Top 6 Reasons EHR Data Holds the Key to Better Patient Outcomes
Written by: Rachana Kolloru, Product Manager, Veradigm
Exploring the Strategic Importance of EHR Data
Electronic health record (EHR) data connects diverse research points across the care continuum. As part of the collection known as “real-world data,” EHRs gather patient data from preventive care sources (screenings, primary care providers), routine healthcare, disease management, all the way to acute care (emergency services, inpatient services), and on to palliative care. The data is instrumental because it can cover each of these areas over time.
Every part of our healthcare system generates data that can be used strategically to inform research toward better outcomes. In this post, we share six reasons why EHR data helps improve patient outcomes.
1. Comprehensive Patient Insights Through EHR Data Integration
Data integration combines data from different sources into a unified, cohesive data set. Data might come from EHRs, claims and billing data, wearable devices, registries, and other sources. The integrated data provides researchers with a comprehensive tool to query and gain insights. Previously, typical research data had come only from randomized clinical trials (RCTs). However, researchers now recognize the importance of a wider data set in producing evidence more quickly and efficiently. EHRs are a particularly good source of data because 88.2% of office-based physicians use EHRs.
The comprehensive view offered by integrated EHR data provides new perspectives on patient demographics and disease progression, which are crucial for clinical trials and personalized medicine.
Clinical Trials Aided by EHR Data Integration
Researchers recruiting for clinical trials can automate the screening process with the help of EHR data integration. Since EHRs contain everything from demographics and medical histories to diagnoses and lab results, researchers can quickly screen large populations for trial eligibility. Researchers can find trial participants who fit their trial criteria by searching for specific medical histories, treatments, and even genetic information.
EHR data can also help researchers include underserved patient populations in clinical trials. Large EHR datasets, such as Veradigm Network EHR Data, include a wide variety of healthcare institutions with patients of diverse backgrounds. Increasing the inclusivity of the clinical trial makes the trial results more generalizable.
Personalized Medicine Aided by EHR Data Integration
Researchers in personalized medicine can study how genetic factors influence patient responses to treatments by reviewing or searching through EHR data integrations. This research allows for personalized treatment plans based on a patient’s unique genetic profile. Because EHRs provide long-term continuous patient data, researchers can track outcomes over time and for specific populations. This critical real-world data helps suggest adjustments for individual patients, improving health outcomes.
2. Data-Driven Decision Making with Real-Time Access
The Role of Real-Time Data in Research
Researchers typically want to reduce the time it takes to generate insights from their research. Real-world data can help with that, and researchers add RWD to the results produced more slowly through RCTs. For example, a study exploring the frequency and severity of myocarditis after COVID-19 vaccinations used real-world evidence (based on RWD) from when COVID-19 began in March 2020 until May 24, 2021. The results were published in the New England Journal of Medicine only four and a half months later.
Real-time EHR data provides deeper insights into disease prevalence and treatment outcomes in the patient population. Researchers can use these insights to:
- Design targeted, efficient clinical trials
- Detect adverse events earlier
- Identify side effects sooner and modify the trial design if necessary
How is Data Extracted from EHRs?
Extracting data from EHR systems can happen in a variety of ways:
- Direct integration with EHR systems using application programming interfaces (APIs) or health information exchanges (HIE)
- Database queries use SQL queries and cohort selection tools available within some EHR systems
- Natural Language Processing (NLP) uses algorithms to sort the valuable information in unstructured data, like the free-text fields of physician notes and discharge summaries
Structured Data Vs. Unstructured Data
Structured data is highly organized and easy to scan. Generally structured data is entered using a predefined format with specific fields and numerical values, often using drop-down menus. Structured data frequently uses standardized formats for clinical terms or lab results, making comparisons more straightforward.
Unstructured data allows for free-form entry as narrative. Clinician notes are typically free-form, as are summaries and reports. In the past, gathering and retrieving unstandardized information for analysis was challenging. Today’s natural language processing (NLP) computer systems extract insights from unstructured data, providing access to information unavailable using traditional software tools.
3. Enhancing Patient Recruitment with EHR Data Conversion
Using EHR data to recruit clinical participants is becoming more common. A 2021 study noted that the majority of their focus group members used EHR as the primary modality for screening potential clinical trial participants. Members said it was also the method that produced the highest yield. In addition, EHR-based recruitment allowed recruiters to create complex queries, resulted in fewer screen failures, and retained the ability to send messages to non-study doctors to recruit potential trial participants.
EHR Data Conversion Explained
For diverse EHR data to be useful, that data must be converted into a standard, usable format. EHR data conversion has seven general steps:
- Data extraction pulls structured and unstructured data from EHRs. Data can be in many formats, such as ICD codes, lab results, clinical notes, and free text. Tools like Fast Healthcare Interoperability Resources (FHIR) and APIs extract data in a common language and structure.
- Data cleansing and preprocessing. Data from different EHRs across many different institutions may use a variety of nomenclatures (for instance, ICD, SNOMED, and other clinical terminology). Extracted data must be normalized to achieve a common standard. After normalization, the data is checked for duplicates, which are removed (de-duplication). After that, missing data is either imputed (inserting an estimated value) or the data set is excluded from the analysis.
- Data standardization. Clinical data from each different system must be mapped to the standardized terminology. Standard terminologies include ICD-10 for diagnoses, LOINC (Logical Observation Identifiers Names and Codes) for lab results, SNOMED CT for clinical concepts, and RxNorm for medications. Unstructured data requires NLP to extract pertinent information like symptoms and diagnosis descriptions.
- Patient cohort identification. Clinical trial eligibility criteria are then applied to the extracted, cleaned, and standardized data. Automated algorithms search for patients who meet the inclusion criteria and not the exclusion criteria. Some systems offer real-time matching of eligible patients for ongoing clinical trials, which can speed recruitment.
- Data integration. Aggregated and standardized data is often stored in clinical data warehouses designed to support research queries. Tools that meet FHIR specifications increase the likelihood of interoperability across platforms and institutions, helping keep data consistent and accessible. Data integration must adhere to patient privacy laws (HIPAA in the U.S. and GDPR in the EU).
- Patient stratification. Data scientists use the cleaned and standardized data to run machine learning models that predict patient outcomes. The results are used to classify patients by risk level and identify those groups or individuals that might benefit from intervention.
- Reporting and data visualization. Reporting tools that can generate comprehensive reports help researchers decide on the appropriateness of potential patients for clinical trials. Some systems provide a dashboard to help visualize trends, eligibility percentages, and recruitment metrics.
How to Improve Accuracy and Reliability of EHR Converted Data
The use of digital data improves data accuracy by importing structured and unstructured data from their original digital sources. This avoids the time-consuming and error-prone re-entering of information. Sometimes data missing from the structured data in an EHR (“data missingness”) can be found in the unstructured sections.
4. Optimizing Clinical Trials Through EHR Data Migration
Data Migration Ensures Integrity and Continuity across Multi-Phase Trials
Data migration is the translating and transferring of data from one validated system to another.
When moving to a new system or across clinical trial phases, data quality, structure, and completeness must be maintained. Data migration processes use validation and verification steps to ensure integrity and accuracy. During the migration process, irrelevant, incomplete, or duplicate data can be identified and corrected.
EHR data migration enhances data reliability and accessibility and generally streamlines trial operations.
5. Long-term Research and Development Support
Retrospective EHR data can play a significant role in long-term research by supporting longitudinal studies, post-market surveillance, and regulatory compliance.
Retrospective EHR Data and Longitudinal Studies
Researchers conducting longitudinal studies can use retrospective EHR data to track patient health in many populations over long periods. Available data includes changes in health status, disease progression, and treatment outcomes. Researchers can study chronic conditions by identifying patterns related to the disease from years—or decades—of data. Risk factors and treatment impacts will also become evident.
Large-scale retrospective cohort analysis is also possible with EHR data. These studies could provide statistically significant information while reducing the time and cost compared to prospective studies. For example, Veradigm Network EHR Data has access to over 152 million patient records1 from a network of EHR systems that could serve as a research base.
EHR Data and Post-Marketing Surveillance (PMS)
EHRs like those in the Veradigm Network can also help assess real-world drug efficacy and safety beyond clinical trials. The EHR data can also be used to detect rare or long-term adverse events, track medication prescribing trends, and follow outcomes related to specific therapies.
Using EHR Data for Regulatory Compliance and Trend Analysis
EHR data can also satisfy the evidence requirement for product safety, efficacy, and performance over time by providing real-world evidence. All of this data can be used to track and analyze trends and search for therapeutic effectiveness over time.
6. Cost Management and Operational Efficiency
Leverage Data to Reduce Research and Development Costs
The strategic use of EHR data can benefit research and development activities and result in lower costs. One way to leverage data is to use linked data, which is the integration of EHR data from many sources. This linked data eliminates the cost of duplicating datasets. These data linkages can also help increase the breadth and depth of data related to individual patients.
Harness the Full Potential of EHR Data
Contact us to learn more about licensing Veradigm Network EHR Data to move your research forward.
References:
- 5yr timeperiod: Q1 2020 - Q4 2024