Why sites and sponsors need a Medical Records API
EMR Integration: The Challenge
In the past several decades, Electronic Medical Record (EMR) systems have become the standard in documenting clinical care in the United States and many other countries, while Electronic Data Capture (EDC) systems have become standard in global pharmaceutical trials. Naturally, the industry has been highly focused on the tantalizing possibility of integrating EMR into EDC.
However, integration has been difficult, costly, and elusive, often involving long and protracted implementations at the health care provider level to connect the EMR to the EDC. These point-to-point integrations are very difficult to scale given the highly fragmented provider market and numerous EDC solutions, resulting in a nearly infinite combination of EMR-to-EDC integration possibilities.
Not only that, but EMR data is usually too inaccurate and incomplete to flow directly into the EDC. Instead, study teams often have to perform extensive curation of the data before entering into the EDC. Therefore, an EMR-to-EDC pipeline, by itself, is not a complete solution since it does not permit the Principal Investigator (PI) to perform the necessary reconciliation, and therefore cannot provide complete traceability between the two systems.
CRIO’s Innovative Solution
CRIO, in partnership with Pluto Health, has solved these two problems. Together, the two companies have developed a scalable program that integrates care and research at the PI level, with EMR and care data flowing into the CRIO eSource system. This innovative partnership has two distinctions from traditional industry approaches.
First, with patient consent, medical data is retrieved, reviewed for health gaps, and interventions recommended through the Pluto Health patient-facing application. The patient is in control of his/her medical records – thus bypassing the need to have point-to-point integrations and enabling the unification of information across multiple health care networks. The patient also benefits from personalized recommendations to support their overall ongoing health and wellness.
Second, the medical data flows into CRIO’s eSource system, and not directly into the EDC. CRIO’s eSource system is a site-facing application where PHI is not visible to Sponsors. CRIO eSource empowers the PI and study team to review, interpret, redact, and curate medical data before “publishing” to the study. Only when the team has published the data is it visible to the Sponsor in CRIO’s integrated EDC application (Reviewer). Thus, CRIO eSource serves as a “middleware” between the EMR and the EDC, empowering the study team to do the necessary curation and reconciliation between the two systems.
[SIDEBAR]
In this innovative approach, medical data flows from multiple EMR systems into a single patient dashboard (Pluto Health), which relays the data, along with customized healthcare analytics and recommendations for interventions, into the CRIO eSource system, and then, after PI publication, to CRIO’s integrated EDC system (Reviewer) for sponsor review.
The pipeline compresses a multi-week process into 15 minutes, integrates care and research, and facilitates faster determination of eligibility and greater safety oversight by the PI.
The rest of this white paper will discuss in more detail the current limitations in using EMR systems for clinical research, and describe how CRIO’s unique partnership with Pluto Health overcomes these limitations.
Limitations of the EMR in Clinical Research
For interventional clinical trials, most clinical research sites utilize their own source system (whether paper-based or electronic, such as CRIO’s) to capture the clinical trial data. They generally do not use the EMR because the EMR is built as a workflow tool for providing care and billing insurance companies and is not optimized for clinical research. CRIO’s survey data confirmed that most sites utilize the EMR as a recruitment tool, not as a source collection tool.
Challenge 1: Medical Records Can Be Difficult and Lengthy to Obtain
EMR systems play a vital role in the data collection process because they house the patient’s incoming clinical history used to determine eligibility. Principal Investigators (PIs) need to reference the EMR to identify past and ongoing medical conditions, medications, procedures, allergies, and other relevant data. Not only that, but the PI needs to determine that the patient’s participation in the trial is safe, and consistent with his or her patient care journey to date.
In many trials, site staff utilize the screening period – often 2 to 6 weeks in length – to obtain the medical records. If the clinical trial subject is a patient of theirs, they can simply obtain a study-specific medical release and print out the medical records onsite for insertion into the source binder. If the subject is not a patient, or the subject is a patient but has data from a different institution, then site staff must obtain a medical release form and submit it to the relevant provider.
Often, this process may take weeks, requiring follow-up calls to track the status of the records. When the records do arrive, the subject is no longer onsite for the screening visit, so the PI often has to reconcile the medical records against the subject’s self-report through a separate phone call. The records may reveal exclusionary conditions the subject neglected to mention, thus creating a Screen Failure that could have been identified upfront. This means that the subject may have done study-specific procedures that could have been avoided.
Challenge 2: EMRs Are Not Accurate or Comprehensive Enough for Clinical Research
In addition, medical records are not complete or fully accurate. For one thing, a patient may have multiple providers and therefore separate records from multiple institutions. Even within a given institution, several providers will have recorded data over multiple encounters, which creates a cumulative record that is rarely pruned or cleaned. The result is an accumulation of provisional diagnoses, conditions that have self-resolved, medications once prescribed but not actively taken, etc.
For this reason, PI reconciliation of the medical record is critical. As a result, the medical conditions and medications recorded in the EMR are rarely transcribed into source (and hence the EDC) in a 1:1 manner. Medical conditions listed in the EMR may be excluded from source, or conditions may appear in source that are not explicitly listed in the EMR (although often consistent and sometimes “mappable”). The best way to think of the EMR is as an input to the source data.
Challenge 3: Medical Record Delays Cause Ongoing Patient Safety Risks
After the subject is randomized, site staff perform ongoing treatment procedures and query the subject about adverse events (AEs) and other medical developments such as changes in medications. Site staff record these changes in the source binder. When the subject reports an AE, staff must once again solicit the underlying medical records so that the Investigator can assess the AE for causality, severity and study treatment implications. This can create yet another laborious workflow to obtain and reconcile records.
The CRIO-Pluto Health Integration
Recent developments have promoted interoperability of medical records and patient access. The Fast Healthcare Interoperability Resource (FHIR) standard has emerged as the platform for interoperability across EMR systems. The 21st Century Cures Act, among others, gives patients the right to receive their medical records without charge. The U.S. is headed towards a world where patients will soon have open, direct control of their medical records on their devices.
Leveraging these evolutions within the healthtech environment, Pluto Health developed a tech-enabled service platform unifying health information from disparate sources such as insurance, medical, and social determinants; integrating trusted clinical guidelines; and creating personalized health recommendations in minutes. The Pluto Health service improves health care access for marginalized communities, enhances decision-making for point of care providers, and incorporates clinical trials as a potential treatment option.
Critically, with patient consent, the application only requires patients to input their name, date of birth, gender, and address to then query the health networks within the patient’s local geography to identify matches across medical records. Pluto Health accesses medical records from the major health information exchanges, which have 90% population reach in the United States.
For the integration with CRIO, the process will start with obtaining the patient’s signed medical release, then a study team can use CRIO to initiate a request to Pluto Health. Pluto Health’s application then creates a unified health record for the patient including a customized health care interventions report, and then transmits both the medical records and custom report to the study team’s account in CRIO. The entire process can take 15 minutes, so it will be completed while the patient is onsite.
The PI can now review a complete, unified set of medical records, along with a synopsis of intervention recommendations. The two data sets, together, enable the PI to assess the benefits and risks of patient participation in the trial, perform a more informed eligibility assessment, and pinpoint any safety risks. Combined with the PI’s interview of the patient and the various study-specific procedures, the PI can now make the best-informed judgment on eligibility. And, regardless of eligibility, the patient receives the benefit of the custom care report.
Once the PI has completed his or her review and reconciliation of the records, the PI can electronically sign the medical records PDF (with annotations and redactions, as required), at which point the PDF becomes a part of the study record, ready for remote monitoring by the sponsor or CRO from CRIO’s integrated Reviewer application.
Critically, the study team can, as needed, use this same process to enable the PI to make rapid assessments of any self-reported Adverse Events that arise during the trial. This empowers PI oversight on the patient’s entire trial participation, not just at Eligibility.
Conclusion
CRIO’s innovative EMR integration with Pluto Health leverages the recent trends towards interoperability and patient centricity. It provides a powerful, scalable way to fold electronic medical data into the PI workflow, while respecting patient control over their records and providing valuable health recommendations to patients. This is the right way to do an EMR-to-EDC integration – not by creating a direct data flow, which would result in duplication of the data quality issues in EMR systems into the EDC – but by creating a pipeline that aggregates data from multiple EMRs into a single record, along with a future-facing care journey recommendation, for PI assessment and reconciliation.
