Pivoting CDI: The World of Healthcare Watches

Is CDI about to embark on a long journey to reinvent Itself?

There is no arguing that artificial intelligence (AI) and natural language processing (NLP) are making inroads in the healthcare revenue cycle, creating better efficiencies with the automation of a multitude of historically manually performed tasks, thereby reducing positions that were once performed by staff.

AI is clearly beginning to take hold and make significant inroads in the clinical documentation integrity (CDI) space. I have noticed serval posts on LinkedIn, as well as in Becker’s Healthcare e-newsletters, discussing the role of AI in the revenue cycle. Just recently, there was a blog post published in KevinMD titled “How an AI bot transformed my EHR experience (KevinMD blog)” centering on how AI streamlined the provider’s documentation and charting in the electronic health record (EHR) by scanning through the documentation as the note is being completed, providing suggested diagnoses with associated ICD-10 codes. I had the privilege of participating in a demo of a clinical AI software solution that is truly physician-centric, geared toward streamlining the physician’s documentation of his/her assessment. The AI system built into the software scans the entire medical record in real time, while the physician is documenting in the EHR, indifferent as to what method of documentation the physician utilizes. Any overlooked diagnoses are identified with appropriate clinical specificity and presented to the physician for consideration in real time.

What blew me away was that not only are diagnoses suggested as they are identified, but the system also pulls in the supporting clinical indicators, relevant clinical findings, tests ordered (with results, treatments, relevant consultant assessments, and recommendations), and other ancillary assessments and recommendations, etc. The physician reviews the suggested diagnosis with accompanied supporting information, can agree, and by a simple click can automatically pull the AI-generated information into the assessment portion of the note. The physician can reject the suggested diagnosis or add to the suggested diagnosis with supporting information, and again by a simple click pull the information into the physician’s note.

Reflecting on the system after the demo and speaking with several CDI directors about the functionality and capabilities of the system, the question must be asked: how will AI CDI systems impact current CDI processes and CDI professionals? Granted, there are other commercially available CDI systems currently available that purport to utilize AI and NLP to process physician documentation and other available data points, serving to prioritize the records for CDI intervention that have the greatest opportunity for documentation improvement (i.e., capture of a CC/MCC that may require a CDI query to the physician). These available CDI systems still require some form of CDI intervention, although some claim to offer the functionality of automated queries, a dangerous precedent, in my opinion. There has now reached a point where CDI can potentially be displaced by CDI AI software solutions; the future of the CDI profession can be rewarding if the profession is committed to pivoting from current processes, goals, and purposes to ensure continued growth and future development that more closely aligns with and supports the hospital’s or health system’s revenue cycle.

Ensuring CDI’s Future: A Critical Turning Point

Since the inception of the CDI profession, the structure and processes of CDI have remained virtually unchanged. Granted, there have been some minor changes, including increased use of various software platforms, expansion into the outpatient CDI arena, and focus upon quality and safety reporting measures. The CDI profession has remained laser-focused upon processes relying on documentation of CCs/MCCs, clarification of principal diagnoses and patient safety indicators (PSIs) and hospital-acquired conditions (HACs), among other quality measures. The profession, in my opinion, has reached a point where it is faced with a critically important decision: either CDI reinvents itself or is met with the reality of experiencing a slow decline in relevance and job opportunities. It simply is inevitable at this point; the handwriting is on the wall, and the decision is in the CDI profession’s court. A state of denial and continued apathy demonstrated by CDI will prove detrimental. Changing the name of the profession to clinical documentation “integrity” or labelling CDI as “population health CDI” or some other variant are shallow approaches to addressing continued shortfalls that are being showcased with progressive AI inroads in CDI, like the software solution I previewed.

So, what steps or path should CDI embark upon in its long journey to reinvent itself? First, the CDI profession must recognize and embrace the notion that the medical record is first and foremost a communication tool versus a reimbursement tool, where dollars are extracted from the record as the primary outcome of CDI’s task-based activities. Having said that, the CDI profession can only start its transformation to a high-value, more purposeful entity with processes that best serve the patient, all physicians, and other healthcare stakeholders, aligning more closely with the overall revenue cycle.

Today’s CDI model emphatically contributes to unnecessary, self-inflicted, nontechnical denials through unrelenting focus upon diagnosis securement, without achievement of meaningful improvement in supportive physician documentation. Such documentation must provide the relevant clinical information, clinical context, and clinical picture necessary to best communicate patient care and alleviate a large portion of payor denials, particularly clinical validation denials. A good number of such denials are legitimate, attributable to insufficient physician documentation.

The next inclusive step on the trek to transformation is a commitment to expanding the CDI specialist’s breadth and depth of knowledge in best-practice standards and principles of physician documentation. An ideal starting point is to understand the principles of evaluation and management (E&M) services, not from a coding perspective, but from a “documentation” perspective. The Centers for Medicare & Medicaid Services (CMS) Evaluation and Management Service Guide (E & M Service Guide) is a great resource to start out with, serving as a strong foundation for knowledge and skill-set building in physician documentation.

Another valuable resource is Bates’ Guide to Physical Examination and History Taking, providing one with the understanding and inner workings of the history & physical documentation. Insufficient documentation in the history and physical contributes to most medical necessity denials, in my own personal experience as a CDI professional reviewing denied cases trying to rehabilitate operations as part of the denials and appeals process. In fact, the Comprehensive Error Rate Testing (CERT) contractor consistently identifies medical necessity and insufficient documentation as accounting for nearly 80 percent of improper payments every year, in the category of short-term acute inpatient care DRG improper payments (CERT Report).

These are just a few of the many steps necessary to reinvent CDI, making it more valuable and relevant as it relates to physician documentation, communication of patient care and achievement of quality care, with engagement of physicians in the important role of patient care communication. Couple this with advancing AI software solutions like the one I described earlier, and the achievement of solid physician documentation without the intrusive query process, and these steps will fully support a high-performing revenue cycle.

Getting Started

Getting started requires a commitment to relevant change, letting go of the natural tendency to cling to the status quo. The future of the CDI profession hinges on transformation of our purpose, mission, and vision of CDI. In my next article, I will outline the next key components to consider driving interest and commitment to this.

Our future depends strongly on willingness and resolve to transform ourselves. Stay tuned!