Characteristics of meta-analyses of systemic opioids for breathlessness
| Characteristic of meta-analysis | Jennings et al3 | Ekström et al6 | Barnes et al7 |
|---|---|---|---|
| Design of included studies (n) | Double-blind RCTs | Double-blind RCTs | Double-blind RCTs |
| N studies | 9 (all crossover trials) | 8 (all crossover trials) | 12 (1 parallel and 11 crossover trials) |
| N trial participants | 102 | 118 | 198 |
| Population (n trial participants) | COPD (n=80) Chronic heart failure (n=12) Cancer (n=10) | COPD (n=113) Other (n=5) | COPD (n=107) CHF (n=47) Cancer (n=41) Other (n=3) |
| Intervention | Oral or parenteral opioid | Oral or parenteral opioid | Oral or parenteral opioid |
| Comparison | Placebo | Placebo | Placebo or any other pharmacological or non-pharmacological interventions that were directly compared with the opioid treatment (only two trials used non-placebo comparator) |
| Duration of treatment (n studies) | Single or few doses (N=5); longer treatment of 1–6 weeks (n=4) | Single dose or 1 day (n=3); 4 days to 6 weeks (n=5) | Single dose or 1–2 days (n=7); 4 days to 6 weeks (n=5) |
| Statistical method for pooling | Random effects model. Change on different scales compared as SMDs | Random effects model. Change on different scales compared as SMDs | Fixed effect models. Changes compared as MD when on the same scale and SMD when on separate scales, and separately for change from baseline and postscores. Random effects model was used in a sensitivity analysis |
| Accounted for crossover designs | Yes | Yes | No (analysed data as from parallel trials) |
| Findings for whole study population | |||
| Pooled effect of opioids (95% CI; I2; n trial participants)* | SMD −0.40 (−0.63 to −0.17; I2=42.3% ; n=102) | SMD −0.34 (−0.58 to −0.10; I2=0%; n=118) | Oral opioid, change from baseline: SMD 0.07 (−0.30 to 0.44; I2=65%; n=116) Oral opioid, postscores: SMD −0.27 (−0.56 to 0.02; I2=0%; n=190) Subcutaneous opioid, change from baseline: MD 0.20 (−2.50 to 2.90; n=20) |
| Stated quality of evidence | Not stated | Moderate (GRADE) | Not stated for systemic opioids For opioids overall: very low for change from baseline and low for postscores (GRADE)† |
| Findings in COPD participants | |||
| Pooled effect of opioids (95% CI; I2; n trial participants)* | SMD −0.26 (−0.44 to 0.08; I2=23.6%; n=80)† | SMD −0.34 (−0.58 to −0.10; I2=0%; n=118) | Change from baseline: SMD −0.49 (−1.08 to 0.10; I2=0%; n=46)† Postscores: SMD −0.21 (−0.45 to 0.04; I2=0%; n=262)† |
| Stated quality of evidence (criteria) | Not stated | Moderate (GRADE) | Not stated |
| Risk of bias assessment | Using Jadad score of methods of randomisation and blinded. Most items were rated as unclear | Using the Cochrane risk of bias tool. Ratings were low or unclear for all items; no item was rated as high | Using the Cochrane risk of bias tool as well as an additional item based on study size: ≥200 (low risk), 50–199 (unclear risk) and <50 (high risk) participants in each treatment arm. All items in the Cochrane risk of bias tool were rated as low or unclear except three items rated as high: performance bias (n=1), detection bias (n=1) and other bias (n=1).† Risk of study size bias was rated as high risk for all studies |
Characteristics are for trials included in each published meta-analysis.3 ,6 ,7
*Negative estimate indicates reduction in breathlessness by opioids compared with placebo.
†Included both trials of systemic and nebulised opioids which were not reported separately.
CHF, congestive heart failure; GRADE, Grading of Recommendations Assessment, Development, and Evaluation; I2, proportion of the total variance in effect estimates that are between studies; MD, mean difference; RCT, randomised controlled trial; SMD, standardised mean difference.