Combining individual patient data from randomized and non-randomized studies to predict real-world effectiveness of interventions.
Ontology highlight
ABSTRACT:
SUBMITTER: Seo M
PROVIDER: S-EPMC9251754 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:ObjectiveWe aimed to investigate bias in applying machine learning to predict real-world individual treatment effects.Materials and methodsUsing a virtual patient cohort, we simulated real-world healthcare data and applied random forest and gradient boosting classifiers to develop prediction models. Treatment effect was estimated as the difference between the predicted outcomes of a treatment and a control. We evaluated the impact of predictors (ie, treatment predictors [X1], confounders [X2], treatment effects modifiers [X3], and other outcome risk factors [X4]) with known effects on treatment and outcome using real-world data, and outcome imbalance on predicting individual outcome. Using counterfactuals, we evaluated percentage of patients with biased predicted individual treatment effects.ResultsThe X4 had relatively more impact on model performance than X2 and X3 did. No effects were observed from X1. Moderate-to-severe outcome imbalance had a significantly negative impact on model performance, particularly among subgroups in which an outcome occurred. Bias in predicting individual treatment effects was significant and persisted even when the models had a 100% accuracy in predicting health outcome.DiscussionInadequate inclusion of the X2, X3, and X4 and moderate-to-severe outcome imbalance may affect model performance in predicting individual outcome and subsequently bias in predicting individual treatment effects. Machine learning models with all features and high performance for predicting individual outcome still yielded biased individual treatment effects.ConclusionsDirect application of machine learning might not adequately address bias in predicting individual treatment effects. Further method development is needed to advance machine learning to support individualized treatment selection.
| S-EPMC7647181 | biostudies-literature
Project description:Using patient derived lung cancer organoids to predict clinical response and find theraputic mechanisms
2022-09-26 | PXD037076 |
Project description:Background: Idarucizumab has been included in guidelines for the management of bleeding or surgical procedure in dabigatran-treated patients without need for biological monitoring. The aim of the study was to assess the prognostic value of dabigatran plasma level before reversal to test the hemostatic efficacy of idarucizumab. The secondary objectives were (i) to analyze plasma dabigatran level according to the risk of rebound and (ii) to evaluate the incidence of post-reversal non-favorable clinical outcomes (including thromboembolism, bleeding, antithrombotic, and death) and antithrombotic resumption. Methods and Results: This was an observational multicentric cohort study, which included all French patients who required idarucizumab for dabigatran reversal. Between May 2016 and April 2019, 87 patients from 21 French centers were enrolled. Patients received idarucizumab for overt bleeding (n = 61), urgent procedures (n = 24), or overdose without bleeding (n = 2). Among patients with major bleeding (n = 57), treatment with idarucizumab was considered effective in 44 (77.2%) of them. Patients who did not achieve effective hemostasis after reversal had a significantly higher mean level of plasma dabigatran at baseline (524.5 ± 386 vs. 252.8 ng/mL ± 235, p = 0.033). Furthermore, patients who did not achieve effective hemostasis after reversal had less favorable outcomes during follow-up (46.2 vs. 81.8%, p = 0.027). ROC curve identified a cutoff of 264 ng/mL for dabigatran level at admission to be predictive of ineffective hemostasis. No plasma dabigatran rebound was observed after reversal in patients with dabigatran plasma level < 264 ng/mL at baseline. Conclusion: This retrospective study shows that dabigatran level before reversal could predict hemostatic effectiveness and dabigatran plasma rebound after idarucizumab injection.
| S-EPMC7772865 | biostudies-literature
Project description:Rapid and robust strategies to evaluate the efficacy and effectiveness of novel and existing pharmacotherapeutic interventions (repurposed treatments) in future pandemics are required. Observational "real-world studies" (RWS) can report more quickly than randomized controlled trials (RCTs) and would have value were they to yield reliable results. Both RCTs and RWS were deployed during the coronavirus disease 2019 (COVID-19) pandemic. Comparing results between them offers a unique opportunity to determine the potential value and contribution of each. A learning review of these parallel evidence channels in COVID-19, based on quantitative modeling, can help improve speed and reliability in the evaluation of repurposed therapeutics in a future pandemic. Analysis of all-cause mortality data from 249 observational RWS and RCTs across eight treatment regimens for COVID-19 showed that RWS yield more heterogeneous results, and generally overestimate the effect size subsequently seen in RCTs. This is explained in part by a few study factors: the presence of RWS that are imbalanced for age, gender, and disease severity, and those reporting mortality at 2 weeks or less. Smaller studies of either type contributed negligibly. Analysis of evidence generated sequentially during the pandemic indicated that larger RCTs drive our ability to make conclusive decisions regarding clinical benefit of each treatment, with limited inference drawn from RWS. These results suggest that when evaluating therapies in future pandemics, (1) large RCTs, especially platform studies, be deployed early; (2) any RWS should be large and should have adequate matching of known confounders and long follow-up; (3) reporting standards and data standards for primary endpoints, explanatory factors, and key subgroups should be improved; in addition, (4) appropriate incentives should be in place to enable access to patient-level data; and (5) an overall aggregate view of all available results should be available at any given time.
| S-EPMC10582658 | biostudies-literature
Project description:The performance of a drug in a clinical trial setting often does not reflect its effect in daily clinical practice. In this third of three reviews, we examine the applications that have been used in the literature to predict real-world effectiveness from randomized controlled trial efficacy data. We searched MEDLINE, EMBASE from inception to March 2014, the Cochrane Methodology Register, and websites of key journals and organisations and reference lists. We extracted data on the type of model and predictions, data sources, validation and sensitivity analyses, disease area and software. We identified 12 articles in which four approaches were used: multi-state models, discrete event simulation models, physiology-based models and survival and generalized linear models. Studies predicted outcomes over longer time periods in different patient populations, including patients with lower levels of adherence or persistence to treatment or examined doses not tested in trials. Eight studies included individual patient data. Seven examined cardiovascular and metabolic diseases and three neurological conditions. Most studies included sensitivity analyses, but external validation was performed in only three studies. We conclude that mathematical modelling to predict real-world effectiveness of drug interventions is not widely used at present and not well validated. © 2016 The Authors Research Synthesis Methods Published by John Wiley & Sons Ltd.
| S-EPMC5129568 | biostudies-literature
Project description:BackgroundComparative-effectiveness studies using real-world data (RWD) can be susceptible to surveillance bias. In solid tumor oncology studies, analyses of endpoints such as progression-free survival (PFS) are based on progression events detected by imaging assessments. This study aimed to evaluate the potential bias introduced by differential imaging assessment frequency when using electronic health record (EHR)-derived data to investigate the comparative effectiveness of cancer therapies.MethodsUsing a nationwide de-identified EHR-derived database, we first analyzed imaging assessment frequency patterns in patients diagnosed with advanced non-small cell lung cancer (aNSCLC). We used those RWD inputs to develop a discrete event simulation model of two treatments where disease progression was the outcome and PFS was the endpoint. Using this model, we induced bias with differential imaging assessment timing and quantified its effect on observed versus true treatment effectiveness. We assessed percent bias in the estimated hazard ratio (HR).ResultsThe frequency of assessments differed by cancer treatment types. In simulated comparative-effectiveness studies, PFS HRs estimated using real-world imaging assessment frequencies differed from the true HR by less than 10% in all scenarios (range: 0.4% to -9.6%). The greatest risk of biased effect estimates was found comparing treatments with widely different imaging frequencies, most exaggerated in disease settings where time to progression is very short.ConclusionsThis study provided evidence that the frequency of imaging assessments to detect disease progression can differ by treatment type in real-world patients with cancer and may induce some bias in comparative-effectiveness studies in some situations.
| S-EPMC9290806 | biostudies-literature
Project description:Despite the availability of various osteoporosis treatments, adherence remains suboptimal. One contributing factor may be patient experience with therapy. This US, multicenter, combined retrospective chart review and patient questionnaire study included postmenopausal women at high risk for fracture and is the first study to describe real-world patient experience with abaloparatide (ABL) injection. Eight geographically diverse secondary care sites in the United States participated (n = 193). Mean ± SD age was 67.4 ±8.62 years. Most patients (86%) were satisfied with the ABL regimen, especially with ease of preparation (82%), ease of storage (87%), and storage convenience (89%), an attribute 83% of the patients thought was important. The majority of patients reported complete satisfaction with the ABL regimen allowing for their ability to conduct daily activities (85%) and convenience to fit into their daily schedule (84%). All reported taking ABL as directed, by injection in the lower abdomen, and 83% of patients reported medium or high adherence. Patients were satisfied with the needle size (76% completely satisfied), and 93% reported never deliberately missing a dose. Although injecting medication (18%) and higher out-of-pocket costs (17%) were deemed the most bothersome attributes, the majority (69%) noted their healthcare team understands how osteoporosis impacts their lives. In multivariable analyses, ease of preparation (OR = 2.62; 95% CI, 1.01-6.81; p = 0.048) and fracture history (OR = 1.72; 95% CI, 1.03-2.86; p = 0.037) were significantly associated with overall satisfaction. Ease of preparation was a predictor of higher satisfaction with treatment convenience (coefficient = 13.60; 95% CI, 8.08-19.12; p = 0.00). Remembering to take the medication was a significant predictor of self-reported adherence (OR = 16.66; 95% CI, 3.30-84.24; p = 0.001). In conclusion, the majority of patients were satisfied with ABL and found it convenient/easy to prepare and store. High self-reported adherence may be associated with positive patient experience including ease of use and adequate support from healthcare providers. © 2020 The Authors. JBMR Plus published by Wiley Periodicals LLC. on behalf of American Society for Bone and Mineral Research.
| S-EPMC7990148 | biostudies-literature
Project description:Despite the availability of a variety of insulins, rates of insulinisation and the acceptance of insulin therapy is suboptimal in real-world clinical settings. Patient and physician concerns with hypoglycaemia and weight gain are the two key issues that serve to impede appropriate insulinisation in patients with diabetes. Recently introduced second-generation basal insulin analogues [for e.g., insulin glargine 300 U/mL (Gla-300) and insulin degludec] are designed to have improved pharmacokinetic profiles with an intention to deliver steady insulin levels over a longer period. Several randomised controlled and real-world studies have proven the resultant advantages of second-generations insulin analogues in lowering intra-individual variability in plasma insulin levels, flexibility in dosing, a sustained glucose-lowering effect, and decreasing the risk of hypoglycaemia. Gla-300 is one of the newer second-generation basal insulin analogues to have been approved for both type 1 and 2 diabetes. In this article, we review the currently available clinical and real-world data of Gla-300.
| S-EPMC7156297 | biostudies-literature
Project description:AimsTo develop and pilot test a taxonomy that empirically estimates health intervention effectiveness from efficacy data.MethodsWe developed a taxonomy to score health interventions across 11 items on a scale from 0-100. The taxonomy was pilot-tested in efficacy and effectiveness diabetes prevention studies identified in two separate systematic reviews; here, the face validity, inter-rater reliability and factor structure of the taxonomy were established. Random effects meta-analyses were used to obtain weight loss and diabetes incidence pooled effects across studies. These effects and taxonomy scores were used to down calibrate efficacy estimates to effectiveness estimates as follows: Efficacy effect*[Efficacy score/highest possible score].ResultsWe scored 82 effectiveness lifestyle modification studies (mean score 49.2), 32 efficacy lifestyle modification studies (mean score 69.8) and 20 efficacy studies testing medications (mean score 77.4). The taxonomy had face validity and good inter-rater reliability (ICC = 0.9 [0.87, 0.93]). The between-groups down calibrated weight loss estimate was similar to that observed in the effectiveness meta-analysis (1.7 and 1.8 kg, respectively). The down calibrated diabetes relative risk reduction was also similar to that observed in the effectiveness meta-analysis (30.6% over 2.7 years and 29% over 2 years, respectively).ConclusionsThe taxonomy is a promising tool to estimate the real-world impact of health interventions.
| S-EPMC6883153 | biostudies-literature
Project description:Randomized controlled trials (RCTs) have been considered as gold standard for establishing the efficacy and safety of investigational new drugs; nonetheless, the generalizability of their findings has been questioned. To address this issue, an increasing number of naturalistic studies and real-world database analyses have been conducted. The question of how much information from these two approaches is congruent or discrepant with each other is of great importance for the clinical practice. To answer this question, we focused on data from the antipsychotic (AP) treatment of schizophrenia. Our aim was two-fold: to conduct a meta-analysis of real-world studies (RWS), and to compare the results of RWS meta-analysis with previously published meta-analyses of RCTs. The principal measure of effectiveness was all-cause treatment discontinuation for both RWS and RCTs (when not available, then drop out for RCTs). We included publications for 8 selected APs (oral formulations of amisulpride, aripiprazole, clozapine, haloperidol, olanzapine, quetiapine, risperidone, and long-acting injectable (LAI) risperidone). We identified 11 RWS and 7 RCT meta-analyses for inclusion. Our results indicated that the RWS yielded statistically conclusive and consistent findings across individual investigations. For the overwhelming majority of the comparisons where both RWS and RCT meta-analyses were available, there was good congruency between the RWS and the RCT results. Our results support that RCTs, despite their limitations, provide evidence which is generalizable to real-world settings. This is an important finding for both regulators and clinicians. RWS can provide guidance for situations where no evidence is available from double-blind clinical trials.
| S-EPMC8494924 | biostudies-literature