A systematic search of the literature published between January 1966 and December 2004 was undertaken using Medline, Embase, CINHAL, PEDro, and Cochrane Library databases. The search included both a general search and an intervention‐specific search. The general search strategy consisted of two basic components: AS in whatever possible terms in the databases (Appendix 1) and types of research in the form of systematic review/meta‐analysis, randomised controlled trial (RCT)/controlled trial, uncontrolled trial, cohort study, case‐control study, cross sectional study, and economic evaluation (Appendix 2). The general search aimed at summarising the current available treatments from the literature for AS. The summary results of this search were reported to the ASAS/EULAR expert committee at the beginning of the recommendation development process.
After the expert group had generated 10 recommendation propositions, an intervention‐specific search was undertaken to identify evidence for each specific treatment. The search strategy included the terms for AS (Appendix 1) and any possible terms for the specific intervention. The results of the two searches were then combined and duplications excluded. Medical subject heading (MeSH) search was used for all databases and keyword search was used if the MeSH search was not available. All MeSH search terms were exploded. The reference lists of reviews or systematic reviews were examined and any additional studies meeting the inclusion/exclusion criteria were included.
The search in the Cochrane Library included a MeSH search of the Cochrane reviews, abstracts of Quality Assessed Systematic Reviews, the Cochrane Controlled Trial Register, NHS Economic Evaluation Databases, Health Technology Assessment Database, and NHS Economic Evaluation Bibliography Details Only. Abstracts of scientific meetings held in 2003 and 2004 and “online first” publications were also hand searched for relevant studies.
Only studies with clinical outcomes for AS were included. The main focus of interest was on systematic reviews, RCTs/controlled trials, uncontrolled trials/cohort studies, case‐control studies, cross sectional studies, and economic evaluations. Studies that included patients with AS in a larger cohort of different spondyloarthritides (for example, psoriatic arthritis) were excluded unless the results for patients with AS were reported separately. Studies of other seronegative spondyloarthritides or other inflammatory joint conditions, animal studies, non‐clinical outcome studies, narrative review articles, commentaries, and guidelines were also excluded.
Evidence was categorised according to study design using a hierarchy of evidence in descending order according to qualities2
(table 1), and the highest level of available evidence for each intervention was reviewed in detail. When very few high level studies were available, the next highest category was also reviewed. The most recent meta‐analysis of RCTs (level Ia) was reviewed for each intervention, and any RCTs published since the meta‐analysis was conducted were also considered.
Table 1Evidence hierarchy and traditional strength of recommendation
Efficacy was assessed specifically for AS, and toxicity data were evaluated specifically for the intervention, irrespective of the musculoskeletal condition.
Estimation of effectiveness and cost effectiveness
Effect size (ES, mean change divided by the standard deviation of the change) and 95% confidence interval (95% CI) compared with placebo or active control were calculated for two predetermined continuous outcomes: pain relief and improvement in function.3
The percentage of patients with an ASAS20 response to treatment, or with moderate to excellent (predefined as more than 50%) pain relief or moderate functional improvement (predefined as more than 20%) was calculated where possible, and the number needed to treat (NNT) was estimated for these dichotomous outcomes.4
The NNT is the number of patients who need to be treated with a given intervention in order to prevent one poor outcome, and is the inverse of the absolute risk reduction. Results from the latest systematic review were used if there was more than one systematic review for the same intervention. Statistical pooling was undertaken as appropriate5
when a systematic review was not available, or when more recent RCTs could be included. The NNT and 95% CI were reported only if statistically significant; otherwise “NS” (not significant) was used to avoid difficulty in interpreting the results.6
Relative risk (RR) was calculated for adverse effects.
When economic evaluation was available, study design, comparator, perspective, time horizon, discounting, total costs, and effectiveness were reviewed. The incremental cost effectiveness ratio was calculated. Quality of life years were used when available, otherwise disease‐specific outcomes such as pain relief and functional improvement were used.
Data were extracted by one investigator (JZ). A customised form was used for the data extraction.