SAN FRANCISCO – Clinicians may have become too enthusiastic about using noninvasive ventilation in lieu of intubation and mechanical ventilation, according to a nationwide database study from the Boston University Pulmonary Center.
Noninvasive ventilation (NIV) use tripled during 2000-2009 for acute respiratory failure (ARF) caused by chronic obstructive pulmonary disease, for which there is good evidence of a mortality benefit (Cochrane Database Syst. Rev. 2004;1:CD004104).
Dr. Allan J. Walkey
But in the same period, use increased 3.4-fold for ARF caused by conditions for which there is little evidence to support its use, including asthma, pneumonia, neurologic diseases, and nonpulmonary sepsis. Currently, about half of ARF patients receiving NIV have COPD, while half have other conditions.
The investigators found that patients are more likely to fail NIV – and subsequently receive mechanical ventilation – when it’s used for those conditions instead of COPD (odds ratio, 1.12; 95% confidence interval, 1.08-1.16). Overall, 12% of COPD patients failed NIV, compared with 18% of non-COPD patients (P less than .001).
Of the study patients who failed NIV, 37% died, compared with 35% who received mechanical ventilation alone (P = .002).
The researchers analyzed approximately 11 million hospital records from 2000-2009 coded for ARF in the Agency for Healthcare Research and Quality’s Nationwide Inpatient Sample (NIS) database, excluding patients with sleep apnea.
"People might have a little too much faith in noninvasive ventilation. They see it work really well in COPD, so they think it might work well in everyone," said lead investigator Dr. Allan J. Walkey, director of pulmonary and critical care education at Boston University.
He said the study’s take-home message is to "think twice if you are going to put a noninvasive ventilator on someone without COPD or acute cardiogenic pulmonary edema," for which there is also strong evidence of benefit (Cochrane Database Syst. Rev. 2008;3:CD005351). In other patients, "there’s generally not good evidence to support use of noninvasive ventilation, and [our study suggests] it may be associated with worse outcomes. [Those patients] need to be monitored closely, and if they are showing signs of failure, they need to be intubated," he said at an international conference of the American Thoracic Society.
It’s unclear why NIV might be problematic for some. Perhaps not being able to suction secretions out of an endotracheal tube leads to problems in pneumonia. Maybe fading sensorium in other conditions increases the likelihood of gastric aspiration if there’s no tube in place to protect the airway, Dr. Walkey said.
Although the NIS database does not include vital signs, lab reports, and other patient-level clinical data, the researchers used an algorithm to assess and control for disease severity. "There is growing evidence that the adjustment of disease severity using data you do have in large databases" – such as information on comorbidities, billing codes, and demographics – "is actually as accurate as those based on disease severity scores," Dr. Walkey said (Crit. Care Med. 2011;39:2425-30).
The researchers also found wide regional variations in NIV use, with the heaviest use in the Northeast and the lowest in the Midwest; use was roughly equal in teaching and nonteaching hospitals.
Insurance claims for acute respiratory failure increased from 818,781 in 2000 to 1,531,352 in 2009, with an associated 25% decrease in ARF mortality over that period.
Dr. Walkey said he had no financial disclosures.