Current issues of ACP Journal Club are published in Annals of Internal Medicine


Association between β-agonist use and life-threatening asthma was not altered by asthma severity

ACP J Club. 1993 Nov-Dec;119:88. doi:10.7326/ACPJC-1993-119-3-088

Source Citation

Ernst P, Habbick B, Suissa S, et al. Is the association between inhaled β-agonist use and life-threatening asthma because of confounding by severity? Am Rev Respir Dis. 1993 Jul 1;148:75-9.



To determine if the association between β-agonist bronchodilator use and life-threatening asthma is confounded by asthma severity.


Case-control study of patients drawn from a cohort who were prescribed medication for asthma.


Saskatchewan, Canada.


784 patients (129 case patients and their 655 control patients) from the matched case-control analysis of 12 301 patients participating in the Saskatchewan Asthma Epidemiology Project. Patients were between 5 and 54 years old and had been dispensed ≥ 10 prescriptions for asthma. Case patients had fatal asthma (FA) or near-fatal asthma (NFA) and control patients were receiving asthma medication without FA or NFA. Controls were matched for region, receipt of social assistance, age, date of entry, and hospitalization.

Assessment of risk factors

Precipitating factors, presenting respiratory symptoms and their duration, signs of a life-threatening asthma attack, lung function, arterial blood gas determinations, and past medical history were extracted from hospital charts and questionnaires completed by the attending physician.

Main outcome measures

FA and NFA (intubation, mechanical ventilation, or a PCO2 > 45 mm Hg during a hospitalization for asthma).

Main results

The analysis was restricted to 75 matched sets for which a hospital record was abstracted and to 108 matched sets for which a physician questionnaire was obtained. For the 108 questionnaire sets, a history of loss of consciousness or seizures during an acute asthma attack (odds ratio [OR] 10.2, 95% Cl 3.9 to 26.7) and a history of severe asthma attacks induced by food (OR 5.1, Cl 2.4 to 11.1) were associated with a life-threatening episode of asthma. For 28 matched sets with blood gas data, respiratory acidosis or hypercarbia (OR 3.1, Cl 1.3 to 7.6) was associated with a life-threatening event. Using 75 matched sets with information from hospitalizations for asthma, each unit increase in the severity score (OR 1.4, Cl 1.2 to 2.6) was associated with a lifethreatening event. After adjustment for these clinical markers of asthma severity, the OR for asthma death or near-death per metered-dose of fenoterol (OR 2.5, Cl 1.7 to 3.8) or albuterol (OR 2.0, Cl 1.5 to 2.7) remained significant.


Markers for asthma severity that could be abstracted from charts or obtained from physician questionnaires did not alter the association between β-agonist use and the risk for death or near-death from asthma.

Source of funding: Boehringer Ingelheim Pharmaceuticals.

For article reprint: Dr. P. Ernst, McGill University, Respiratory Epidemiology Unit, Department of Epidemiology and Biostatistics, 1110 Pine Avenue West, Montreal, Quebec H3A 1A3, Canada. FAX 514398-8981.


The study by Ernst and colleagues is a follow-up to the original Saskatchewan Asthma Epidemiology Project. Commentators on the previous report questioned whether the link between increased β-agonist use and NFA or FA was possibly caused by increased severity of asthma, rather than by an independent toxic effect of the β-agonist. This study explores this issue by examining the actual patient records of the participants. The absence of objective measurements, such as spirometry or blood gas results, in most participants reflects previous and, in many instances, continued poor evaluation of patients with acute asthma (1) and weakens the results of the current study.

The results show that differences exist in the underlying severity of asthma between case patients and control patients but that these differences do not explain the relation between excess β-agonist use and NFA and FA events. This study also highlights 2 additional prognostic markers for patients with asthma. The previous occurrence of a syncopal episode in association with an acute asthma attack or attacks related to food was associated with a poor prognosis. Based on a review of clinical records, the food reactions did not appear to be anaphylactoid in most patients.

Another study using the same database (2) found that the prescription of at least 1 inhaler of beclomethasone per month for the previous 12 months appears to protect a patient from having a NFA or FA event. This adds to the accumulating evidence supporting the use of regular anti-inflammatory therapy with symptomatic bronchodilator therapy as a rescue medication for breakthrough symptoms.

J. Mark FitzGerald, MD
University of British ColumbiaVancouver, British Columbia, Canada


1. Fitzgerald JM, Hargreave FE. Emergency department management of acute asthma and prospective evaluation of outcome. Can Med Assoc J. 1990;142: 591-5.

2. Ernst P, Spitzer Wo, Suissa S, et al. Risk of fatal and near fatal asthma in relation to inhaled corticosteroid use. JAMA. 1992;268:3462-4.