Dataset Information

Generating real-world evidence from unstructured clinical notes to examine clinical utility of genetic tests: use case in BRCAness.

ABSTRACT:

Background

Next-generation sequencing provides comprehensive information about individuals' genetic makeup and is commonplace in oncology clinical practice. However, the utility of genetic information in the clinical decision-making process has not been examined extensively from a real-world, data-driven perspective. Through mining real-world data (RWD) from clinical notes, we could extract patients' genetic information and further associate treatment decisions with genetic information.

Methods

We proposed a real-world evidence (RWE) study framework that incorporates context-based natural language processing (NLP) methods and data quality examination before final association analysis. The framework was demonstrated in a Foundation-tested women cancer cohort (N?=?196). Upon retrieval of patients' genetic information using NLP system, we assessed the completeness of genetic data captured in unstructured clinical notes according to a genetic data-model. We examined the distribution of different topics regarding BRCA1/2 throughout patients' treatment process, and then analyzed the association between BRCA1/2 mutation status and the discussion/prescription of targeted therapy.

Results

We identified seven topics in the clinical context of genetic mentions including: Information, Evaluation, Insurance, Order, Negative, Positive, and Variants of unknown significance. Our rule-based system achieved a precision of 0.87, recall of 0.93 and F-measure of 0.91. Our machine learning system achieved a precision of 0.901, recall of 0.899 and F-measure of 0.9 for four-topic classification and a precision of 0.833, recall of 0.823 and F-measure of 0.82 for seven-topic classification. We found in result-containing sentences, the capture of BRCA1/2 mutation information was 75%, but detailed variant information (e.g. variant types) is largely missing. Using cleaned RWD, significant associations were found between BRCA1/2 positive mutation and targeted therapies.

Conclusions

In conclusion, we demonstrated a framework to generate RWE using RWD from different clinical sources. Rule-based NLP system achieved the best performance for resolving contextual variability when extracting RWD from unstructured clinical notes. Data quality issues such as incompleteness and discrepancies exist thus manual data cleaning is needed before further analysis can be performed. Finally, we were able to use cleaned RWD to evaluate the real-world utility of genetic information to initiate a prescription of targeted therapy.

SUBMITTER: Zhao Y

PROVIDER: S-EPMC7789545 | biostudies-literature | 2021 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Generating real-world evidence from unstructured clinical notes to examine clinical utility of genetic tests: use case in BRCAness.

Zhao Yiqing Y Weroha Saravut J SJ Goode Ellen L EL Liu Hongfang H Wang Chen C

BMC medical informatics and decision making 20210106 1

<h4>Background</h4>Next-generation sequencing provides comprehensive information about individuals' genetic makeup and is commonplace in oncology clinical practice. However, the utility of genetic information in the clinical decision-making process has not been examined extensively from a real-world, data-driven perspective. Through mining real-world data (RWD) from clinical notes, we could extract patients' genetic information and further associate treatment decisions with genetic information.< ...[more]

PMID: 33407429

Similar Datasets

Project description:IntroductionInternational guidelines advocate regular surveillance of patients following urothelial carcinoma (UC). A validated molecular diagnostic non-invasive urine test, Cxbladder Monitor, correctly identifies patients with a UC history who have low-probability of recurrence. The present study assesses the clinical utility of Cxbladder Monitor in reducing the number and frequency of urologic procedures ordered without missing detection of recurrent UC.MethodsData from 828 physician-patient assessments were generated from 18 participant physicians who each evaluated the same real-world clinical case data for 30 patients undergoing surveillance for recurrent UC. Each physician ordered tests and procedures and their timing, following review of the patient's demographic data, pre-existing conditions, risk factors and clinical history before and after disclosure of Cxbladder Monitor results. Changes in the number, type and timing of procedures ordered were assessed.ResultsThe addition of Cxbladder Monitor significantly reduced the overall number of tests ordered by 38.7%, including flexible cystoscopy by 43%, for patients whose Cxbladder Monitor result was low-probability. When the result was elevated-probability, the number of procedures ordered, including cystoscopy, was increased consistent with the increased risk of recurrent UC. Importantly, based on the tests ordered by each physician for each of the patients, all cases of recurrent UC would have been detected.ConclusionThe increase in clinical utility of Cxbladder Monitor for the management of patients undergoing surveillance for recurrent UC was shown to be driven by the reduction in procedures ordered for low-probability patients and for the more invasive procedures ordered for elevated-probability patients. In this study, the total number of procedures ordered, including the number of cystoscopies, was reduced especially in patients with low-probability of UC. The invasive procedures were ordered in a more targeted fashion for elevated-probability patients, without compromising the detection of recurrent UC. CLINICALTRIALS.Gov identifierNCT02700659.FundingPacific Edge Limited.

Project description:Background & aimsNon-invasive tests (NITs) offer a practical solution for advanced fibrosis identification in non-alcoholic fatty liver disease (NAFLD). Despite increasing implementation, their use is not standardised, which can lead to inconsistent interpretation and risk stratification. We aimed to assess the types of NITs and the corresponding cut-offs used in a range of healthcare settings.MethodsA survey was distributed to a convenience sample of liver health experts who participated in a global NAFLD consensus statement. Respondents provided information on the NITs used in their clinic with the corresponding cut-offs and those used in established care pathways in their areas.ResultsThere were 35 respondents from 24 countries, 89% of whom practised in tertiary level settings. A total of 14 different NITs were used, and each respondent reported using at least one (median = 3). Of the respondents, 80% reported using FIB-4 and liver stiffness by vibration-controlled transient elastography (Fibroscan®), followed by the NAFLD fibrosis score (49%). For FIB-4, 71% of respondents used a low cut-off of <1.3 (range <1.0 to <1.45) and 21% reported using age-specific cut-offs. For Fibroscan®, 21% of respondents used a single liver stiffness cut-off: 8 kPa in 50%, while the rest used 7.2 kPa, 7.8 kPa and 8.7 kPa. Among the 63% of respondents who used lower and upper liver stiffness cut-offs, there were variations in both values (<5 to <10 kPa and >7.5 to >20 kPa, respectively).ConclusionsThe cut-offs used for the same NITs for NAFLD risk stratification vary between clinicians. As cut-offs impact test performance, these findings underscore the heterogeneity in risk-assessment and support the importance of establishing consistent guidelines on the standardised use of NITs in NAFLD management.Lay summaryOwing to the high prevalence of non-alcoholic fatty liver disease (NAFLD) in the general population it is important to identify those who have more advanced stages of liver fibrosis, so that they can be properly treated. Non-invasive tests (NITs) provide a practical way to assess fibrosis risk in patients. However, we found that the cut-offs used for the same NITs vary between clinicians. As cut-offs impact test performance, these findings highlight the importance of establishing consistent guidelines on the standardised use of NITs to optimise clinical management of NAFLD.

Project description:Randomized clinical trials (RCTs) are essential to guide medical practice; however, their generalizability to a given population is often uncertain. We developed a statistically informed Generative Adversarial Network (GAN) model, RCT-Twin-GAN, that leverages relationships between covariates and outcomes and generates a digital twin of an RCT (RCT-Twin) conditioned on covariate distributions from a second patient population. We used RCT-Twin-GAN to reproduce treatment effect outcomes of the Systolic Blood Pressure Intervention Trial (SPRINT) and the Action to Control Cardiovascular Risk in Diabetes (ACCORD) Blood Pressure Trial, which tested the same intervention but had different treatment effect results. To demonstrate treatment effect estimates of each RCT conditioned on the other RCT patient population, we evaluated the cardiovascular event-free survival of SPRINT digital twins conditioned on the ACCORD cohort and vice versa (SPRINT-conditioned ACCORD twins). The conditioned digital twins were balanced by the intervention arm (mean absolute standardized mean difference (MASMD) of covariates between treatment arms 0.019 (SD 0.018), and the conditioned covariates of the SPRINT-Twin on ACCORD were more similar to ACCORD than a sprint (MASMD 0.0082 SD 0.016 vs. 0.46 SD 0.20). Most importantly, across iterations, SPRINT conditioned ACCORD-Twin datasets reproduced the overall non-significant effect size seen in ACCORD (5-year cardiovascular outcome hazard ratio (95% confidence interval) of 0.88 (0.73-1.06) in ACCORD vs median 0.87 (0.68-1.13) in the SPRINT conditioned ACCORD-Twin), while the ACCORD conditioned SPRINT-Twins reproduced the significant effect size seen in SPRINT (0.75 (0.64-0.89) vs median 0.79 (0.72-0.86)) in ACCORD conditioned SPRINT-Twin). Finally, we describe the translation of this approach to real-world populations by conditioning the trials on an electronic health record population. Therefore, RCT-Twin-GAN simulates the direct translation of RCT-derived treatment effects across various patient populations with varying covariate distributions.

Project description:Melanoma currently lacks a reliable blood-based biomarker of disease activity, although circulating tumor DNA (ctDNA) may fill this role. We investigated the clinical utility (i.e., impact on clinical outcomes and interpretation of radiographic data) of measuring ctDNA in patients with metastatic or high-risk resected melanoma. Patients were prospectively accrued into ≥ 1 of three cohorts, as follows. Cohort A: patients with radiographically measurable metastatic melanoma who underwent comparison of ctDNA measured by a BEAMing digital PCR assay to tissue mutational status and total tumor burden; when appropriate, determinations about initiation of targeted therapy were based on ctDNA data. Cohorts B and C: patients with BRAF- or NRAS-mutant melanoma who had either undergone surgical resection of high-risk disease (cohort B) or were receiving or had received medical therapy for advanced disease (cohort C). Patients were followed longitudinally with serial ctDNA measurements with contemporaneous radiographic imaging to ascertain times to detection of disease activity and progressive disease, respectively. The sensitivity and specificity of the ctDNA assay were 86.8% and 100%, respectively. Higher tumor burden and visceral metastases were found to be associated with detectable ctDNA. In two patients in cohort A, ctDNA test results revealed a targetable mutation where tumor testing had not; both patients experienced a partial response to targeted therapy. In four of 30 patients with advanced melanoma, ctDNA assessments indicated evidence of melanoma activity that predicted radiographic evidence of disease progression by 8, 14, 25, and 38 weeks, respectively. CtDNA was detectable in three of these four patients coincident with radiographic evaluations that alone were interpreted as showing no evidence of neoplastic disease. Our findings provide evidence for the clinical utility of integrating ctDNA data in managing patients with melanoma in a real-world setting.

Dataset Information

Generating real-world evidence from unstructured clinical notes to examine clinical utility of genetic tests: use case in BRCAness.

Background

Methods

Results

Conclusions

Publications

Generating real-world evidence from unstructured clinical notes to examine clinical utility of genetic tests: use case in BRCAness.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets