Skip to main content

Table 2 Numerical scoring system used to allocate a score-based grade for the overall risk of bias in each of the twenty-six reviewed studies

From: Treatment in canine epilepsy – a systematic review

Studies Randomisation sequence generation Allocation concealment Blinding of participants and personnel Blinding of outcome assessment Incomplete outcome data Selective reporting Other bias Overall `risk of bias' category
Boothe et al. [11] 1 1 1 2 2 2 2 Low/moderate (11)
        Primary investigator could potentially influence the treatment.  
Chung et al. [24] 3 3 3 3 1 2 2 Moderate/high (17)
        Research support but unclear if it was financial.  
Cunningham et al. [29] 3 3 3 3 2 2 2 Moderate/high (18)
        Conference abstract  
Dewey et al. [18] 3 3 3 3 2 2 2 Moderate/high (18)
        Less than 6 months study duration.  
Dewey et al. [19] 3 3 3 3 2 2 1 Moderate/high (17)
EMEA pseudo-trial [13] 1 2 1 3 1 2 2 Low/moderate (12)
        The follow up assessment of efficacy was not blinded. Different drug formulations were used compared to the final formulation.  
Govendir et al. [21] 3 3 3 3 2 2 3 High (19)
        A few cases were treated by the referring vets. The study had financial support. Less than 6 months study duration.  
Heynold et al. [36] 3 3 3 3 2 2 3 High (19)
        The study had financial support but unclear if it influenced the results. Less than 6 months study duration.  
Kiviranta [17] 3 3 3 3 2 2 1 Moderate/high (17)
Löscher et al. [26] 3 3 3 3 2 2 3 High (19)
        Part of the study was retrospective  
Morton et al. [31] 3 3 3 3 2 2 3 High (19)
        A few cases were treated by the referring vets. The study had financial support but unclear if it influenced the results.  
Muñana et al. [12] 1 1 1 1 2 2 2 Low/moderate (10)
        The study had financial support but unclear if it influenced the results.  
Nafe [25] 3 3 3 3 2 2 3 High (19)
        Less than 6 months study duration.  
Pearce [32] 3 3 3 3 2 2 2 Moderate/high (18)
Platt et al. [22] 3 3 3 3 1 2 3 Moderate/high (18)
        Less than 6 months study duration.  
Podell et al. [33] 3 3 3 3 2 2 3 High (19)
        Retrospective nature of study.  
Rieck et al. [27] 3 3 3 3 2 2 3 High (19)
        Part of the study was retrospective  
Ruehlmann et al. [35] 3 3 3 3 1 2 3 Moderate/high (18)
        Part of the study was retrospective. No clarification of statistical analysis  
Schwartz-Porsche [28] 3 3 3 3 2 2 2 Moderate/high (18)
Schwartz-Porsche et al. [15] 2 2 3 3 2 2 2 Moderate/high (16)
        The study had research support but unclear if it influenced the results. No clarification of statistical analysis  
Schwartz-Porsche et al. [30] 3 3 3 3 2 2 1 Moderate/high (17)
Steinberg [23] 3 3 3 3 2 2 2 Moderate/high (18)
        Conference abstract  
Tipold et al. [14] 1 2 1 1 1 2 3 Low/moderate (11)
        Statistical analysis was conducted before unblindingonly on the per-protocol population and not on the intent-to-treat population. A high and unbalanced population of animals was excluded. The reasons for exclusion were in many cases treatmet-related (post-randomization bias). Conflict of interest about imepitoin reported.  
Trepanier et al. [34] 3 3 3 3 2 2 2 Moderate/high (18)
        Some samples were submitted by the referring vets.  
Volk et al. [16] 3 3 3 3 1 2 3 Moderate/high (18)
        The study had financial support but unclear if it influenced the results. Part of the study was retrospective  
Von Klopmann et al. [20] 3 3 3 3 2 2 3 High (19)
        Less than 6 months study duration.  
  1. Each aspect the risk of bias was categorised as `high’, `moderate’, `low’ or `unclear’. These categories were assigned a numerical score as follows: High risk of bias =3, moderate or unclear risk of bias =2, low risk of bias =1. Within each study these seven scores were summed to form a total score. This score translates to an overall estimated risk of bias associated with the findings of the study in question, as follows: Score 19–21 = overall high risk of bias, score 16 – 18 = overall moderate/high risk of bias, score 13 – 15 = overall moderate risk of bias, score 10 – 12 = overall low/moderate risk of bias, score 7 – 9 = overall low risk of bias.