RISK FACTORS VERSUS MECHANISMS OF DEATH IN SUBJECTS WITH ASTHMA

To the Editor:

Turner and colleagues have recently reported some risk factors associated with near-fatal asthma (1). The study included smokers, yet it was well designed and well controlled. However, in the DISCUSSION section the authors compare their study to one of our studies (2) and make several statements that are not totally accurate and need clarification.

First, Turner and colleagues state that they confirm and extend the results of our study. It is important to emphasize that the underlying hypotheses tested in both studies are fundamentally different, making comparisons between the two studies impossible. The purpose of their study was to describe the risk factors associated with near-fatal asthma, while the purpose of our study was to know the mechanisms of death from asthma (2).

Second, Turner and coworkers state that we have studied relatively few patients, yet the number of mechanically ventilated subjects was similar in both studies. Irrespective of the sample size the major difference between these two studies was the inclusion criteria. While Turner included smoking (and ex-smoking) patients with severe airflow limitation showing an FEV₁ slightly lower than 1 L, the vast majority of our asthmatic subjects were unconscious and in respiratory arrest on arrival. Indeed, in two of our subjects, but in none of Turner's study, cardiac disturbances secondary to hypoxemia were detected and successfully treated during resuscitation maneuvers. This in itself points to a major difference in the severity of the asphyxia found in our subjects on arrival as compared to the subjects enrolled by Turner and colleagues. In addition, our subjects suffered from sudden, overwhelming deterioration while Turner and coworkers studied subjects with more protracted crises, and in whom the lack of electrocardiographic abnormalities was not surprising. Thus, there seem to be differences between the two populations studied.

Third, the article in question emphasized that we did not study a control group. Because the objective of our study was to find a mechanistic answer to how asthmatics die (or could die), we included patients on the verge of dying. A control group (dead subjects on arrival) would not have given the physiological and biochemical parameters needed to draw any valid conclusion. Thus, to answer our research question, a control group was considered at that moment unfeasible, impractical, and indeed not needed. On the other extreme of the spectrum, asthmatics without respiratory failure, as proposed by Turner and colleagues in their DISCUSSION section, would have shown less airflow limitation and better acid-base status and, again, would have provided irrelevant data for testing our hypothesis.

Fourth, Turner and colleagues incorrectly stated that our study was retrospective.

In conclusion, because of major differences in the hypotheses tested and some differences in the patient population, the data from Turner and coworkers do not confirmlet alone extendour results. Although Turner and coworkers have described once again some of the already-known risk factors associated with near-fatal asthma, their data do not help to ascertain the mechanisms of death from asthma.

NÉSTOR A. MOLFINO

Department of Medicine

McGill University

Montreal, Quebec, Canada

	References

1. Turner, M. O., K. Noertjojo, S. Vedal, T. Bai, S. Crump, and J. M. FitzGerald. 1998. Risk factors for near-fatal asthma: a case-control study in hospitalized patients with asthma. Am. J. Respir. Crit. Care Med. 157: 1804-1809 [Abstract/Free Full Text].

2. Molfino, N. A., L. J. Nannini, A. N. Martelli, and A. S. Slutsky. 1991. Respiratory arrest in near-fatal asthma. New Eng. J. Med. 324: 285-288 [Abstract].

From the Authors:

Our case-control study of near-fatal asthma (NFA) was designed to assess risk factors for an NFA episode by prospective identification of hospitalized patients and standardized data collection (1). Molfino and colleagues published a descriptive, case series of ten patients with eleven episodes of NFA to evaluate possible mechanisms for their respiratory failure (2). Of the 19 patients who met our a priori definition of respiratory failure, 10 required mechanical ventilation. The mean initial pH was 7.10 and Pa_CO2 83 for these 10 patients. We agree that Molfino's patients were more severe at presentation because the majority (8 of 11 episodes) were in respiratory arrest before treatment could even be initiated. However, despite the differences in study objectives and severity of patients at presentation to the hospital, there are some common observations that merit further discussion.

The duration of symptoms before hospitalization was similar in both our NFA cases and controls. In the patients reported by Molfino and colleagues, only 18% (2 of 11 episodes) presented with symptoms of 6 h or less (2). The mean duration of symptoms in his patients was 72 h. Our data support these findings and extend them by comparison with a control group. Kolbe and colleagues also found rapid onset attacks (< 6 h duration) to be an uncommon (8.5%, 27 of 316 hospital admissions) manifestation of severe asthma (3).

Our study design also allowed us to compare the use of -agonists during the period before hospitalization. In the 24 h before admission, the control patients used more short-acting -agonists than the NFA cases. Although Molfino and colleagues wanted to investigate mechanisms of death from asthma, they were unable to document patterns of utilization and dosing of short-acting -agonists before the near-fatal event. The absence of these relevant and important data led us to believe their study was retrospective.

The compliance with inhaled corticosteroid prescriptions in our NFA cases and hospitalized control subjects was uniformly poor (29 and 28%, respectively). These data emphasize that undertreatment with antiinflammatory medications may contribute to the severity of the exacerbation. However, this finding is not unique to NFA patientsa finding that does extend Molfino's conclusions (2).

Another difference between the patients described by Molfino and those in our study is that we included current and former smokers. Unfortunately, many patients with asthma have smoked or continue to smoke. In an asthma education study of outpatients, we found 22.5% to be smokers, and 41.5% to be ex-smokers (4). In a randomized, controlled trial of community-based asthma education, Garrett and colleagues (5) enrolled patients who had recently been hospitalized for a severe exacerbation of asthma and found 34% to be smokers. They had little success helping these patients stop smoking, despite a smoking-cessation component to their educational intervention (5). Since we wanted to assess risks for NFA, we would have been remiss not to explore the possibility of smoking as a significant risk factor. Therefore, our data are generalizable to all asthma patients (smokers and nonsmokers) who may present to an emergency department with a severe exacerbation.

We documented no serious cardiac arrhythmias in either cases or control subjects at the time of presentation. Our data nicely complement the observations of Molfino and colleagues in that 100% of our patients and 82% of his cases had no serious cardiac arrhythmias. The two arrhythmias observed by Molfino responded rapidly to manual ventilation with 100% oxygen (2).

In summary, our study identifies important risk factors for NFA from prospectively collected data. We confirm many of the observations from Molfino's descriptive study and extend their relevance by virtue of our case-control study design. Our data show that many near-fatal and severe exacerbations are not sudden and overwhelming, but develop gradually, therefore providing a window of opportunity to intervene therapeutically. These findings should assist physicians and other health care workers caring for asthma patients in the emergency department, hospital, and outpatient setting.

MARK O. TURNER

Department of Medicine

Respiratory Division

University of British Columbia

Vancouver, British Columbia, Canada

	References

1. Turner, M. O., K. Noertjojo, S. Vedal, T. Bai, S. Crump, and J. M. FitzGerald. 1998. Risk factors for near-fatal asthma: a case-control study in hospitalized patients with asthma. Am. J. Respir. Crit. Care Med. 157: 1804-1809 .

2. Molfino, N. A., L. J. Nannini, A. N. Martelli, and A. S. Slutsky. 1991. Respiratory arrest in near-fatal asthma. N. Engl. J. Med. 324: 285-288 .

3. Kolbe, J., W. Fergusson, and J. Garrett. 1998. Rapid onset asthma: a severe but uncommon manifestation. Thorax 53: 241-247 [Abstract/Free Full Text].

4. Turner, M. O., D. Taylor, R. Bennett, and J. M. FitzGerald. 1998. A randomized trial comparing peak expiratory flow and symptom self-management plans for patients with asthma attending a primary care clinic. Am. J. Respir. Crit. Care Med. 157: 540-546 [Abstract/Free Full Text].

5. Garrett, J., J. M. Fenwick, G. Taylor, E. Mitchell, J. Stewart, and H. Rea. 1994. Prospective controlled evaluation of the effect of a community-based asthma education centre in a multi-racial working class neighborhood. Thorax 49: 976-983 [Abstract/Free Full Text].