Findings from a recent EHR usability study conducted by the National Institute of Standards and Technology (NIST) once brought to the fore the problem of clinical documentation in the digital age of healthcare.
The study of EHR use, particularly copy-and-paste functionality, led to three major findings. First, clinicians participating in the study were concerned about EHR data integrity as a result of copying and pasting information. Second, clinicians identified entering the wrong information into the wrong record as a high potential risk. Third, participants reported that over documentation introduced challenges to accessing “accurate, relevant and timely information on a patient” at the point of care.
Despite its intended purpose to improve the ease and efficiency of clinical documentation, NIST concluded that the copy-and-paste functionality “has introduced overwhelming and unintended safety-related issues into the clinical environment.”
Concerns about the accuracy and quality of EHR documentation are nothing new. In a 2013 update to 2007 guidance on EHR documentation integrity, a workgroup convened by the American Health Information Management Association (AHIMA) called for safeguards to ensure electronic documentation did not undermine patient care.
“Without safeguards in place, records could reflect an inaccurate picture of the patient’s condition, either at admission or as it changes over time,” the AHIMA workgroup wrote. “The provider must understand the necessity of reviewing and editing all defaulted data to ensure that only patient-specific data for that visit is recorded, while all other irrelevant data pulled in by the default template is removed.”
What’s more, the authors of the EHR documentation guidance emphasized the urgency of addressing how the use of automated EHR functions could compromise the integrity of clinical health data.
“Data quality and record integrity issues must be addressed now, before widespread deployment of health information exchange (HIE),” they maintained. “Poor data quality will be amplified with HIE if erroneous, incomplete, redundant, or untrustworthy data and records are allowed to cascade across the healthcare system.”
More recently, research has pointed to a potential disconnect between patient-reported and provider-record health data. Researchers at the University of Michigan sought to investigate whether patient-reported eye symptoms were recorded as part of clinical documentation in EHR systems. Comparing eye symptom questionnaire (ESQ) and EHR documentation, Valikodath et al. found a “substantial discrepancy” between the two.
“Discordance in symptom reporting could be because of differences in terminology of symptoms between the patient and clinician or errors of omission, such as forgetting or choosing not to report or record a symptom,” they wrote. “Perhaps a more bothersome symptom is the focus of the clinical encounter, and other less onerous symptoms (e.g., glare) are not discussed (or documented). However, even for the exclusive sensitivity analysis, we show that the ESQ and the EMR are inconsistently documented.”
While discrepancies in patient- and provider-reported documentation were relatively harmless and did not directly impact patient safety, their existence does raise questions about data accuracy and completeness.
ROOT CAUSES OF EHR DOCUMENTATION PROBLEMS
The benefits of EHR use more generally range from timely access to clinical data and alerts to avoid medical errors to care coordination and improved billing and coding. EHR documentation is the means of realizing these benefits.
“Documentation is often the communication tool used by and between providers. Documenting a patient’s record with all relevant and important facts, and having that information readily available, allows providers to furnish correct and appropriate services that can improve quality, safety, and efficiency,” the Centers for Medicare & Medicaid stated in a 2015 fact sheet on EHR technology.
In that same guidance, the federal agency identified a handful of common EHR challenges that healthcare organizations and providers need to address. For EHR documentation in particular, these challenges include an inability to log clinicians entering data, cloning data from record to record, and upcoding to receive higher payment.
A year later, CMS released guidance focused specifically on preserving EHR documentation integrity with an emphasis on helping prevent fraud, abuse, and improper payments.
“Providers and others should use program integrity-related EHR software features and capabilities to ensure the integrity of the EHR documentation. Some EHR features may create information integrity concerns; however, providers and others can mitigate these concerns by implementing proper policies and processes,” the federal agency concluded.
Where the CMS recommendations for EHR documentation integrity fall short is in identifying and remediating the root cause of these inaccurate, incomplete, or unreliable information within a patient’s EHR.
Around the same time, both the American Medical Association (AMA) and American Medical Informatics Association (AMIA) (among others) set out to improve future EHR use by recommending changes to EHR design that address the causes of poor EHR documentation.
For the latter, the first area of EHR improvement necessary involved simplifying and speeding documentation and included two recommendations germane to EHR documentation improvement.
The first was decreasing data entry burden on clinicians by allowing other members of the care team to enter data into the EHR:
AMIA cited the increasing documentation burden on clinicians as the impetus behind the use of copy-and-paste functionality and the resulting bloat in EHR documentation.
“Clinicians remain uncertain regarding who can and cannot enter data into each patient’s record, placing a tremendous data entry burden on providers, the most expensive members of the care team. Clinician time is better spent diagnosing and treating patients,” the association argued. “Regulatory guidance that stipulates that data may be populated by others on the care team, including patients, would reduce this burden.”
The second recommendation called for separating data entry from data reporting:
According to AMIA, EHR documentation requirements were responsible for making structured data preferable to unstructured data.
Similarly, AMA released guidance for improving EHR usability that included an emphasis on reducing clinical documentation demands on clinicians.
The first of eight total recommendations centered on improving physician-patient interactions by removing EHR technology and data entry as an obstacle to face-to-face communication. AMA traces the problem to EHR design not based on clinical workflows:
In a similar vein, AMA also called on EHR developers to focus on EHR designs that help reduce the cognitive workload existing systems impose on end users:
The solution to these and the other EHR usability challenges comes down to user-centered design with clinicians provide substantial input into how developers go about designing their EHR technology.
SOLUTIONS FOR IMPROVING EHR DOCUMENTATION
Solutions to clinical documentation problems fall into two general categories: technical and administrative.
Beginning with the latter, healthcare organizations of any side need to establish business practices supporting quality clinical documentation. And the guidance from AHIMA guidance from 2013 still holds true in laying out for primary conditions necessary for maintaining EHR documentation integrity:
- Desire and commitment to conduct business and provide care in an ethical manner
- Purchasing systems that include functions and capabilities to prevent or discourage fraudulent activity
- Implementing and using policies, procedures, and system functions and capabilities to prevent fraud
- Inclusion of an HIM professional such as a record content expert on the IT design and EHR implementation team to ensure the end product is compliant with all billing, coding, documentation, regulatory, and payer guidelines
As for technical solutions to EHR documentation pitfalls, the AHIMA recommendations align well with those identified by CMS in 2016, both of which call for capabilities in EHR technology that require user authentication and access management, allow for the tracking of user activities (e.g., audit logs), and restrict alterations to user auditing files.
These technical recommendations, however, fail to address EHR documentation challenges to clinical productivity. A growing body of evidence points to the potential for natural language processing technologies to reduce burdens associated with clinical documentation.
“We found that a pure protocol of NLP Entry as well as hybrid protocols (involving both NLP Entry and Standard Entry) showed promise for EHR documentation, relative to Standard Entry alone (Standard-Standard Entry),” Kaufman et al. observed in JMIR Medical Informatics.
In an interview CHRISTUS Health in early 2016, CMIO Luke Webster credited natural language processing as a key component of clinical documentation improvement efforts at the Texas health system. In many cases, physicians were unaware that their dictation was running through a voice recognition engine and being transcribed before a human transcriptionist reviewed and finalized the documentation.
“One of the hopes (and obviously the plans that we have in place) is to leverage that foundational technology to improve clinical documentation real-time, such as prompting providers as they document,” said Webster.
While natural language processing could ease EHR documentation burdens on providers, it requires the implementation of more technology. An easier solution could be in the form of rethinking workflows for clinical documentation as Mississippi's Memorial Hospital of Gulfport CMIO David Northington, MD, explained last year:
A combination of these administrative and technical solutions will be necessary to ensure that EHR documentation benefits patients in the here and now as well as those receiving care in the future.