Efficacy of noninvasive ventilation and hospital outcomes of acute chronic obstructive pulmonary disease in the ICU: a question of volume alone?

Critical Care Medicine www.ccmjournal.org e249 Did physician staffing in some ICUs consist primarily of specialist physicians (critical care and/pulmonary) and others without or with few specialty physicians, possibly contributing to a lower propensity to use NIV? We would think physician staffing would have had more of an influence on NIV utilization than the number of AECOPD seen by any hospital. If the expertise does not exist at a center, that treatment would not be provided, irrespective of the number of patients seen with that condition. Fourth, they should have been able to determine if there were individual differences in the use of NIV among their categories of low-, medium-, and high-volume ICUs. They should have performed subgroup analysis on the use of NIV within their categories of ICUs. One would expect the ICUs with the highest use of NIV to be in the high-volume group and lowest use in the low-volume group. If sites with high use were in the low-volume group and vice versa, this would invalidate their hypothesis. Finally, better characterization of the severity of illness is essential for proper extrapolation of their results (4). If the severity of illness such as initial pH is not factored in analysis, comparisons based on patient volume are skewed as comparable patients may not be included. The corollary to this involves admission criteria between ICUs and whether this changed during the observation period. The location of treatment may have also changed since NIV use may have expanded beyond the ICU to other beds such as emergency units or respiratory care units, not considered ICU beds, but with NIV expertise (1–4). Differences between sites would skew their findings about high-, medium-, and low-use ICUs. The authors have disclosed that they do not have any potential conflicts of interest.