Dataset Information

Evaluating predictive modeling algorithms to assess patient eligibility for clinical trials from routine data.

ABSTRACT:

Background

The necessity to translate eligibility criteria from free text into decision rules that are compatible with data from the electronic health record (EHR) constitutes the main challenge when developing and deploying clinical trial recruitment support systems. Recruitment decisions based on case-based reasoning, i.e. using past cases rather than explicit rules, could dispense with the need for translating eligibility criteria and could also be implemented largely independently from the terminology of the EHR's database. We evaluated the feasibility of predictive modeling to assess the eligibility of patients for clinical trials and report on a prototype's performance for different system configurations.

Methods

The prototype worked by using existing basic patient data of manually assessed eligible and ineligible patients to induce prediction models. Performance was measured retrospectively for three clinical trials by plotting receiver operating characteristic curves and comparing the area under the curve (ROC-AUC) for different prediction algorithms, different sizes of the learning set and different numbers and aggregation levels of the patient attributes.

Results

Random forests were generally among the best performing models with a maximum ROC-AUC of 0.81 (CI: 0.72-0.88) for trial A, 0.96 (CI: 0.95-0.97) for trial B and 0.99 (CI: 0.98-0.99) for trial C. The full potential of this algorithm was reached after learning from approximately 200 manually screened patients (eligible and ineligible). Neither block- nor category-level aggregation of diagnosis and procedure codes influenced the algorithms' performance substantially.

Conclusions

Our results indicate that predictive modeling is a feasible approach to support patient recruitment into clinical trials. Its major advantages over the commonly applied rule-based systems are its independency from the concrete representation of eligibility criteria and EHR data and its potential for automation.

SUBMITTER: Kopcke F

PROVIDER: S-EPMC4029400 | biostudies-literature | 2013 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Evaluating predictive modeling algorithms to assess patient eligibility for clinical trials from routine data.

Köpcke Felix F Lubgan Dorota D Fietkau Rainer R Scholler Axel A Nau Carla C Stürzl Michael M Croner Roland R Prokosch Hans-Ulrich HU Toddenroth Dennis D

BMC medical informatics and decision making 20131209

<h4>Background</h4>The necessity to translate eligibility criteria from free text into decision rules that are compatible with data from the electronic health record (EHR) constitutes the main challenge when developing and deploying clinical trial recruitment support systems. Recruitment decisions based on case-based reasoning, i.e. using past cases rather than explicit rules, could dispense with the need for translating eligibility criteria and could also be implemented largely independently fr ...[more]

PMID: 24321610

Similar Datasets

Project description:Benchmarking surveillance systems requires realistic simulations of disease outbreaks. However, obtaining these data in sufficient quantity, with a realistic shape and covering a sufficient range of agents, size and duration, is known to be very difficult. The dataset of outbreak signals generated should reflect the likely distribution of authentic situations faced by the surveillance system, including very unlikely outbreak signals. We propose and evaluate a new approach based on the use of historical outbreak data to simulate tailored outbreak signals. The method relies on a homothetic transformation of the historical distribution followed by resampling processes (Binomial, Inverse Transform Sampling Method-ITSM, Metropolis-Hasting Random Walk, Metropolis-Hasting Independent, Gibbs Sampler, Hybrid Gibbs Sampler). We carried out an analysis to identify the most important input parameters for simulation quality and to evaluate performance for each of the resampling algorithms. Our analysis confirms the influence of the type of algorithm used and simulation parameters (i.e. days, number of cases, outbreak shape, overall scale factor) on the results. We show that, regardless of the outbreaks, algorithms and metrics chosen for the evaluation, simulation quality decreased with the increase in the number of days simulated and increased with the number of cases simulated. Simulating outbreaks with fewer cases than days of duration (i.e. overall scale factor less than 1) resulted in an important loss of information during the simulation. We found that Gibbs sampling with a shrinkage procedure provides a good balance between accuracy and data dependency. If dependency is of little importance, binomial and ITSM methods are accurate. Given the constraint of keeping the simulation within a range of plausible epidemiological curves faced by the surveillance system, our study confirms that our approach can be used to generate a large spectrum of outbreak signals.

Project description:More than half of males in China are current smokers and evidence from western countries tells us that an unprecedented number of smoking-attributable deaths will occur as the Chinese population ages. We used the China Lung Cancer Policy Model (LCPM) to simulate effects of computed tomography (CT)-based lung cancer screening in China, comparing the impact of a screening guideline published in 2015 by a Chinese expert group to a version developed for the United States by the U.S. Centers for Medicare & Medicaid Services (CMS). The China LCPM, built using an existing lung cancer microsimulation model, can project population outcomes associated with interventions for smoking-related diseases. After calibrating the model to published Chinese smoking prevalence and lung cancer mortality rates, we simulated screening from 2016 to 2050 based on eligibility criteria from the CMS and Chinese guidelines, which differ by age to begin and end screening, pack-years smoked, and years since quitting. Outcomes included number of screens, mortality reduction, and life-years saved for each strategy. We projected that in the absence of screening, 14.98 million lung cancer deaths would occur between 2016 and 2050. Screening with the CMS guideline would prevent 0.72 million deaths and 5.8 million life-years lost, resulting in 6.58% and 1.97% mortality reduction in males and females, respectively. Screening with the Chinese guideline would prevent 0.74 million deaths and 6.6 million life-years lost, resulting in 6.30% and 2.79% mortality reduction in males and females, respectively. Through 2050, 1.43 billion screens would be required using the Chinese screening strategy, compared to 988 million screens using the CMS guideline. In conclusion, CT-based lung cancer screening implemented in 2016 and based on the Chinese screening guideline would prevent about 20,000 (2.9%) more lung cancer deaths through 2050, but would require about 445 million (44.7%) more screens than the CMS guideline.

Project description:Well-powered genomic studies require genome-wide marker coverage across many individuals. For non-model species with few genomic resources, high-throughput sequencing (HTS) methods, such as Genotyping-By-Sequencing (GBS), offer an inexpensive alternative to array-based genotyping. Although affordable, datasets derived from HTS methods suffer from sequencing error, alignment errors, and missing data, all of which introduce noise and uncertainty to variant discovery and genotype calling. Under such circumstances, meaningful analysis of the data is difficult. Our primary interest lies in the issue of how one can accurately infer or impute missing genotypes in HTS-derived datasets. Many of the existing genotype imputation algorithms and software packages were primarily developed by and optimized for the human genetics community, a field where a complete and accurate reference genome has been constructed and SNP arrays have, in large part, been the common genotyping platform. We set out to answer two questions: 1) can we use existing imputation methods developed by the human genetics community to impute missing genotypes in datasets derived from non-human species and 2) are these methods, which were developed and optimized to impute ascertained variants, amenable for imputation of missing genotypes at HTS-derived variants? We selected Beagle v.4, a widely used algorithm within the human genetics community with reportedly high accuracy, to serve as our imputation contender. We performed a series of cross-validation experiments, using GBS data collected from the species Manihot esculenta by the Next Generation (NEXTGEN) Cassava Breeding Project. NEXTGEN currently imputes missing genotypes in their datasets using a LASSO-penalized, linear regression method (denoted 'glmnet'). We selected glmnet to serve as a benchmark imputation method for this reason. We obtained estimates of imputation accuracy by masking a subset of observed genotypes, imputing, and calculating the sample Pearson correlation between observed and imputed genotype dosages at the site and individual level; computation time served as a second metric for comparison. We then set out to examine factors affecting imputation accuracy, such as levels of missing data, read depth, minor allele frequency (MAF), and reference panel composition.

Dataset Information

Evaluating predictive modeling algorithms to assess patient eligibility for clinical trials from routine data.

Background

Methods

Results

Conclusions

Publications

Evaluating predictive modeling algorithms to assess patient eligibility for clinical trials from routine data.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets