OBJECTIVES: To identify and evaluate the instruments used to measure the effect of pharmacological intervention on symptoms of chronic obstructive pulmonary disease (COPD) in clinical trials.
DESIGN: An extensive literature search was conducted for articles published in English in a peer-reviewed journal from 1995 to March 2002 which described a randomised controlled clinical trial measuring symptoms of COPD in response to pharmacological interventions.
PATIENTS: Patients with any severity of COPD.
INTERVENTIONS: Any pharmacological intervention for treatment of COPD.
MEASUREMENTS AND RESULTS: A total of 43 eligible articles were identified. The individual symptoms most frequently measured were dyspnoea/breathlessness, chest tightness or discomfort and exacerbations. There was considerable variability in the methods, terminology and symptom measurement instruments used. The most widely used instruments for measuring dyspnoea were the Borg scale, the Baseline Dyspnoea Index and the Transitional Dyspnoea Index. None of the instruments used had published evidence of rigorous psychometric testing.
CONCLUSIONS: Numerous methods have been employed to assess the symptoms of COPD in clinical trials, making it difficult to compare the results of different trials. No single measurement instrument predominates, and none of the measures identified in the review have undergone rigorous psychometric testing in this patient population. There is a clear need for a fully developed and validated tool for measuring the effects of therapeutic interventions on symptoms in COPD in clinical trials.
Jadad, Alejandro and Carols Rizo, Paola Cubillos, Elisabeth Stahl
Disease Category: Lungs & airways
Disease Name: Chronic obstructive pulmonary disease (COPD)
Age Range: Unknown
Sex: Either
Nature of Intervention: Drug
- None
- Systematic review of outcome measures/measurement instruments
- Systematic review of outcomes measured in trials
- Systematic review
Searches were conducted using MEDLINE, CINAHL, HealthStar, EMBASE and the Cochrane Library, using the terms (as Medical Sub-Headings and text words, whenever appropriate): (lung diseases, obstructive OR bronchitis OR pulmonary emphysema OR COPD OR chronic bronchitis OR emphysema OR chronic obstructive pulmonary disease) AND (randomized controlled trial OR clinical trial OR controlled clinical trial). This database search was limited to the period 1995 to March 2002. The search was complemented by searching the personal collections of the investigators and scanning the reference lists of identified articles, meta-analyses and systematic reviews.
Articles meeting all the following criteria were selected by two of us
• report published as a full report in English in a peer-reviewed journal
• randomised controlled trial (parallel or crossover design)
• trial comparing one or more pharmacological intervention(s) in COPD, or attempting to validate one or more tool(s) to assess symptoms of COPD
• trial including patients with COPD of any severity and presenting data on the impact of pharmacological intervention(s) on COPD symptoms, and if possible data on the psychometric properties of tools to assess the symptoms of COPD
• trial presenting data on patients with COPD of any severity, but not aggregated with data from patients with other respiratory disorders. Articles were excluded if they presented pharmacokinetic or pharmaco-dynamic data only, or if they did not provide separate data on symptoms.
Data extraction was performed by three independent reviewers with clinical and methodological expertise. They resolved any differences by consensus, referring to the information in the original report. The following items were extracted and included in the present analysis: citation details; symptoms measured; tools used for symptom measurement; frequency of measurement and any mention of efforts made to validate the tools. Symptoms were defined as events that can only be measured by the person experiencing them; i.e. dyspnoea/breathlessness is a symptom, whereas cough is not. However, where events were rated by the person experiencing them (e.g. a patient rating for cough severity), they were also included as symptoms. Where the validation of a tool was by reference to other papers, these articles were obtained and reviewed.
The methodological quality of the trials was assessed using a 3-item validated scale.
Due to substantial heterogeneity between trials, a meta-analysis was not performed.