Improve predicted mortality rates through documentation

Association of Clinical Documentation Improvement Specialists, January 1, 2010

An 89-year-old female is admitted to the hospital from home with a ST segment elevation myocardial infarction (STEMI). She also has stage 3 chronic kidney disease and refuses catheterization due to risk of dialysis. The patient enacts her living will. She is given appropriate medical therapy, but Sunday night at 3 a.m., her MI extends suddenly, and the blocked artery of the STEMI has re-stenosed. The woman passes away by 4 a.m.

The house officer enters the room, pronounces her death, and lists acute myocardial infarction (AMI) as the sole diagnosis. The chart is processed in HIM, and the business office drops a bill within 36 hours, to the delight of the revenue cycle metrics dashboard team. It’s happened again—the mortality of a predictable moribund patient of short length of stay damages another physician’s profile.

These deaths often tend to be assigned the lower of the MS-DRG “triplets” (without a CC or MCC). They tend to transpire overnight or on the weekends and are not caught by the CDI team. The MS-DRG system seems to account for these events by design. These triplets leave an open door to many questions when we look at case mix relative to mortality. It is reasonable to ask why a patient without a CC or MCC
would expire. Other questions include:

  • Are these patients less sick than other patients who passed away?
  • Will their mortalities appear unjustified in the eyes of the Medicare Provider Analysis and Review or other database?
  • Will this event translate into a publicly available profile indicating that mortalities are higher than predicted?
  • Are physicians aware of their own profiles and how easy it is to ruin a profile?

DRGs that are associated with a higher frequency of mortality yet are frequently underdocumented by the physician in regard to severity of illness include congestive heart failure, pneumonias, urinary tract infections, and malignancies.

Analyze dying process, not just cause of death
Physicians live in a world of ambiguity. Outpatient requirements do not recognize rule-out diagnoses, whereas inpatient documentation not only allows it but encourages its use. A similar paradox is found when we evaluate the manner in which physicians are allowed to document death.

Death certificate documentation does not allow for documentation of the dying process or the mode of death. Instead, death certificates require that a cause of death be noted (generally the principal diagnosis). Ironically, documenting the cause of death is not enough under the MS-DRG system. Instead, we need to focus on the dying process.

I serve as the senior vice president of quality and utilization management at my hospital. What I have learned while evaluating mortality charts with lower severity of illness is that there are three common elements to these events:

In these cases, it is easy to predict the impending demise of the patient. To the clinician, it is obvious early on that the patient most likely will not survive.

All care provided by the facility is of an appropriate level of quality and utilization.

The patient passes away after a generally short length of stay—usually equated to a brief expiration note that, according to bylaws, suffices as the discharge summary.

Establish straightforward solutions
One simple intervention can fix all these problems. The medical staff must adopt a culture of  documenting the dying process. Very simple entries will elevate the severity of illness to the maximum value under the MS-DRG. For example, CDI specialists should ask physicians to document the following:

Cardiac arrest

Respiratory failure

Hypotensive shock


CDI specialists need to educate physicians that the primary diagnosis (e.g., AMI) would not be improved with a cardiac complication, but that respiratory failure would change the MS-DRG and improve predicted mortality measures.

Most physicians will perceive a request for the mode of death as another intrusion by the magnanimous CDI or case management department. But once a CDI specialist provides an explanation of predicted mortality measures, there is virtually no resistance.

The HIM department needs to be prepared to apply bill holds to deaths that do not have MCCs. Everyone needs to realize that it is unreasonable to believe that any patient would die without a MCC. In fact, they don’t. These patients have respiratory failure and cardiac arrest. They have hypotension. Most go into a coma. Inherently, the MS-DRG system penalizes rushed documentation, but fewer scenarios are more damaging for a hospital and physician than the lack of documentation of the dying process.

Editor’s note: Arnott is a physician consultant at Wellspring Partners, a Huron Consulting Group Practice, based in Indianapolis. He is senior physician consultant at Physician Motivators (www.physicianmotivators.com) in Canfield, OH; senior vice president of quality and utilization  management and cochief medical officer at Northside Medical Center in Youngstown, OH; and maintains a private office in Canfield. You can contact him via e-mail at jarnott@huronconsultinggroup.com.

Most Popular