Should DNI Equal DN-NPPV?

Abstract & Commentary

Synopsis: This prospective multicenter observational study demonstrates that patients with do-not-intubate (DNI) status who receive non-invasive positive pressure ventilation (NPPV) for respiratory failure have high hospital mortality. Presence of cough, being awake, or having a primary diagnosis of COPD or CHF, is associated with improved outcome.

Source: Levy M, et al. Crit Care Med. 2004;32(10): 2002-2007.

This study’s objectives were to describe the percentage of patients receiving NPPV who have DNI orders, the outcomes in such patients, and the utility of primary diagnosis and clinical observation in predicting these outcomes. All patients receiving NPPV at 4 New England hospitals were screened; patients with a written DNI order were included in the study. Those receiving continuous positive airway pressure (CPAP) alone were excluded. Patient data were gathered by respiratory therapists, who also administered the NPPV. Recorded data included age, admitting diagnoses, patient location, timing of DNI order, arterial blood gas, mask type, NPPV settings, cough (absent, weak or strong), secretions (present or absent) and mental status (not awake, awake or agitated). Measured outcomes were duration of NPPV, survival to hospital discharge and discharge placement. Usual statistical analyses were performed with stepwise logistic regression approach to identifying predictors of outcomes.

During a 10-month period, 114 patients with DNI orders (2 of whom had requested comfort measures only) received NPPV. These patients represented 9.4% of all patients receiving NPPV. (Only 20.2% of these patients had DNI orders in place prior to admission.) Median age of the study patients was 78 years and 37% were male. All had hypercapnic respiratory failure: some also had a component of hypoxemia. The 3 most common primary diagnoses were chronic obstructive pulmonary disease (COPD), congestive heart failure (CHF), and pneumonia. The majority of patients (67.5%) were treated on general medical wards. Oronasal masks were used for 97.4% of patients and the mean applied airway pressures were 13 ± 3 cm H2O during inspiration and 5.3 ± 1.2 cm H2O during exhalation. Most patients (83.3%) were judged to tolerate NPPV and mean use was 13.2 ± 2.4 hours. No serious complications (skin ulceration, aspiration) occurred. Overall, 49 patients (43%) survived to hospital discharge: of these, 17 of these went home. A primary diagnosis of COPD or CHF was associated with the highest survival (50-70% compared to < 30% for other diagnoses) and likelihood of discharge to home. Presence of cough and being awake were associated with survival but the presence of secretions was not.

Comment by Saadia R. Akhtar, MD, MSC

Levy and colleagues’ study has several limitations, including its observational nature, the potential biases introduced by having the same respiratory therapists collect study data and administer the NPPV, absence of a specific definition of DNI (was NPPV included in the discussion?) and the lack of clear measures of NPPV tolerance, patient and family satisfaction, and quality of life (before, during and after the experience). Despite these issues, I believe this work is an important contribution to the critical care community’s continuing efforts to define the most appropriate applications for NPPV. It expands the very limited literature on use of NPPV for patients with DNI status and raises a number of valuable points.

The results of this study reinforce some things that are already known, such as that NPPV is effective and indicated in respiratory failure secondary to COPD exacerbation and that it is more effective in patients who are awake.1,2 The study adds to data suggesting that NPPV may be beneficial for CHF and that it may be more useful for hypercapnic rather than purely hypoxemic acute respiratory failure.3,4 It reminds us of (but does not try to answer) the ethical questions surrounding NPPV use in patients with DNI status and acute respiratory failure due to irreversible conditions: does NPPV prolong suffering or offer comfort for patients who are dying?

Finally, perhaps the most interesting and important finding in Levy et al’s report is the fact that only about 20% of study patients had a DNI order in place prior to the current admission for acute respiratory failure. Providing the best possible care for our patients means giving them the information necessary to understand their health status, prognosis and care options as well as offering time and opportunities for open discussion so that decisions about level of care may be made before acute crises occur. The study findings suggest that health care providers must make a much greater effort to have such conversations with patients. Furthermore, they must explicitly describe and address NPPV, with inclusion of data from studies such as this.

For patients who are willing to consider NPPV but not endotracheal intubation, I suggest using NPPV for standard indications: acute respiratory failure secondary to COPD or (as a second-line after CPAP) cardiogenic pulmonary edema. Otherwise, until there is a large, well-done, randomized, controlled trial of NPPV in patients with DNI status and acute respiratory failure which includes usual medical as well as quality of life measures, the only firm recommendation I can make is to encourage thoughtful, well-informed decision-making at all levels.

Saadia R. Akhtar, MD, MSc, Pulmonary and Critical Care Medicine Yale University School of Medicine, is Associate Editor for Critical Care Alert.


1. Conti G, et al. Intens Care Med. 2002;28(12): 1701-1707.

2. Nava S, Ceriana P. Respir Care. 2004;49(3):295-303.

3. Nava S, et al. Am J Respir Crit Care Med. 2003. 168(12):1432-1437.

4. Celikel T, et al. Chest. 1998. 114(6):1636-1642.