Brian C. Sauer

Guidelines for Good Database Selection and use in Pharmacoepidemiology Research

The use of healthcare databases in research provides advantages such as increased speed, lower costs and limitation of some biases. However, database research has its own challenges as studies must be performed within the limitations of resources, which often are the product of complex healthcare systems. The primary purpose of this document is to assist in the selection and use of data resources in pharmacoepidemiology, highlighting potential limitations and recommending tested procedures. This guidance is presented as a detailed text with a checklist for quick reference and covers six areas: selection of a database, use of multiple data resources, extraction and analysis of the study population, privacy and security, quality and validation procedures and documentation. Copyright

Merging Veterans Affairs Rheumatoid Arthritis Registry and Pharmacy Data to Assess Methotrexate Adherence and Disease Activity in Clinical Practice

The Veterans Affairs (VA) Rheumatoid Arthritis (VARA) registry and the VA Pharmacy Benefits Management database were linked to determine the association of methotrexate (MTX) adherence with rheumatoid arthritis (RA) disease activity.

A review of covariate selection for non-experimental comparative effectiveness research†

This paper addresses strategies for selecting variables for adjustment in non‐experimental comparative effectiveness research and uses causal graphs to illustrate the causal network that relates treatment to outcome. Variables in the causal network take on multiple structural forms. Adjustment for a common cause pathway between treatment and outcome can remove confounding, whereas adjustment for other structural types may increase bias. For this reason, variable selection would ideally be based on an understanding of the causal network; however, the true causal network is rarely known. Therefore, we describe more practical variable selection approaches based on background knowledge when the causal structure is only partially known. These approaches include adjustment for all observed pretreatment variables thought to have some connection to the outcome, all known risk factors for the outcome, and all direct causes of the treatment or the outcome. Empirical approaches, such as forward and backward selection and automatic high‐dimensional proxy adjustment, are also discussed. As there is a continuum between knowing and not knowing the causal, structural relations of variables, we recommend addressing variable selection in a practical way that involves a combination of background knowledge and empirical selection and that uses high‐dimensional approaches. This empirical approach can be used to select from a set of a priori variables based on the researchers knowledge to be included in the final analysis or to identify additional variables for consideration. This more limited use of empirically derived variables may reduce confounding while simultaneously reducing the risk of including variables that may increase bias. Copyright

Developing indicators of inpatient adverse drug events through nonlinear analysis using administrative data.

Background:Because of uniform availability, hospital administrative data are appealing for surveillance of adverse drug events (ADEs). Expert-generated surveillance rules that rely on the presence of International Classification of Diseases, 9th Revision Clinical Modification (ICD-9-CM) codes have limited accuracy. Rules based on nonlinear associations among all types of available administrative data may be more accurate. Objectives:By applying hierarchically optimal classification tree analysis (HOCTA) to administrative data, derive and validate surveillance rules for bleeding/anticoagulation problems and delirium/psychosis. Research Design:Retrospective cohort design. Subjects:A random sample of 3987 admissions drawn from all 41 Utah acute-care hospitals in 2001 and 2003. Measures:Professional nurse reviewers identified ADEs using implicit chart review. Pharmacists assigned Medical Dictionary for Regulatory Activities codes to ADE descriptions for identification of clinical groups of events. Hospitals provided patient demographic, admission, and ICD9-CM data. Results:Incidence proportions were 0.8% for drug-induced bleeding/anticoagulation problems and 1.0% for drug-induced delirium/psychosis. The model for bleeding had very good discrimination and sensitivity at 0.87 and 86% and fair positive predictive value (PPV) at 12%. The model for delirium had excellent sensitivity at 94%, good discrimination at 0.83, but low PPV at 3%. Poisoning and adverse event codes designed for the targeted ADEs had low sensitivities and, when forced in, degraded model accuracy. Conclusions:Hierarchically optimal classification tree analysis is a promising method for rapidly developing clinically meaningful surveillance rules for administrative data. The resultant model for drug-induced bleeding and anticoagulation problems may be useful for retrospective ADE screening and rate estimation.

Overestimation of the effects of adherence on outcomes: a case study in healthy user bias and hypertension

Background The healthy user bias is usually overlooked as an explanation in studies in which a strong association is found between poor patient medication adherence and worse disease outcomes. Such studies are increasing in frequency across disease states and influence clinical practice. Adherence to antihypertensive medications was studied to illustrate confounding in such studies. Methods Using data from veterans with hypertension starting antihypertensive treatment, causal models were developed that predicted the risks of hospitalisation, myocardial infarction (MI) and death associated with poor adherence (<80%) while adjusting for patient demographics, baseline disease severity and disease comorbidity. In a second set of otherwise identical models, adjustment was made for time-varying blood pressure (BP), thus controlling for adherence effects that were mediated through the main pharmacological effects of the drugs. It was hypothesised that the second set of models would reveal a positive association between poor adherence and adverse disease outcomes that is largely explained by unmeasured confounders, including health-related behaviours. Results The models that did not adjust for time-varying BP levels showed that patients with poor adherence had statistically significantly increased risks of 3.7% for hospitalisation, 28.1% for MI and 23.3% for death. These estimates exceed the benefits of these drugs demonstrated by clinical trials. When controlling for time-varying BP, the increased risks were similar (3.4% for hospitalisation, 27.7% for MI and 23.4% for death). The findings were consistent across a range of adherence thresholds (50–90%) and when allowing disease status variables to vary. Conclusions The associations between poor adherence and outcomes are largely independent of the pharmacological effects of the drugs on BP control as well as commonly measured patient covariates. This finding suggests that even carefully designed observational adherence studies using rich clinical data are impossibly confounded and probably overestimate the true magnitude of the effect. Clinical practice guidelines based on reported adherence effects should be reconsidered.

A simulation-based evaluation of methods to estimate the impact of an adverse event on hospital length of stay

Introduction:We used agent-based simulation to examine the problem of time-varying confounding when estimating the effect of an adverse event on hospital length of stay. Conventional analytic methods were compared with inverse probability weighting (IPW). Methods:A cohort of hospitalized patients, at risk for experiencing an adverse event, was simulated. Synthetic individuals were assigned a severity of illness score on admission. The score varied during hospitalization according to an autoregressive equation. A linear relationship between severity of illness and the logarithm of the discharge rate was assumed. Depending on the model conditions, adverse event status was influenced by prior severity of illness and, in turn, influenced subsequent severity. Conditions were varied to represent different levels of confounding and categories of effect. The simulation output was analyzed by Cox proportional hazards regression and by a weighted regression analysis, using the method of IPW. The magnitude of bias was calculated for each method of analysis. Results:Estimates of the population causal hazard ratio based on IPW were consistently unbiased across a range of conditions. In contrast, hazard ratio estimates generated by Cox proportional hazards regression demonstrated substantial bias when severity of illness was both a time-varying confounder and intermediate variable. The direction and magnitude of bias depended on how severity of illness was incorporated into the Cox regression model. Conclusions:In this simulation study, IPW exhibited less bias than conventional regression methods when used to analyze the impact of adverse event status on hospital length of stay.

Factors associated with screening or treatment initiation among male United States veterans at risk for osteoporosis fracture.

Male osteoporosis continues to be under-recognized and undertreated in men. An understanding of which factors cue clinicians about osteoporosis risk in men, and which do not, is needed to identify areas for improvement. This study sought to measure the association of a providers recognition of osteoporosis with patient information constructs that are available at the time of each encounter. Using clinical and administrative data from the Veterans Health Administration system, we used a stepwise procedure to construct prognostic models for a combined outcome of osteoporosis diagnosis, treatment, or a bone mineral density (BMD) test order using time-varying covariates and Cox regression. We ran separate models for patients with at least one primary care visit and patients with only secondary care visits in the pre-index period. Some of the strongest predictors of clinical osteoporosis identification were history of gonadotropin-releasing hormone (GnRH) agonist exposure, fragility fractures, and diagnosis of rheumatoid arthritis. Other characteristics associated with a higher likelihood of having osteoporosis risk recognized were underweight or normal body mass index, cancer, fall history, and thyroid disease. Medication exposures associated with osteoporosis risk recognition included opioids, glucocorticoids, and antidepressants. Several known clinical risk factors for fracture were not correlated with osteoporosis risk including smoking and alcohol abuse. Results suggest that clinicians are relying on some, but not all, clinical risk factors when assessing osteoporosis risk.

Persistence and Dose Escalation of Tumor Necrosis Factor Inhibitors in US Veterans with Rheumatoid Arthritis

Objective. Limited evidence exists comparing the persistence, effectiveness, and costs of biologic therapies for rheumatoid arthritis in clinical practice. Comparative effectiveness studies are needed to understand real-world experience with these agents. We evaluated treatment patterns, costs, and effectiveness of tumor necrosis factor inhibitor (TNFi) agents in patients enrolled in the Veterans Affairs Rheumatoid Arthritis (VARA) registry. Methods. Observational data from the VARA registry and linked administrative databases were analyzed. Longitudinal data from VARA patients initiating adalimumab (ADA), etanercept (ETN), or infliximab (IFX) from 2003 (the date all agents were available within the Veteran Affairs) to 2010 were analyzed. Outcomes included Disease Activity Score using 28 joints (DAS28), treatment persistence, dose escalation, and direct costs of drugs and drug administration. Results. For 563 eligible patients, baseline DAS28, DAS28 improvements, and persistence on initial treatment were similar across agents. Fewer patients receiving ETN (n = 5/290; 2%) underwent dose escalation than did patients taking ADA (n = 32/204; 16%) or IFX (n = 44/69; 64%). Annual costs for first course of TNFi therapy were lower for injectable ADA (

Comparison of opioid-related deaths by work-related injury

13,100 US) and ETN (

Opioid prescribing knowledge and practices: Provider survey following promulgation of guidelines–Utah, 2011

13,500 US) than for intravenously administered IFX (