Unlocking the potential of EHRs for clinical decision-making

© fedrunovan - stock.adobe.com

Despite initial promises to streamline care and documentation, many physicians find EHRs often fall short in providing meaningful clinical value. Rather than interactive tools, EHRs are seen as data entry systems lacking comprehensive clinical guidance.

While EHRs provide alerts for potential errors and side effects, they fail to offer the broader clinical context physicians need.

What healthcare providers need is an interactive workspace that diagnostically connects patient data to reveal insights. This goes beyond documentation by mapping information to quality measures, public health reporting, and coding in real time. By anticipating needs and providing immediate support, this approach eliminates the need for post-visit corrections. To achieve this, EHRs should integrate natural language processing (NLP) to extract data, utilize standards-based terminologies to structure it, and incorporate clinical intelligence to filter it.

With such a system, doctors gain visibility into the clinical rationale behind recommendations, improving trust and productivity. Workflows can be designed to directly support high-value care activities, not just data capture.

While some physicians want technologies that automatically generate documentation through ambient listening, it's important to recognize documentation is just the starting point. Like self-driving cars, automated documentation often falls short. Regardless of the input method - whether NLP, speech recognition, or document scanning - human analysis is still needed to determine relevance and identify next steps.

In essence, what doctors need is a clinically interactive workspace that diagnostically connects data so that appropriate action can be taken, at the point of care, for each of a patient’s clinical issues. Such a system should go beyond documentation and assist in meeting quality measures, case reporting, E&M, HCC requirements, nursing care plans, and more, all in a single interactive tool. When a provider is considering a diagnosis, the system should proactively trigger intelligent workflows that ensure sufficient documentation, enable quality reporting, and guide appropriate actions for that specific condition.

Ambient listening technologies, combined with speech-to-text conversion, and NLP, can enable providers to capture the history of present illness (HPI) and other relevant information. This is often referred to as “telling the patient story” and it is something that many providers are now requesting. The challenge is in converting the “story” into actionable data to enable the provider to complete the encounter, address all clinical issues, satisfy all relevant quality measures, generate correct diagnosis and billing codes, meet documentation requirements, update the patient’s problem list, and actually have time to think about what they are doing.

Before widespread EHR adoption, providers scanned charts, met with patients, took notes, thought, documented later, and moved on. In the past decade, they became data entry clerks feeding EHRs, and getting little in return. Now, technology provides solutions.

Once HPI and other relevant information is captured with speech and converted to text, large language models like GPT-4 and Bard can assist documentation. But "clinical guardrails" or "sources of truth" are then needed to avoid AI "hallucinations." Providers also need tools that can take AI output and diagnostically connect it to data for downstream care and ever-increasing reporting, coding, and documentation requirements.

A critical part of an integrated, interactive clinical workspace is the ability to connect information from a clinical encounter to the rest of the patient record. This requires a diagnostic relevancy engine that organizes information aligned with clinician thinking, supports terminology standards, and customizes the presentation for each patient and clinician.

This evolution of EHRs from passive data collection to active clinical engagement with interactive workspaces has the potential to fulfill the long-awaited promise of EHRs in enhancing patient care. As healthcare becomes increasingly data-driven, clinicians require more than just robust data capture; they need interactive tools that can transform information into clinical value. An EHR workspace must move beyond simple data inputs and provide personalized, diagnostically connected responses that empower physicians to deliver high-value, patient-centered care.

David Lareau is the chief executive officer of Medicomp Systems, which provides physician-driven, point-of-care solutions that fix EHRs.