The authors concluded that salmeterol reduced withdrawal and exacerbations and improved lung function and health status over 1 year compared with placebo/usual care in chronic obstructive pulmonary disease. The conclusions appear to follow from the results presented, but the limited search and lack of reporting of review methods mean it is difficult to assess the reliability of these conclusions.

To determine the effect of salmeterol on a number of clinically relevant outcomes in patients with chronic obstructive pulmonary disease (COPD).

MEDLINE was searched for relevant papers; the search terms were reported. The GlaxoSmithKline clinical trial tracking system was also searched using the same search terms. Studies that were completed by January 2007 were eligible for inclusion.

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) that were double-blind, parallel-group, placebo-controlled and of at least 12 weeks' duration were eligible for inclusion.

Specific interventions included in the review

Studies that compared salmeterol (50 microg twice a day) with placebo (with or without usual therapy) were eligible for inclusion. In addition to study medication and placebo, usual COPD therapy was continued in most of the included trials. Antibiotics for acute exacerbations were allowed in all trials. Short-acting bronchodilators were allowed in all trials but not in combination with other drugs (except in 1 trial). Regular treatment with oral and/or inhaled corticosteroids and methylxanthines was allowed in the majority of the included trials. Anticholinergic therapy was allowed in 4 included trials. Salmeterol was administered by pressurised metered dose inhaler or Diskus/Accuhaler dry powder inhaler. The duration of the intervention was between 12 and 52 weeks.

Participants included in the review

Studies of non-asthmatic adults with stable COPD and no recent infections, exacerbations or hospitalisations in the previous 4 weeks were eligible for inclusion. Patients with other severe conditions (such as cardiac, liver and renal disease) were excluded. The participants in the included studies were aged between 35 and 79 years (mean 63.8 years), and 72% were male. Most of them had a diagnosis of COPD for 5 or more years. The mean baseline forced expiratory volume in 1 second (FEV1) was 1.29 L, representing 45.2% of the predicted value, and the average baseline reversibility was 5.96% of predicted.

Outcomes assessed in the review

Studies assessing withdrawal rates, moderate-to-severe exacerbations, pre-bronchodilator FEV1 or health status were eligible for inclusion. The included studies assessed health status using St George's Respiratory Questionnaire or the Chronic Respiratory Disease Questionnaire. The outcomes were assessed at 3, 6 and 12 months.

How were decisions on the relevance of primary studies made?

The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.

The authors did not state that they assessed validity.

The authors did not state how the data were extracted for the review, or how many reviewers performed the data extraction.

The data were extracted separately for the intention-to-treat (ITT) population (defined as all randomised participants taking at least 1 dose of study medication) and the poorly reversible population (defined as participants with reversibility less than 10% of predicted FEV1, following inhalation of albuterol 400 microg or equivalent). Consistent definitions were applied to exacerbations data by examining case records. For each study the proportion of patients achieving a clinically significant improvement in health status was extracted (this was defined in the text). The hazard ratio (HR) was used to calculate the time to first exacerbation for each study.

How were the studies combined?

The studies were combined using a meta-analysis. A fixed-effect model was used where no heterogeneity was found, otherwise a random-effects model was used. Summary estimates were calculated separately for the ITT population and for the population with poor reversibility. The time to first exacerbation and the time to withdrawal were calculated using proportional hazards modelling.

How were differences between studies investigated?

The authors stated that statistical heterogeneity was assessed using a test of interaction. The influence of the following variables was partially examined: body mass index, disease severity, smoking history, duration of COPD, current therapy and age. It was not reported whether these analyses were selected post hoc or determined in advance.

Nine RCTs (3,580 participants) were included in the review.

Withdrawal.

A significant reduction in the percentage of early withdrawal was found with salmeterol compared with placebo/usual care at 1 to 3 months (HR 0.67, 95% confidence interval, CI: 0.55, 0.81), 1 to 6 months (HR 0.71, 95% CI: 0.61, 0.84), and 1 to 12 months (HR 0.73, 95% CI: 0.63, 0.84). The authors reported that similar results were found in the poorly reversible population.

Exacerbations.

Salmeterol was found to reduce the risk of exacerbation compared with placebo/usual care at 3 months (28%), 6 months (24%) and 12 months (22%), (p<0.0001). Disease severity, smoking history, duration of COPD or current therapy did not significantly alter these findings. Similar results were found in the poorly reversible population in which a 20 to 25% reduction in risk of exacerbation was shown (p=0.0002).

Lung function.

A significant increase in average pre-bronchodilator FEV1 as a percentage of predicted FEV1 was found with salmeterol compared with placebo/usual care at 3 months (3.2%), 6 months (3.0%) and 12 months (3.0%), (p<0.0001). Comparable results were found in the poorly reversible population.

Health status.

A significant improvement in health status was found with salmeterol compared with usual care for weeks 8 to 28 (difference in percentage of meaningful change 7.7%, 95% CI: 4.6, 10.7), and for weeks 8 to 52 (difference in percentage of meaningful change 7.9%, 95% CI: 5.1, 10.7). Similar results were found for the poorly reversible population. In the ITT population, the greatest benefits with salmeterol were found in younger participants and those with greatest reversibility.

The authors stated that heterogeneity was only detected for the analysis of FEV1 at 1 to 6 months.

Salmeterol was superior to placebo across a broad range of outcomes, demonstrating a reduction in withdrawal rate, exacerbation rate, improvement in lung function and health status, and no evidence of tachyphylaxis or bronchodilation over 1 year.

The review question was supported by clear inclusion criteria in terms of the intervention, population, outcomes and study design. The literature search made no attempt to locate unpublished material nor was it clear whether the search was restricted by language, thus it is possible that relevant studies might have been missed. The authors did not report the methodology used to select the studies or extract the data, thus it is not possible to assess the likelihood of bias having been introduced at these stages, they do not appear to have assessed the validity of the included trials. The analysis appears appropriate. The conclusions appear to follow from the results presented, but the limited search and the lack of reporting of review methods means it is difficult to assess the reliability of these conclusions.

The authors did not state any implications for practice or further research.

GlaxoSmithKline Research and Development.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.