Drug exposure misclassification in pharmacoepidemiology: Sources and relative impact.
Ontology highlight
ABSTRACT: Background
Drug exposure assessment based on dispensing data can be misclassified when patients do not adhere to their therapy or when information about over-the-counter drugs is not captured in the study database. Previous research has considered hypothetical sensitivity and specificity values, whereas this study aims to assess the impact of literature-based real values of exposure misclassification.Methods
A synthetic cohort study was constructed based on the proportion of exposure theoretically captured in a database (range 0.5-1.0) and the level of adherence (0.5-1.0). Three scenarios were explored: nondifferential misclassification, differential misclassification (misclassifications dependent on an unmeasured risk factor doubling the outcome risk), and nondifferential misclassification in a comparative effectiveness study (RRA and RRB both 2.0 compared to nonuse, RRA-B 1.0).Results
For the scenarios with nondifferential misclassification, 25% nonadherence or 25% uncaptured exposure changed the RR from 2.0 to 1.75, and 1.95, respectively. Applying different proportions of nonadherence or uncaptured use (20% vs. 40%) for subgroups with and without the risk factor, an RR of 0.95 was observed in the absence of a true effect (i.e., true RR = 1). In the comparative effectiveness study, no effect on RR was seen for different proportions of uncaptured exposure; however, different levels of nonadherence for the drugs (20% vs. 40%) led to an underestimation of RRA-B (0.89).Discussion
All scenarios led to biased estimates, but the magnitude of the bias differed across scenarios. When testing the robustness of findings of pharmacoepidemiologic studies, we recommend using realistic values of nonadherence and uncaptured exposure based on real-world data.
SUBMITTER: Hempenius M
PROVIDER: S-EPMC9292927 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:The definition of a new case is a vital step in incidence studies in both epidemiology and pharmacoepidemiology, although with significant differences in methodology between the fields. We define and apply a framework for two different types of new cases of drug use, first-ever and recurrent, and show how the associated misclassifications related to length of run-in period can be expressed by the positive predictive value (PPV). In the study, we consider individual-level dispensations of statins 2006-2019 for 1,017,058 individuals with at least one dispensation in 2019 in Sweden. The incidence proportion for statins for both sexes of all ages in Sweden 2019 varied from 17.4/1000 with a run-in of 8 months, 9.45/1000 with 5 years and 8.4/1000 with 10 years. The PPV was 49% with 8 months and 89% for 5 years using 10 years as gold standard. We conclude that the interpretation of incidence and thus the selection of an appropriate run-in period, in pharmacoepidemiology, depends on whether first-ever use, recurrent treatment or both together (new cases) is the focus of the research question studied. At least five different misclassifications can be introduced depending on how incidence is defined.
| S-EPMC9320840 | biostudies-literature
Project description:PURPOSE:Real-world data for observational research commonly require formatting and cleaning prior to analysis. Data preparation steps are rarely reported adequately and are likely to vary between research groups. Variation in methodology could potentially affect study outcomes. This study aimed to develop a framework to define and document drug data preparation and to examine the impact of different assumptions on results. METHODS:An algorithm for processing prescription data was developed and tested using data from the Clinical Practice Research Datalink (CPRD). The impact of varying assumptions was examined by estimating the association between 2 exemplar medications (oral hypoglycaemic drugs and glucocorticoids) and cardiovascular events after preparing multiple datasets derived from the same source prescription data. Each dataset was analysed using Cox proportional hazards modelling. RESULTS:The algorithm included 10 decision nodes and 54 possible unique assumptions. Over 11 000 possible pathways through the algorithm were identified. In both exemplar studies, similar hazard ratios and standard errors were found for the majority of pathways; however, certain assumptions had a greater influence on results. For example, in the hypoglycaemic analysis, choosing a different variable to define prescription end date altered the hazard ratios (95% confidence intervals) from 1.77 (1.56-2.00) to 2.83 (1.59-5.04). CONCLUSIONS:The framework offers a transparent and efficient way to perform and report drug data preparation steps. Assumptions made during data preparation can impact the results of analyses. Improving transparency regarding drug data preparation would increase the repeatability, reproducibility, and comparability of published results.
| S-EPMC6055712 | biostudies-literature
Project description:AimsDrug exposure status based on routinely collected data might be misclassified when the database contains only prescriptions from 1 type of prescriber (e.g. general practitioner and not specialist). This study aims to quantify the impact of such exposure misclassification on the risk of major bleeding and stroke/transient ischaemic attack (TIA)associated with direct oral anticoagulants (DOACs) vs. vitamin K antagonists (VKAs).MethodsIncident anticoagulant users (>12 mo free of anticoagulation use) in the Dutch PHARMO Database Network between 2008 and 2017 were included. Drug exposure was assessed using pharmacy dispensing information. The risks of hospital admission of major bleeding for DOAC vs. VKA users was assessed with Cox regression analysis, where exposure was based on all dispensings, on general practitioner (GP)-prescribed dispensings only or on specialist-prescribed dispensings only. Hazard ratios (HRs) were estimated also for hospitalization for gastrointestinal bleeding, intracranial bleeding and stroke/TIA.ResultsWe included 99 182 VKA-initiators and 21 795 DOAC-initiators. Use of DOAC was associated with a lower risk of major bleeding compared to VKA use; HR 0.79 (95% confidence interval 0.70-0.90), 0.78 (0.68-0.91) and 0.62 (0.50-0.76), for exposure based on complete dispensing information, only GP- and only specialist-prescribed dispensings, respectively. Similar results were found for the other bleeding outcomes. For stroke/TIA the HRs were 0.96 (0.84-1.09), 1.00 (0.84-1.18) and 0.72 (0.58-0.90), respectively.ConclusionIncluding only GP-prescribed anticoagulant dispensings in this case did not materially impact the effect estimates compared to including all anticoagulant dispensings. Including only specialist-prescribed dispensings, however, strengthened the effect estimates.
| S-EPMC8451929 | biostudies-literature
Project description:With limited funding and biological specimen availability, choosing an optimal sampling design to maximize power for detecting gene-by-environment (G-E) interactions is critical. Exposure-enriched sampling is often used to select subjects with rare exposures for genotyping to enhance power for tests of G-E effects. However, exposure misclassification (MC) combined with biased sampling can affect characteristics of tests for G-E interaction and joint tests for marginal association and G-E interaction. Here, we characterize the impact of exposure-biased sampling under conditions of perfect exposure information and exposure MC on properties of several methods for conducting inference. We assess the Type I error, power, bias, and mean squared error properties of case-only, case-control, and empirical Bayes methods for testing/estimating G-E interaction and a joint test for marginal G (or E) effect and G-E interaction across three biased sampling schemes. Properties are evaluated via empirical simulation studies. With perfect exposure information, exposure-enriched sampling schemes enhance power as compared to random selection of subjects irrespective of exposure prevalence but yield bias in estimation of the G-E interaction and marginal E parameters. Exposure MC modifies the relative performance of sampling designs when compared to the case of perfect exposure information. Those conducting G-E interaction studies should be aware of exposure MC properties and the prevalence of exposure when choosing an ideal sampling scheme and method for characterizing G-E interactions and joint effects.
| S-EPMC4256150 | biostudies-literature
Project description:Measurement error is common in epidemiology, but few studies use quantitative methods to account for bias due to mismeasurement. One potential barrier is that some intuitive approaches that readily combine with methods to account for other sources of bias, like multiple imputation for measurement error (MIME), rely on internal validation data, which are rarely available. Here, we present a reparameterized imputation approach for measurement error (RIME) that can be used with internal or external validation data. We illustrate the advantages of RIME over a naive approach that ignores measurement error and MIME using a hypothetical example and a series of simulation experiments. In both the example and simulations, we combine MIME and RIME with inverse probability weighting to account for confounding when estimating hazard ratios and counterfactual risk functions. MIME and RIME performed similarly when rich external validation data were available and the prevalence of exposure did not vary between the main study and the validation data. However, RIME outperformed MIME when validation data included only true and mismeasured versions of the exposure or when exposure prevalence differed between the data sources. RIME allows investigators to leverage external validation data to account for measurement error in a wide range of scenarios.
| S-EPMC7608057 | biostudies-literature
Project description:IntroductionMisclassification errors have been reported in rapid diagnostic HIV tests (RDTs) in sub-Saharan African countries. These errors can lead to missed opportunities for prevention-of-mother-to-child-transmission (PMTCT), early infant diagnosis and adult HIV-prevention, unnecessary lifelong antiretroviral treatment (ART) and wasted resources. Few national estimates or systematic quantifications of sources of errors have been produced. We conducted a comprehensive assessment of possible sources of misclassification errors in routine HIV testing in Zimbabwe.MethodsRDT-based HIV test results were extracted from routine PMTCT programme records at 62 sites during national antenatal HIV surveillance in 2017. Positive- (PPA) and negative-percent agreement (NPA) for HIV RDT results and the false-HIV-positivity rate for people with previous HIV-positive results ("known-positives") were calculated using results from external quality assurance testing done for HIV surveillance purposes. Data on indicators of quality management systems, RDT kit performance under local climatic conditions and user/clerical errors were collected using HIV surveillance forms, data-loggers and a Smartphone camera application (7 sites). Proportions of cases with errors were compared for tests done in the presence/absence of potential sources of errors.ResultsNPA was 99.9% for both pregnant women (N = 17224) and male partners (N = 2173). PPA was 90.0% (N = 1187) and 93.4% (N = 136) for women and men respectively. 3.5% (N = 1921) of known-positive individuals on ART were HIV negative. Humidity and temperature exceeding manufacturers' recommendations, particularly in storerooms (88.6% and 97.3% respectively), and premature readings of RDT output (56.0%) were common. False-HIV-negative cases, including interpretation errors, occurred despite staff training and good algorithm compliance, and were not reduced by existing external or internal quality assurance procedures. PPA was lower when testing room humidity exceeded 60% (88.0% vs. 93.3%; p = 0.007).ConclusionsFalse-HIV-negative results were still common in Zimbabwe in 2017 and could be reduced with HIV testing algorithms that use RDTs with higher sensitivity under real-world conditions and greater practicality under busy clinic conditions, and by strengthening proficiency testing procedures in external quality assurance systems. New false-HIV-positive RDT results were infrequent but earlier errors in testing may have resulted in large numbers of uninfected individuals being on ART.
| S-EPMC8059712 | biostudies-literature
Project description:ObjectiveA wealth of data is generated through Australia's universal health care arrangements. However, use of these data has been hampered by different federal and state legislation, privacy concerns and challenges in linking data across jurisdictions. A series of data reforms have been touted to increase population health research capacity in Australia, including pharmacoepidemiology research. Here we catalogued research leveraging Australia's Pharmaceutical Benefits Scheme (PBS) data (2014-2018) and discussed these outputs in the context of previously implemented and new data reforms.MethodsWe conducted a systematic review of population-based studies using PBS dispensing claims. Independent reviewers screened abstracts of 4,996 articles and 310 full-text manuscripts. We characterised publications according to study population, analytical approach, data sources used, aims and medicines focus.ResultsWe identified 180 studies; 133 used individual-level data, 70 linked PBS dispensing claims with other health data (66 across jurisdictions). Studies using individual-level data focussed on Australians receiving government benefits (87 studies) rather than all PBS-eligible persons. 63 studies examined clinician or patient practices and 33 examined exposure-outcome relationships (27 evaluated medicines safety, 6 evaluated effectiveness). Medicines acting on the nervous and cardiovascular system account for the greatest volume of PBS medicines dispensed and were the most commonly studied (67 and 40 studies, respectively). Antineoplastic and immunomodulating agents account for approximately one third of PBS expenditure but represented only 10% of studies in this review.ConclusionsThe studies in this review represent more than a third of all population-based pharmacoepidemiology research published in the last three decades in Australia. Recent data reforms have contributed to this escalating output. However, studies are concentrated among specific subpopulations and medicines classes, and there remains a limited understanding of population benefits and harms derived from medicines use. The current draft Data Availability and Transparency legislation should further bolster efforts in population health research.
| S-EPMC8107783 | biostudies-literature
Project description:In observational studies, misclassification of exposure is ubiquitous and can substantially bias the estimated association between an outcome and an exposure. Although misclassification in a single observational study has been well studied, few papers have considered it in a meta-analysis. Meta-analyses of observational studies provide important evidence for health policy decisions, especially when large randomized controlled trials are unethical or unavailable. It is imperative to account properly for misclassification in a meta-analysis to obtain valid point and interval estimates. In this paper, we propose a novel Bayesian approach to filling this methodological gap. We simultaneously synthesize two (or more) meta-analyses, with one on the association between a misclassified exposure and an outcome (main studies), and the other on the association between the misclassified exposure and the true exposure (validation studies). We extend the current scope for using external validation data by relaxing the "transportability" assumption by means of random effects models. Our model accounts for heterogeneity between studies and can be extended to allow different studies to have different exposure measurements. The proposed model is evaluated through simulations and illustrated using real data from a meta-analysis of the effect of cigarette smoking on diabetic peripheral neuropathy.
| S-EPMC6779051 | biostudies-literature
Project description:BackgroundSystematic processing noise, which includes batch effects, is very common in microarray experiments but is often ignored despite its potential to confound or compromise experimental results. Compromised results are most likely when re-analysing or integrating datasets from public repositories due to the different conditions under which each dataset is generated. To better understand the relative noise-contributions of various factors in experimental-design, we assessed several Illumina and Affymetrix datasets for technical variation between replicate hybridisations of Universal Human Reference (UHRR) and individual or pooled breast-tumour RNA.ResultsA varying degree of systematic noise was observed in each of the datasets, however in all cases the relative amount of variation between standard control RNA replicates was found to be greatest at earlier points in the sample-preparation workflow. For example, 40.6% of the total variation in reported expressions were attributed to replicate extractions, compared to 13.9% due to amplification/labelling and 10.8% between replicate hybridisations. Deliberate probe-wise batch-correction methods were effective in reducing the magnitude of this variation, although the level of improvement was dependent on the sources of noise included in the model. Systematic noise introduced at the chip, run, and experiment levels of a combined Illumina dataset were found to be highly dependent upon the experimental design. Both UHRR and pools of RNA, which were derived from the samples of interest, modelled technical variation well although the pools were significantly better correlated (4% average improvement) and better emulated the effects of systematic noise, over all probes, than the UHRRs. The effect of this noise was not uniform over all probes, with low GC-content probes found to be more vulnerable to batch variation than probes with a higher GC-content.ConclusionsThe magnitude of systematic processing noise in a microarray experiment is variable across probes and experiments, however it is generally the case that procedures earlier in the sample-preparation workflow are liable to introduce the most noise. Careful experimental design is important to protect against noise, detailed meta-data should always be provided, and diagnostic procedures should be routinely performed prior to downstream analyses for the detection of bias in microarray studies.
| S-EPMC3269440 | biostudies-literature
Project description:PURPOSE:To investigate the ability of the propensity score (PS) to reduce confounding bias in the presence of nondifferential misclassification of treatment, using simulations. METHODS:Using an example from the pregnancy medication safety literature, we carried out simulations to quantify the effect of nondifferential misclassification of treatment under varying scenarios of sensitivity and specificity, exposure prevalence (10%, 50%), outcome type (continuous and binary), true outcome (null and increased risk), confounding direction, and different PS applications (matching, stratification, weighting, regression), and obtained measures of bias and 95% confidence interval coverage. RESULTS:All methods were subject to substantial bias toward the null due to nondifferential exposure misclassification (range: 0%-47% for 50% exposure prevalence and 0%-80% for 10% exposure prevalence), particularly if specificity was low (<97%). PS stratification produced the least biased effect estimates. We observed that the impact of sensitivity and specificity on the bias and coverage for each adjustment method is strongly related to prevalence of exposure: as exposure prevalence decreases and/or outcomes are continuous rather than categorical, the effect of misclassification is magnified, producing larger biases and loss of coverage of 95% confidence intervals. PS matching resulted in unpredictably biased effect estimates. CONCLUSIONS:The results of this study underline the importance of assessing exposure misclassification in observational studies in the context of PS methods. Although PS methods reduce confounding bias, bias owing to nondifferential misclassification is of potentially greater concern.
| S-EPMC5854503 | biostudies-literature