Methodological choices in the area of ‘study design’
| Topics | Results from studies in inflammatory arthritis (n=58) | Aspects identified by authors of SLRs in other chronic diseases (n=24) | ||
| n/N (%) | Detailed information | n* (%) | Comments | |
| 1. The included population aligned with the work-related study objective | 16/58 (28%) | The included population specifically aligned with the work-related study objective in 16/58 (28%) studies (16/28 (57%) studies with work as primary outcome)†—n/N§:
| 3 (12%) 3 (12%) 3 (12%) | Lack of clarity on the recruitment procedure98 115 120 Study population not representative100 114 120 Study population too heterogeneous114 115 119 |
| 2. Sample size calculated for the work-related outcome | 8/58 (14%) | The sample size for the work-related outcome was calculated in eight studies (8/28 (29%) studies with work as primary outcome)†—n/N§:
| 3 (12%) 4 (17%) 21 (87%) | No sample size calculation in included studies115–117 Study population too small99 100 103 114 No mention to the sample size calculation (if performed or not by included studies)97–114 118–120 |
| 3. Time horizon accurate for the work outcome of interest | 56/58 (97%) | The time-horizon aligned with the work outcome domain of interest in 57/58 (98%) studies—n/N§:
Work status‡: 2/17 (12%)21 52 Sick leave and/or presenteeism: 13/52 (25%)19 28 31 32 35 37 43 48 52 54 58 67 71
Work status‡: 5/17 (29%)21 30 36 42 51 Sick leave and/or presenteeism: 13/52 (25%)18 20 21 36 38 42 51 55 56 61 63 64 68 | 2 (8%) 3 (12%) 4 (14%) | Follow-up was reported as: Highly heterogeneous across studies97 114 Too short to assess work outcomes111 117 120 Not done/not described106 108 114 115 |
| Topics | Results from studies in inflammatory arthritis (n=58) | Aspects identified by authors of SLRs in other chronic diseases (n=24) | ||
| n/N (%) | Detailed information | n* (%) | Comments | |
| 3. Time horizon accurate for the work outcome of interest (continuation) |
Sick leave and/or presenteeism: 26/52 (51%)16 17 22 24–27 29 33 34 39 44–47 49 50 53 59 60 62 65 66 69 72 73 | |||
| 4. Use of a comparator | 26/58 (47%) | A comparator was used in 26/58 studies (47%)—n/N§:
| 3 (12%) 1 (4%) | Most studies lacked a control group97 102 114 Unmatched control groups114 |
*Number of systematic literature reviews reporting on the corresponding topic.
†7/28 (25%) studies with work as primary outcome included unselected patients from registries.
‡Emery et al (2016) have two different time horizons because this is a post hoc analyses of two trials: time horizon of 26 and 24 weeks for the Optimal Protocol for Treatment Initiation with Methotrexate and Adalimumab (OPTIMA) and PRevention Of Work Disability (PROWD) trials, respectively.
§The denominator may vary according to the type of intervention, work outcome of interest or type of study.
n/N, number of original studies in which the methodological choice was identified/number of studies in which the topic was possible to assess; OBS, observational longitudinal study; RCT, randomised controlled trial.