An old adage about clinical documentation states, "If you didn't document it, it didn't happen." Despite the importance of documentation as the data source for quality metrics, reimbursement and continuity of care, clinicians are frustrated and overwhelmed by administrative work. Responding to clinical documentation integrity queries is the last thing that many clinicians want to deal with.
Fortunately, 3M Health Information Systems has an advanced artificial intelligence-based CDI solution, 3M™ M*Modal CDI Engage One™, that streamlines the documentation process and highlights CDI program improvement opportunities through new, user-friendly dashboards.
Becker's Healthcare recently spoke with Kaitlyn Crowther, chief product owner at 3M HIS, to learn more.
CDI affects more than the bottom line
The goal of CDI is to accurately show all of the resources used to care for a patient, ranging from treatments and labs to clinical decision-making. Accurate documentation is the foundation of appropriate reimbursement for services, but it also ensures continuity of care as patients transition from the inpatient setting to outpatient care.
Hierarchical condition categories, for example, provide comprehensive insight into the acuity of patients across different care environments. They are a measurable way to capture and recapture chronic conditions for an individual throughout the year.
"If a patient is being treated in an acute care setting for heart failure or stroke, those conditions must be documented to represent the severity of the patient," Ms. Crowther said. "Let's say after discharge, the patient visits a doctor for diabetes. There may be things that happened during their inpatient stay that would affect their outpatient diabetes treatment."
Technology can enhance CDI without increasing administrative burden
Technology, and more specifically, AI, can augment CDI work by automatically identifying documentation improvement opportunities at the point of care. The key is to make requests in a way that enables clinicians to respond easily.
"Our solution uses a rules-based approach that follows how a human CDI specialist would gather more information," Ms. Crowther said. "We see if clinical documentation exists to support a clinical condition in the chart. Based on that information, we can determine whether there is supporting evidence to document further specificity and acuity and present that to clinicians in real time. Providers can either accept, ignore or dismiss requests for additional documentation."
3M CDI Engage One leverages Natural Language Understanding (NLU) technology to identify when concepts and evidence are documented with a positive certainty and subject, such as whether information pertains to the patient or are related to family history. The technology also focuses on current temporality, so issues are addressed that are specific to a particular episode of care.
Dashboards & analytics measure success and identify CDI-related improvement opportunities
To help organizations quantify the effectiveness of their CDI work, 3M CDI Engage One offers comprehensive reporting.
"We measure how clinicians interact with automated CDI opportunities," Ms. Crowther said. "They might resolve the opportunity and provide documentation that satisfies our rules-based engine. We can also see if they ignore an opportunity because it isn't pertinent at that time, or if they physically dismiss the opportunity in the user interface."
Based on that data, the solution analyzes how clinicians' behaviors impact the final documentation for encounters. For example, it can show whether the organization missed out on capturing a more specific diagnosis or a procedure.
In addition, 3M CDI Engage One features new dashboards that slice and dice data and help pinpoint opportunities to improve operational performance. They show CDI trends across different subspecialty groups, facilities, individual clinicians, different medical conditions and more. Users can also drill down into reports containing greater detail.
"Trends in medical condition data are very powerful," Ms. Crowther said. "If there is a low documentation rate associated with a particular condition, opportunities may exist for the CDI team to educate clinicians about how and why to fully document."
Getting started with a technology-enabled CDI program
For organizations interested in launching a technology-enabled CDI program, a good first step is to analyze the CDI team's current query volumes. The goal is to begin with a few high-impact medical conditions that may be good initial candidates for automation through AI, and build from there.
Another best practice is to do peer comparisons using standard data like MEDPAR. "If you are serving similar patient populations as the provider down the street, but the data suggests that they are treating individuals who are sicker or need more acute care, documentation may be the reason for that difference," Ms. Crowther said.
When 3M HIS introduces 3M CDI Engage One to an organization, adoption specialists deploy the technology in a very measured way. "We nudge clinicians about CDI opportunities that are important to the health system," Ms. Crowther said. "They understand exactly what and why they are being nudged on certain areas."
A third important success factor is having a CDI physician champion. This individual educates their peers about the importance of documentation and why information must be documented in certain ways. "The physician champion is a key role when rolling out a CDI technology and it ensures that we have buy-in from the clinician population," Ms. Crowther said.