Automatic Selection of Clinical Trials Based on A Semantic Web Approach.

Recruitment of patients in clinical trials is nowadays preoccupying, as the inclusion rate is particularly low. The main identified factors are the multiplicity of open clinical trials, the high number and complexity of eligibility criteria, and the additional workload that a systematic search of the clinical trials a patient could be enrolled in for a physician. The principal objective of the ASTEC project is to automate the prescreening phase during multidisciplinary meetings (MDM). This paper presents the evaluation of a computerized recruitment support systems (CRSS) based on semantic web approach. The evaluation of the system was based on data collected retrospectively from a 6 month period of MDM in Urology and on 4 clinical trials of prostate cancer. The classification performance of the ASTEC system had a precision of 21%, recall of 93%, and an error rate equal to 37%. Missing data was the main issue encountered. The system was designed to be both scalable to other clinical domains and usable during MDM process.

Improving the pre-screening of eligible patients in order to increase enrollment in cancer clinical trials.

BACKGROUND: The performance of randomized controlled trials (RCTs) is often hindered by recruitment difficulties. This study aims to explore the pre-screening phase of four prostate cancer RCTs to identify the impact of a systematic pre-selection of eligible patients for RCT recruitment. METHODS: The pre-screening of four RCTs opened at the Comprehensive Cancer Center in Rennes was analyzed retrospectively (French Genitourinary Tumor Group (GETUG) 14, 15, 16, and 17). Data were extracted from electronic multidisciplinary cancer (MDC) reports and manually completed by physicians and medical secretaries. These data were the main source of information for clinicians to discuss treatment alternatives during MDC sessions. The pre-screening decisions made by the clinicians during these MDC meetings were compared with those made after a systematic review of the MDC reports by a clinical research assistant (CRA). Any inconsistencies in decisions between the CRA and the MDC physicians were corrected by the principal investigator (PI). RESULTS: The pre-screening rate was 9.1% during the MDC meetings, while it was estimated to be 12.9% after the final review by the PI, and 29% after the systematic review by the CRA. The study showed that 77% and 67% of the MDC reports did not mention clinical and pathological Tumor, lymph node and metastasis classification of malignant tumors (TNM) staging, respectively, and that 35 of the CRA's 47 proposals rejected by the PI concerned implicit information (not specified in the MDC reports). Only one patient was proposed by the PI, and none by the CRA. CONCLUSIONS: These results confirm that pre-screening could be improved by a systematic review of the medical reports. They also highlight the fact that missing data in electronic MDC reports leads to over-enrollment of non-eligible patients, but not to over-exclusion of eligible patients. Thus, our study confirms the potential gain in using semi-automated pre-selection of MDC reports, in order to avoid missing out on patients eligible for RCTs. TRIAL REGISTRATION: The trials evaluated in this study were previously registered with clinicaltrials.gov (registration number: NCT00104741 on 3 March 2005; NCT00104715 on 3 March 2005; NCT00423475 on 16 January 2007; and NCT00667069 on 24 April 2008).

OWL model of clinical trial eligibility criteria compatible with partially-known information.

BACKGROUND: Clinical trials are important for patients, for researchers and for companies. One of the major bottlenecks is patient recruitment. This task requires the matching of a large volume of information about the patient with numerous eligibility criteria, in a logically-complex combination. Moreover, some of the patient's information necessary to determine the status of the eligibility criteria may not be available at the time of pre-screening. RESULTS: We showed that the classic approach based on negation as failure over-estimates rejection when confronted with partially-known information about the eligibility criteria because it ignores the distinction between a trial for which patient eligibility should be rejected and trials for which patient eligibility cannot be asserted. We have also shown that 58.64% of the values were unknown in the 286 prostate cancer cases examined during the weekly urology multidisciplinary meetings at Rennes' university hospital between October 2008 and March 2009.We propose an OWL design pattern for modeling eligibility criteria based on the open world assumption to address the missing information problem. We validate our model on a fictitious clinical trial and evaluate it on two real clinical trials. Our approach successfully distinguished clinical trials for which the patient is eligible, clinical trials for which we know that the patient is not eligible and clinical trials for which the patient may be eligible provided that further pieces of information (which we can identify) can be obtained. CONCLUSIONS: OWL-based reasoning based on the open world assumption provides an adequate framework for distinguishing those patients who can confidently be rejected from those whose status cannot be determined. The expected benefits are a reduction of the workload of the physicians and a higher efficiency by allowing them to focus on the patients whose eligibility actually require expertise.