Health Information Management

Improve your CDI program with severity adjusted data

JustCoding News: Inpatient, October 28, 2009

Want to receive articles like this one in your inbox? Subscribe to JustCoding News: Inpatient!

The days of capturing a patient’s condition solely through complications and comorbidities (CC) and major CCs (MCC) are slowly fading. There is a new focus on capturing severity of illness (SOI) and risk of mortality (ROM), and clinical documentation improvement (CDI) programs are taking notice. And why shouldn’t they? SOI and ROM programs provide a higher level of detail about a patient’s condition and the care provided. They also strengthen hospitals’ quality data and physician report cards, which in turn improve revenue and reduce compliance risk.

During a September 18 HCPro audio conference, “Severity of Illness and Risk of Mortality: Sharpen Your CDI Focus with New Measures of Success,” speaker Garri L. Garrison, RN, CPC, CMC, CPUR, defined SOI as “the extent of physiological decomposition or organ system loss of function experienced by the patient.” Or in simpler terms, “how sick is the patient?” In self-explanatory terms, ROM is meant to indicate the patient’s likelihood of dying.

These two terms should catapult your CDI program to new heights of clinical specificity. A severity adjusted program is an ideal method to carry out your new and improved CDI efforts. A severity adjustment program allows hospitals, consumers, payers, and regulators to understand the patients being treated, the costs incurred and, within reasonable limits, the services and outcomes expected.

Understand SOI/ROM levels

In the audio conference, Garrison used the 3M™ All Patient Refined DRG (APR DRG) severity adjusted classification system to illustrate her examples, although there are several different severity adjustment systems available on the market.

Severity adjustment systems use a similar structure to that which you find with Medicare severity diagnosis related groups (MS-DRG), having one global category—the DRG. However, a severity adjustment system differs from MS-DRGs in that it further adjusts the data into four subclass levels for SOI, and another four subclass levels for ROM. Both subclasses exist through a numbered ranking system:

  • 1: Minor
  • 2: Moderate
  • 3: Major
  • 4: Extreme

SOI indicates how sick the patient is, and ROM refers to probability of death. Here, Garrison provided a clear SOI/ROM scenario:

“I can have a condition that has a severity of illness level or subclass level of a three. But that does not mean under risk of mortality that I will be a three. So I could be complex, costly to treat, but I’m not going to die from this condition. So I could have a subclass in mortality that’s a 1. So the scores are not always equal and often are not,” she said.

Never assume that your SOI and ROM numbers are the same, Garrison said. The designated numbers will climb when a disease progresses.

For example, the SOI of uncomplicated diabetes retains a one on the adjustment system, and progresses to a two when the patient experiences diabetes with renal manifestation. The same progression can be said for ROM: The more complications, the higher the risk of death.

Compare two patients with same SOI but different relative weight

A recurring problem for CDI specialists occurs when they have to address a particular physician argument that ends with, “but my patient is sicker.” Using the severity adjusted system will resolve this argument because now, “you’re comparing cases with the same complexity, even drilled down to a particular diagnosis or to a particular procedure,” explained Garrison. The system becomes a tool within physician groups or across facilities to compare and contrast the patients being treated, the costs that are incurred, and the expected outcome of the cases.

The problem becomes a little more complex, though, once separate APR DRGs are taken into consideration. In one case, a physician has a pneumonia patient for whom the SOI ranking is a two, and another physician has a patient with acute renal failure for whom the SOI is also a two. These cases will have different costs and lengths of stay (LOS) expectations as well as have different SOI weights.

This checks and balances process exists for those physicians who feel they’re being penalized for having a ‘less sick’ patient. A relative weight is assigned to each subclass, just as with MS-DRGs, and this allows you to calculate an APR DRG case mix index (CMI).

“The CMI lets us know that we have higher complex cases when we’re using this severity system,” explained Garrison. “When the third-party payers use these types of systems for payment, they have a weight for each of those subclasses so that they can calculate payment based on the complexity.”

Severity adjustment programs, like all CDI systems, depend on complete documentation. Reputations and reimbursements are at risk when a hospital’s death rate far exceeds ROM data. Using a severity adjustment system can ensure that the full complexity of an organization’s patient population is captured. When used correctly, a severity adjusted classification system can help facilities avoid misclassified information and improper coding due to lack of specificity in the clinical documentation.

Examine one case study

During the audio conference, speaker Shelia Bullock, RN, BSN, MBA, CCM, CCDS, presented a personal experience that she and her CDI team from the University of Mississippi Medical Center (UMMC) encountered. Bullock’s goal at her institution was to reduce the rate of mortality from above benchmark to average or below. “So one of the things we initiated was a mortality review for all payers,” Bullock said.

In the concurrent review CDI program, Bullock would review only her DRG payers. “So for the mortalities project, we reviewed all deaths regardless of payer type,” she said. “We reviewed any death chart that was not a severity of illness four and a risk of mortality of four.”

Bullock said she and her team met with physicians one-on-one, and worked vigorously with her inpatient coders “to guarantee a one-day turnaround” so as not to affect timely coding.

In 2007, through their research, UMMC expected only 1.94% of their patients to die. They observed 2.47%. That left them with a mortality index rating of 1.28. The goal of this program is to reach a mortality index rating of 1.00 or below. This occurs when the percent of expected deaths equal the percent of observed deaths. To improve this index, the facility launched the mortality review program and in 2008 UMMC’s mortality index decreased to 1.02. Within the first six months of 2009, UMMC reached its goal of matching observed and expected death rates (a rating of 1.00 or less).

The key to keeping numbers at or below average is consistency. The best CDI programs meet frequently with physicians; attend mortality and morbidity conferences; and attend monthly physician performance improvement meetings, said Bullock.

CDI programs aren’t the only facet that gains from a severity adjusted system. Physicians also benefit handily through this system. Physicians will have public data reports that can be shared among their peers. It also enables them to facilitate “their [evaluation and management] billings because they’re documenting the complexity of the patient and that helps to explain the time they’ve been with their patients,” said Bullock. Physicians will be able to compare patients’ average LOS and the charges incurred to determine how their performance compares to that of their peers.

The SOI/ROM system requires a thorough documentation process that will ultimately bring higher specificity to diagnoses and a more complete CDI program.

Editor’s note: E-mail Garrison at

E-mail Bullock at

To learn more about other issues that affect severity adjusted data, purchase a copy of HCPro’s September 18 audio conference “Severity of Illness and Risk of Mortality: Sharpen Your CDI Focus with New Measures of Success.”

Want to receive articles like this one in your inbox? Subscribe to JustCoding News: Inpatient!

Most Popular