Guest Post: AMI 30-day mortality measures and CDI – Is your approach ischemic?

by Shannon Newell, RHIA, CCS

The 30-day all cause acute myocardial infarction (AMI) mortality outcome measure has been linked to hospital payments since the inception of the Hospital Value-Based Purchasing Program (HVBP) in fiscal year 2013. In February 2016, CMS announced that 70% of commercial payers agreed to use the measure as one of the cardiology outcomes linked to payment.

The Medicare Episode Payment Bundle for AMIs, proposed to begin in July 2017, will further link reimbursement to measure performance. Traditional CDI approaches to strengthen data quality in the name of risk of mortality are not sufficient enough to capture the documentation needed under the CMS mortality risk adjustment methodology.

CMS includes any discharges with the principal diagnosis of an AMI in the mortality outcome measure (except for a few exclusions outlined in the methodology, such as discharged against medical advice). The final group of included discharges is referred to as the “cohort.” Consider the following case studies:

Case Study 1

• Scenario:
  
  • Admitted with severe unstable angina and diaphoresis with the finding of ST elevation in the anterior leads on an EKG with the initial set of cardiac enzymes being negative
  • Because of the suspicion of acute transmural myocardial infarction (MI), receives TPA infusion
  • Subsequent cardiac enzymes are negative, and at the time of discharge, the physician documents an aborted transmural MI
• Principal diagnosis:
  
  • The principal diagnosis would be unstable angina because the AMI was aborted
  • If coronary artery disease is documented, coding guidelines would require selection of this condition as the principal diagnosis
• AMI mortality cohort:
  
  • This discharge would be excluded from the AMI mortality cohort with the correct application of coding guidelines for principal diagnosis selection

Case Study 2

• Scenario:
  
  • Admitted with subendocardial MI with acute systolic heart failure and cardiogenic pulmonary edema
  • Treated with supplemental oxygen and IV Lasix therapy, and because of the patient’s age, the MI is treated conservatively
• Principal diagnosis: Two principal diagnosis options exist:
  
  • Option 1: Acute systolic heart failure
  • Option 2:  Subendocardial MI
• AMI mortality cohort:
  
  • Coding guidelines permit the selection of either option as the principal diagnosis
  • Today’s CDI teams typically select the option which will result in the highest MS-DRG relative weight, which will result in attribution of the discharge to the AMI mortality cohort in this case

Once a hospital claim goes out the door with the principal diagnosis of AMI, CMS analyzes claims databases to determine if that Medicare beneficiary died within 30 days of discharge (for any reason). If a mortality is identified, CMS attributes the death to the hospital that reported a claim with AMI as the principal diagnosis within 30 days preceding the death.

Each discharge included in the AMI mortality cohort is then risk adjusted to determine the expected rate of mortality. The CMS methodology uses CCs/MCCs, not hierarchical condition categories, to determine the likelihood of death for each discharge.

• Each CC is comprised of related ICD-10 codes
• There are a variety of CCs which affect AMI mortality risk adjustment
  
  • These categories are weighted; some have a more significant effect on risk adjustment than others
  • 12 categories drive 90% of strengthened risk adjustment opportunity
  • Only about 25% of these conditions serve as MS-DRG CCs/MCCs
• CMS looks for the presence of these conditions in the AMI admission as well as all Part A and (face to face) Part B claims for the 12 months prior to that admission
• Some of the conditions must be documented prior to the patient’s admission for AMI management or there is no “credit” for risk adjustment

Performance for the CMS 30-day AMI mortality measure adjusts payments received for the Medicare and commercial patient population. This reimbursement is not limited to hospitals; under the Episode Payment Model other providers will be affected as well. Although this is a hospital-centric measure, meaning the measure assesses the hospital’s management quality for the AMI population, documentation and reported codes across the continuum affect the expected risk of mortality, and in turn impact measure performance.

A few questions for CDI and quality programs to consider as they shift initiatives to support value based payments:

• Does your principal diagnosis selection process consider attribution of discharges to quality measures in addition to MS-DRG relative weights?
• Does your record review process to support the capture of CCs/MCCs related to mortality only on actual deaths or on the entire denominator included in the measure?
• Are you focused on the capture of CCs/MCCs which have an effect on CMS’ mortality methodology?
  
  • The comorbidities in this algorithm differ from those used in other mortality risk adjustment methodologies commonly used in CDI programs today
  • The broad number of comorbid categories requires the CDI and quality teams focus on point of care capture
• Do the clinical documentation and reported codes in your system’s outpatient settings support the capture of comorbid conditions related to risk adjustment?

Editor’s note: Newell is a Managing Director with CCDI-DQ with extensive operational and consulting expertise in coding and clinical documentation improvement, performance improvement, case management, and health information management.