The strength of evidence provided by a study type is dependent upon the clinical question being addressed as indicated in Table 1. It is also dependent upon how well the study was designed and implemented. Factors to be considered in the study design may include the sample size, bias, blinding, control of variables, appropriate use of statistical tests, the power of the study, the accuracy and precision of any measurements made, the sample population and other components that may reduce the strength of evidence provided by the study.
When composing the clinical bottom line it is important that the strength of the body evidence provided by the studies is assessed and categorised according to Table 2 below. The outcomes from the studies should then be clearly stated. Conclusions and additional comments based upon the strength of evidence and the outcomes reported should then be made.
Table 1: Level of evidence table, adapted from the Oxford Centre for Evidence-Based Medicine’s levels of evidence
|
Strength of evidence |
Clinical question being addressed |
|||||
|
Treatment |
Prognosis |
Risk |
Diagnosis |
Prevalence |
Incidence |
|
|
1 (strongest) |
Systematic review and meta-analysis |
Systematic review and meta-analysis |
Systematic review and meta-analysis |
Systematic review and meta-analysis |
Systematic review and meta-analysis |
Systematic review and meta-analysis |
|
2 |
Randomised controlled trial |
Cohort study |
Cohort study |
Diagnostic test evaluation study |
Cross-sectional study |
Cohort study |
|
3 |
Cohort study |
– |
Case-control study |
– |
– |
– |
|
4 |
Case report or case study |
Case report or case study |
Case report or case study |
Case report or case study |
Case report or case study |
Case report or case study |
|
5 (weakest) |
Opinion consensus |
Opinion consensus |
Opinion consensus |
Opinion consensus |
Opinion consensus |
Opinion consensus |
Modified from Rees Gwen (2019)
Table 2: Significance of the four levels of collective evidence used in the clinical bottom line
|
Strength of evidence provided by the study designs |
Definition |
|
Strong |
High level of confidence that the estimate of the effect reported by the studies lies close to the true effect. |
|
Moderate |
Moderate confidence that the estimate of effect reported by the studies lies close to the true effect. |
|
Weak |
Limited confidence that the estimate of effect reported by the studies lies close to the true effect. Additional appropriate studies are required. |
|
Zero |
No studies available. |
Modified from Balshem et al (2011)
When writing a Knowledge Summary, authors will be asked to fill in the below section within the submission template:
Figure 1: Clinical bottom line submission template
|
Question (In PICO format)
Clinical bottom line
Indicate the category of research question that was addressed
Indicate the number and type of study designs which were critically appraised
Indicate the strength of evidence
Indicate the summarised collective outcome(s) from the studies
The conclusion should provide an answer to the Knowledge Summary question Additional comments and caveats can be added if required
|
Below is an example using the Knowledge Summary by Natasha A Jocelyn (2018) (http://dx.doi.org/10.18849/ve.v3i2.139).
Figure 2: Example of completed Clinical bottom line
|
Question In an Adult Horse With Severe Asthma (Previously Recurrent Airway Obstruction) Does Using Inhaled Corticosteroids Result in an Equal Improvement in Clinical Signs When Compared to Systemic Corticosteroids?
Clinical bottom line
Treatment.
Four papers were critically reviewed. There were 3 prospective crossover design clinical studies and a randomised design clinical study.
Strong.
Inhaled corticosteroids (fluticasone and beclomethasone) when used at an appropriate dose can have equivalent effects on severe equine asthma as systemic intravenous dexamethasone. Inhaled corticosteroids can take longer to have the desired effects.
In an adult horse with severe asthma (previously recurrent airway obstruction) inhaled corticosteroids result in an equal improvement in clinical signs when compared to systemic corticosteroids.
|