Showing papers in "Pharmacoepidemiology and Drug Safety in 2012"

One-to-many propensity score matching in cohort studies

Abstract: Background Among the large number of cohort studies that employ propensity score matching, most match patients 1:1. Increasing the matching ratio is thought to improve precision but may come with a trade-off with respect to bias. Objective To evaluate several methods of propensity score matching in cohort studies through simulation and empirical analyses. Methods We simulated cohorts of 20 000 patients with exposure prevalence of 10%–50%. We simulated five dichotomous and five continuous confounders. We estimated propensity scores and matched using digit-based greedy (“greedy”), pairwise nearest neighbor within a caliper (“nearest neighbor”), and a nearest neighbor approach that sought to balance the scores of the comparison patient above and below that of the treated patient (“balanced nearest neighbor”). We matched at both fixed and variable matching ratios and also evaluated sequential and parallel schemes for the order of formation of 1:n match groups. We then applied this same approach to two cohorts of patients drawn from administrative claims data. Results Increasing the match ratio beyond 1:1 generally resulted in somewhat higher bias. It also resulted in lower variance with variable ratio matching but higher variance with fixed. The parallel approach generally resulted in higher mean squared error but lower bias than the sequential approach. Variable ratio, parallel, balanced nearest neighbor matching generally yielded the lowest bias and mean squared error. Conclusions 1:n matching can be used to increase precision in cohort studies. We recommend a variable ratio, parallel, balanced 1:n, nearest neighbor approach that increases precision over 1:1 matching at a small cost in bias. Copyright © 2012 John Wiley & Sons, Ltd.

The U.S. Food and Drug Administration's Mini-Sentinel program: status and direction.

Abstract: The Mini-Sentinel is a pilot program that is developing methods, tools, resources, policies, and procedures to facilitate the use of routinely collected electronic healthcare data to perform active surveillance of the safety of marketed medical products, including drugs, biologics, and medical devices. The U.S. Food and Drug Administration (FDA) initiated the program in 2009 as part of its Sentinel Initiative, in response to a Congressional mandate in the FDA Amendments Act of 2007. After two years, Mini-Sentinel includes 31 academic and private organizations. It has developed policies, procedures, and technical specifications for developing and operating a secure distributed data system comprised of separate data sets that conform to a common data model covering enrollment, demographics, encounters, diagnoses, procedures, and ambulatory dispensing of prescription drugs. The distributed data sets currently include administrative and claims data from 2000 to 2011 for over 300 million person-years, 2.4 billion encounters, 38 million inpatient hospitalizations, and 2.9 billion dispensings. Selected laboratory results and vital signs data recorded after 2005 are also available. There is an active data quality assessment and characterization program, and eligibility for medical care and pharmacy benefits is known. Systematic reviews of the literature have assessed the ability of administrative data to identify health outcomes of interest, and procedures have been developed and tested to obtain, abstract, and adjudicate full-text medical records to validate coded diagnoses. Mini-Sentinel has also created a taxonomy of study designs and analytical approaches for many commonly occurring situations, and it is developing new statistical and epidemiologic methods to address certain gaps in analytic capabilities. Assessments are performed by distributing computer programs that are executed locally by each data partner. The system is in active use by FDA, with the majority of assessments performed using customizable, reusable queries (programs). Prospective and retrospective assessments that use customized protocols are conducted as well. To date, several hundred unique programs have been distributed and executed. Current activities include active surveillance of several drugs and vaccines, expansion of the population, enhancement of the common data model to include additional types of data from electronic health records and registries, development of new methodologic capabilities, and assessment of methods to identify and validate additional health outcomes of interest. Copyright © 2012 John Wiley & Sons, Ltd. key words—Mini-Sentinel; FDA; U.S. Food and Drug Administration; Sentinel Initiative; FDA Amendments Act of 2007

A systematic review of validated methods for identifying atrial fibrillation using administrative data.

Abstract: Purpose The objectives of this study were to characterize the validity of algorithms to identify AF from electronic health data through a systematic review of the literature and to identify gaps needing further research. Methods Two reviewers examined publications during 1997–2008 that identified patients with atrial fibrillation (AF) from electronic health data and provided validation information. We abstracted information including algorithm sensitivity, specificity, and positive predictive value (PPV). Results We reviewed 544 abstracts and 281 full-text articles, of which 18 provided validation information from 16 unique studies. Most used data from before 2000, and 10 of 16 used only inpatient data. Three studies incorporated electronic ECG data for case identification or validation. A large proportion of prevalent AF cases identified by ICD-9 code 427.31 were valid (PPV 70%–96%, median 89%). Seven studies reported algorithm sensitivity (range, 57%–95%, median 79%). One study validated an algorithm for incident AF and reported a PPV of 77%. Conclusions The ICD-9 code 427.31 performed relatively well, but conclusions about algorithm validity are hindered by few recent data, use of nonrepresentative populations, and a disproportionate focus on inpatient data. An optimal contemporary algorithm would likely draw on inpatient and outpatient codes and electronic ECG data. Additional research is needed in representative, contemporary populations regarding algorithms that identify incident AF and incorporate electronic ECG data. Copyright © 2012 John Wiley & Sons, Ltd.

A systematic review of validated methods for identifying cerebrovascular accident or transient ischemic attack using administrative data.

Abstract: Purpose To perform a systematic review of the validity of algorithms for identifying cerebrovascular accidents (CVAs) or transient ischemic attacks (TIAs) using administrative and claims data. Methods PubMed and Iowa Drug Information Service searches of the English language literature were performed to identify studies published between 1990 and 2010 that evaluated the validity of algorithms for identifying CVAs (ischemic and hemorrhagic strokes, intracranial hemorrhage, and subarachnoid hemorrhage) and/or TIAs in administrative data. Two study investigators independently reviewed the abstracts and articles to determine relevant studies according to pre-specified criteria. Results A total of 35 articles met the criteria for evaluation. Of these, 26 articles provided data to evaluate the validity of stroke, seven reported the validity of TIA, five reported the validity of intracranial bleeds (intracerebral hemorrhage and subarachnoid hemorrhage), and 10 studies reported the validity of algorithms to identify the composite endpoints of stroke/TIA or cerebrovascular disease. Positive predictive values (PPVs) varied depending on the specific outcomes and algorithms evaluated. Specific algorithms to evaluate the presence of stroke and intracranial bleeds were found to have high PPVs (80% or greater). Algorithms to evaluate TIAs in adult populations were generally found to have PPVs of 70% or greater. Conclusions The algorithms and definitions to identify CVAs and TIAs using administrative and claims data differ greatly in the published literature. The choice of the algorithm employed should be determined by the stroke subtype of interest. Copyright © 2012 John Wiley & Sons, Ltd.

A systematic review of validated methods for identifying heart failure using administrative data

Abstract: Purpose To identify and describe the validity of algorithms used to detect heart failure (HF) using administrative and claims data sources Methods A systematic review of PubMed and Iowa Drug Information Service searches of the English language was performed to identify studies published between 1990 and 2010 that evaluated the validity of algorithms for the identification of patients with HF using and claims data Abstracts and articles were reviewed by two study investigators to determine their relevance on the basis of predetermined criteria Results The initial search strategy identified 887 abstracts Of these, 499 full articles were reviewed and 35 studies included data to evaluate the validity of identifying patients with HF Positive predictive values (PPVs) were in the acceptable to high range, with most being very high (>90%) Studies that included patients with a primary hospital discharge diagnosis of International Classification of Diseases, Ninth Revision, code 428X had the highest PPV and specificity for HF PPVs for this algorithm ranged from 84% to 100% This algorithm, however, may compromise sensitivity because many HF patients are managed on an outpatient basis The most common ‘gold standard’ for the validation of HF was the Framingham Heart Study criteria Conclusions The algorithms and definitions used to identify HF using administrative and claims data perform well, particularly when using a primary hospital discharge diagnosis Attention should be paid to whether patients who are managed on an outpatient basis are included in the study sample Including outpatient codes in the described algorithms would increase the sensitivity for identifying new cases of HF Copyright © 2012 John Wiley & Sons, Ltd key words—congestive heart failure; validation; administrative data

Design considerations, architecture, and use of the Mini-Sentinel distributed data system

Abstract: Purpose We describe the design, implementation, and use of a large, multiorganizational distributed database developed to support the Mini-Sentinel Pilot Program of the US Food and Drug Administration (FDA). As envisioned by the US FDA, this implementation will inform and facilitate the development of an active surveillance system for monitoring the safety of medical products (drugs, biologics, and devices) in the USA. Methods A common data model was designed to address the priorities of the Mini-Sentinel Pilot and to leverage the experience and data of participating organizations and data partners. A review of existing common data models informed the process. Each participating organization designed a process to extract, transform, and load its source data, applying the common data model to create the Mini-Sentinel Distributed Database. Transformed data were characterized and evaluated using a series of programs developed centrally and executed locally by participating organizations. A secure communications portal was designed to facilitate queries of the Mini-Sentinel Distributed Database and transfer of confidential data, analytic tools were developed to facilitate rapid response to common questions, and distributed querying software was implemented to facilitate rapid querying of summary data. Results As of July 2011, information on 99 260 976 health plan members was included in the Mini-Sentinel Distributed Database. The database includes 316 009 067 person-years of observation time, with members contributing, on average, 27.0 months of observation time. All data partners have successfully executed distributed code and returned findings to the Mini-Sentinel Operations Center. Conclusion This work demonstrates the feasibility of building a large, multiorganizational distributed data system in which organizations retain possession of their data that are used in an active surveillance system. Copyright © 2012 John Wiley & Sons, Ltd.

Guidelines for good database selection and use in pharmacoepidemiology research.

Abstract: 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 © 2011 John Wiley & Sons, Ltd. key words—Health databases; research design; epidemiology; pharmaceuticals; guidance

A systematic review of validated methods for identifying venous thromboembolism using administrative and claims data

Abstract: Background Venous thromboembolism (VTE) is a serious complication. Large claims databases can potentially identify the effects that medications have on VTE. The purpose of this study is to evaluate the evidence supporting the validity of VTE codes. Methods A search of MEDLINE database is supplemented by manual searches of bibliographies of key relevant articles. We selected all studies in which a claim code was validated against a medical record. We reported the positive predictive value (PPV) for the VTE claim compared to the medical record. Results Our search strategy yielded 345 studies, of which only 19 met our eligibility criteria. All of the studies reported on ICD-9 codes, but only two studies reported on pharmacy codes, and one study reported on procedure codes. The highest PPV (65%–95%) was reported for the combined use of ICD-9 codes 415 (pulmonary embolism), 451, and 453 (deep vein thrombosis) as a VTE event. If a specific event like DVT (PPV 24%–92%) or PE (PPV 31%–97%) was evaluated, the PPV was lower than when the combined events were examined. Studies that included patients after orthopedic surgery reported the highest PPV (96%–100%). Conclusions The use of ICD-9 415, 451, and 453 are appropriate for the identification of VTE in claims databases. The codes performed best when codes were evaluated in patients at higher risk of VTE. Copyright © 2012 John Wiley & Sons, Ltd.

Frequency of adverse drug reactions in hospitalized patients: a systematic review and meta‐analysis

Abstract: Purposes To perform a comprehensive systematic review of prospective studies about frequency of adverse drug reactions (ADRs) occurring during hospitalization (ADRIn), including a thorough study quality assessment, meta-analysis and heterogeneity evaluation. Methods Systematic review of several databases: Pubmed, EMBASE, CINAHL, Cochrane, ISI, International Pharmaceutical Abstracts, Scirus, NHS economic, and others, as well as manual search. Inclusion criteria were: prospective studies (assessing all patients before discharge, by a specialized team, at least once a week); with data about ADRs occurring during hospitalization, using WHO's or similar definition of ADR. Two independent reviewers assessed eligibility criteria, extracted data, and evaluated risk of bias. Results From 4139 studies initially found, 22 were included. Meta-analysis indicate that ADRs may occur in 16.88% (CI95%: 13.56,20.21%) of patients during hospitalization; however, this estimate has to be viewed with caution because there was significant heterogeneity (I2 = 99%). The most significant moderators of heterogeneity were risk of bias, population, ward, and methodology for ADR identification. Low risk of bias studies adjusted for population (pediatric versus adult) had I2 = 0%. Conclusions These data are useful as a broad characterization of in-hospital ADRs and their frequency. However, due to heterogeneity, our estimates are crude indicators. The wide variation in methodologies was one of the most important moderators of heterogeneity (even among studies using intensive monitoring). We suggest criteria to standardize methodologies and reduce the risk of bias. Copyright © 2012 John Wiley & Sons, Ltd.

When should case-only designs be used for safety monitoring of medical products?

Abstract: Purpose To assess case-only designs for surveillance with administrative databases. Methods We reviewed literature on two designs that are observational analogs to crossover experiments: the self-controlled case series (SCCS) and the case-crossover (CCO) design. Results SCCS views the ‘experiment’ prospectively, comparing outcome risks in windows with different exposures. CCO retrospectively compares exposure frequencies in case and control windows. The main strength of case-only designs is they entail self-controlled analyses that eliminate confounding and selection bias by time-invariant characteristics not recorded in healthcare databases. They also protect privacy and are computationally efficient, as they require fewer subjects and variables. They are better than cohort designs for investigating transient effects of accurately recorded preventive agents, for example, vaccines. They are problematic if timing of self-administration is sporadic and dissociated from dispensing times,forexample, analgesics. Theytendtohave lessexposure misclassificationbiasandtime-varyingconfounding ifexposures are brief. Standard SCCS designs are bidirectional (using time both before and after the first exposure event), so they are more susceptible than CCOstoreverse-causalitybias,includingimmortal-timebias.Thisistruealsoforsequencesymmetryanalysis,asimplifiedSCCS.Unidirectional CCOs use only time before the outcome, so they are less affected by reverse causality but susceptible to exposure-trend bias. Modifications of SCCS and CCO partially deal with these biases. The head-to-head comparison of multiple products helps to control residual biases. Conclusion The case-only analyses of intermittent users complement the cohort analyses of prolonged users because their different biases compensate for one another. Copyright © 2012 John Wiley & Sons, Ltd. key words—methods; safety monitoring; self-controlled designs; crossover Pharmacoepidemiologists who monitor the safety of medical products using healthcare administrative databases are increasingly interested to know when case-only designs can or cannot be used. To address this question, we (i) defined case-only designs in relation to each other; (ii) examined their main strength: self-controlled comparisons; (iii) discussed the major difference among the designs: directionality; (iv) described the range of medical products assessed with these designs in relation to their susceptibility to exposure misclassification; and (v) made recommendations to safety surveillance programs.

Validation of three coding algorithms to identify patients with end-stage liver disease in an administrative database.

Abstract: PURPOSE: Use of administrative or population-based databases for post-marketing pharmacoepidemiology research in patients with end-stage liver disease (ESLD) has been limited by the difficulty of accurately identifying such patients. Algorithms to identify patients with ESLD using ICD-9-CM codes have not been developed outside of the Veterans Affairs healthcare setting. METHODS: We queried electronic medical records at two tertiary care hospitals to identify patients with ICD-9-CM codes indicative of ESLD. Coding algorithms were developed to identify patients with confirmed ESLD, and these were tested to determine their positive predictive value (PPV). RESULTS: The presence of one inpatient or outpatient ICD-9-CM code for: (i) cirrhosis; (ii) chronic liver disease, and (iii) a hepatic decompensation event yielded a PPV of 85.2% (167/196; 95% CI: 79.4%-89.9%). The PPV increased to 89.3% (150/168; 95% CI: 83.6%-93.5%) when the algorithm required two or more ICD-9-CM codes for a hepatic decompensation. However, an algorithm requiring only one ICD-9-CM code for (i) cirrhosis and (ii) a hepatic decompensation event, in the absence of a chronic liver disease code, yielded a PPV of 85.7% (30/35; 95% CI: 69.7%-95.2%). CONCLUSIONS: A coding algorithm that includes at least one ICD-9-CM code for cirrhosis plus one ICD-9-CM code for a hepatic decompensation event has a high PPV for identifying patients with ESLD. The inclusion of at least two codes indicative of a hepatic decompensation event increased the PPV. This algorithm can be used in future epidemiologic studies to examine the outcomes of a variety of long-term medical therapies in patients with ESLD. Copyright © 2012 John Wiley & Sons, Ltd.

The US Food and Drug Administration's Sentinel Initiative: Expanding the horizons of medical product safety

Abstract: Office of Medical Policy, Center for Drug Evaluation and Research, Food and Drug Administration, Silver Spring, MD 20993, USA Office of Surveillance and Biometrics, Center for Devices and Radiological Health, Food and Drug Administration, Silver Spring, MD 20993, USA Office of Biostatistics and Epidemiology, Center for Biologics Evaluation and Research, Food and Drug Administration, Rockville, MD 20852, USA Office of Surveillance and Epidemiology, Center for Drug Evaluation and Research, Food and Drug Administration, Silver Spring, MD 20993 USA Center for Biologics Evaluation and Research, Food and Drug Administration, Rockville, MD 20852, USA Center for Drug Evaluation and Research, Food and Drug Administration, Silver Spring, MD 20993, USA

Nature and frequency of drug-related problems in self-medication (over-the-counter drugs) in daily community pharmacy practice in Germany.

Abstract: PURPOSE: To quantify drug-related problems (DRPs) in self-medication (over-the-counter [OTC] drug use) identified by community pharmacists (CPs) in Germany at the time the drug is dispensed. METHODS: One hundred CPs were asked to document 100 consecutive customers presenting symptoms or requesting OTC drugs using a standardized documentation form. The number of 10 000 encounters seemed reasonable to evaluate the set objective. For each encounter, data such as age, sex, and first or repeated request and the availability of a patient file in the pharmacy including drug history were documented. Furthermore, identified DRPs, problem descriptions, and solutions were documented. Data were transcribed electronically, coded, checked for validity, and analyzed. RESULTS: In total, 109 CPs documented 12 567 encounters identifying DRPs in 17.6% of all cases. Four indications comprised more than 70% of all DRPs: pain, respiratory, gastrointestinal, and skin disorders. Four DRPs were responsible for almost 75% of all DRPs identified: self-medication inappropriate (29.7%), requested product inappropriate (20.5%), intended duration of drug use too high including abuse (17.1%), and wrong dosage (6.8%). If a drug history was available, significantly more cases with wrong dosage (p Language: en

Using high‐dimensional propensity scores to automate confounding control in a distributed medical product safety surveillance system

Abstract: Distributed medical product safety monitoring systems such as the Sentinel System, to be developed as a part of Food and Drug Administration's Sentinel Initiative, will require automation of large parts of the safety evaluation process to achieve the necessary speed and scale at reasonable cost without sacrificing validity. Although certain functions will require investigator intervention, confounding control is one area that can largely be automated. The high-dimensional propensity score (hd-PS) algorithm is one option for automated confounding control in longitudinal healthcare databases. In this article, we discuss the use of hd-PS for automating confounding control in sequential database cohort studies, as applied to safety monitoring systems. In particular, we discuss the robustness of the covariate selection process, the potential for over- or under-selection of variables including the possibilities of M-bias and Z-bias, the computation requirements, the practical considerations in a federated database network, and the cases where automated confounding adjustment may not function optimally. We also outline recent improvements to the algorithm and show how the algorithm has performed in several published studies. We conclude that despite certain limitations, hd-PS offers substantial advantages over non-automated alternatives in active product safety monitoring systems.

Role of disease risk scores in comparative effectiveness research with emerging therapies.

Abstract: Special challenges apply to the control of confounding in studies of the safety and effectiveness of new and evolving therapies Many covariates can influence choices among alternative therapies, and prescriber preferences often evolve quickly during the period of early experience with a specific drug or dose1 Especially in early follow-up, there are typically relatively few study outcomes This setting of evolving prescriber preferences and relatively few outcomes can limit the use of both traditional multivariable models and propensity scores as approaches to obtain unbiased estimates of relative treatment effects Whether one uses a case-control or prospective study design to compare outcomes across treatment groups, the number of potential confounders included in a standard regression approach is limited by the number of study endpoints For example, reliable estimation in both logistic regression and proportional hazards models requires no more than one covariate (counting separately interaction and higher-order terms) for every 10 study outcomes2 This can lead investigators to prioritize potential confounders and exclude some of theoretical relevance from multivariable analyses, leading to suboptimal confounder control Propensity scores are a valuable strategy to reduce the dimension of potential confounding variables and can be particularly useful when there are relatively few study endpoints3,4 However, with new and evolving therapies, a prescriber’s preference regarding the characteristics of patients who indicate a specific drug choice is likely to change, and at varying rates across providers Patients may also have varying attitudes about use of new therapies Such evolving relationships of specific characteristics with treatment choices imply absence of a sharply defined propensity score, and can lead investigators to consideration of time-varying propensity scores5 Additionally, if some variables are related to treatment choice but not study outcomes, these instruments are best not included in the propensity score,6–8 but their identification in the setting of newly evolving therapies is challenging With these challenges to multivariable analysis and propensity score estimation, a disease risk score can be a useful tool for confounder control Here we provide some background on the use of the disease risk score in epidemiology, consider controversies regarding its estimation, note the balancing properties of this score, and illustrate the development of a disease risk score with examples from the use of statins, including high-intensity statin regimens, after myocardial infarction

A systematic review of validated methods for identifying suicide or suicidal ideation using administrative or claims data

Abstract: PURPOSE: As part of the Mini-Sentinel pilot program, under contract with the Food and Drug Administration, an effort has been made to evaluate the validity of algorithms useful for identifying health outcomes of interest, including suicide and suicide attempt. METHOD: Literature was reviewed to evaluate how well medical episodes associated with these events could be identified in administrative or claims data sets from the USA or Canada. RESULTS: Six studies were found to include sufficient detail to assess performance characteristics of an algorithm on the basis of International Classification of Diseases, Ninth Revision, E-codes (950-959) for intentional self-injury. Medical records and death registry information were used to validate classification. Sensitivity ranged from 13.8% to 65%, and positive predictive value range from 4.0% to 100%. Study comparisons are difficult to interpret, however, as the studies differed substantially in many important elements, including design, sample, setting, and methods. Although algorithm performance varied widely, two studies located in prepaid medical plans reported that comparisons of database codes to medical charts could achieve good agreement. CONCLUSIONS: Insufficient data exist to support specific recommendations regarding a preferred algorithm, and caution should be exercised in interpreting clinical and pharmacological epidemiological surveillance and research that rely on these codes as measures of suicide-related outcomes. Copyright © 2012 John Wiley & Sons, Ltd. Language: en

The Food and Drug Administration's Post‐Licensure Rapid Immunization Safety Monitoring program: strengthening the federal vaccine safety enterprise

Abstract: In 2009, the Department of Health and Human Services created the new Post-Licensure Rapid Immunization Safety Monitoring (PRISM) program, which used data from national health insurance plans and immunization registries to monitor the safety of the H1N1 influenza vaccine. PRISM has now been integrated into the FDA's Mini-Sentinel pilot program. It strengthens the federal vaccine safety enterprise in two important ways. First, PRISM monitors the largest US general population cohort designated for active surveillance of vaccine safety. Second, PRISM links data from health plans with data from state and city immunization registries, which were a crucial source of exposure data in the H1N1 vaccine evaluation. The Mini-Sentinel data that support PRISM are updated quarterly, and PRISM can conduct medical record review for validation of computerized data. The FDA has structured PRISM as a program that includes specific vaccine evaluations, development of an operational framework to guide the design of vaccine safety evaluations, and development of new statistical methods. A human papillomavirus vaccine, Gardasil, and two rotavirus vaccines, RotaTeq and Rotarix, have been chosen for surveillance in the current cycle because their evaluations would benefit most from PRISM's large cohort size. The PRISM program creates important opportunities by offering a robust, responsive new surveillance program with features complementary to existing systems. Methodological and logistical lessons can be shared among PRISM and other surveillance systems, offering potential synergies. FDA and PRISM will work to maximize the program's unique strengths and contributions to a unified federal vaccine safety enterprise.

A systematic review of validated methods for identifying depression using administrative data

Abstract: review. Each abstract was reviewed independently by the first and second authors (L.T. and J.W.) to determine whether the full-text article should be reviewed. The following abstract exclusion criteria were applied: (i) The abstract did not mention depression or dysthymia; (ii) the study did not use an administrative database (eligible sources included insurance claims databases and other secondary databases that identify health outcomes using billing codes); and (iii) the data source was not from the USA or Canada. Exclusion criteria were documented sequentially (i.e., if one exclusion criterion was met, then the other criteria were not documented). If the reviewers disagreed on whether the full text should be reviewed, then it was selected for review. Interrater agreement on whether to include or exclude an abstract was calculated using Cohen’s kappa. Full-text review. Full-text articles were reviewed independently by the first and second authors (L.T. and J.W.), with the goal of identifying validation studies. The full-text review included examination of articles’ reference sections as an additional means of capturing relevant citations. Citations from the references were selected for full-text review if they were cited as a source for a depression algorithm or were otherwise deemed likely to be relevant. Full-text studies were excluded from the final evidence table if they met one or more of the following criteria: (i) The article contained a poorly described or difficult to operationalize depression algorithm defined by the absence of Diagnostic and Statistical Manual of Mental Disorder (DSM) depression diagnosis codes (296.2, 296.3, 300.4, or 311) or the International Classification of Diseases (ICD) diagnosis codes for depression (296.2, 296.3, 300.4, 311, 298.0, or 309.1), and (ii) the article provided no validation measure of depression or did not report validity statistics. Full-text review exclusion criteria were applied sequentially. If there was disagreement on whether a study should be included, the two reviewers (L.T. and J.W.) attempted to reach consensus on inclusion by discussion. If they could not agree, an additional investigator (M.O.) was consulted to make the final decision. All studies that survived the exclusion screen were included in the final evidence table. A single investigator abstracted each study for the table. A second investigator confirmed the accuracy of the abstracted data. A clinician or topic expert was consulted to review the results of the evidence table and to evaluate how the findings compared with the findings of diagnostic methods used in clinical practice. This assessment helped to determine whether the algorithms excluded any depression diagnosis codes commonly used in clinical practice and the appropriateness of the validation measures in relation to clinical diagnostic criteria.

Mini-Sentinel's systematic reviews of validated methods for identifying health outcomes using administrative and claims data: methods and lessons learned.

Abstract: Purpose To overview the methods used in the Mini-Sentinel systematic reviews of validation studies of algorithms to identify health outcomes in administrative and claims data and to describe lessons learned in the development of search strategies, including their ability to identify articles from previous systematic reviews which used different search strategies. Methods Literature searches were conducted using PubMed and the citation database of the Iowa Drug Information Service. Embase was searched for some outcomes. The searches were based on a strategy developed by the Observational Medical Outcomes Partnership (OMOP) researchers. All citations were reviewed by two investigators. Exclusion criteria were applied at abstract and full-text review stages to ultimately identify algorithm validation studies that used data sources from the USA or Canada, as the results of these studies were considered most likely to generalize to Mini-Sentinel data. Nonvalidated algorithms were reviewed if fewer than five algorithm validation studies were identified. Results The results of this project are described in the separate articles and reports written on algorithms to identify each outcome of interest. Conclusions The Mini-Sentinel systematic reviews of algorithms to identify health outcomes in administrative and claims data are expected to be relatively complete, despite some limitations. Algorithm validation studies are inconsistently indexed in PubMed, creating challenges in conducting systematic reviews of these studies. Google Scholar searches, which can perform text word searches of electronically available articles, are suggested as a strategy to identify studies that are not captured through searches of standard citation databases. Copyright © 2012 John Wiley & Sons, Ltd.

A systematic review of validated methods for identifying seizures, convulsions, or epilepsy using administrative and claims data.

Abstract: Purpose To systematically review algorithms to identify seizure, convulsion, or epilepsy cases in administrative and claims data, with a focus on studies that have examined the validity of the algorithms. Methods A literature search was conducted using PubMed and the Iowa Drug Information Service database. Reviews were conducted by two investigators to identify studies using data sources from the USA or Canada because these data sources were most likely to reflect the coding practices of Mini-Sentinel data partners. Results Eleven studies that validated seizure, convulsion, or epilepsy cases were identified. All algorithms included International Classification of Diseases, Ninth Revision, Clinical Modification code 345.X (epilepsy) and either code 780.3 (convulsions) or code 780.39 (other convulsions). Six studies included 333.2 (myoclonus). In populations that included children, 779.0 (convulsions in newborn) was also fairly common. Positive predictive values (PPVs) ranged from 21% to 98%. Studies that used nonspecific indicators such as presence of an electroencephalogram or anti-epileptic drug (AED) level monitoring had lower PPVs. In studies focusing exclusively on epilepsy as opposed to isolated seizure events, sensitivity ranged from 70% to 99%. Conclusions Algorithm performance was highly variable, so it is difficult to draw any strong conclusions. However, the PPVs were generally best in studies where epilepsy diagnoses were required. Using procedure codes for electroencephalograms or prescription claims for drugs possibly used for epilepsy or convulsions in the absence of a diagnostic code is not recommended. Many newer AEDs require no drug level monitoring, so requiring an AED level monitoring procedure in algorithms to identify epilepsy is not recommended. Copyright © 2012 John Wiley & Sons, Ltd.

Electronic healthcare databases for active drug safety surveillance: is there enough leverage?

Abstract: Purpose To provide estimates of the number and types of drugs that can be monitored for safety surveillance using electronic healthcare databases. Methods Using data from eight European databases (administrative claims, medical records) and in the context of a cohort study, we determined the amount of drug exposure required for signal detection across varying magnitudes of relative risk (RR). We provide estimates of the number and types of drugs that can be monitored as a function of actual use, minimal detectable RR, and empirically derived incidence rates for the following adverse events: (i) acute myocardial infarction; (ii) acute renal failure; (iii) anaphylactic shock; (iv) bullous eruptions; (v) rhabdomyolysis; and (vi) upper gastrointestinal bleeding. We performed data simulation to see how expansion of database size would influence the capabilities of such system. Results Data from 19647452 individuals (59594132person-years follow-up) who used 2289 drugs in the EU-ADR network show that for a frequent event such as acute myocardial infarction, there are 531 drugs (23% of total) for which an association with RR=2, if present, can be investigated.For a rareevent suchas rhabdomyolysis, thereare 19 drugs (1%)for which an association ofsame magnitude canbe investigated. Conclusion Active surveillance using healthcare data-based networks for signal detection is feasible, although the leverage to do so may be low for infrequently used drugs and for rare outcomes. Extending database network size to include data from heterogeneous populations and increasing follow-up time are warranted to maximize leverage of these surveillance systems. Copyright © 2012 John Wiley & Sons, Ltd. key words—active drug safety surveillance; drug safety monitoring; signal detection; electronic healthcare records; electronic healthcare databases; EU-ADR

Uses of proton pump inhibitors and hypomagnesemia.

Abstract: Purpose Hypomagnesemia associated with proton pump inhibitor (PPI) therapy has been documented in case reports. We performed a cross-sectional study to examine the association between PPI use and serum magnesium (Mg) levels or hypomagnesemia. Methods Data were extracted from hospitalized adults with basic metabolic panels and/or serum magnesium levels available during the hospital stays. The first Mg value was used for data analysis. Hypomagnesemia is defined as levels less than 1.7 mg/dL (or 0.70 mmol/L). Multiple linear and logistic regression analyses were used to assess the association between PPI use and Mg levels or hypomagnesemia, respectively. Results Among study patients, PPI users (n = 207) had a mean Mg level of 1.91[SD = 0.34] mg/dL, and non-users (n = 280) 2.00 (0.30) mg/dL, p = 0.004. PPI use was associated with lower serum Mg levels (adjusted coefficient β = −0.10, 95%CI = [−0.16, −0.04]) after adjusting for confounders. PPI use was associated with risk of hypomagnesemia after adjusting for confounders (adjusted OR = 2.50, 95%CI = [1.43, 4.36]). Both standard (1) and high (2 or higher) defined daily dose units of PPI therapy were associated with hypomagnesemia. Conclusions PPI use was associated with lower serum Mg levels and hypomagnesemia in a population of hospitalized adult patients. Our study supports the general notion that long-term PPI use could be associated with sub-clinical Mg insufficiency or deficiency status. Copyright © 2012 John Wiley & Sons, Ltd.

Applying propensity scores estimated in a full cohort to adjust for confounding in subgroup analyses

Abstract: A correctly-specified propensity score (PS) estimated in a cohort (“cohort PS”) should in expectation remain valid in a subgroup population. We sought to determine whether using a cohort PS can be validly applied to subgroup analyses and thus add efficiency to studies with many subgroups or restricted data. In each of 3 cohort studies we estimated a cohort PS, defined 5 subgroups, and then estimated subgroup-specific PSs. We compared difference in treatment effect estimates for subgroup analyses adjusted by cohort PSs versus subgroup-specific PSs. Then, 10M times, we simulated a population with known characteristics of confounding, subgroup size, treatment interactions, and treatment effect, and again assessed difference in point estimates. We observed that point estimates in most subgroups were substantially similar with the two methods of adjustment. In simulations, the effect estimates differed by a median of 3.4% (interquartile [IQ] range 1.3% to 10.0%). The IQ range exceeded 10% only in cases where the subgroup had <1000 patients or few outcome events. Our empirical and simulation results indicated that using a cohort PS in subgroup analyses was a feasible approach, particularly in larger subgroups.

Identification of metastatic cancer in claims data

Abstract: Purpose To develop algorithms to identify metastatic cancer in claims data, using tumor stage from an oncology electronic medical record (EMR) data warehouse as the gold standard. Methods Data from an outpatient oncology EMR database were linked to medical and pharmacy claims data. Patients diagnosed with breast, lung, colorectal, or prostate cancer with a stage recorded in the EMR between 2004 and 2010 and with medical claims available were eligible for the study. Separate algorithms were developed for each tumor type using variables from the claims, including diagnoses, procedures, drugs, and oncologist visits. Candidate variables were reviewed by two oncologists. For each tumor type, the selected variables were entered into a classification and regression tree model to determine the algorithm with the best combination of positive predictive value (PPV), sensitivity, and specificity. Results A total of 1385 breast cancer, 1036 lung, 727 colorectal, and 267 prostate cancer patients qualified for the analysis. The algorithms varied by tumor type but typically included International Classification of Diseases-Ninth Revision codes for secondary neoplasms and use of chemotherapy and other agents typically given for metastatic disease. The final models had PPV ranging from 0.75 to 0.86, specificity 0.75–0.97, and sensitivity 0.60–0.81. Conclusions While most of these algorithms for metastatic cancer had good specificity and acceptable PPV, a tradeoff with sensitivity prevented any model from having good predictive ability on all measures. Results suggest that accurate ascertainment of metastatic status may require access to medical records or other confirmatory data sources. Copyright © 2012 John Wiley & Sons, Ltd.

A policy framework for public health uses of electronic health data.

Abstract: Successful implementation of a program of active safety surveillance of drugs and medical products depends on public trust. This article summarizes how the initial pilot phase of the FDA's Sentinel Initiative, Mini-Sentinel, is being conducted in compliance with applicable federal and state laws. The article also sets forth the attributes of Mini-Sentinel that enhance privacy and public trust, including the use of a distributed data system (where identifiable information remains at the data partners) and the adoption by participants of additional mandatory policies and procedures implementing fair information practices. The authors conclude by discussing the implications of this model for other types of secondary health data uses. Copyright © 2012 John Wiley & Sons, Ltd.

Proton pump inhibitors and traditional nonsteroidal anti-inflammatory drugs and the risk of acute interstitial nephritis and acute kidney injury.

Abstract: Purpose This study aims to examine the associations between proton pump inhibitors (PPIs), traditional nonsteroidal anti-inflammatory drugs (tNSAIDs), PPI + tNSAID co-exposure, and the development of the following: (i) acute interstitial nephritis (AIN), a specific kidney injury often attributed to these drugs, and (ii) acute kidney injury (AKI), a general kidney injury encompassing AIN. Methods Two retrospective case–control studies were conducted, one for each outcome, within the General Practice Research Database. Cases were diagnostic-coded AIN (primary outcome) or AKI (secondary outcome) events. Controls were matched on age, sex, and general practitioner practice. Exposures were defined by the presence/absence of the following mutually exclusive therapies on the index date: (i) PPI alone; (ii) tNSAID alone; (iii) PPI + tNSAID; or (iv) neither PPI nor tNSAID (referent). Results Sixty-eight AIN cases and 3347 controls were identified. The adjusted odds ratios (ORs) for PPI and tNSAID exposures alone were 3.20 (0.80–12.79) and 1.90 (0.65–5.51), respectively. Numerous sensitivity analyses produced adjusted ORs for AIN between 3.0 and 7.7, and 1.6 and 1.9, respectively. We identified 27 982 AKI cases and 1 323 850 controls. The adjusted ORs for PPI alone, tNSAID alone, and PPI + tNSAID exposures were 1.05 (0.97–1.14), 1.31 (1.25–1.37), and 1.33 (1.07–1.64), respectively. Numerous sensitivity analyses produced adjusted ORs for AKI between 1.0 and 1.1, 1.1 and 1.3, and 1.3 and 1.4, respectively. Conclusions Proton pump inhibitor exposure may increase the odds of AIN, but this result was not definitive and should be confirmed in a dataset with more AIN cases to allow for increased statistical precision. tNSAIDs, yet not PPIs, were associated with a significantly increased odds of AKI. Copyright © 2012 John Wiley & Sons, Ltd.

Incidence of osteonecrosis of the jaw among users of bisphosphonates with selected cancers or osteoporosis

Abstract: Purpose To quantify the incidence of osteonecrosis of the jaw (ONJ) by bisphosphonate exposure among two cohorts of patients. Methods In a retrospective cohort study, we identified cohort members via health insurance claim diagnosis codes and identified potential cases of ONJ that were confirmed with medical record review. One cohort included patients aged ≥40 years with breast or prostate cancer or multiple myeloma; the other cohort included men aged ≥60 years and women ≥50 years with osteoporosis. For each cohort, we calculated sex- and age-standardized incidence of ONJ by exposure to oral bisphosphonates and intravenous bisphosphonates. Results In the cancer cohort (n = 46 542), sex- and age-standardized incidence of ONJ (n = 26 probable or possible cases) adjusted for abstraction proportion was 0.29 per 1000 person-years (95% confidence interval [CI], 0.07–0.52) among those unexposed to bisphosphonates and 5.3 (95%CI, 1.9–8.7) after intravenous bisphosphonate use. Controlling for covariates, the rate ratio for intravenous use versus no use was 8.8 (95%CI, 2.0–38). Patients with multiple myeloma had a rate 4.5 times that of patients with breast cancer. In the osteoporosis cohort (n = 31 244), sex- and age-standardized ONJ (n = 11 probable or possible cases) incidence was 0.26 per 1000 person-years (95%CI, 0.06–0.47) among those unexposed to bisphosphonate and 0.15 (95%CI, 0.00–0.36) after oral bisphosphonate use. Conclusion Among patients with selected cancers, incidence of ONJ was higher among those with multiple myeloma and users of intravenous bisphosphonates. Copyright © 2012 John Wiley & Sons, Ltd.

A systematic review of validated methods for identifying ventricular arrhythmias using administrative and claims data.

Abstract: Background Drug-induced pro-arrhythmia is a serious and unexpected event. Large administrative and claims databases can potentially identify drugs or interactions leading to cardiac arrhythmias. The purpose of this study is to evaluate the evidence supporting the validity of algorithms or codes to identify ventricular arrhythmias using administrative and claims data. Methods A search of MEDLINE database is supplemented by manual searches of bibliographies of key relevant articles. We selected all studies in which an administrative and claims data algorithm or code was validated against a medical record. We report the positive predictive value (PPV) for ICD-9 codes compared to medical records. Results Our search strategy yielded 664 studies, of which only seven met our eligibility criteria. Two additional studies were identified by peer reviewers. The most commonly included databases were Medicare and Medicaid, and the most commonly evaluated ICD-9 codes were 426.x and 427.x. The individual use of ICD-9 codes 427.x yielded a high PPV (78%–100%). The highest PPV was seen when both ICD-9 codes 427.x and 798.x were used (92%). The same codes yielded the highest PPV when found in the principal diagnosis position (100%). Conclusions The use of ICD-9 codes 427.x, alone or in combination with code 798.x, in the principal position is appropriate for the identification of ventricular arrhythmias in administrative and claims databases. Copyright © 2012 John Wiley & Sons, Ltd.

Mini-Sentinel's systematic reviews of validated methods for identifying health outcomes using administrative data: summary of findings and suggestions for future research.

Abstract: Purpose The validity of findings from surveillance activities, which use administrative and claims data to link exposures to adverse events, depends in part on the validity of algorithms to identify health outcomes using these data. This review provides a high level overview of the findings of 19 systematic reviews of studies, which have examined the validity of algorithms to identify health outcomes using these data. The author categorized outcomes on the basis of the strength of evidence supporting valid algorithms to identify acute or incident events and suggested priorities for future validation studies. Methods The 19 reviews were evaluated, and key findings and suggestions for future research were summarized by a single reviewer. Outcomes with algorithms that consistently identified acute events or incident conditions with positive predictive values of greater than 70% across multiple studies and populations are described as low priority for future algorithm validation studies. Results Algorithms to identify cerebrovascular accidents, transient ischemic attacks, congestive heart failure, deep vein thrombosis, pulmonary embolism, angioedema, and total hip arthroplasty revision performed well across multiple studies and are considered low priority for future validation studies. Other outcomes were generally thought to require additional validation studies or algorithm refinement to be confident in algorithms. Few studies examined the validity of International Classification of Diseases, 10th Revision, codes. Conclusion Users of these reviews need to consider the generalizability of findings to their study populations. For some outcomes with poorly performing codes, it may always be necessary to validate cases. Copyright © 2012 John Wiley & Sons, Ltd.

Terms used in patient safety related to medication: a literature review.

Abstract: PURPOSE: There is a lack of homogeneity in the terminology used in the context of patient safety related to medication. The aim of this review was to identify the terms and definitions used in patient safety related to medication within the scientific literature. METHODS: Original and review articles that were indexed between 1998 and 2008 in MEDLINE and EMBASE and contained terms used in patient safety related to medication were included. Terms and definitions were extracted and categorised according to whether its definition referred to the process of medication use, or to the clinical outcome of medication use, or both. RESULTS: Of 2564 articles, 147 were included. Sixty terms used in patient safety related to medication with 189 different definitions were identified. Among terms that referred only to the process of medication use (n = 23), medication error provided the greatest number of definitions (n = 29). Among terms that referred only to the clinical outcome of medication use (n = 31), adverse drug event provided the greatest number of definitions (n = 15). Finally, among terms that referred both to the process of use and to the clinical outcome of medication use (n = 13), drug-related problem provided the greatest number of definitions (n = 7). CONCLUSIONS: A multitude of terms and definitions are used in patient safety related to medication. This heterogeneity makes it difficult to compare the results among studies and to appreciate the true magnitude of the problem. Classifying and unifying the terminology is necessary to advance in patient safety strategies. Copyright © 2012 John Wiley & Sons, Ltd. Language: en