high grade complication had a lower preoperative median SF-12 mental composite score (44.8 vs 49.8, p=0.004) but no difference in physical composite score (39.2 vs 43.8, p=0.06). SF-12 mental composite score was also a significant predictive variable when added to our expert model (p=0.01). CONCLUSIONS: Preoperative patient reported mental health was independently associated with high grade complications after radical cystectomy. Therefore, patient self-assessment of health status before surgery through validated questionnaires may provide additional information useful in predicting short-term postoperative outcomes.

Office Cystoscopy Urinary Tract Infection Rate and Cost before and after Implementing New Handling and Storage Practices.

INTRODUCTION: Based on 2010 American Urological Association recommendations our practice transitioned from sterile to high level disinfection flexible cystoscope reprocessing and from sterile to clean handling practices. We examined symptomatic urinary tract infection rate and cost before and after policy implementation. METHODS: We retrospectively reviewed 30-day outcomes following 1,888 simple cystoscopy encounters that occurred from 2007 to 2010 (sterile, 905) and 2012 to 2015 (high level disinfection, 983) at the Malcom Randall Veterans Affairs Medical Center. We excluded veterans who had recent instrumentation, active or recent urinary tract infection, performed intermittent catheterization, or had complicated cystoscopy (dilation, biopsy etc). Patient/procedural factors and cost were collected and compared between groups. RESULTS: Both cohorts had similar age (mean 68 years), race (Caucasian, 82%), comorbidities (cancer history, 62%; diabetes mellitus, 36%; tobacco use, 24.5%), and cystoscopy procedural indications (cancer surveillance, 50%; hematuria, 34%). Urological complication rate was low between groups (1.43%) with no significant difference in symptomatic urinary tract infection events (0.99% sterile vs 0.51% high level disinfection, p=0.29) or unplanned clinic/emergency department visits (0.66% sterile vs 0.71% high level disinfection, p=0.91). Roughly 95% of the cohorts were given prophylactic antibiotics, most commonly fluoroquinolones (91%). High level disinfection was $82 cheaper per procedure than sterile with most cost disparity stemming from reprocessing. Total savings for our facility by switching to high level disinfection was more than $100,000 annually. CONCLUSIONS: In an older, morbid veteran population receiving centralized care and prophylactic antibiotics we found no difference in symptomatic urinary tract infection or unplanned visits between sterile or high level disinfection techniques. However, high level disinfection was associated with a sizable cost savings, improved clinic workflow, and reduced use of personal protective equipment.

Safety signals for QT prolongation or Torsades de Pointes associated with azithromycin with or without chloroquine or hydroxychloroquine.

Background: Combinations of hydroxychloroquine (HCQ) and azithromycin have been promoted as treatments for COVID-19 based on small, uncontrolled clinical trials that have not assessed potential risks. Risks of treatment include QT segment prolongation, Torsades de Pointes (TdP), and death. This comparative pharmacovigilance analysis evaluated the risk of these events. Methods: Data from the U.S. Food and Drug Administration's Adverse Event Reporting System (FAERS) (>13 million total reports) were used. Queries extracted reports based on exposures of HCQ/chloroquine (CQ) alone, azithromycin alone, HCQ/CQ + azithromycin, amoxicillin alone, HCQ/CQ + amoxicillin alone. Amoxicillin served as a control. Events of interest included death and TdP/QT prolongation as well as accidents/injuries and depression as control events. Proportional Reporting Ratios (PRR) and 95% confidence intervals (CI) were calculated where a lower limit of the of 95% CI (Lower95CI) value of ≥2.0 is interpreted as a potential safety signal. Results: Lower95CIs for HCQ/CQ alone showed no potential safety signals for TdP/QT prolongation, death, or any of the control events included. The PRRs and 95% CIs for TdP/QT prolongation was 1.43 (1.29-2.59) with HCQ/CQ use alone and 4.10 (3.80-4.42) for azithromycin alone. For the combined HCQ/CQ + azithromycin group, the PRR and 95% CI was 3.77 (1.80-7.87). For the control of amoxicillin, there were no safety signals when used alone or in combination with HCQ/CQ. Conclusions: HCQ/CQ use was not associated with a safety signal in this analysis of FAERS data. However, azithromycin used alone was associated with TdP/QT prolongation events and should be used with caution.

Comparative pharmacovigilance assessment of mortality with pimavanserin in Parkinson disease-related psychosis.

BACKGROUND: Pimavanserin is approved for treatment of Parkinson disease (PD)-related psychosis, but its use has been associated with an increased risk of death during clinical trials, as well as during postmarketing surveillance. Previous reports on the association between pimavanserin and mortality have not taken into account limitations of data sources nor included comparable populations or comparisons to relevant treatment alternatives. OBJECTIVE: To conduct a comparative pharmacovigilance assessment of pimavanserin vs treatment alternatives and by restricting surveillance data to more representative populations. METHODS: This was a retrospective analysis of adverse event case reports submitted to the FDA's Adverse Event Reporting System (FAERS) from 2016 through 2019 quarter 3 (Q3). FAERS data are collected from the full population, were further restricted to only those with PD, and were based on PD medication use. Reports were assessed for exposure to pimavanserin, clozapine, quetiapine, haloperidol, and other atypical antipsychotics. The outcome of interest was all-cause death. A proportional reporting ratio (PRR) and 95% confidence limits were calculated for each 2 by 2 contingency of outcome (death) and exposure (pimavanserin and others). For each outcome/exposure pair, the baseline population was altered to include the full FAERS sample, only reports with PD, reports with PD treated with levodopa, and reports with PD treated with multiple PD medications. The sample was also stratified by time period before April 2018 and after September 2018 to capture periods of public knowledge and federal response. A lower 95% CI (Lower95CI) ≥ 2 for the PRR was considered as the accepted threshold for a drug safety signal. RESULTS: As of 2019 Q3, there were 2,287 reports of death associated with pimavanserin. Compared within the full FAERS base population, pimavanserin yielded a PRR Lower95CI = 2.08 but was smaller when restricted to comparison among only a base population with PD (Lower95CI = 1.09), PD treated with levodopa (Lower95CI = 1.15), or PD treated with multiple PD medications (Lower95CI = 1.63). Metrics for quetiapine, clozapine, and other atypical antipsychotics were similar in magnitude. Stratification by time showed a possible reporting bias associated with pimavanserin, since no safety signal was detected before April 2018; however, a signal was present thereafter. CONCLUSIONS: Compared in context with treatment alternatives for patients with PD, pimavanserin was not associated with excess reports of death in the FAERS data. This information should be used in shared decision making between physicians and PD patients to balance the risks and benefits of pimavanserin and other treatments for PD psychosis. DISCLOSURES: No outside funding supported this study. The authors report no disclosures or conflicts of interest relevant to this study. Armstrong receives research support from the NIA (P30AG047266, R01AG068128) and the Florida Department of Health (grant 20A08). She is the local principal investigator of a Lewy Body Dementia Association Research Center of Excellence. She also receives compensation from the American Academy of Neurology for work as an evidence-based medicine methodology consultant. She is on the level of evidence editorial board for Neurology and related publications (uncompensated), receives publishing royalties for Parkinson's Disease: Improving Patient Care (Oxford University Press, 2014), and has received an honorarium for presenting for Medscape CME in 2018. Okun serves as a consultant for the Parkinson's Foundation and has received research grants from NIH, Parkinson's Foundation, the Michael J. Fox Foundation, the Parkinson Alliance, Smallwood Foundation, the Bachmann-Strauss Foundation, the Tourette Syndrome Association, and the UF Foundation. Okun has participated as a site principal investigator and/or co-investigator for several NIH-, foundation-, and industry-sponsored trials over the years but has not received honoraria. Malaty has participated in research funded by the Parkinson Foundation, Tourette Association, Dystonia Coalition, Abbvie, Boston Scientific, Eli Lilly, Neuroderm, Pfizer, Revance, and Teva. She has received travel compensation and/or honoraria from the Tourette Association of America, NeuroChallenge Foundation and NIH/Neurobiology of Disease in Children, Parkinson Foundation, Medscape, International Association of Parkinsonism and Related Disorders, and Cleveland Clinic, and royalties from Robert Rose publishers. The other authors have no disclosures.

Real-World Data Approaches for Early Detection of Potential Safety and Effectiveness Signals for Generic Substitution: A Metoprolol Extended-Release Case Study.

Real-world spontaneous adverse event reports and administrative health care data were utilized as one part of a multipronged approach to verify surveillance signals related to generic drug formulations. This study used metoprolol succinate extended release as a historic case example from which to develop an analytic framework. The US Food and Drug Administration Adverse Event Reporting System was utilized for disproportionality analyses and to identify outcomes of interest. Claims data were analyzed for generic uptake, proportion of prescriptions with "dispense as written" orders, time to discontinuation or switching, and relative rates of clinical events. Adverse Event Reporting System data showed that the Medical Dictionary for Regulatory Activities terms for product quality were higher for generic metoprolol cases and that a number of clinical events were increased that could be side effects of high or low variability in drug levels. Compared to amlodipine-benazepril, which also had a first-approved generic at the same time, market share data showed that metoprolol succinate had lower utilization and more prescriptions written as dispense as written. Switching and discontinuation were generally higher for metoprolol users compared to amlodipine-benazepril users. Finally, clinical event rates were generally higher for generic versus brand metoprolol but lower for the same comparison for amlodipine-benazepril users. In the claims-based analyses, the 90-day period immediately after generic entry provided stronger signal capture than using the entire study period. This analytic framework can be implemented to actively monitor new generic formulations for potential bioequivalence failures. Signals from these analyses require confirmation (eg, via pharmacometric analyses) to be informative for regulatory action.

Common inpatient hypoglycemia phenotypes identified from an automated electronic health record-based prediction model.

PURPOSE: Common inpatient hypoglycemia risk factor patterns (phenotypes) from an electronic health record (EHR)-based prediction model and preventive strategies were identified. METHODS: Patients admitted to 2 large academic medical centers who were in the top fifth percentile of a previously developed hypoglycemia risk score and developed hypoglycemia (blood glucose [BG] of <50mg/dL) were included in the study. Frequencies of all combinations of ≥4 risk factors contributing to the risk score among these patients were determined to identify common risk patterns. Clinical pharmacists developed clinical vignettes for each common pattern and formulated medication therapy management recommendations for hypoglycemia prevention. RESULTS: A total of 401 admissions with a hypoglycemic event were identified among 1,875 admissions whose hypoglycemic risk was in the top fifth percentile among all admissions that received antihyperglycemic drugs and evaluated. Five distinct phenotypes emerged: (1) frail patients with history of hypoglycemia receiving insulin on hospital day 1, (2) a rapid downward trend in BG values in patients receiving an insulin infusion or with a history of hypoglycemia, (3) administration of insulin in the presence of an active nothing by mouth order in frail patients, (4) repeated low BG level in frail patients, and (5) inadequate night-time BG monitoring for patients on long-acting insulin. The 5 themes jointly described 53.0% of high-risk patients who experienced hypoglycemia. CONCLUSION: Five distinct phenotypes that are prevalent in patients at greatest risk for inpatient hypoglycemia were identified.

Development and validation of a dynamic inpatient risk prediction model for clinically significant hypokalemia using electronic health record data.

Purpose: The purpose of this study was to develop a dynamic risk prediction model for inpatient hypokalemia and evaluate its predictive performance. Methods: A retrospective cohort included all admissions aged 18 years and above from 2 large tertiary hospitals in Florida over a 22-month period. Hypokalemia was defined as a potassium value of less than 3 mmol/L, and subsequent initiation of potassium supplements. Twenty-five risk factors (RF) identified from literature were operationalized using discrete electronic health record (EHR) data elements. For each of the first 5 hospital days, we modeled the probability of developing hypokalemia at the subsequent hospital day using logistic regression. Predictive performance of our model was validated with 100 bootstrap datasets and evaluated by the C statistic and Hosmer-Lemeshow goodness-of-fit test. Results: A total of 4511 hypokalemia events occurred over 263 436 hospital days (1.71%). Validated C statistics of the prediction model ranged from 0.83 (Day 1 model) to 0.86 (Day 3), while p-values for the Hosmer-Lemeshow test spanned from 0.005 (Day 1) to 0.27 (Day 4 and 5). For the Day 3 prediction, 9.9% of patients with risk scores in the 90th percentile developed hypokalemia and accounted for 60.4% of all hypokalemia events. After controlling for baseline potassium values, strong predictors included diabetic ketoacidosis, increased mineralocorticoid activity, polyuria, use of kaliuretics, use of potassium supplements and watery stool. Conclusion: This is the first risk prediction model for hypokalemia. Our model achieved excellent discrimination and adequate calibration ability. Once externally validated, this risk assessment tool could use real-time EHR information to identify individuals at the highest risk for hypokalemia and support proactive interventions by pharmacists.

Development and validation of an automated algorithm for identifying patients at higher risk for drug-induced acute kidney injury.

PURPOSE: Using information from institutional electronic health records, we aimed to develop dynamic predictive models to identify patients at high risk of acute kidney injury (AKI) among those who received a nephrotoxic medication during their hospital stay. METHODS: Candidate predictors were measured for each of the first 5 hospital days where a patient received a nephrotoxic medication (risk model days) to predict an AKI, using logistic regression with reduced backward variables elimination in 100 bootstrap samples. An AKI event was defined as an increase of serum creatinine ≥ 200% of a baseline SCr within 5 days after a risk model day. Final models were internally validated by replication in 100 bootstrap samples and a risk score for each patient was calculated from the validated model. As performance measures, the area under the receiver operation characteristic curves (AUC) and the number of AKI events among patients who had high risk scores were estimated. RESULTS: The study population included 62,561 admissions followed by 1,212 AKI events (1.9 events/100 admissions). We constructed 5 risk models corresponding to the first 5 hospital days where patients were exposed to at least one nephrotoxic medication. Validated AUCs of the 5 models ranged from 0.78 to 0.81. Depending on risk model day, admissions ranked in the 90th percentile of the risk score captured between 43% to 49% of all AKI events. CONCLUSION: A dynamic prediction model was built successfully for inpatient AKI with excellent discriminative validity and good calibration, allowing clinicians to focus on a select high-risk population that captures the majority of AKI events.

Development of a predictive model for drug-associated QT prolongation in the inpatient setting using electronic health record data.

PURPOSE: We aimed to construct a dynamic model for predicting severe QT interval prolongation in hospitalized patients using inpatient electronic health record (EHR) data. METHODS: A retrospective cohort consisting of all adults admitted to 2 large hospitals from January 2012 through October 2013 was established. Thirty-five risk factors for severe QT prolongation (defined as a Bazett's formula-corrected QT interval [QTc] of ≥500 msec or a QTc increase of ≥60 msec from baseline) were operationalized for automated EHR retrieval; upon univariate analyses, 26 factors were retained in models for predicting the 24-hour risk of QT events on hospital day 1 (the Day 1 model) and on hospital days 2-5 (the Days 2-5 model). RESULTS: A total of 1,672 QT prolongation events occurred over 165,847 days of risk exposure during the study period. C statistics were 0.828 for the Day 1 model and 0.813 for the Days 2-5 model. Patients in the upper 50th percentile of calculated risk scores experienced 755 of 799 QT events (94%) allocated in the Day 1 model and 804 of 873 QT events (92%) allocated in the Days 2-5 model. Among patients in the 90th percentile, the Day 1 and Days 2-5 models captured 351 of 799 (44%) and 362 of 873 (41%) QT events, respectively. CONCLUSION: The risk models derived from EHR data for all admitted patients had good predictive validity. All risk factors were operationalized from discrete EHR fields to allow full automation for real-time identification of high-risk patients. Further research to test the models in other health systems and evaluate their effectiveness on outcomes and patient care in clinical practice is recommended.

Predicting medication-associated altered mental status in hospitalized patients: Development and validation of a risk model.

PURPOSE: This study presents a medication-associated altered mental status (AMS) risk model for real-time implementation in inpatient electronic health record (EHR) systems. METHODS: We utilized a retrospective cohort of patients admitted to 2 large hospitals between January 2012 and October 2013. The study population included admitted patients aged ≥18 years with exposure to an AMS risk-inducing medication within the first 5 hospitalization days. AMS events were identified by a measurable mental status change documented in the EHR in conjunction with the administration of an atypical antipsychotic or haloperidol. AMS risk factors and AMS risk-inducing medications were identified from the literature, drug information databases, and expert opinion. We used multivariate logistic regression with a full and backward eliminated set of risk factors to predict AMS. The final model was validated with 100 bootstrap samples. RESULTS: During 194,156 at-risk days for 66,875 admissions, 262 medication-associated AMS events occurred (an event rate of 0.13%). The strongest predictors included a history of AMS (odds ratio [OR], 9.55; 95% confidence interval [CI], 5.64-16.17), alcohol withdrawal (OR, 3.34; 95% CI, 2.18-5.13), history of delirium or psychosis (OR, 3.25; 95% CI, 2.39-4.40), presence in the intensive care unit (OR, 2.53; 95% CI, 1.89-3.39), and hypernatremia (OR, 2.40; 95% CI, 1.61-3.56). With a C statistic of 0.85, among patients scoring in the 90th percentile, our model captured 159 AMS events (60.7%). CONCLUSION: The risk model was demonstrated to have good predictive ability, with all risk factors operationalized from discrete EHR fields. The real-time identification of higher-risk patients would allow pharmacists to prioritize surveillance, thus allowing early management of precipitating factors.

A dynamic risk model for inpatient falls.

PURPOSE: Construction and validation of a fall risk prediction model specific to inpatients receiving fall risk-increasing drugs (FRIDs) are described. METHODS: In a retrospective cohort study of 75,036 admissions to 2 hospitals over a designated 22-month period that involved FRID exposure during the first 5 hospital days, factors influencing fall risk were investigated via logistic regression. The resultant risk prediction model was internally validated and its performance compared with that of a model based on Morse Fall Scale (MFS) scores. RESULTS: A total of 220,904 patient-days of FRID exposure were evaluated. The three most commonly administered FRIDs were oxycodone (given on 79,697 patient-days, 36.08%), morphine (52,427, 23.73%) and hydromorphone (42,063, 19.04%). Within the 90th percentile of modeled risk scores, 144 of the 466 documented falls (30.9%) were captured by the developed risk prediction model (unbiased C statistic, 0.69), as compared with 94 falls (20.2%) captured using the MFS model (unbiased C statistic, 0.62). Strong predictors of inpatient falls included a history of falling (odds ratio [OR], 1.99; 95% confidence interval (CI), 1.42-2.80); overestimation of ability to ambulate (OR, 1.53; 95% CI, 1.12-2.09); and "comorbidity predisposition," a composite measure encompassing a history of falling and 11 past diagnoses (OR, 1.60; 95% CI, 1.30-1.97). CONCLUSION: The proposed risk model for inpatient falls achieved superior predictive performance when compared with the MFS model. All risk factors were operationalized from discrete electronic health record fields, allowing full automation of real-time identification of high-risk patients.

Development and validation of an automated algorithm for identifying patients at high risk for drug-induced hypoglycemia.

PURPOSE: Hypoglycemia is one of the most concerning adverse drug events in hospitalized patients. Using information from institutional electronic health records, we aimed to develop dynamic predictive models to identify patients at high risk for hypoglycemia during antihyperglycemic therapy. METHODS: The study population consisted of 21,840 patients who received antihyperglycemic medication on any of the first 5 hospital days (the "risk model days") at 2 large hospitals. Data on candidate predictors were extracted from discrete electronic health record fields to construct models for predicting hypoglycemia within 24 hours after each risk model day. Final models were internally validated by replication in 100 bootstrap samples and reapplying model parameters to the original risk population. RESULTS: The development and validation sample included 60,762 risk model days followed by 1,256 days with hypoglycemic events (2.07 events per 100 risk model days). The days 3, 4, and 5 models presented similar associations between predictors and the risk of hypoglycemia and were therefore collapsed into a single model. The strongest hypoglycemia risk factors across all 3 risk periods (day 1, day 2, and days 3-5) were blood glucose (BG) fluctuations, BG trend, history of hypoglycemia, lower body weight, lower creatinine clearance, use of long-acting or high-dose insulin, and sulfonylurea use. C statistics for the 3 models ranged from 0.844 to 0.887. Depending on the model used, risk scores in the upper 90th percentile predicted 48.5-63.1% of actual hypoglycemic events. It was estimated that by targeting only patients in the upper 90th percentile, providers would need to intervene during fewer than 9 admissions to prevent 1 hypoglycemic event. CONCLUSION: The developed prediction models were found to have excellent discriminative validity and good calibration, allowing clinicians to focus interventions on a select high-risk population in which the majority of hypoglycemic events occur.

Image Quality and Patient-Specific Organ Doses in Stone Protocol CT: A Comparison of Traditional CT to Low Dose CT with Iterative Reconstruction.

OBJECTIVE: To compare organ specific radiation dose and image quality in kidney stone patients scanned with standard CT reconstructed with filtered back projection (FBP-CT) to those scanned with low dose CT reconstructed with iterative techniques (IR-CT). MATERIALS AND METHODS: Over a one-year study period, adult kidney stone patients were retrospectively netted to capture the use of noncontrasted, stone protocol CT in one of six institutional scanners (four FBP and two IR). To limit potential CT-unit use bias, scans were included only from days when all six scanners were functioning. Organ dose was calculated using volumetric CT dose index and patient effective body diameter through validated conversion equations derived from previous cadaveric, dosimetry studies. Board-certified radiologists, blinded to CT algorithm type, assessed stone characteristics, study noise, and image quality of both techniques. RESULTS: FBP-CT (n=250) and IR-CT (n=90) groups were similar in regard to gender, race, body mass index (mean BMI = 30.3), and stone burden detected (mean size 5.4 ± 1.2 mm). Mean organ-specific dose (OSD) was 54-62% lower across all organs for IR-CT compared to FBP-CT with particularly reduced doses (up to 4.6-fold) noted in patients with normal BMI range. No differences were noted in radiological assessment of image quality or noise between the cohorts, and intrarater agreement was highly correlated for noise (AC2=0.873) and quality (AC2=0.874) between blinded radiologists. CONCLUSIONS: Image quality and stone burden assessment were maintained between standard FBP and low dose IR groups, but IR-CT decreased mean OSD by 50%. Both urologists and radiologists should advocate for low dose CT, utilizing reconstructive protocols like IR, to reduce radiation exposure in their stone formers who undergo multiple CTs.

Measurement of selected preventable adverse drug events in electronic health records: Toward developing a complexity score.

PURPOSE: The defining of a select number of high-priority preventable adverse drug events (pADEs) for measurement in the electronic health record (EHR) and the estimation of pADE incidences in two tertiary care facilities are described. METHODS: This study was part of a larger effort aimed at developing an automated electronic health record (EHR)-based complexity-score (C-score) that ranks hospitalized patients according to their risk for pADEs for clinical intervention. We developed measures for 16 high-priority pADEs often deemed preventable using discrete clinical and administrative EHR data. For each pADE we specified inclusion and exclusion criteria that were used to define risk populations for each specific pADE. The incidence of each type of pADE was then measured during a designated follow-up period considering all adult admissions to 2 large academic tertiary care hospitals, who were eligible for the pADE-specific risk populations during any of their first 5 hospital days. RESULTS: Utilizing the data from 83,787 admissions who were at risk for at least one pADE during at least one of their first five hospital days, we found that 27,193 admissions (32.5%) developed at least one pADE. Uncontrolled postsurgical pain, uncontrolled pneumonia, and drug-associated hypotension had the highest incidences with the following number of days with pADE per number of patients at risk: 13,484 of 19,640; 527 of 1,530; and 13,394 of 43,630, while drug-associated falls (446 of 75,036), drug-associated acute mental status changes (262 of 66,875) and venous thromboembolism (214 of 74,283) had the lowest incidence rates. CONCLUSION: EHR-based definitions of clinically important pADEs were developed, and the incidence of the pADEs was estimated. These definitions will be advanced for the creation of prediction models to develop a C-score for identifying patients at risk for pADEs to prioritize pharmacist intervention.

Development and validation of a complexity score to rank hospitalized patients at risk for preventable adverse drug events.

PURPOSE: The development of risk models for 16 preventable adverse drug events (pADEs) and their aggregation into the final complexity score (C-score) are described. METHODS: Using data from 2 tertiary care facilities, logistic regression models were constructed for the first 5 hospital days that admissions were at risk for each of 16 pADEs. The best model for each pADE was validated in 100 bootstrap samples. The C-score was then aggregated and predicted individual pADE risk as the probability to develop at least 1 pADE. Using the 100 bootstrap samples for each pADE, 100 C-scores for validation were generated. RESULTS: We utilized electronic health records (EHR) data from 65,518 admissions to UF Health Shands and 18,269 admissions to UF Health Jacksonville to develop risk models for 16 pADEs. Most models had very strong discriminant validity (C-statistic > 0.8), with the highest predicted decile representing about half of manifest pADEs. Among admissions in the highest C-score decile, about two thirds experienced at least 1 pADE (C-statistic, 0.838; 95% confidence interval, 0.838-0.839). C-score precision, defined as the percentage of patients consistently (i.e., at least 95 of 100 samples) ranked in the 90th percentile, was 80-84%. CONCLUSION: The C-score was developed and validated for the identification of hospitalized patients at highest risk for pADEs. Aggregation of individual prediction models into a single score reduced its predictive power for most pADEs, compared with the individual risk models, but concentrated in the highest C-score decile a patient group more than two thirds of whom experienced at least 1 pADE.

Prediction of single procedure success rate using S.T.O.N.E. nephrolithometry surgical classification system with strict criteria for surgical outcome.

OBJECTIVE: To evaluate the S.T.O.N.E. nephrolithometry scoring system for percutaneous nephrolithotomy using computerized tomography (CT) imaging with strict criteria for stone clearance. MATERIALS AND METHODS: We analyzed a cohort of 122 patients who consecutively underwent primary percutaneous nephrolithotomy from July 2010 to March 2012 at our university-based referral hospital. All patients routinely have preoperative and postoperative CT imaging for stone burden determination. Primary outcome (residual stone) was scored as 0-2, 3-4, and >4 mm. All S.T.O.N.E. nephrolithometry parameters were recorded and scored as per published definition. The t test was used for continuous variables, and the chi-square testing or the Fisher exact test (when counts were small) was used for categorical covariates. S.T.O.N.E. score correlation with stone-free status was analyzed by logistic regression. RESULTS: Nephrolithometry score ranged from 5 to 13 with a mean of 9.5. Postoperative CT for residual stone showed 67 (54.9%), 26 (21.3%), and 29 (23.8%) patients had 0-2, 3-4, and >4 mm residual stone, respectively. Mean nephrolithometry scores for residual stone of 0-2, 3-4, and >4 mm were 8.87, 9.73, and 10.79 respectively (P <.0001). There were 11 (9.8%) complications. CONCLUSION: With use of strict CT imaging criteria for assessment of residual stone status, the S.T.O.N.E. scoring system is reproducible and predictive of treatment success. Further investigation is required to both validate this model and to determine if other predictive parameters will improve it as a predictive model.