The team found a strong relationship between the acuity of the procedures and the performance indicators. High-acuity procedures were those involving reconstruction, such as mandibulectomies and pharyngolaryngectomies. Low-acuity procedures were those that did not involve reconstruction, such as lymphadenectomies, minor glossectomies, and parotidectomies.
Explore This IssueNovember 2009
In every area, from hospital stays to infections, results were significantly better in low-acuity cases than in high-acuity cases. The acuity of the procedure was a major determinant, as it influenced these performance measures, Dr. Weber said.
Infections within 30 days, for example, occurred at a rate of 10.5% for high-acuity cases and 1.2% for low-acuity cases. Similarly, 13.1% of high-acuity patients had to be readmitted to the hospital within 30 days of the procedure, whereas just 5.4% of the low-acuity patients had to be readmitted so soon. There were two or more negative indicators in 53.4% of the high-acuity cases, but in only 4.6% of the low-acuity cases.
At M. D. Anderson, the number of high-acuity procedures went up slightly from 2004 to 2007. In 2004, there were 63 such procedures, accounting for about 13% of the total. Three years later, there were 81, accounting for about 15%.
The team also looked at the role of comorbidities. Predictably, the high-acuity patients were significantly more likely to have some comorbid condition. Those conditions-which included cardiovascular disease, diabetes, liver disease, prior heart failure, and renal disease-also had a strong relationship to the performance measures.
For every comorbid condition reviewed, except diabetes, there were significant correlations to at least a majority of the performance indicators. Cardiovascular disease and prior heart failure were the strongest signals that problems were likely.
The team also created rankings for the physicians by tallying, for each, the number of their patients with two or more negative indicators, and then balancing that with the number of their patients who had comorbid conditions.
Overall, the risk estimate for negative surgical quality factors for a high-acuity procedure versus a low-acuity procedure was 331.0-by far the most powerful determinant of a problem following a procedure.
The acuity of the procedure had the greatest impact, Dr. Weber said. You can see the hazard ratio is very high. There was also an influence by the physician ranking, more so for the high-acuity procedures than the low-acuity procedures. And the comorbid conditions also had a significant influence.
Dr. Weber said that the data gathered will be useful outside M. D. Anderson, as well as inside the institution. Extramurally, these data may serve as a comparator between providers and institutions and serve as a resources for health care purchasers and consumers, he said.
Creating Performance Quality Parameters
Not stopping there, surgeons at M. D. Anderson also turned to early-stage tongue cancer and created performance quality parameters. They then examined their own cases to see whether those parameters were being met.