Measuring CDI Performance: A Truthful Conclusion

Task-based, outcomes measurement versus process improvement generally does not support sustainable long-term results.

I was recently asked by a chief financial officer (CFO) what other clinical documentation integrity (CDI) key performance indicators (KPIs) should be utilized, aside from the traditional ones. The CFO was attempting to evaluate the efficacy and effectiveness of his CDI program and the achieved return on investment, after it had been in place for seven years. He was referring to the traditional approach of measuring CDI program outcomes and overall return on investment using KPIs consisting of the following:

• Number of initial charts reviewed;

• Number of second charts reviewed;

• Number of queries issued;

• Number of queries responded to by the physician;

• Physician query agreement rate;

• CC/MCC (complication and comorbidity/major CC) capture rate;

• Case mix index; and

• Coder CDI DRG congruence rate.

These measures of CDI performance merely scratch the surface, with so many far more effective measurements of long-term CDI success available – ones that every CFO should consider and adopt. Relying on the above-mentioned measures provides a misguided and inaccurate picture of any program’s effectiveness of performance. So, how should overall program effectiveness be approximated and measured?

Measuring CDI Performance Validly and Reliably

While not downplaying the significance and importance of the traditional CDI KPIs, they must be utilized as a second-level tool. These KPIs only represent “task-based activities” that indicate if CDI staff are performing what is expected of them, per their job descriptions. A quote by Eliyahu M. Goldratt applies, in this instance: “tell me how you will measure me and I will tell you how I will behave.”

Task-based activity measurement geared toward outcomes versus process improvement generally does not support sustainable long-term results. Short-term gain is achieved while foregoing long-term success. An illustrative example involves measuring CC/MCC capture rate. The mere fact that a CC/MCC is captured, allowing for coding and billing of a higher-weighted DRG with increased reimbursement, does not necessarily mean that the end result will be an improved financial picture – payer clinical validation denials could always be looming.
 
In other words, just because the physician said it is so does not mean it is so. Payers are issuing these denials at an increasing rate, and while some of them may be egregious, others are accurate, based upon the clinical facts, clinical context, and clinical story as documented in the record. Often there is conflicting information in the record, the diagnosis is “dropped” in the record (secured through a one-and-done query), or the diagnosis appears only in the discharge summary. Capture of CCs/MCCs can be akin to a cat-and-mouse game, with the payer representing the cat and the provider the mouse. The so-called “sinister seven” diagnoses that payers frequently challenge under the auspices of clinical validation are the common CC/MCC diagnoses CDI professionals often search for in the record, generating a query. The medical record coder then assigns the diagnosis, yet payers have a rule in their claims processing program, meaning the record is requested and reviewed, and the payer issues a clinical validation denial. The payer may be making up their own clinical criteria used in their validation decision; just the same, the provider is now faced with the additional cost to appeal and hopefully overturn the denial. Making matters more difficult is that too often, the diagnosis in question is what I refer to as a “naked diagnosis,” thereby lending to the payer’s clinical validation denial determination. The “sinister seven” diagnoses include the following:

• Non-ST-elevation myocardial infarction (NSTEMI);
• Protein calorie malnutrition;
• Sepsis;
• Acute hypoxemic respiratory failure;
• Pneumonia;
• Acute renal failure/tubular necrosis; and
• Encephalopathy.

As can clearly be seen, CC/MCC capture rate is not necessarily a valid and reliable measure of individual CDI performance. Neither is the KPI of the number of charts reviewed, since reviewing a record does not correlate with achieving true documentation integrity. The number of records reviewed is analogous to the number of rocks looked under during an Easter egg hunt. A CFO focusing primarily on CC/MCC capture and number of charts reviewed is certainly getting a false sense of actual CDI program performance. He or she is focusing upon feel-good money versus real money that drops to the bottom line.

A Better Way to Measure CDI Performance

There are a host of different ways to validly and reliably measure CDI performance and overall success, providing the CFO with more objective financial information. A byproduct of complete and accurate documentation is solid financial results that generate net patient revenue versus gross revenue (which CDI consulting companies and CDI leadership sometimes claim represent CDI performance results). I submit to all CFOs and CDI leadership that the following measures represent more useful points indicating whether their CDI program is moving in the right direction:

• Level-of-care downgrades;
• Medical necessity denials;
• Denials CDI reviewed;
• Denials CDI reviewed and queried;
• Continued stay denials;
• Observations changed to inpatient denials;
• DRG downgrades; and
• Clinical validation denials.

Most non-technical denials in the inpatient hospital setting occur due to medical necessity and insufficient documentation, as pointed out in the last few years’ Comprehensive Error Rate Testing (CERT) Annual Reports’ Supplemental Improper Payment Data Report. The report for the 2021 fiscal year highlighted that nearly 80 percent of hospital inpatient improper payments were accounted for in just two categories, insufficient documentation and medical necessity. Medical necessity and insufficient documentation are virtually the same, since the latter inarguably contributes to the former.

The 2022 report shows that of the total improper payments made to hospital inpatient hospitals, 44.4 percent were associated with medical necessity, 29.5 percent were associated with insufficient documentation, while 24.6 percent were associated with coding. Nearly three-quarters of improper payments made to inpatient hospitals were based on two categories, attributable to physician documentation.

I encourage you to peruse the 2022 Report (2022 CERT Report), and pay particular attention to the section highlighting improper payments to Part A inpatient hospitals. You certainly will want to review Table D4 and Table E4, which showcase the top 20 DRGs with the highest improper payments by dollar amount and improper payment rate, respectively.

With improper rates like those outlined therein, it is more than safe to say that current CDI processes and associated KPI measurements are not noticeably moving the needle on assisting physicians in achieving better documentation. I submit to the CDI profession and CFOs that the only alternative is to build and expand upon current CDI processes. The time to transform CDI as a profession is long overdue, migrating away from task-based activities to a more effective role-based environment, whereby CDI works with physicians, physician advisors, case managers, utilization review/utilization management, and denials and appeals staff. A multidisciplinary approach to physician documentation improvement will be more efficient and effective in achieving clinical documentation integrity. Today’s current strong financial headwinds faced by all hospitals dictate that a new approach to CDI is paramount to helping address negative operating margins facing many hospitals.

Making a True Difference

There is always resistance to change, with a relative comfort level associated with the status quo. It is readily apparent, as outlined and highlighted above, that present CDI operations and outcomes-based measures of performance, generated by task-based activities, leaves staff ill-prepared to make a noticeable change in the quality and completeness of physician documentation, as defined by Medicare as follows (Complete & Accurate):

• All entries in the medical record must be complete. A medical record is considered complete if it contains sufficient information to identify the patient; support the diagnosis/condition; justify the care, treatment, and services; document the course and results of care, treatment, and services; and promote continuity of care among providers. With these criteria in mind, an individual entry into the medical record must contain sufficient information on the matter that is the subject of the entry to permit the medical record to satisfy the completeness standard.

For CFOs and the CDI profession, changes in CDI operations are inevitable, and long overdue. Physician documentation and communication of patient care must be clear, concise, consistent, contextually correct, and consensus-driven. There is little if any improvement in physician documentation presently being seen, despite the many activities embedded in CDI programs. Current measures of CDI performance are perpetuating the limited success of most CDI programs.

The financial health of hospitals and quality patient care are dependent upon better physician documentation as the foundation of medicine and the revenue cycle.