How Quality Improvement Links to Real Change in Patient Care

The era of accountable care and pay for performance is here, and physicians will have to embrace these novel reimbursement models. In a plenary session, Rubin Cohen, MD, FCCP, a member of the American College of Chest Physician's quality improvement committee, discussed the relationship between quality improvement and outcomes.

In a plenary session at the 2014 meeting of the American College of Chest Physicians (ACCP) entitled “Linking Payment to Performance: What Do the Data Tell Us?”, Rubin Cohen, MD, FCCP, a member of the ACCP quality improvement committee, discussed the relationship between quality improvement and outcomes.

Based on healthcare data collected in 2012 from the Organisation for Economic Co-operation and Development, says Dr Cohen, the United States spends more than other countries with similar market economies—approximately “two and a half times the average for other developed countries”—on healthcare.

The cost of healthcare in the United States has risen dramatically since 1945. If the cost of food had climbed at the same rate as the cost of healthcare between 1945 and today, a dozen eggs would cost $55, a dozen oranges would cost $134, and a gallon of milk would cost $48. Although these are probably unfair comparisons, as healthcare is different from farming, Dr Cohen noted, “This brings it home how much costs have increased.”

In spite of the high spending on healthcare in the United States, residents other industrialized nations that spend less on healthcare have a higher life expectancy. One major reason for the higher cost of healthcare in the United States is regional disparities in care. For instance, patients in the city of Elyria, Ohio, receive 3 times as many angioplasty procedures as patients in Cleveland, which is only 30 miles away. Likewise, the number of days spent in hospitals during the last 6 months of life varies across academic medical centers throughout the country. For example, the number of in-hospital days during end-of-life care at NewYork-Presbyterian Hospital is 24 days on average, but at UCSF Medical Center in San Francisco, the average is 13 days.

In a study published in the Journal of the American Medical Association in 2003, Jencks et al measured changes in the quality of care delivered to Medicare beneficiaries. As part of the analysis, investigators graphed the amount of money spent on care versus quality, and found that the places with the highest healthcare costs also had the lowest quality of care.1

According to Dr Cohen, the main reason for this relationship is a lot of unwanted variations between medical systems. Although it is impossible to turn all of medicine into a series of protocols, reducing unwanted variability is a reasonable goal. CMS, by far the largest payer in the United States, is incentivizing performance by penalizing hospitals that do not publicly report quality data. Another method of incentivizing performance is the accountable care organization model, which rewards healthcare professionals for meeting certain preestablished targets for care.

The Hospital Readmissions Reduction Program (HRRP) is one vehicle for CMS to impose financial penalties on hospitals. Currently, the HRRP program penalizes hospitals for substandard outcomes in terms of pneumonia readmissions, and readmissions for congestive heart failure, myocardial infarction, and (as of October 2014) chronic obstructive pulmonary disease.

Paying for performance, however, does not always improve care. For example, when CMS stopped reimbursement for central line infections in 2008, the rate of central line infections appeared to drop precipitously by 2009. “That's not what happened,” stated Dr Cohen. Because physicians could no longer receive reimbursement for central line infections, physicians no longer took blood cultures in patients with a central line. In other words, the physicians stopped detecting central line infections because identifying such an infection would lead to a lack of payment.

In 1996, CMS instituted a heart failure tracking program. Hospitals were required to submit performance metrics on left ventricular function, use of certain medications, disease counseling, and smoking cessation. It was expected that quality would improve and mortality rates would change once the program was implemented. This expectation was never met. In 2009, Fonarow and colleagues analyzed the data and found no change in 30-day mortality—including adjusted and unadjusted mortality rates. By all other measures, however, the program was a success. Use of angiotensin-converting enzyme inhibitors increased, left ventricular dysfunction was measured more often, but no changes in mortality occurred.2

In a similar study by Lee et al published in the New England Journal of Medicine in 2012, investigators measured rates of in-hospital pneumonia. After incentives were put in place, the authors found no significant difference in pneumonia rates. Nonpayment for pneumonia did not affect quality of care.3

In a meta-analysis of CMS measures, a single measure was found to be supported by high-quality evidence: influenza vaccination. The rest of the measures were logical and sensible, noted Dr Cohen, but were not supported by a high level of evidence. The difference between improving actual outcomes and improving process measures is that processes may not affect patient outcomes. Some proposed solutions to this disconnect include increasing incentives. However, a 0.5% or 1% increase in reimbursement may not be enough money to stimulate change.4

Importantly, although outcomes may not be changing clearly, culture change is occurring. For instance, incentives are prompting nurses to pay more attention to reducing the amount of time that patients spend with a central line inserted.

The era of accountable care and pay for performance is here, and physicians will have to embrace these novel reimbursement models, according to Dr Cohen. New strategies for improving outcome measures—not just process measures—will be necessary as quality performance measurement techniques and quality improvement interventions continue to evolve.


  1. Jencks SF, Huff ED, Cuerdon T. Change in the quality of care delivered to Medicare beneficiaries, 1998-1999 to 2000-2001. JAMA. 2003;289(3):305-312.
  2. Fonarow GC, Peterson ED. Heart failure performance measures and outcomes: real or illusory gains. JAMA. 2009;302(7):792-794.
  3. Lee GM, Kleinman K, Soumerai SB, et al. Effect of nonpayment for preventable infections in U.S. hospitals. N Engl J Med. 2012;367(15):1428-1437.
  4. Wilson KC, Schünemann HJ. An appraisal of the evidence underlying performance measures for community-acquired pneumonia. Am J Respir Crit Care Med. 2011;183(11):1454-1462.