- Introduce Basic Principles and practice of grading of evidence - Grading of Evidence is Outcome Focused - Grade Outcomes from a Range of studies - Focus on Consistency and Transparency   - GRADE is an approach - This Approach helps to frame questions - This is Outcomes Focused   - Ask an explicit question, including - Specify all important outcomes. *Explicitly rate the quality of evidence - Recommend the Evidence Pertaining to the Outcome - Summarise evidence in succinct, transparent informative summary of findings   - Quality of Evidence - Magnitude of Relative and Absolute Effects - Reason for the Quality Rating   - Study design, - Risk of bias, - Levels of imprecision, - Levels of inconsistency, - Extent of indirectness - Magnitude of effect.   - Recommendations Labelled as Strong or Weak - Quality of Evidence - Balance between Desirable and Undesirable Consequences   - We have learned about risk of bias - We have learned about appraising individual studies - We have learned to assess studies individually across outcomes - This is different in the sense that we are focusing on   - Different from Focus on Studies Across Outcomes [Schematic] - EP provides a record of the judgments - Judgments that were made by review or guideline authors - Intended for Review Authors and SoF table - Maintains Transparency   - Broader Audience - End Users of Systematic Reviews and Guidelines - Concise Summary of Key Information - Use GRADEPro (Online) or SoF Table Generator [upordowngrade] - Specify Relevant Setting - Specify What Population You will be Working on - The Relevant Intervention - The specified outcomes - Population, Interventions and Outcomes should be as similar as possible   - Specify importance of relative outcomes before starting the exercise - Respecify after completion of the exercise - If you work with surrogate outcomes …   - When an outcome cannot be directly measured but measured with some other means - Test of a Diagnostic Test as an indicator for Survival - Measure of Bone Density for Risk of Bone Fracture in Osteoporosis in Post menopausal Women   - How important are Patient specific outcomes? - In that case, Rate Down - If the Settings Do Not Translate Directly, Rate Down - These are Mapped to our Notions of Generalisability   - Essentially qualitative: High, Moderate, Low, Very Low - Statt with an initial rating and then upgrade or downgrade - Apply to a Body of Evidence, NOT to individual studies - For SRs: are the effects of estimate CORRECT? - For Recommendations: Are the effect estimates ADEQUATE?   - RCT: High - Observational Studies: Moderate to Low - Quality is more than Risk of Bias   - Design - Risk of Bias - Imprecision - Inconsistency - Indirectness - Publication Bias   - This is also Evidence   - Why?   - Include Expert Opinion - Experience with patients and colleagues - Understanding of biology - Understanding of preclinical research - Use Expert Opinion to Understand - Rate the Quality of that Evidence (see above) - Do not Rate Interpretation   - Grading is about a BODY of Evidence - SR is about individual studies pooled together - Grading is about an outcome across studies - SR is about across outcomes across studies   - Risk of Bias: Internal Validity - Chance (under powered? Correctly Powered?) - Bias (Selection? Response? Randomisation? Blinding?) - Confounding (?Multivariate Analysis, Matching?)   - Say an SR on body pain treatment with Salicylates showed pooled RR = 6 (4 - 13) - Start with RR and 95% CI, and population selection, and risk of bias, - How many studies were included? How many patients included? - What was the level of publication bias? - Was the duration too short? - Which body part was studied? - These are all relevant questions for GRADE [Study Limitations of RCTs] - Grading of Studies is Distinct and Outcome Focused - Grading is integral part of both SRs and Guideline Development - Judge Studies across designs and for each outcome - Provisions for up or downgrading studies - Consider design, risk of bias, precision, consistency, directness, publication bias - More than just notions of quality in SRs and single studies
Screenshot 2015 07 18 22.21.37

Arindam Basu

and 2 more

SUMMARY OR ABSTRACT OF THE THESIS The summary or abstract of the thesis needs more work! Smoking is a leading cause of preventable diseases and death world wide and within New Zealand.To date, a lot of previous research has been conducted and this suggests that Internet and Cell phone based interventions are effective to achieve cessation of smoking however the effects are short term.The purpose of this Meta Analysis was to investigate the effectiveness of Internet and Cell phone based interventions to achieve longer term cessation of smoking among adolescent and adult smokers.The analysis was based on the assessment of Randomied Controlled Trials whose interventions included Internet and Cellphone components and reported outcomes at six months or longer. Furthermore, this analysis was based on English Language articles published in the previous 10 years and whose comparison group received either any other intervention or an intervention inclusive of but not limited to Internet and Cell phone based interventions delivered at a lower frequency. Both a Fixed effects and a Random effects Meta Analysis between all studies were conducted to assess the length of abstinence.Morever the individual studies were grouped using an outcome theme and five subgroup analyses were conducted. Findings from both, the Fixed effects and Random effects analysis suggest great heterogeneity among the studies.Additionally some heterogeneity was found among the five subgroup but overall findings suggest that Internet and Cell phone based interventions used in smoking cessation are effective in achieving longer term cessation of smoking. Findings from the subgroup analyses suggest that both Internet and Cell phone based interventions combined with an additional intervention Nicotine Replacement Therapy are most effective in achieving longer term cessation of smoking among adolescent and adult smokers.