Search

To develop a claims-based model to predict persistent high-dose opioid use among patients undergoing total knee replacement (TKR).Using Medicare claims (2010-2014), we identified patients ages ≥65 years who underwent TKR with no history of high-dose opioid use (mean 25 morphine milligram equivalents [MMEs]/day) in the year prior to TKR. We used group-based trajectory modeling to identify distinct opioid use patterns. The primary outcome was persistent high-dose opioid use in the year after TKR. We split the data into training (2010-2013) and test (2014) sets and used logistic regression with least absolute shrinkage and selection operator regularization, utilizing a total of 83 preoperative patient characteristics as candidate predictors. A reduced model with 10 prespecified variables, which included demographic characteristics, opioid use, and medication history was also considered.The final study cohort included 142,089 patients who underwent TKR. The group-based trajectory model identified 4 distinct trajectories of opioid use (group 1: short-term, low-dose; group 2: moderate-duration, low-dose; group 3: moderate-duration, high-dose; and group 4: persistent high-dose). The model predicting persistent high-dose opioid use achieved high discrimination (receiver operating characteristic area under the curve [AUC] 0.85 [95% confidence interval (95% CI) 0.84-0.86]) in the test set. The reduced model with 10 predictors performed equally well (AUC 0.84 [95% CI 0.84-0.85]).In this cohort of older patients, 10.6% became persistent high-dose (mean 22.4 MME/day) opioid users after TKR. Our model with 10 readily available clinical factors may help identify patients at high risk of future adverse outcomes from persistent opioid use after TKR.

The high-dimensional propensity score (HDPS) is a semi-automated procedure for confounder identification, prioritisation and adjustment in large healthcare databases that requires investigators to specify data dimensions, prioritisation strategy and tuning parameters. In practice, reporting of these decisions is inconsistent and this can undermine the transparency, and reproducibility of results obtained. We illustrate reporting tools, graphical displays and sensitivity analyses to increase transparency and facilitate evaluation of the robustness of analyses involving HDPS.Using a study from the UK Clinical Practice Research Datalink that implemented HDPS we demonstrate the application of the proposed recommendations.We identify seven considerations surrounding the implementation of HDPS, such as the identification of data dimensions, method for code prioritisation and number of variables selected. Graphical diagnostic tools include assessing the balance of key confounders before and after adjusting for empirically selected HDPS covariates and the identification of potentially influential covariates. Sensitivity analyses include varying the number of covariates selected and assessing the impact of covariates behaving empirically as instrumental variables. In our example, results were robust to both the number of covariates selected and the inclusion of potentially influential covariates. Furthermore, our HDPS models achieved good balance in key confounders.The data-adaptive approach of HDPS and the resulting benefits have led to its popularity as a method for confounder adjustment in pharmacoepidemiological studies. Reporting of HDPS analyses in practice may be improved by the considerations and tools proposed here to increase the transparency and reproducibility of study results.

Real-world data (RWD) source such as insurance claims data, provides the opportunity to draw on the experience of thousands of patients treated in routine clinical practice to better understand the relationships between exposures and outcomes. Nonetheless, conducting studies on atopic dermatitis (AD) using such data is challenging because it is an episodic condition with a heterogeneous clinical presentation. Furthermore, although there are ICD-10-CM diagnosis codes for AD, the codes do not specify disease severity. Consequently, the potential for misclassification of AD diagnosis and severity is high. To develop and validate a claims-based algorithm for moderate-to-severe AD (MTS-AD) using medical records. This study was conducted within the Optum Research Database, a U.S. claims database from a large, commercial health plan affiliated with Optum. The study protocol was approved by the BRANY Institutional Review Board. Patients with at least one diagnosis code for AD (ICD-10-CM: L20*) between 28 March 2017 and 30 November 2019 were identified. Within this source population, several candidates claims-based algorithms were applied to identify patients with MTS-AD. The candidate algorithms included proxy measures for MTS-AD, such as a number of AD diagnoses, the number and types of AD therapies, and the specialty of the treating provider. Among patients who met one or more candidate algorithms for MTS-AD, a subset was randomly selected for medical record review. Additionally, charts of a random sample of patients from the source population were reviewed to assess the sensitivity. Following deidentification, the charts were adjudicated by a dermatologist with expertise in AD diagnosis and treatment to confirm the presence of MTS-AD. Using the adjudicated medical charts as the ‘gold standard’, the positive predictive value (PPV) of each candidate algorithm was calculated and compared to a pre-specified threshold of ≥70%. Overall, 278 medical records were sought and 200 were adjudicated, including 100 records among patients meeting a candidate algorithm for MTS-AD and 100 records among patients in the random sample. For the initial set of MTS-AD algorithms, which included ≥1 AD diagnosis as a criterion in each candidate algorithm, the PPVs ranged from 38% to 100%. For the algorithms with PPVs that met or exceeded the threshold of 70%, only six patients met the algorithm criteria. The MTS-AD algorithms were then modified to require ≥2 AD diagnoses as a criterion; the PPVs of the revised algorithms ranged from 57% to 100%. The final algorithm selected for MTS-AD had a PPV of 76% (95% CI: 53–90%) and a sensitivity of 35% (95% CI: 21–52%). This final algorithm included the following criteria: (i) ≥ 2 claims with an ICD-10-CM code for AD from any physician and either ≥1 dispensing of dupilumab or ≥2 dispensings of high-potency topical corticosteroids1 or systemic immunosuppressants2; or (ii) ≥ 2 claims with an ICD-10-CM code for AD from a dermatologist or allergist and ≥3 dispensings of medium potency topical corticosteroids,3 topical tacrolimus, phototherapy or oral/parenteral corticosteroids. A claims-based algorithm for identifying patients with moderate-to-severe AD was successfully developed and validated using RWD from a U.S. claims database. This validated moderate-to-severe AD algorithm is available for use in future studies within administrative databases, including those seeking to evaluate the safety or effectiveness of AD therapies in the post-marketing setting.

BackgroundNon-interventional large-scale research on real-world patients who had a stroke requires the use of multiple data sources ensuring access to longitudinal data from large populations with clinically-detailed information. We sought to establish a framework for longitudinal research on patients hospitalised with stroke by linking information-rich, deidentified inpatient data from the Paul Coverdell National Acute Stroke Program (PCNASP) to commercial and Medicare Advantage longitudinal claims data.MethodsAll stroke admissions in PCNASP between 2008 and 2015 were evaluated for linkage to longitudinal claims from a commercial insurer using an algorithm based on six available common data fields (patient age, gender, admission date, discharge date, discharge diagnosis and state) and a hospital match. We evaluated the linkage quality (via the percentage of unique records in the linked dataset) and the representativeness of the linked population. We also described medical history, stroke severity and patterns of medication use among the PCNASP-claims linked cohort.ResultsThe linkage produced uniqueness equal to 99.1%. We identified 5644 linked and 98 896 unlinked patients who had an ischaemic stroke hospitalisation in claims data. Linked patients were younger than unlinked (69.7 vs 72.5 years), but otherwise similar by medical history, prestroke medication use or lab values. Stroke severity was mild and most patients were discharged home. Prestroke and discharge use of antihypertensive and statins in the PCNASP were greater than their use as measured by filled prescriptions in claims.ConclusionsHigh-quality linkage between the PCNASP and commercial claims data is feasible. This linkage identified differences between reported or recommended versus actual out-of-hospital medication utilisation, highlighting the importance of longitudinal data availability for research aimed to improve the care of patients who had a stroke.

Resistant hypertension (RH) has been poorly studied due to the difficulty in distinguishing it from nonadherence-the exclusion of which is necessary to accurately diagnose RH. Therefore, little is known about the prevalence, predictors, and outcomes of true RH. We evaluated 1838 patients from the standard blood pressure (BP) arm of the Action to Control Cardiovascular Risk in Diabetes Trial. We classified patients into three groups: "true RH", "pseudo-RH" (i.e., patients with BP levels that would classify them as RH but who were non-adherent), and "other" (i.e., those who could not be classified as having "true RH" or "pseudo-RH"). We examined predictors of true and pseudo-RH and the relationship between true RH and the composite outcome of nonfatal MI, nonfatal stroke, or cardiovascular death. Among 1838 participants with complete information, 489 (26.6%) met the definition of true RH, and 94 (16.1%) RH patients had "pseudo-RH" on ≥1 visit over the first 12 months. Predictors of RH included: baseline SBP ≥ 160 mmHg (OR = 8.79; 95% CI: 5.70-13.68) and baseline SBP between 140-159 (OR = 2.91; 95% CI: 2.13-4.00) compared to SBP < 140, additional baseline BP medication (OR = 3.40; 95% CI: 2.83-4.11), macroalbuminuria (OR = 2.35; 95% CI: 1.50-3.67), CKD (OR = 1.53; 95% CI: 0.99-2.33), history of stroke (OR = 1.73; 95% CI: 1.04-2.82), and black race (OR = 1.39; 95% CI: 1.02-1.88); the cross-validated C-statistic was 0.80. "True RH" patients had a 65% increased hazard in composite outcome (HR = 1.65; 95% CI: 1.13-2.42). In conclusion, the majority of patients classified as having RH had "true RH," which was more common among those who are black, have macroalbuminuria, CKD, stroke, higher baseline SBP, and are taking more baseline antihypertensives. These patients are at increased risk for cardiovascular and mortality events.

Many real-word evidence (RWE) studies that utilize existing healthcare data to evaluate treatment effects incur substantial but avoidable bias from methodologically flawed study design; however, the extent of preventable methodological pitfalls in current RWE is unknown. To characterize the prevalence of avoidable methodological pitfalls with potential for bias in published claims-based studies of medication safety or effectiveness, we conducted an English-language search of PubMed for articles published from January 1, 2010 to May 20, 2019 and randomly selected 75 studies (10 case-control and 65 cohort studies) that evaluated safety or effectiveness of cardiovascular, diabetes, or osteoporosis medications using US health insurance claims. General and methodological study characteristics were extracted independently by two reviewers, and potential for bias was assessed across nine bias domains. Nearly all studies (95%) had at least one avoidable methodological issue known to incur bias, and 81% had potentially at least one of the four issues considered major due to their potential to undermine study validity: time-related bias (57%), potential for depletion of outcome-susceptible individuals (44%), inappropriate adjustment for postbaseline variables (41%), or potential for reverse causation (39%). The median number of major issues per study was 2 (interquartile range (IQR), 1-3) and was lower in cohort studies with a new-user, active-comparator design (median 1, IQR 0-1) than in cohort studies of prevalent users with a nonuser comparator (median 3, IQR 3-4). Recognizing and avoiding known methodological study design pitfalls could substantially improve the utility of RWE and confidence in its validity.

Background: Regulators are evaluating the use of noninterventional real-world evidence (RWE) studies to assess the effectiveness of medical products. The RCT DUPLICATE initiative (Randomized, Controlled Trials Duplicated Using Prospective Longitudinal Insurance Claims: Applying Techniques of Epidemiology) uses a structured process to design RWE studies emulating randomized, controlled trials (RCTs) and compare results. We report findings of the first 10 trial emulations, evaluating cardiovascular outcomes of antidiabetic or antiplatelet medications. Methods: We selected 3 active-controlled and 7 placebo-controlled RCTs for replication. Using patient-level claims data from US commercial and Medicare payers, we implemented inclusion and exclusion criteria, selected primary end points, and comparator populations to emulate those of each corresponding RCT. Within the trial-mimicking populations, we conducted propensity score matching to control for >120 preexposure confounders. All study measures were prospectively defined and protocols registered before hazard ratios and 95% CIs were computed. Success criteria for the primary analysis were prespecified for each replication. Results: Despite attempts to emulate RCT design as closely as possible, differences between the RCT and corresponding RWE study populations remained. The regulatory conclusions were equivalent in 6 of 10. The RWE emulations achieved a hazard ratio estimate that was within the 95% CI from the corresponding RCT in 8 of 10 studies. In 9 of 10, either the regulatory or estimate agreement success criteria were fulfilled. The largest differences in effect estimates were found for RCTs where second-generation sulfonylureas were used as a proxy for placebo regarding cardiovascular effects. Nine of 10 replications had a standardized difference between effect estimates of Conclusions: Agreement between RCT and RWE findings varies depending on which agreement metric is used. Interim findings indicate that selection of active comparator therapies with similar indications and use patterns enhances the validity of RWE. Even in the context of active comparators, concordance between RCT and RWE findings is not guaranteed, partially because trials are not emulated exactly. More trial emulations are needed to understand how often and in what contexts RWE findings match RCTs. Registration: URL: https://www.clinicaltrials.gov ; Unique identifiers: NCT03936049, NCT04215523, NCT04215536, NCT03936010, NCT03936036, NCT03936062, NCT03936023, NCT03648424, NCT04237935, NCT04237922.

Background We explored whether clinically relevant baseline characteristics of patients with type 2 diabetes can modify the effect of glucagon-like peptide-1 receptor agonists (GLP-1 RA) or sodium-glucose cotransporter-2 inhibitors (SGLT-2i) on the risk of major adverse cardiovascular events (MACE). Methods We investigated Medline and EMBASE through June 2019. We included randomized clinical trials reporting the effect of GLP-1 RA or SGLT-2i on MACE in subgroups of patients with type 2 diabetes, identified through key baseline factors: established cardiovascular disease; heart failure; chronic kidney disease; uncontrolled diabetes; duration of diabetes; hypertension; obesity; age; gender and race. Hazard ratios (HRs) and 95% confidence intervals (CIs) from trials were meta-analyzed using random-effects models. Results Ten trials enrolling 89,790 patients were included in the analyses. Subgroup meta-analyses showed a 14% risk reduction of MACE in patients with established cardiovascular disease [GLP1-RA: HR, 0.86 (95% CI, 0.80–0.93); SGLT-2i: 0.86 (0.80–0.93)], and no effect in at-risk patients without history of cardiovascular events [GLP1-RA: 0.94 (0.82–1.07); SGLT-2i: 1.00 (0.87–1.16)]. We observed a trend toward larger treatment benefits with SGLT-2i among patients with chronic kidney disease [0.82 (0.69–0.97)], and patients with uncontrolled diabetes for both GLP1-RA or SGLT-2i [GLP1-RA: 0.82 (0.71–0.95); SGLT-2i: 0.84 (0.75–0.95)]. Uncontrolled hypertension, obesity, gender, age and race did not appear to modify the effect of these drugs. Conclusions In this exploratory analysis, history of cardiovascular disease appeared to modify the treatment effect of SGLT2i or GLP1-RA on MACE. Chronic kidney disease and uncontrolled diabetes should be further investigated as potential effect modifiers.

Supplemental Digital Content is available in the text., We use simulated data to examine the consequences of depletion of susceptibles for hazard ratio (HR) estimators based on a propensity score (PS). First, we show that the depletion of susceptibles attenuates marginal HRs toward the null by amounts that increase with the incidence of the outcome, the variance of susceptibility, and the impact of susceptibility on the outcome. If susceptibility is binary then the Bross bias multiplier, originally intended to quantify bias in a risk ratio from a binary confounder, also quantifies the ratio of the instantaneous marginal HR to the conditional HR as susceptibles are depleted differentially. Second, we show how HR estimates that are conditioned on a PS tend to be between the true conditional and marginal HRs, closer to the conditional HR if treatment status is strongly associated with susceptibility and closer to the marginal HR if treatment status is weakly associated with susceptibility. We show that associations of susceptibility with the PS matter to the marginal HR in the treated (ATT) though not to the marginal HR in the entire cohort (ATE). Third, we show how the PS can be updated periodically to reduce depletion-of-susceptibles bias in conditional estimators. Although marginal estimators can hit their ATE or ATT targets consistently without updating the PS, we show how their targets themselves can be misleading as they are attenuated toward the null. Finally, we discuss implications for the interpretation of HRs and their relevance to underlying scientific and clinical questions. See video Abstract: http://links.lww.com/EDE/B727.

Randomized clinical trials (RCTs) are the gold standard in producing clinical evidence of efficacy and safety of medical interventions. More recently, a new paradigm is emerging—specifically within the context of preauthorization regulatory decision‐making—for some novel uses of real‐world evidence (RWE) from a variety of real‐world data (RWD) sources to answer certain clinical questions. Traditionally reserved for rare diseases and other special circumstances, external controls (eg, historical controls) are recognized as a possible type of control arm for single‐arm trials. However, creating and analyzing an external control arm using RWD can be challenging since design and analytics may not fully control for all systematic differences (biases). Nonetheless, certain biases can be attenuated using appropriate design and analytical approaches. The main objective of this paper is to improve the scientific rigor in the generation of external control arms using RWD. Here we (a) discuss the rationale and regulatory circumstances appropriate for external control arms, (b) define different types of external control arms, and (c) describe study design elements and approaches to mitigate certain biases in external control arms. This manuscript received endorsement from the International Society for Pharmacoepidemiology (ISPE).

Background: Many new targeted cancer drugs have received FDA approval based on durable responses in nonrandomized controlled trials (non-RCTs). The goal of this study was to evaluate whether the response rates (RRs) and durations of response (DoRs) of targeted cancer drugs observed in non-RCTs are consistent when these drugs are tested in RCTs. Methods: We used the FDA’s Table of Pharmacogenomic Biomarkers in Drug Labeling to identify cancer drugs that were approved based on changes in biomarker endpoints through December 2017. We then identified the non-RCTs and RCTs for these drugs for the given indications and extracted the RRs and DoRs. We compared the RRs and median DoR in non-RCTs versus RCTs using the ratio of RRs and the ratio of DoRs, defined as the RRs (or DoRs) in non-RCTs divided by the RRs (or DoRs) in RCTs. The ratio of RRs or DoRs was pooled across the trial pairs using random-effects meta-analysis. Results: Of the 21 drug–indication pairs selected, both non-RCTs and RCTs were available for 19. The RRs and DoRs in non-RCTs were greater than those in RCTs in 63% and 87% of cases, respectively. The pooled ratio of RRs was 1.06 (95% CI, 0.95–1.20), and the pooled ratio of DoRs was 1.17 (95% CI, 1.03–1.33). RRs and DoRs derived from non-RCTs were also poor surrogates for overall survival derived from RCTs. Conclusions: The RRs were not different between non-RCTs and RCTs of cancer drugs approved based on changes to a biomarker, but the DoRs in non-RCTs were significantly higher than in RCTs. Caution must be exercised when approving or prescribing targeted drugs based on data on durable responses derived from non-RCTs, because the responses could be overestimates and poor predictors of survival benefit.

The randomized controlled trial (RCT) is the gold standard for evaluating the causal effects of medications. Limitations of RCTs have led to increasing interest in using real-world evidence (RWE) to augment RCT evidence and inform decision making on medications. Although RWE can be either randomized or nonrandomized, nonrandomized RWE can capitalize on the recent proliferation of large healthcare databases and can often answer questions that cannot be answered in randomized studies due to resource constraints. However, the results of nonrandomized studies are much more likely to be impacted by confounding bias, and the existence of unmeasured confounders can never be completely ruled out. Furthermore, nonrandomized studies require more complex design considerations which can sometimes result in design-related biases. We discuss questions that can help investigators or evidence consumers evaluate the potential impact of confounding or other biases on their findings: Does the design emulate a hypothetical randomized trial design? Is the comparator or control condition appropriate? Does the primary analysis adjust for measured confounders? Do sensitivity analyses quantify the potential impact of residual confounding? Are methods open to inspection and (if possible) replication? Designing a high-quality nonrandomized study of medications remains challenging and requires broad expertise across a range of disciplines, including relevant clinical areas, epidemiology, and biostatistics. The questions posed in this paper provide a guiding framework for assessing the credibility of nonrandomized RWE and could be applied across many clinical questions.

Actionable real-world evidence (RWE) requires accurate estimation of causal treatment effects. Calibration of RWE against randomized controlled trials (RCTs) is sometimes done to demonstrate that RWE can support the same causal conclusion as RCTs. Disagreements can occur when studies in each pair asked different questions in different populations or due to the presence of residual bias. Distinguishing among reasons for differences will impact the level of confidence in RWE. Several projects, such as the RCT DUPLICATE initiative funded by US FDA, NIH, and others, have launched in the last few years with the aim of assessing whether nonrandomized database studies can in some circumstances produce conclusions on the effectiveness of medications that are similar to those provided by RCTs.1,2 While comparison of randomized and nonrandomized findings is not new, heightened current interest is in part spurred by new initiatives at several regulatory agencies focused on assessing the role real-world evidence (RWE) can play in regulatory decision-making.3,4 FDA defines RWE as evidence on the benefits and risks of medications derived from routinely collected healthcare data, although other data sources such as patient registries may also be used. A key challenge for any project attempting to calibrate RWE findings against RCT findings is that differences between treatment effect estimates from the two study types can be driven by bias due to lack of randomization in RWE and/or by other differences in the design, such as inclusion/exclusion criteria, outcome measurement, or motivations for patients to adhere to study medications. Even if RWE studies are designed to match the corresponding RCT as closely as possible, emulation of all study components is typically impossible. Emulating target trials In clinical epidemiology, it has been recommended for several decades to contemplate how a randomized trial would be designed to answer a specific question before designing the nonrandomized counterpart to study the same question. Hernan and Robins call the former a hypothetical “target trial,” which would then be emulated by a nonrandomized study.5 This process has proven very useful in clarifying design choices for nonrandomized research on medications. It is also a highly flexible process, as specification of the target trial is often iterative. Realities of data collection, patient access, and other practical considerations impose constraints to the nonrandomized emulation; the hypothetical target trial can thus be adjusted so that the design that is feasible in a given healthcare database can accord with the design of the target trial. In calibrating RWE studies against existing RCTs, the process of emulating a target RCT in nonrandomized data is similar, except that in this case the target trial is already underway or completed. Thus, adjusting the design of the target RCT to improve the feasibility of the design in existing data is not possible, making exact emulation more difficult. Instead, investigators must adapt the design elements that they can from the trial, given the constraints of the database, and note the trial specifications that cannot be completely emulated. This process will highlight unavoidable emulation differences between a completed or in-progress RCT and the RWE replication. When assembling a series of RWE replications of RCTs, as in the RCT DUPLICATE initiative, there will be some trials where RWE specifications can closely emulate the RCT and others that cannot be emulated as closely. Examples of the latter are run-in periods during which non-adherent or drug intolerant patients are excluded before randomization, or, to homogenize patients, run-in periods that place all patients on a common medication before randomization. Therefore, enumerating – and quantifying to the extent possible – such emulation differences can provide insight into whether and to what extent differences in treatment effect estimates between RWE and corresponding RCTs are due to bias related to a lack of randomization versus other differences in design between the two study types. Specifically, attempting to correlate measures of emulation difference, such as those suggested in Table 1, with the magnitude of differences in treatment effect estimates may provide understanding of which emulation differences are most important in contributing to the “efficacy-effectiveness gap” between RWE and RCT findings, as long as bias due to confounding is not also correlated with the emulation difference of interest. Once better understood, such metrics could possibly be transformed into a simple three-point emulation scale. Table 1. Challenges in calibrating RWE against RCTs and measures of differences between the study types.

Coarsened exact matching (CEM) is a matching method proposed as an alternative to other techniques commonly used to control confounding. We compared CEM with 3 techniques that have been used in pharmacoepidemiology: propensity score matching, Mahalanobis distance matching, and fine stratification by propensity score (FS). We evaluated confounding control and effect-estimate precision using insurance claims data from the Pharmaceutical Assistance Contract for the Elderly (1999–2002) and Medicaid Analytic eXtract (2000–2007) databases (United States) and from simulated claims-based cohorts. CEM generally achieved the best covariate balance. However, it often led to high bias and low precision of the risk ratio due to extreme losses in study size and numbers of outcomes (i.e., sparse data bias)—especially with larger covariate sets. FS usually was optimal with respect to bias and precision and always created good covariate balance. Propensity score matching usually performed almost as well as FS, especially with higher index exposure prevalence. The performance of Mahalanobis distance matching was relatively poor. These findings suggest that CEM, although it achieves good covariate balance, might not be optimal for large claims-database studies with rich covariate information; it might be ideal if only a few (

Background The US Food and Drug Administration’s accelerated approval and later withdrawal of bevacizumab in patients with metastatic breast cancer (mBC) is a seminal case for ongoing debates about the validity of using progression-free survival (PFS) as a surrogate measure for overall survival (OS) in cancer drug approvals. We systematically reviewed and meta-analyzed the evidence around bevacizumab’s regulatory approval and withdrawal in mBC. Methods We searched for all published phase II or III clinical trials testing bevacizumab as a first-line therapy for patients with mBC. Data were extracted on trial demographics, interventions, and outcomes. Descriptive analysis was stratified by whether the trial was initiated before, during, or after the accelerated approval. We used a cumulative random-effects meta-analysis to assess the evolution of evidence of the effect of bevacizumab on PFS and OS. We estimated the association between the trial-level PFS and OS effect using a nonlinear mixed-regression model. Results Fifty-two studies were included. Trial activity dramatically dropped after the accelerated approval was withdrawn. Eight clinical trials reported hazard ratios (hazard ratios) and were meta-analyzed. The cumulative hazard ratio for PFS was 0.72 (95% CI = 0.65 to 0.79), and the cumulative hazard ratio for OS was 0.90 (95% CI = 0.80 to 1.01). The regression model showed a statistically nonsignificant association between PFS benefit and OS benefit (β = 0.43, SE = 0.81). Conclusion The US Food and Drug Administration’s decision-making in this case was consistent with the evolving state of evidence. However, the fact that seven clinical trials are insufficient to conclude validity (or lack thereof) for a trial-level surrogate suggests that it would be more efficient to conduct trials using the more clinically meaningful endpoints.

OBJECTIVE Using real-world data (RWD) from three U.S. claims data sets, we aim to predict the findings of the CARdiovascular Outcome Trial of LINAgliptin Versus Glimepiride in Type 2 Diabetes (CAROLINA) comparing linagliptin versus glimepiride in patients with type 2 diabetes (T2D) at increased cardiovascular risk by using a novel framework that requires passing prespecified validity checks before analyzing the primary outcome. RESEARCH DESIGN AND METHODS Within Medicare and two commercial claims data sets (May 2011–September 2015), we identified a 1:1 propensity score–matched (PSM) cohort of T2D patients 40–85 years old at increased cardiovascular risk who initiated linagliptin or glimepiride by adapting eligibility criteria from CAROLINA. PSM was used to balance >120 confounders. Validity checks included the evaluation of expected power, covariate balance, and two control outcomes for which we expected a positive association and a null finding. We registered the protocol (NCT03648424, ClinicalTrials.gov) before evaluating the composite cardiovascular outcome based on CAROLINA’s primary end point. Hazard ratios (HR) and 95% CIs were estimated in each data source and pooled with a fixed-effects meta-analysis. RESULTS We identified 24,131 PSM pairs of linagliptin and glimepiride initiators with sufficient power for noninferiority (>98%). Exposure groups achieved excellent covariate balance, including key laboratory results, and expected associations between glimepiride and hypoglycemia (HR 2.38 [95% CI 1.79–3.13]) and between linagliptin and end-stage renal disease (HR 1.08 [0.66–1.79]) were replicated. Linagliptin was associated with a 9% decreased risk in the composite cardiovascular outcome with a CI including the null (HR 0.91 [0.79–1.05]), in line with noninferiority. CONCLUSIONS In a nonrandomized RWD study, we found that linagliptin has noninferior risk of a composite cardiovascular outcome compared with glimepiride.

Background: The EMPA-REG OUTCOME trial (Empagliflozin Cardiovascular Outcome Event Trial in Type 2 diabetes Mellitus Patients) showed that empagliflozin, a sodium-glucose cotransporter-2 inhibitor, reduces the risk of hospitalization for heart failure (HHF) by 35%, on top of standard of care in patients with type 2 diabetes mellitus (T2D) and established cardiovascular disease. The EMPRISE (Empagliflozin Comparative Effectiveness and Safety) study aims to assess empagliflozin’s effectiveness, safety, and healthcare utilization in routine care from August 2014 through September 2019. In this first interim analysis, we investigated the risk of HHF among T2D patients initiating empagliflozin versus sitagliptin, a dipeptidyl peptidase-4 inhibitor. Methods: Within 2 commercial and 1 federal (Medicare) claims data sources in the United States, we identified a 1:1 propensity score–matched cohort of T2D patients ≥18 years old initiating empagliflozin or sitagliptin from August 2014 through September 2016. The HHF outcome was defined as a HF discharge diagnosis in the primary position (HHF-specific); a broader definition was based on a HF discharge diagnosis in any position (HHF-broad). Hazard ratios (HRs) and 95% CIs were estimated controlling for over 140 baseline characteristics in each data source and pooled by fixed-effects meta-analysis. Results: After propensity-score matching, we identified 16,443 patient pairs who initiated empagliflozin or sitagliptin. Average age was approximately 59 years, almost 54% of the participants were males, and approximately 25% had records of existing cardiovascular disease. Compared with sitagliptin, the initiation of empagliflozin decreased the risk of HHF-specific by 50% (HR, 0.50; 95% CI, 0.28–0.91), and the risk of HHF-broad by 49% (HR, 0.51;95% CI, 0.39–0.68), over a mean follow-up of 5.3 months. The results were consistent in patients with and without baseline cardiovascular disease, and for empagliflozin at both the 10- and 25-mg daily doses; analyses comparing empagliflozin versus the dipeptidyl peptidase-4 inhibitor class, and comparing sodium-glucose cotransporter-2 inhibitor versus dipeptidyl peptidase-4 inhibitor classes also produced consistent findings. Conclusions: The first interim analysis from EMPRISE showed that compared with sitagliptin, the initiation of empagliflozin was associated with a decreased risk of HHF among patients with T2D as treated in routine care, with and without a history of cardiovascular disease. Clinical Trial Registration: URL: https://www.clinicaltrials.gov . Unique identifier: NCT03363464.

AIM: Recent years have witnessed a proliferation of observational research examining the association between glucose-lowering medications and cancer, paired with an increasing concern on the quality of the produced evidence. Our objective was to review published studies in this area to identify the most common methodological challenges and sources of bias. METHODS: We systematically searched PubMed to identify observational studies on glucose-lowering medications and cancer published between January 2000 and January 2016. We assessed the design and analytical methods used in each study, with a focus on their ability to achieve study validity, and further evaluated the prevalence of major methodological choices over time. RESULTS: Out of 155 studies evaluated, only 26% implemented a new user design, 41% used an active comparator, 33% implemented a lag or latency period, 51% adjusted for diabetes duration. Potential for immortal person-time bias was identified in 63% of the studies; 55% of the studies adjusted for variables measured during the follow-up without appropriate statistical methods. Aside from a decreasing trend in adjusting for variables measured during the follow-up, we observed no trends in methodological choices over time. CONCLUSIONS: The prevalence of well-known design and analysis flaws that may lead to biased results remains high among observational studies on glucose-lowering medications and cancer, limiting the conclusions that can be drawn from these studies. Avoiding known pitfalls could substantially improve the quality and validity of real-world evidence in this field.

PURPOSE Bootstrapping can account for uncertainty in propensity score (PS) estimation and matching processes in 1:1 PS-matched cohort studies. While theory suggests that the classical bootstrap can fail to produce proper coverage, practical impact of this theoretical limitation in settings typical to pharmacoepidemiology is not well studied. METHODS In a plasmode-based simulation study, we compared performance of the standard parametric approach, which ignores uncertainty in PS estimation and matching, with two bootstrapping methods. The first method only accounted for uncertainty introduced during the matching process (the observation resampling approach). The second method accounted for uncertainty introduced during both PS estimation and matching processes (the PS reestimation approach). Variance was estimated based on percentile and empirical standard errors, and treatment effect estimation was based on median and mean of the estimated treatment effects across 1000 bootstrap resamples. Two treatment prevalence scenarios (5% and 29%) across two treatment effect scenarios (hazard ratio of 1.0 and 2.0) were evaluated in 500 simulated cohorts of 10 000 patients each. RESULTS We observed that 95% confidence intervals from the bootstrapping approaches but not the standard approach, resulted in inaccurate coverage rates (98%-100% for the observation resampling approach, 99%-100% for the PS reestimation approach, and 95%-96% for standard approach). Treatment effect estimation based on bootstrapping approaches resulted in lower bias than the standard approach (less than 1.4% vs 4.1%) at 5% treatment prevalence; however, the performance was equivalent at 29% treatment prevalence. CONCLUSION Use of bootstrapping led to variance overestimation and inconsistent coverage, while coverage remained more consistent with parametric estimation.

During 2017, opioids were associated with 47,600 deaths in the United States, approximately one third of which involved a prescription opioid (1). Amid concerns that overprescribing to patients with acute pain remains an essential factor underlying misuse, abuse, diversion, and unintentional overdose, several states have restricted opioid analgesic prescribing (2,3). To characterize patterns of opioid analgesic use for acute pain in primary care settings before the widespread implementation of limits on opioid prescribing (2,3), patients filling an opioid analgesic prescription for acute pain were identified from a 2014 database of commercial claims. Using a logistic generalized additive model, the probability of obtaining a refill was estimated as a function of the initial number of days supplied. Among 176,607 patients with a primary care visit associated with an acute pain complaint, 7.6% filled an opioid analgesic prescription. Among patients who received an initial 7-day supply, the probability of obtaining an opioid analgesic prescription refill for nine of 10 conditions was

Only an estimated 5 percent of rare pediatric diseases have a treatment, although collectively they affect more than ten million children in the US. To stimulate drug development for rare pediatric diseases, Congress expanded the priority review voucher (PRV) program in 2012. A pediatric PRV, which can be sold to another manufacturer, requires the FDA to provide priority six-month review rather than the standard ten-month review to another drug of the company's choosing. We compared rare pediatric disease drugs eligible for a PRV and rare adult disease drugs (which are not eligible for a PRV). We found that compared to drugs for rare adult diseases, drugs for rare pediatric diseases progressed more quickly through all phases of clinical testing and were more likely to be first-in-class. The voucher program was not associated with a change in the rate of new pediatric drugs starting or completing clinical testing, but there was a significant increase in the rate of progress from Phase I to Phase II clinical trials after the program was implemented. New policies may be needed to expand the pipeline of therapies for rare pediatric diseases.

Resistant hypertension (RH) has been poorly studied due to the difficulty in distinguishing it from nonadherence-the exclusion of which is necessary to accurately diagnose RH. Therefore, little is known about the prevalence, predictors, and outcomes of true RH. We evaluated 1838 patients from the standard blood pressure (BP) arm of the Action to Control Cardiovascular Risk in Diabetes Trial. We classified patients into three groups: "true RH", "pseudo-RH" (i.e., patients with BP levels that would classify them as RH but who were non-adherent), and "other" (i.e., those who could not be classified as having "true RH" or "pseudo-RH"). We examined predictors of true and pseudo-RH and the relationship between true RH and the composite outcome of nonfatal MI, nonfatal stroke, or cardiovascular death. Among 1838 participants with complete information, 489 (26.6%) met the definition of true RH, and 94 (16.1%) RH patients had "pseudo-RH" on ≥1 visit over the first 12 months. Predictors of RH included: baseline SBP ≥ 160 mmHg (OR = 8.79; 95% CI: 5.70-13.68) and baseline SBP between 140-159 (OR = 2.91; 95% CI: 2.13-4.00) compared to SBP 140, additional baseline BP medication (OR = 3.40; 95% CI: 2.83-4.11), macroalbuminuria (OR = 2.35; 95% CI: 1.50-3.67), CKD (OR = 1.53; 95% CI: 0.99-2.33), history of stroke (OR = 1.73; 95% CI: 1.04-2.82), and black race (OR = 1.39; 95% CI: 1.02-1.88); the cross-validated C-statistic was 0.80. "True RH" patients had a 65% increased hazard in composite outcome (HR = 1.65; 95% CI: 1.13-2.42). In conclusion, the majority of patients classified as having RH had "true RH," which was more common among those who are black, have macroalbuminuria, CKD, stroke, higher baseline SBP, and are taking more baseline antihypertensives. These patients are at increased risk for cardiovascular and mortality events.

ObjectiveTo determine whether assessment tools for non-randomised studies (NRS) address critical elements that influence the validity of NRS findings for comparative safety and effectiveness of medications.DesignSystematic review and Delphi survey.Data sourcesWe searched PubMed, Embase, Google, bibliographies of reviews and websites of influential organisations from inception to November 2019. In parallel, we conducted a Delphi survey among the International Society for Pharmacoepidemiology Comparative Effectiveness Research Special Interest Group to identify key methodological challenges for NRS of medications. We created a framework consisting of the reported methodological challenges to evaluate the selected NRS tools.Study selectionChecklists or scales assessing NRS.Data extractionTwo reviewers extracted general information and content data related to the prespecified framework.ResultsOf 44 tools reviewed, 48% (n=21) assess multiple NRS designs, while other tools specifically addressed case–control (n=12, 27%) or cohort studies (n=11, 25%) only. Response rate to the Delphi survey was 73% (35 out of 48 content experts), and a consensus was reached in only two rounds. Most tools evaluated methods for selecting study participants (n=43, 98%), although only one addressed selection bias due to depletion of susceptibles (2%). Many tools addressed the measurement of exposure and outcome (n=40, 91%), and measurement and control for confounders (n=40, 91%). Most tools have at least one item/question on design-specific sources of bias (n=40, 91%), but only a few investigate reverse causation (n=8, 18%), detection bias (n=4, 9%), time-related bias (n=3, 7%), lack of new-user design (n=2, 5%) or active comparator design (n=0). Few tools address the appropriateness of statistical analyses (n=15, 34%), methods for assessing internal (n=15, 34%) or external validity (n=11, 25%) and statistical uncertainty in the findings (n=21, 48%). None of the reviewed tools investigated all the methodological domains and subdomains.ConclusionsThe acknowledgement of major design-specific sources of bias (eg, lack of new-user design, lack of active comparator design, time-related bias, depletion of susceptibles, reverse causation) and statistical assessment of internal and external validity is currently not sufficiently addressed in most of the existing tools. These critical elements should be integrated to systematically investigate the validity of NRS on comparative safety and effectiveness of medications.Systematic review protocol and registrationhttps://osf.io/es65q.

The emergence and global spread of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has resulted in an urgent need for evidence on medical interventions and outcomes of the resulting disease, coronavirus disease 2019 (COVID-19). Although many randomized controlled trials (RCTs) evaluating treatments and vaccines for COVID-19 are already in progress, the number of clinical questions of interest greatly outpaces the available resources to conduct RCTs. Therefore, there is growing interest in whether nonrandomized real-world evidence (RWE) can be used to supplement RCT evidence and aid in clinical decision making, but concerns about nonrandomized RWE have been highlighted by a proliferation of RWE studies on medications and COVID-19 outcomes with widely varying conclusions. The objective of this paper is to review some clinical questions of interest, potential data types, challenges, and merits of RWE in COVID-19, resulting in recommendations for nonrandomized RWE designs and analyses based on established RWE principles.

There is increasing interest in utilizing real-world data (RWD) to produce real-world evidence (RWE) on the benefits and risks of medical products that could support regulatory approval decisions. The field of pharmacoepidemiology has a long history of focusing on data and evidence that would now be termed "real-world," including evidence from healthcare claims, registries, and electronic health records. However, several emerging trends over the past decade are converging to support the use of these and other RWD sources for approval decisions, and there are several recent examples and ongoing research that demonstrate how RWE may be used to support regulatory approval of new or expanded indications. The goal of this article is to review the current landscape and future directions of the use of RWE in this context. This manuscript is endorsed by the International Society for Pharmacoepidemiology (ISPE).

PURPOSE: To evaluate recent trends in inpatient postoperative utilization of opioid and non-opioid analgesics in US hospitals. METHODS: Using Premier Research database (October 2007 – September 2017), we identified adults who were hospitalized for inpatient surgical procedures (N = 6,068,133). For each month, we calculated proportion of patients admitted that month who were administered (1) opioids, (2) acetaminophen, (3) non-steroidal anti-inflammatory drugs (NSADs), and (4) gabapentinoids (gabapentin or pregabalin) during the postoperative period, defined as inpatient postoperative days 1–7, unless discharged earlier. For patients administered opioids, we estimated total and average daily postoperative opioid dose in morphine milligrams equivalents (MMEs). Monthly measures were standardized to the distribution of surgeries and the length of postoperative stay within each surgery during the last year of available data. RESULTS: Overall, 90.8% of patients were administered opioids postoperatively; mean total postoperative dose was 304 MMEs (median 130). Both the frequency and the amount of opioids administered remained stable over 2007–2017. Postoperative use of acetaminophen increased from mean standardized monthly prevalence of 78% in 2007–2008 to 85% in 2017, while the use of NSAIDs remained stable at a standardized mean of 37%. The use of gabapentinoids increased from below 10% in 2007–2008 to the mean standardized monthly prevalence of 23% in 2017. CONCLUSION: Despite growing awareness of risks associated with postoperative opioid use, we observed no change in postoperative utilization of opioids in US hospitals. Increasing the use of non-opioid pain management approaches, could constitute an important target in our efforts to curtail US opioid epidemic.

Key Points Question Is the adjuvant use of gabapentinoids with opioids associated with increased risk of opioid-related adverse events in patients undergoing surgery? Findings In this cohort study of 5 547 667 US surgical admissions, the addition of gabapentinoids to opioids was associated with an increased risk of opioid overdose and other opioid-related adverse events; the absolute risk of adverse events was low. Meaning These findings suggest that patients who receive gabapentinoids with opioids for postoperative analgesia should be closely monitored for possible respiratory depression., This cohort study evaluates the association between perioperative coadministration of gabapentinoids and opioids with inpatient opioid-related adverse events in US surgical patients., Importance The use of gabapentinoids in multimodal postoperative analgesia is increasing; however, when coadministered with opioids, these drugs may potentiate central nervous system and respiratory depression. Objective To evaluate the association between perioperative coadministration of gabapentinoids and opioids with inpatient opioid-related adverse events in surgical patients. Design, Setting, and Participants This cohort study used propensity score trimming, stratification, and weighting of adults admitted for a major surgery between October 2007 and December 2017 who were treated with opioids on the day of surgery and included in the Premier Research database. Data analysis was conducted from February to April 2020. Exposure Gabapentinoids (gabapentin or pregabalin) coadministered with opioids starting the day of surgery vs opioid therapy without gabapentinoids. Main Outcomes and Measures Primary outcome was opioid overdose. Secondary outcomes included respiratory complications, unspecified adverse effects of opioid use, and a composite of these 3 outcomes. Patients were followed up for as long as 30 days from the day of surgery until deviation from the initial treatment regimen or discharge. Results Gabapentinoids with opioids were administered to 892 484 of 5 547 667 eligible admissions (16.1%; mean [SD] age, 63.5 [11.8] years; 353 315 [39.6%] men). Among the 4 655 183 patients who received opioids only, the mean (SD) age was 63.7 (14.7) years, and 1 913 284 (41.1%) were men. Overall, 441 overdose events were identified, with absolute risks of 1.4 per 10 000 patients with gabapentinoid exposure and 0.7 per 10 000 patients receiving opioids only. Following propensity score trimming, the cohort included 737 383 patients exposed to gabapentinoids and 3 002 480 patients receiving opioids only. The primary analysis yielded the adjusted hazard ratio of 1.95 (95% CI, 1.49-2.55), and the number needed to treat for an additional overdose to occur was 16 914 patients (95% CI, 11 556-31 537 patients). Adjusted hazard ratios for secondary outcomes were 1.68 (95% CI, 1.59-1.78) for respiratory complications, 1.77 (95% CI, 1.61-1.93) for unspecified adverse effects of opioids, and 1.70 (95% CI, 1.62-1.79) for the composite outcome. The results were consistent across sensitivity analyses and subgroups identified by key clinical factors. Conclusions and Relevance In this real-world cohort study of patients who underwent major surgery, concomitant use of gabapentinoids with opioids was associated with increased risk of opioid overdose and other opioid-related adverse events; however, the absolute risk of adverse events was low.

Introduction: Resistant hypertension (RH) has been poorly studied due to the difficulty distinguishing RH from medication non-adherence—exclusion of which is necessary to accurately diagnose RH. Therefore, little is known about prevalence, predictors, and outcomes of true RH, particularly for patients with diabetes. Objective: To characterize prevalence, predictors, and outcomes of true RH under a definition accounting for adherence in a study with strict adherence monitoring. Methods: We analyzed data from 1838 patients in the standard BP arm of ACCORD. Patients were classified into 3 mutually-exclusive groups: “true RH”, “pseudo-RH” (i.e., patients with BP treatment levels that would classify them as having RH, but who were non-adherent), and “Other” (i.e., those without “true RH” or “pseudo-RH” classifications). Predictors of true- and pseudo-RH were identified with logistic and boosted regression models. Relationship between true RH and primary composite outcome of nonfatal MI, nonfatal stroke, or cardiovascular death was evaluated with cox models. Results: Among 1838 participants with complete information, 489 (26.6%) met the definition of true RH, and 94 (16.1%) of RH patients had “pseudo-RH” on ≥1 visit over the first 12 months. Multivariable predictors of RH included: baseline SBP ≥160 mmHg (OR=8.79; 95CI:5.70-13.68), baseline SBP between 140-159 (OR=2.91; 95CI:2.13-4.00) compared to SBP Conclusions: The majority of patients classified as RH had “true RH,” with only 16% being “pseudo-RH” due to non-adherence. True RH is more common among those who are black, have macroalbuminuria, CKD, history of stroke, higher baseline SBP, and are taking more baseline anti-hypertensives. “True RH” patients are at increased risk for cardiovascular and mortality events.

Background Event-free survival (EFS) has been listed on the FDA Table of Surrogate Endpoints as a surrogate measure that can be considered for accelerated or traditional approval in breast cancer. However, no studies have evaluated the correlation between the treatment effects on EFS and treatment effects on overall survival (OS). Methods We performed a systematic search of the literature until May 2020 according to the PRISMA guideline for all published randomized controlled trials (RCTs) in early breast cancer in the neoadjuvant setting. Data on EFS and OS, including the hazard ratio (HR) and 95% confidence intervals (CI), were extracted from each study and the association between the trial-level EFS HR and the trial-level OS HR was estimated using a linear mixed-effects model on the log scale. Findings Of the 7 RCTs (N = 2211) included in the analysis, 5 included patients with HER2 positive tumor type. The estimated linear association between log HR EFS and log HR OS indicated a positive slope ( β = 0.58 [95% CI: −0.32–1.48]) and the coefficient of determination confirmed a moderate trial-level association between log HRs for OS and EFS (R² 0.76 [95% CI 0.34–1.00], but with wide confidence intervals. Interpretation Treatment effects in EFS are moderately correlated with treatment effects in OS in early breast cancer in the neoadjuvant setting, but the association was not significant. Thus, there is currently insufficient evidence to support EFS for use as a surrogate endpoint for traditional approval, although it may be considered for accelerated approval. Funding Arnold Ventures.

The optimal learner for prediction modeling varies depending on the underlying data-generating distribution. Super Learner (SL) is a generic ensemble learning algorithm that uses cross-validation to select among a "library" of candidate prediction models. The SL is not restricted to a single prediction model, but uses the strengths of a variety of learning algorithms to adapt to different databases. While the SL has been shown to perform well in a number of settings, it has not been thoroughly evaluated in large electronic healthcare databases that are common in pharmacoepidemiology and comparative effectiveness research. In this study, we applied and evaluated the performance of the SL in its ability to predict treatment assignment using three electronic healthcare databases. We considered a library of algorithms that consisted of both nonparametric and parametric models. We also considered a novel strategy for prediction modeling that combines the SL with the high-dimensional propensity score (hdPS) variable selection algorithm. Predictive performance was assessed using three metrics: the negative log-likelihood, area under the curve (AUC), and time complexity. Results showed that the best individual algorithm, in terms of predictive performance, varied across datasets. The SL was able to adapt to the given dataset and optimize predictive performance relative to any individual learner. Combining the SL with the hdPS was the most consistent prediction method and may be promising for PS estimation and prediction modeling in electronic healthcare databases.

INTRODUCTION: Sodium glucose co-transporter-2 inhibitors (SGLT2) are commonly prescribed to patients with type 2 diabetes mellitus, but can increase the risk of diabetic ketoacidosis. Identifying patients prone to diabetic ketoacidosis may help mitigate this risk. METHODS: We conducted a population-based cohort study of adults initiating SGLT2 inhibitor use from 2013 through 2017. The primary objective was to identify potential predictors of diabetic ketoacidosis. Two machine-learning methods were applied to model high-dimensional pre-exposure data: gradient boosted trees and least absolute shrinkage and selection operator (LASSO) regression. We rank ordered the variables produced from LASSO by the size of their estimated coefficient (largest to smallest). With gradient boosted trees, a relative importance measure for each variable is provided rather than a coefficient. The “top variables” were identified after reviewing the distributions of the effect estimates from LASSO and gradient boosted trees to identify where there was a substantial decrease in variable importance. The identified predictors were then assessed in a logistic regression model and reported as odds ratios (ORs) with 95% confidence intervals (CIs). RESULTS: We identified 111,442 adults who started SGLT2 inhibitor use. The mean age was 57 years, 44% were female, the mean hemoglobin A1C was 8.7%, and the mean creatinine was 0.89 mg/dL. During a mean follow-up of 180 days, 192 patients (0.2%, i.e., 2 per 1000) were diagnosed and hospitalized with diabetic ketoacidosis (DKA) and 475 (0.4%, i.e., 4 per 1000) were diagnosed in either an inpatient or outpatient setting. Using gradient boosted trees, the strongest predictors were prior DKA, baseline hemoglobin A1C level, baseline creatinine level, use of medications for dementia, and baseline bicarbonate level. Using LASSO regression not including laboratory test results due to missing data, the strongest predictors were prior DKA, digoxin use, use of medications for dementia, and recent hypoglycemia. The logistic regression model incorporating the variables identified from gradient boosted trees and LASSO regression suggested the following pre-exposure characteristics had the strongest association with a hospitalization for DKA: use of dementia medications (OR = 7.76, 95% CI 2.60, 23.1), prior intracranial hemorrhage (OR = 11.5, 95% CI 1.46, 91.1), a prior diagnosis of hypoglycemia (OR = 5.41, 95% CI 1.92,15.3), prior DKA (OR = 2.45, 95% CI 0.33, 18.0), digoxin use (OR = 4.00, 95% CI 1.21, 13.2), a baseline hemoglobin A1C above 10% (OR = 3.14, 95% CI 1.95, 5.06), and baseline bicarbonate below 18 mmol/L (OR 5.09, 95% CI 1.58, 16.4). CONCLUSION: Diabetic ketoacidosis affected approximately 2 per 1000 patients starting to use an SGLT2 inhibitor. We identified both anticipated, e.g., low baseline serum bicarbonate, and unanticipated, e.g., digoxin, dementia medications, risk factors for SGLT2 inhibitor-induced DKA. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11606-020-06561-z.

EMPRISE aims to study the effectiveness, safety and healthcare utilization of empagliflozin (EMPA) in routine care patients. In an interim analysis, we assessed the comparative cardiovascular (CV) effectiveness of EMPA in older adults. Using data from Medicare (2014-2017), we identified 11,579 pairs of 1:1 propensity score-matched patients ≥66 years with type 2 diabetes initiating EMPA or a DPP-4 inhibitor (DPP-4i), and 17,500 pairs initiating EMPA or a GLP-1 receptor agonist (GLP-1RA). We assessed hospitalization for heart failure (HHF), defined as a HF discharge diagnosis in the primary (HHF-Specific) or in any position (HHF-Broad) and a modified MACE outcome (MI, stroke, or all-cause mortality). We estimated HR and 95% CI adjusting for >140 baseline covariates. Compared to DPP-4i, EMPA was associated with a decreased risk of HHF [HHF-Specific: 0.43 (0.30-0.63); HHF-Broad: 0.57 (0.47-0.69)] and MACE [HR (95% CI) = 0.63 (0.50-0.79)]. Compared to GLP-1RA, EMPA had a reduced risk of HHF [HHF-Specific: 0.67 (0.50-0.91); HHF-Broad: 0.83 (0.71-0.96)] and a similar risk of MACE [1.02 (0.85-1.23)](Table 1). Results for the individual components of MACE were consistent and robust to competing risks. These findings, addressing the comparative effectiveness of EMPA in routine care patients with mean age of 72 years, complement available information from RCTs. Disclosure E. Patorno: Other Relationship; Self; Boehringer Ingelheim International GmbH. A. Pawar: None. L.G. Bessette: None. M. Najafzadeh: None. D.J. Wexler: Other Relationship; Self; Novo Nordisk A/S. J. Franklin: None. A. Deruaz-Luyet: Employee; Self; Boehringer Ingelheim International GmbH. Employee; Spouse/Partner; Sanofi-Aventis Deutschland GmbH. K. Brodovicz: Employee; Self; Boehringer Ingelheim Pharmaceuticals, Inc. L. Koeneman: Employee; Self; Eli Lilly and Company. Stock/Shareholder; Self; Eli Lilly and Company. S. Schneeweiss: None. Funding Boehringer Ingelheim

IntroductionWith increasing rates of opioid overdoses in the US, a surveillance tool to identify high-risk patients may help facilitate early intervention.ObjectiveTo develop an algorithm to predict overdose using routinely-collected healthcare databases.MethodsWithin a US commercial claims database (2011-2015), patients with ≥1 opioid prescription were identified. Patients were randomly allocated into the training (50%), validation (25%), or test set (25%). For each month of follow-up, pooled logistic regression was used to predict the odds of incident overdose in the next month based on patient history from the preceding 3-6 months (time-updated), using elastic net for variable selection. As secondary analyses, we explored whether using simpler models (few predictors, baseline only) or different analytic methods (random forest, traditional regression) influenced performance.ResultsWe identified 5,293,880 individuals prescribed opioids; 2,682 patients (0.05%) had an overdose during follow-up (mean: 17.1 months). On average, patients who overdosed were younger and had more diagnoses and prescriptions. The elastic net model achieved good performance (c-statistic 0.887, 95% CI 0.872-0.902; sensitivity 80.2, specificity 80.1, PPV 0.21, NPV 99.9 at optimal cutpoint). It outperformed simpler models based on few predictors (c-statistic 0.825, 95% CI 0.808-0.843) and baseline predictors only (c-statistic 0.806, 95% CI 0.787-0.26). Different analytic techniques did not substantially influence performance. In the final algorithm based on elastic net, the strongest predictors were age 18-25 years (OR: 2.21), prior suicide attempt (OR: 3.68), opioid dependence (OR: 3.14).ConclusionsWe demonstrate that sophisticated algorithms using healthcare databases can be predictive of overdose, creating opportunities for active monitoring and early intervention.

Regulators wish to understand whether real world evidence can be used for secondary indications of biologics. Using the secondary indication of adalimumab for ulcerative colitis (UC) as an example, we aimed to replicate the ULTRA-2 randomized controlled trial finding on the effectiveness of adalimumab in patients with UC using realworld data analyses. Adalimumab, a TNF-alpha receptor inhibitor initially approved for Crohn's disease, was approved for moderate to severe UC in 2012. The ULTRA-2 trial had shown improved remission against placebo in patients with UC. Using claims data (2006-2012), we conducted a cohort study of patients with UC who initiated adalimumab and compared them with (i) nonusers and (ii) new users of infliximab using propensity score matching. The coprimary end points were corticosteroid (CS) discontinuation within 8 weeks and 1 year of treatment. We computed hazard ratios (HRs) and 95% confidence intervals (CIs). We identified 398 matched pairs of adalimumab users vs. nonusers and 326 pairs of adalimumab vs. infliximab users. Adalimumab users were 28% more likely to achieve CS-discontinuation compared with nonusers over 1 year (HR = 1.28; 95% CI 0.94-1.73). However, unlike in ULTRA-2, this effect was not observed in the first 8 weeks (HR = 0.79; 95% CI 0.65-0.97). Compared with infliximab, adalimumab initiators showed no incremental benefit over 1 year (HR = 1.08; 95% CI 0.80-1.04), but showed a 22% reduction (HR = 0.78; 95% CI 0.64-0.95) during the first 8 weeks of treatment. In summary, our results highlight opportunities and some limitations of database analysis to identify treatment effects for secondary indications.

Numerous initiatives over the past decade have targeted the problem of antibiotic overuse in the US; however, the cumulative impact of such initiatives upon recent patterns of use is not known. The aims of this study were to (1) describe general trends in outpatient antibiotic use among adults over the period 2006–2015; and (2) identify rapid shifts in use during this time period as potential indicators for key events. This was an observational study set in the ambulatory setting. Patients ≥ 18 years of age were selected from the Optum Clinformatics Datamart™, a commercial insurance claims database. The outcome measures of interest were prescriptions filled/1000 enrolled individuals, by year or quarter. We used linear regression to identify trends in use over multiple years, and change-point regression to identify rapid shifts in use within individual years. From 2006 to 2015, antibiotic use declined significantly, decreasing by 12% for adults younger than 65 years of age (913–807 prescriptions/1000 individuals, p = 0.0001) and by 5% for adults ≥ 65 years of age (991–943 prescriptions/1000 individuals, p = 0.018). With change-point regression, we identified a number of rapid shifts in the use of specific antibiotic classes, such as downward shifts in the use of quinolones and macrolides during the second quarter of 2008 and 2013, respectively. Over the period 2006–2015 outpatient use of antibiotics decreased substantially among adults. Rapid shifts in use occurring in 2008 and 2013 may reflect the presence of key drivers of change, such as abrupt changes in access to care or perceived antibiotic safety.

Efforts at predicting long-term adherence to medications have been focused on patients filling typical month-long supplies of medication. However, prediction remains difficult for patients filling longer initial supplies, a practice that is becoming increasingly common as a method to enhance medication adherence.To (a) extend methods involving short-term filling behaviors and (b) develop novel variables to predict adherence in a cohort of patients receiving longer initial prescriptions.In this retrospective cohort study, we used claims from a large national insurer to identify patients initiating a 90-day supply of oral medications for diabetes, hypertension, and hyperlipidemia (i.e., statins). Patients were included in the cohort if they had continuous database enrollment in the 180 days before and 365 days after medication initiation. Adherence was measured in the subsequent 12 months using the proportion of days covered metric. In total, 125 demographic, clinical, and medication characteristics at baseline and in the first 30-120 days after initiation were used to predict adherence using logistic regression models. We used 10-fold cross-validation to assess predictive accuracy by discrimination (c-statistic) measures.In total, 32,249 patients met the inclusion criteria, including 14,930 patients initiating statins, 12,887 patients initiating antihypertensives, and 4,432 patients initiating oral hypoglycemics. Prediction using only baseline variables was relatively poor (cross-validated c-statistic = 0.644). Including indicators of acute clinical conditions, health resource utilization, and short-term medication filling in the first 120 days greatly improved predictive ability (0.823). A model that incorporated all baseline characteristics and predictors within the first 120 days after medication initiation more accurately predicted future adherence (0.832). The best performing model that included all 125 baseline and postbaseline characteristics had strong predictive ability (0.837), suggesting the utility of measuring these novel postbaseline variables in this population.We demonstrate that long-term, 12-month adherence in patients filling longer supplies of medication can be strongly predicted using a combination of clinical, health resource utilization, and medication filling characteristics before and after treatment initiation.This work was supported by an unrestricted grant from CVS Health to Brigham and Women's Hospital. Shrank and Matlin were employees and shareholders at CVS Health at the time of this study; they report no financial interests in products or services that are related to this subject. Spettell is an employee of, and shareholder in, Aetna. This research was previously presented at the 2016 Annual Conference of the International Society for Pharmacoepidemiology; August 25-28, 2016; Dublin, Ireland.

BACKGROUND: A run-in phase is often employed prior to randomization in a clinical trial to exclude non-adherent patients, placebo responders, active drug non-responders, or patients who do not tolerate the active drug. This may impact the generalizability of trial results. OBJECTIVE: To determine if clinical outcomes differed between randomized controlled trials with run-in phases compared with randomized controlled trials of the same medication without run-in phases. DESIGN, PARTICIPANTS: From 2006 to 2014, the Food and Drug Administration approved 258 new medications. Sitaglitpin, saxagliptin, linagliptin, and alogliptin were among the only drugs with a common mechanism of action that each had multiple clinical trials, some of which had run-in phases and some of which did not. We identified all published randomized controlled trials for these four medications from MEDLINE and EMBASE as well as prior systematic reviews. MAIN MEASURES: We extracted key measures of medication efficacy (reduction in hemoglobin A1C) and safety (serious adverse events) from qualifying trials. Study results were pooled for each medication using random effects meta-analysis. KEY RESULTS: We identified 106 qualifying trials for DPP4 inhibitors, of which 88 had run-in phases and 18 did not. The average run-in phase duration was 4.0 weeks (range 1–21), and 73% of run-in phases administered placebo rather than active drug. The reduction in hemoglobin A1C compared to baseline was similar for trials with and without run-in phases (0.70%, 95% confidence interval [CI] 0.65–0.75 vs 0.76%, 95% CI 0.69–0.84, p = 0.27). The proportion of patients with serious adverse events was also similar for trials with and without run-in phases (4%, 95% CI: 3–5% vs 3%, 95% CI: 1–4%, p = 0.35). CONCLUSION: Trials with run-in phases provided similar estimates for medication efficacy and safety compared to trials without run-in phases. Because run-in phases are costly and time-consuming, these results call their utility into question for clinical trials of short duration.

Purpose To explore generalized boosted modeling (GBM) as a method for identifying subgroups with greater benefit or harm with dabigatran versus warfarin for treatment of atrial fibrillation. Methods We identified new initiators of warfarin or dabigatran with nonvalvular atrial fibrillation in 2 healthcare claims databases (2009-2013) and used GBM within 1 data source (development cohort) to explore subgroups where their effect on thromboembolism and major bleeding may differ. Identified subgroups were evaluated in the second data source (validation cohort) with stabilized-inverse-probability-of-treatment weights to adjust for confounding. Results Development and validation cohorts included 13 624 (28% dabigatran) and 62 596 (29% dabigatran) initiators, respectively. In development data, the strongest exposure interactions were prior thromboembolism and renal disease. In validation data, reduction in thromboembolism with dabigatran was greater for patients with versus without a history of thromboembolism by 2.8 (95% CI, -0.5 to 5.4) events per 100 patient-years. Major bleeding was reduced by 1.6/100 patient-years for dabigatran compared to warfarin initiators, without evidence of variation by renal disease. Conclusions We explored use of GBM to identify potential subgroups with different treatment effect. Dabigatran's superiority to warfarin at prevention of thromboembolism may be greater in secondary than primary prevention. In practice, secondary prevention patients are more often treated with warfarin.

AIM To evaluate the extent to which balance in unmeasured characteristics of patients with type 2 diabetes (T2DM) was achieved in claims data, by comparing against more detailed information from linked electronic health records (EHR) data. METHODS Within a large US commercial insurance database and using a cohort design, we identified patients with T2DM initiating linagliptin or a comparator agent within class (ie, another dipeptidyl peptidase-4 inhibitor) or outside class (ie, pioglitazone or a sulphonylurea) between May 2011 and December 2012. We focused on comparators used at a similar stage of diabetes to linagliptin. For each comparison, 1:1 propensity score (PS) matching was used to balance >100 baseline claims-based characteristics, including proxies of diabetes severity and duration. Additional clinical data from EHR were available for a subset of patients. We assessed representativeness of the claims-EHR-linked subset, evaluated the balance of claims- and EHR-based covariates before and after PS-matching via standardized differences (SDs), and quantified the potential bias associated with observed imbalances. RESULTS From a claims-based study population of 166 613 patients with T2DM, 7219 (4.3%) patients were linked to their EHR data. Claims-based characteristics in the EHR-linked and EHR-unlinked patients were similar (SD

Aim: To assess heterogeneity in adherence to medications in two example comparative effectiveness research studies. Patients & methods: We analyzed data from commercially insured patients initiating a statin or anticoagulant during 2005–2012. We calculated the cross-validated R 2 from a series of hierarchical linear models to assess variation in 1-year adherence. Results: There was less heterogeneity in adherence in the statin cohort compared with the anticoagulant cohort, where patient characteristics explained 7.2% of variation in adherence, and adding therapy and provider characteristics increased the proportion of variation explained to 8.0 and 8.5%, cumulatively. Random effects provided essentially no explanatory power, even in the statin cohort with large numbers of patients clustered within each pharmacy, prescriber and provider. Conclusion: The dependence of adherence on the healthcare system was stronger when the healthcare system influenced treatment choice and patient access to medication and when indications for treatment were strong.

Regulators consider randomized controlled trials (RCTs) as the gold standard for evaluating the safety and effectiveness of medications, but their costs, duration, and limited generalizability have caused some to look for alternatives. Real world evidence based on data collected outside of RCTs, such as registries and longitudinal healthcare databases, can sometimes substitute for RCTs, but concerns about validity have limited their impact. Greater reliance on such real world data (RWD) in regulatory decision making requires understanding why some studies fail while others succeed in producing results similar to RCTs. Key questions when considering whether RWD analyses can substitute for RCTs for regulatory decision making are WHEN one can study drug effects without randomization and HOW to implement a valid RWD analysis if one has decided to pursue that option. The WHEN is primarily driven by externalities not controlled by investigators, whereas the HOW is focused on avoiding known mistakes in RWD analyses.