Learning from past mistakes: assessing trial quality, power and eligibility in non-renal systemic lupus erythematosus randomized controlled trials
Open Access
- 4 April 2008
- journal article
- review article
- Published by Oxford University Press (OUP) in Rheumatology
- Vol. 47 (9), 1367-1372
- https://doi.org/10.1093/rheumatology/ken230
Abstract
Objectives. To evaluate the post hoc study power of randomized controlled trials (RCTs) in the treatment of non-renal SLE and to determine the generalizability of these RCTs using an SLE database. Methods. RCTs in non-renal SLE were identified using PubMed (1975–2007). Inclusion/exclusion criteria, trial quality (5-point scale) and results of each study were recorded. The inclusion/exclusion criteria were compared with an SLE database to determine the proportion of patients from the database who would theoretically be eligible for these trials. For each negative study, we calculated the post hoc study power. We also looked for temporal improvements of trials in the literature and examined if pharmaceutical involvement influenced trial quality. Results. Sixty-four articles were included; the mean power of 30 negative studies was 24.6 ± s.e.m. 3.9% (range 2.5–81.1%). Only one study had a power > 80%. Overall, potential eligibility of SLE patients in the database was 45.1 ± s.e.m. 3.6%. Only 14 studies (21.9%) were of good quality. Fortunately, RCT quality is improving over time (trials 2003; P < 0.001). Trials with pharmaceutical involvement had a significantly higher number of enrollees and better study quality. Conclusions. Negative RCTs in SLE were mostly underpowered but the generalizability of these trials was high. Determination of study power and the impact of eligibility criteria on generalizability of study results are crucial in the design of clinical trials to ensure applicability to clinical practice.This publication has 74 references indexed in Scilit:
- Systemic lupus erythematosus trials: successes and issuesCurrent Opinion in Rheumatology, 2004
- Type II Error in the Spine Surgical LiteratureSpine, 2004
- Sample size considerations for superiority trials in systemic lupus erythematosus (SLE).Lupus, 1999
- Methodological issues of corticosteroid use in SLE clinical trials.Lupus, 1999
- The problem of the type II statistical errorObstetrics & Gynecology, 1995
- Statistical power, sample size, and their reporting in randomized controlled trialsJAMA, 1994
- Sample Size and Power in Psychiatric Research*The Canadian Journal of Psychiatry, 1990
- The beta error and sample size determination in clinical trials in emergency medicineAnnals of Emergency Medicine, 1987
- The 1982 revised criteria for the classification of systemic lupus erythematosusArthritis & Rheumatism, 1982
- Hypothesis testing in neurosurgical trialsJournal of Neurosurgery, 1981