Are Benchmarks the Problem for High VAP Rates?

CHICAGO – Payers are relying ever more on tying physician and hospital payments to quality measures, but what happens if the benchmarks they use vary among institutions or are flawed? Such may be the case with ventilator-associated pneumonia and large trauma centers, as a recent study shows VAP rates at such facilities exceed national benchmarks, which some say are inadequate for comparison.

Dr. Christopher P. Michetti of Inova Fairfax Hospital in Falls Church, Va., presented a retrospective study designed to determine VAP rates at major trauma centers and to lay groundwork for more accurate benchmarking that relies less on National Health Safety Network data, he said. He spoke at the annual meeting of the American Association for Surgery in Trauma. The study was performed through the AAST Multi-Institutional Trials Committee.

"Hospitals are under pressure to reduce their VAP rates, yet a direct association between VAP rates and quality of care or outcomes has not been demonstrated," he remarked.

"VAP rates ...are remarkably variable," Dr. Michetti said. "It is not appropriate to measure all trauma centers against a single benchmark, nor does an actual benchmark appear to exist at this point." Comparing VAP rates between different trauma centers is "like comparing apples and oranges," he said.

The study looked at VAP rates at 47 level I and II trauma centers for 2008 and 2009 with an average of 3,000 trauma evaluations a year. The average VAP rate for the study group was 17.2/1,000 ventilator days, compared with 8.1/1,000 for NHSN data. "In fact, the 90th-percentile rate for NHSN was still below the mean rate from our study," Dr. Michetti said. Across all 47 centers in the study, VAP rates ranged from a low of 1.8/1,000 ventilator days to a high of 57.6/1,000 ventilator days.

The case mix at the trauma centers did not auger well for lower VAP rates, as 88% of the cases were blunt trauma, Dr. Michetti noted. "VAP rates are generally higher for blunt-trauma patients, at about 17/1,000 ventilator days, compared with penetrating trauma at 11/1,000," he said.

Most other variables among the centers in the study – such as having a closed or open ICU, or using a bacteriologic vs. a clinical strategy to diagnose VAP – showed little impact on the pneumonia rates. VAP rates did not correlate with the size or level of trauma center, injury severity or type of ICU, he said.

Among the problems he noted with the NHSN data on VAP rates are the lack of source hospital identification, population risk, or injury severity stratification. "In addition, the NHSN rates are substantially lower than other published rates among trauma patients," he said.

However, the investigators did isolate a few variables that may influence VAP rates: Among centers where the trauma service alone made the diagnosis, the average VAP rate was 27.5/1,000 ventilator days. When the infection control, quality, or epidemiology department made the call, the average VAP rate was 11.9/1,000 days. Centers that excluded patients also had rates about 30% lower than those that did not. This variability raises questions about using VAP as a quality measure, Dr. Michetti said. "Before we take that leap, diagnostic and reporting standards are necessary."

The heightened attention on VAP as a quality measure for critical care is having other implications, he said. "As pressure to reduce VAP rates grows, an increasing number of patients are being labeled as having ventilator-associated tracheobronchitis or excluded for reasons such as aspiration," he said.

Discussant Dr. Karen J. Brasel of the Medical College of Wisconsin in, Milwaukee, acknowledged the need for the study, but raised the question: "Are the benchmarks the problem, or are we the problem?"

"I think the answer is yes, both," Dr. Michetti said. "I’m not sure that an adequate benchmark exists probably because no representative sample of trauma centers has been done to set that benchmark." He noted that the CDC’s Healthcare Infection Control Practices Advisory Committee does not recommend reporting of VAP, which argues against using that as a benchmark. Meanwhile, across individual centers no reporting standards exist, "so centers can’t agree on what is VAP," Dr. Michetti said.

Dr. Michetti had no disclosures, and the study received no outside funding.

CHICAGO – Payers are relying ever more on tying physician and hospital payments to quality measures, but what happens if the benchmarks they use vary among institutions or are flawed? Such may be the case with ventilator-associated pneumonia and large trauma centers, as a recent study shows VAP rates at such facilities exceed national benchmarks, which some say are inadequate for comparison.

Dr. Christopher P. Michetti of Inova Fairfax Hospital in Falls Church, Va., presented a retrospective study designed to determine VAP rates at major trauma centers and to lay groundwork for more accurate benchmarking that relies less on National Health Safety Network data, he said. He spoke at the annual meeting of the American Association for Surgery in Trauma. The study was performed through the AAST Multi-Institutional Trials Committee.

"Hospitals are under pressure to reduce their VAP rates, yet a direct association between VAP rates and quality of care or outcomes has not been demonstrated," he remarked.

"VAP rates ...are remarkably variable," Dr. Michetti said. "It is not appropriate to measure all trauma centers against a single benchmark, nor does an actual benchmark appear to exist at this point." Comparing VAP rates between different trauma centers is "like comparing apples and oranges," he said.

The study looked at VAP rates at 47 level I and II trauma centers for 2008 and 2009 with an average of 3,000 trauma evaluations a year. The average VAP rate for the study group was 17.2/1,000 ventilator days, compared with 8.1/1,000 for NHSN data. "In fact, the 90th-percentile rate for NHSN was still below the mean rate from our study," Dr. Michetti said. Across all 47 centers in the study, VAP rates ranged from a low of 1.8/1,000 ventilator days to a high of 57.6/1,000 ventilator days.

The case mix at the trauma centers did not auger well for lower VAP rates, as 88% of the cases were blunt trauma, Dr. Michetti noted. "VAP rates are generally higher for blunt-trauma patients, at about 17/1,000 ventilator days, compared with penetrating trauma at 11/1,000," he said.

Most other variables among the centers in the study – such as having a closed or open ICU, or using a bacteriologic vs. a clinical strategy to diagnose VAP – showed little impact on the pneumonia rates. VAP rates did not correlate with the size or level of trauma center, injury severity or type of ICU, he said.

Among the problems he noted with the NHSN data on VAP rates are the lack of source hospital identification, population risk, or injury severity stratification. "In addition, the NHSN rates are substantially lower than other published rates among trauma patients," he said.

However, the investigators did isolate a few variables that may influence VAP rates: Among centers where the trauma service alone made the diagnosis, the average VAP rate was 27.5/1,000 ventilator days. When the infection control, quality, or epidemiology department made the call, the average VAP rate was 11.9/1,000 days. Centers that excluded patients also had rates about 30% lower than those that did not. This variability raises questions about using VAP as a quality measure, Dr. Michetti said. "Before we take that leap, diagnostic and reporting standards are necessary."

The heightened attention on VAP as a quality measure for critical care is having other implications, he said. "As pressure to reduce VAP rates grows, an increasing number of patients are being labeled as having ventilator-associated tracheobronchitis or excluded for reasons such as aspiration," he said.

Discussant Dr. Karen J. Brasel of the Medical College of Wisconsin in, Milwaukee, acknowledged the need for the study, but raised the question: "Are the benchmarks the problem, or are we the problem?"

"I think the answer is yes, both," Dr. Michetti said. "I’m not sure that an adequate benchmark exists probably because no representative sample of trauma centers has been done to set that benchmark." He noted that the CDC’s Healthcare Infection Control Practices Advisory Committee does not recommend reporting of VAP, which argues against using that as a benchmark. Meanwhile, across individual centers no reporting standards exist, "so centers can’t agree on what is VAP," Dr. Michetti said.

Dr. Michetti had no disclosures, and the study received no outside funding.

CHICAGO – Payers are relying ever more on tying physician and hospital payments to quality measures, but what happens if the benchmarks they use vary among institutions or are flawed? Such may be the case with ventilator-associated pneumonia and large trauma centers, as a recent study shows VAP rates at such facilities exceed national benchmarks, which some say are inadequate for comparison.

Dr. Christopher P. Michetti of Inova Fairfax Hospital in Falls Church, Va., presented a retrospective study designed to determine VAP rates at major trauma centers and to lay groundwork for more accurate benchmarking that relies less on National Health Safety Network data, he said. He spoke at the annual meeting of the American Association for Surgery in Trauma. The study was performed through the AAST Multi-Institutional Trials Committee.

"Hospitals are under pressure to reduce their VAP rates, yet a direct association between VAP rates and quality of care or outcomes has not been demonstrated," he remarked.

"VAP rates ...are remarkably variable," Dr. Michetti said. "It is not appropriate to measure all trauma centers against a single benchmark, nor does an actual benchmark appear to exist at this point." Comparing VAP rates between different trauma centers is "like comparing apples and oranges," he said.

The study looked at VAP rates at 47 level I and II trauma centers for 2008 and 2009 with an average of 3,000 trauma evaluations a year. The average VAP rate for the study group was 17.2/1,000 ventilator days, compared with 8.1/1,000 for NHSN data. "In fact, the 90th-percentile rate for NHSN was still below the mean rate from our study," Dr. Michetti said. Across all 47 centers in the study, VAP rates ranged from a low of 1.8/1,000 ventilator days to a high of 57.6/1,000 ventilator days.

The case mix at the trauma centers did not auger well for lower VAP rates, as 88% of the cases were blunt trauma, Dr. Michetti noted. "VAP rates are generally higher for blunt-trauma patients, at about 17/1,000 ventilator days, compared with penetrating trauma at 11/1,000," he said.

Most other variables among the centers in the study – such as having a closed or open ICU, or using a bacteriologic vs. a clinical strategy to diagnose VAP – showed little impact on the pneumonia rates. VAP rates did not correlate with the size or level of trauma center, injury severity or type of ICU, he said.

Among the problems he noted with the NHSN data on VAP rates are the lack of source hospital identification, population risk, or injury severity stratification. "In addition, the NHSN rates are substantially lower than other published rates among trauma patients," he said.

However, the investigators did isolate a few variables that may influence VAP rates: Among centers where the trauma service alone made the diagnosis, the average VAP rate was 27.5/1,000 ventilator days. When the infection control, quality, or epidemiology department made the call, the average VAP rate was 11.9/1,000 days. Centers that excluded patients also had rates about 30% lower than those that did not. This variability raises questions about using VAP as a quality measure, Dr. Michetti said. "Before we take that leap, diagnostic and reporting standards are necessary."

The heightened attention on VAP as a quality measure for critical care is having other implications, he said. "As pressure to reduce VAP rates grows, an increasing number of patients are being labeled as having ventilator-associated tracheobronchitis or excluded for reasons such as aspiration," he said.

Discussant Dr. Karen J. Brasel of the Medical College of Wisconsin in, Milwaukee, acknowledged the need for the study, but raised the question: "Are the benchmarks the problem, or are we the problem?"

"I think the answer is yes, both," Dr. Michetti said. "I’m not sure that an adequate benchmark exists probably because no representative sample of trauma centers has been done to set that benchmark." He noted that the CDC’s Healthcare Infection Control Practices Advisory Committee does not recommend reporting of VAP, which argues against using that as a benchmark. Meanwhile, across individual centers no reporting standards exist, "so centers can’t agree on what is VAP," Dr. Michetti said.

Dr. Michetti had no disclosures, and the study received no outside funding.