Digitization of Clinical Research

Soroka Medical Center, YonaLink Ltd.

Automatic retrieval in FHIR of information from the hospital’s database to a developer file who conducts clinical research in the hospital.

The data will be transferred to YonaLink and will be converted again to a standard used to submit to CDISC (FDA).

As part of the project a conversion of lab tests and medication in hospitalization is planned, while the POC is planned upon a research that is now taking place at Soroka.

The group’s study case

Clinical research, which are the basis of drug approval, technology or a new treatment, are taking place in medical centers, and are a source for advancing the clinical knowledge, HealthCare innovation and even for income of hospitals in Israel.

The world of data management in clinical research is stiff and unified and is under the strict demands of the FDA. The data that is gathered in clinical research is not big but very clean and of high quality. As for today, in all hospitals in Israel and in most of the hospitals around the world, this process is done manually, as the patient's clinical data is being entered by research coordinators. As the nature of a manual process, it is slow and inefficient, raises the clinical research costs and raises the probability of mistakes in data reporting.

YonaLink Ltd. has developed a FHIR based platform for managing clinical prospective research. The system automatically exports clinical research data from the electronic medical center systems to the clinical research pool without compromising the information security and the patients privacy. By doing so, the system eliminates the need for manual copying of data that is taking place today, and prevents mistakes in the data, increases its credibility, and allows significant savings in time and money (about ⅔ of data management expenses). In addition, the system enables online tracking after the research and notify of any deviations from the research protocol or of data deficits, in order to assure an optimal conduction of the trial process.

Soroka Medical Center is the only hospital in the Negev, as such it facilitates clinical data regarding all the population of the Negev. Moreover, since the hospital is associated with Clalit HMO, the database covers all Clalit insured patients in the Negev (approximately 70% of the Negev population) taken from the hospital and the community and therefore is even extensive.

As the first step in building Soroka Medical Center’s ability (and along the way, all HMO’s ability) to access the data for needs of clinical research, two worlds of content were chosen, for which FHIR compliance will be made- The laboratory test and Medications. Inside the hospital, a FHIR facade was set up, which allows reading this data using API query which is sent from the YonaLink system. The FHIR facade returns a FHIR response which contains the details for the blood test or medication requested, for the patient and date requested. The option of making the clinical research database management automatic or even semi-automatic, FHIR aligned, is expected to make the medical centers which will adapt this ability - much more attractive for pharma and drug companies to conduct their clinical research.

The project’s world of content

Structured data that is used for cynical research:

Blood tests data - at this point only Chemistry, Hematology and Endocrinology
Medications actually taken by the patient

Solution’s Architecture

Automatic retrieval in FHIR of information from the hospital’s database to a developer file who conducts clinical research in the hospital. The data will be transferred to YonaLink and will be converted again to a standard used to submit to CDISC (FDA).

First step - Automatic SQL procedures for retrieval of information from Soroka’s clinical database. At this point, different parameters regarding the research population, the course of the research and the output results can be defined.
Second step - Mapping Clalit’s internal codes, such as identifying lab tests (ind) for universal coding (LOINC).
Third step - Adding data to the right tables in cache of the facade by internal database population script of FHIR facade which will run periodically. The script will make the mapping needed between the data on the output file to the internal facade table architecture (cache). At the end of the research period, the data will be deleted from the facade.
Fourth step - Returning results in real time for the end user who raised a FHIR request. The input is a FHIR query (API) from the Yonalink server which is installed inside Soroka (due to information security). The output is Search FHIR Bundle which is suitable for the API query from Yonalink.

Profiles to be characterized in the project

Regarding blood tests - at this point only Chemistry, Hematology and Endocrinology
Medication Resource - Medications taken by the patient at ER / hospitalization only. We will focus on Medication Administration and Medication Statement.