PRA Prism | Prism eSource

Background

For decades, our industry has struggled with the inefficiencies that result from moving data from their source to a database that can support the needs of a clinical trial. Being able to access the eSource data directly has been a goal for sponsor companies, sites and service providers like PRA Health Sciences. In addition, this paradigm has the broad support of regulatory agencies and has resulted in multiple eSource guidances from the FDA and the EMA.

FDA Guidance for Industry: Electronic Source Data in Clinical Investigations
FDA Guidance for Industry: Use of Electronic Health Record Data in Clinical Investigations
EMA Reflection paper on expectations for electronic source data and data transcribed to electronic data collection tools in clinical trials

PRA Health Sciences believes that for this new modality to significantly impact research, it needs to be an industry-wide solution, not a one-off proprietary product. For this reason, the solution was built on the standards developed by a joint collaboration between IHE (Integrating the Healthcare Enterprise), the standards group for the healthcare industry, and CDISC, the standards group for the life sciences industry. These standards have been vetted by the FDA and EMA and accommodate both HIPAA and 21 CFR Part 11. Prism eSource was further enhanced to accommodate the Fast InterOperability Healthcare Resources (FHIR) standard. This HL-7 standard significantly expands the number of EHR systems that are now compatible with Prism eSource.

A decade in the making

With its development over 10+ years, Prism is the leader in EHR-based eSource. Every year, IHE sponsors a software validation conference called the Connectathon where companies across the healthcare enterprise come and validate their integration with other systems. The Prism eSource solution has been tested in every Connectathon since 2009 and has passed validation with the following EHR systems which represent > 90% of the hospital market:

Cerner	Allscripts
Greenway	Tiani Spirit (Cisco)
eMDs	GE Healthcare

Prism eSource Features

The IHE/CDISC standards provide the backbone for our EHR integration. It allows subjects to be enrolled directly from the EHR system and for data to be migrated from the EHR to the clinical trial database. While this meets the basic needs of most clinical trials, our goal is to provide the functionality needed for sites to handle all research data needs from within their EHR and have developed features beyond those defined by the standards. These extended features are built to work with the Prism EDC system, but are also valuable as features of a more generic application of our eSource solution.

Core Features

Data Migration This is the solution’s primary function. Clinical research data already captured in the EHR can be auto-populated into the clinical database. The study eCRF is displayed within the EHR interface and data fields with corresponding data in the EHR are already completed with the EHR data. The study coordinator can accept the EHR data or override with manually entered data. Data fields without a match in the EHR are completed manually. All data entry by the site, whether auto-populated or manually entered, generates an audit trail captured by the clinical research database. (Mapping of the EHR data to the eCRFs is performed as part of the study database definition.)
Data Quality Checks during Data Entry Pre-programmed data quality checks run when the data are submitted to the clinical database. The coordinator is alerted that the data has failed a data quality check and can enter new data in the field, if appropriate. Changes in the data will be recorded in the audit trail.
Flag Data in Clinical Database as eSource Any data that is auto-populated from the EHR will contain an “eSource flag” as part of the metadata that is transmitted to the clinical database. This will identify the data as originating from eSource and therefore not requiring source data verification (SDV).

Filter Multiple Data Points The clinical trial protocol may want to limit the amount of data collected. For example, the protocol may require only recording medications taken 30 days before enrollment into the study while the EHR may contain medication history going back years. The solution allows the study coordinator to filter the data displayed to only the data of interest.
Select from Multiple Data Points Fields on an eCRF may have multiple matches in an EHR record for that patient. For example, the form may ask for vital signs on a particular date, but the EHR has multiple sets of vital signs for the patient on that date. The solution will display all the possible data matches and the study coordinator can select which data are submitted to the clinical database.
Store Certified Copy of Source Document A key feature of the EHR integration standards vetted by the FDA is the ability to store a copy of the EHR data as it existed at the time the data were submitted to the clinical trial database. This copy is considered a certified copy of the source and meets the applicable regulatory requirements. The solution stores a copy of the source in an independent Prism Archiver program.

Prism eSource Benefits

EHR integration using the Prism eSource solution results in a number of process efficiencies and potential cost savings. Because sites rely on their EHR systems for patient care and insurance reimbursement, data entry into the EHR system is timelier than what is typically seen with EDC systems. That means the sponsor receives the data faster. And since many of the data points are migrated directly from the source (EHR), the data are cleaner. The cost savings of faster, cleaner data are hard to quantify but no less real and no less significant.

The major cost savings are related to source data verification (SDV) – one of the most expensive components of a clinical trial. Data that are migrated from the EHR in a validated solution like Prism eSource do not need to be source data verified. This process improvement will not eliminate site visits, but it can dramatically reduce the number of visits and allow site monitors to focus on study conduct, not just data collection.

The solution also has a source document-viewing feature – the Archiver. This feature is defined in the CDISC/IHE standards underlying our solution. When data are transmitted from the EHR to the clinical database, a copy of the migrated data included in the patient’s CCD (Continuity of Care Document) is stored in the Archiver and is considered by the FDA to represent a certified copy of the source document. The Archiver allows the end user to view this copy of the source data remotely while comparing it to the data captured in the EDC system.

The site also experiences significant benefit through the reduction of duplicate data entry in the EHR and then in the EDC system. In a survey of investigative sites, coordinators reported that 80% - 100% of the required study data are already recorded in the EHR system. Since the data migration is cleaner, the site coordinator spends less time responding to data queries that result from transcription errors. Our solution also results in less time spent on training. Site coordinators can perform all the necessary functions within the EHR system they use every day. They do not need to learn a new software program. As an additional benefit, the decreased workload associated with doing clinical trials will help reverse the trend of investigators dropping out of clinical research due to the burden of doing trials and increase site satisfaction.

Efficiencies Seen with Duke University Pilot

A landmark demonstration pilot of the technology was conducted by the Duke University Clinical Research Institute recently and was published in a peer reviewed journal.† While they did not use the Prism eSource solution, their tool was based on the same technology and same standards. The study compared eSource and traditional workflows used for clinical trials. The results were very positive with several key findings:

37% reduction in time need to enter data
Error rate was reduced from 9% to 0% for auto-populated data
Benefits were seen with auto-populated and non-autopopulated data
eSource workflow reduced site requirements by 1 FTE

†Nordo, Eisenstein, Hawley, et. al, A comparative effectiveness study of eSource used for data capture for a clinical research registry, International Journal of Medical Informatics, 103 (2017) 89-94.

Prism eSource