Respiratory Failure: Malnutrition and Pulmonary Function are Key!


Synopsis: Mechanical ventilation resulting from acute COPD exacerbation was most closely associated with malnutrition, poor pulmonary function, and APACHE II score. Such patients also had worse pulmonary function and arterial blood gases on presentation.

Source: Vitacca M, et al. Eur Respir J 1996;9:1487-1493.

Exacerbation of copd is an important public health problem that results in significant morbidity, mortality, and use of health care resources. This is especially true in patients requiring mechanical ventilation. The source of this problem is multifactorial and therefore often difficult to study. Respiratory failure in COPD frequently results from upper and lower airway infection, concomitant heart failure, and bronchospasm. Vitacca et al, in a report from the European Respiratory Journal, evaluated bedside respiratory physiological parameters and admission laboratory data obtained in combination with nutritional indices to predict need for mechanical ventilation in COPD patients.

Thirty-nine consecutive patients admitted to a respiratory unit with an established diagnosis of COPD (American Thoracic Society definition) were studied, although patients with pneumonia, other co-morbid medical illnesses, and those unable to perform respiratory maneuvers were excluded. The group was over the age of 60, and 55% were on home oxygen.

Patients were defined as failing medical therapy if they required either intubation or noninvasive mechanical ventilation. Medical failures were malnourished, had reduced body weight, and a higher Nutritional Prognostic Index (NPI). (See Table.) In addition, measures of respiratory function and gas exchange were also more impaired with higher PaCOs, lower saturations (82% vs 89%), and reduced FVC (37% vs 58%), and FEV1 (26% vs 41%). The APACHE II score integrating both acute physiological impairment and chronic health status was also quite elevated in those patients requiring mechanical ventilation (20 vs 12). Respiratory mechanic measures demonstrated significantly elevated work of breathing and auto PEEP at the time of admission. Using a discriminate analysis, the NPI, APACHE II, and FEV1/FVC ratio showed the greatest ability to distinguish between successful or unsuccessful medical therapy. All told, NPI and FVC could correctly predict outcome in 76% of patients. Six patients died, two with severe pneumonia and two with multi-organ failure. The duration of a trial of unsuccessful medical therapy was 26 hours on average. Noninvasive ventilation was implemented for an average of 4 ± 2 days compared to 12 ± 5 days for standard mechanical ventilation.

Admission findings of COPD patients

Medical Medical
Failure Success P value
% ideal weight 86 109 < 0.01
APACHE II 20 12 < 0.001
SaO2% 82 89 < 0.001
FVC% 37 53 < 0.001
Work of breathing 1.74 1.22 < 0.001

Adapted from: Vitacca M, et al. Eur Respir J 1996;9:1487-1493.


What does this trial tell us? First, the sicker the patient on presentation, the more likely the patient will be to have a complicated course. The patients in the "failure" group had very high APACHE II scores, low vital capacities, and impaired arterial blood gases. They were significantly more malnourished by physical examination and biochemical measures. Such patients need to be managed in the intensive care unit from the start. Since there were six deaths out of 14 patients, an intensive approach might have reduced ultimate mortality, perhaps even in a less intensive and probably a less costly alternative to ICU, like a respiratory care unit or specialty pulmonary floor. While the differences between success and failure were stark, many patients will fall into a gray zone where clinical judgment must prevail.

Several other points regarding respiratory failure in COPD are worth noting. While the authors excluded pneumonia, cancer, and heart failure in this study, COPD is often a significant co-morbid complication in other medical illnesses. Influenza and pneumococcal vaccination, which are underused in eligible patients, may play a role in limiting the number and severity of these episodes. Second, early antibiotic treatment directed at H. influenza, Moraxella catarrhalis, and S. pneumonia are also potential cost-effective strategies that may limit progression to pneumonia and multi-organ failure. Additionally, "asthma" complicates many of these episodes, and the use of moderate doses of corticosteroids early in the episode may decrease the need for mechanical ventilation and shorten length of stay.

In summary, identifying those COPD patients who will likely need intubation does not appear to be difficult at the bedside if nutritional status, pulmonary function, and physiological impairment are taken into account. The difficulty will be to provide scientifically validated clinical strategies to limit the number and duration of this all too common problem.