Difference between screening and diagnosis

Population Health Assignment 2 – Structured Critical Appraisal and Evidence Synthesis Aim: The aim of this assignment is to conduct a critical appraisal of the three articles that you located in Assignment 1. Assignment 2 will involve applying concepts from Topics 4-10 in assessing the quality of the selected articles and synthesising the evidence. Length: Up to 1800 words (+/-10%), including tables but not reference list. Penalties will apply over 1880 words. Note: Every 50 words over the word limit will incur a 0.5 mark penalty Weighting: 60% (individual assignment). Due date: Midnight Friday 4 September. Assignments should be submitted through the Turnitin portal in vUWS under the 2h > Population Health 1 folder. Please remember to include your name and student ID in the header or footer of the document. Late submissions will have 10% marks deducted for every day late, unless prior arrangements have been made with the convenor. Please use Vancouver referencing style and ensure your assignment is double spaced, 12 point, Times New Roman or Arial or Calibri font. Task: Reflect upon the content of the Population Health topics we have covered in this course, and critically appraise the articles you selected in Assignment one by answering the following questions: Required Structure: Question 1: Describe the characteristics of the three papers you selected. Specifically address the following questions: -What was the exposure or intervention (this should be the same in all three studies)? -What was the outcome (this should be the same in all three studies)? -What were the study designs? -Who were the study population? (Length: ~300 words) Question 2: To what extent can the observed association between the exposure and outcome be attributed to bias and confounding?

Consider the following questions in your critique of your three selected studies: -Are the results likely to be affected by selection and/or measurement bias? -Are the results likely to be affected by confounding? -Are the results likely to be affected by chance variation? 1 (Length: ~500 words) Question 3: Do the findings of the three articles accord with each other? Specifically: -Are the findings of your three selected studies consistent with each other? Pay particular attention to the study designs and how they may contribute to your overall synthesis of this evidence. (Length: ~300 words) Question 4: What is the summary of the evidence? – Synthesise the results of the three studies together. – Provide a brief summary of the overall results of these studies, ensuring you discuss the results in the context of the strength of the evidence. (Length: ~500 words) Question 5: Are the findings externally valid, that is generalisable? Specifically: – Can the findings of your three papers be applied to the source populations from which the study populations were derived? – Using your synthesis of results from Question 3, can the results be applied to the Greater Western Sydney population? (Length: ~200 words) Difference between screening and diagnosis

A note on referencing Consistently apply a recognised academic referencing system, such as Vancouver. Make sure you cite everything that is not your own idea, and list all sources in full at the end. Check the Library website for guides and library staff can help you if you are unsure. You can also use bibliographic software such as EndNote or Mendeley for storing references in a database and applying style templates to Word documents. Your reference list is NOT part of the assignment word count. 2 Assignment Part 2 – Structured critical appraisal (60%) Marking Rubric FAIL <50% PASS 50-64% CREDIT 65-74% DIST 75-84% HIGH DIST >=85% Description of the selected paper characteristics Little demonstrated attempt to describe exposure, outcome, study design, or study population. Some elements of the exposure, outcome, study design, and study population presented. Exposition of exposure, outcome, study design, and study population. Detailed exposition of exposure, outcome, study design, and study population. Extensive and relevant exposition of exposure, outcome, study design, and study population. Little relevant critique of methodological limitations of the three articles presented. Critique is descriptive, and there is little attempt, if any, to incorporate elements of bias, confounding, and chance variation. Some relevant critique of methodological limitations of the three articles presented. Critique is descriptive. There is some attempt to incorporate elements of bias and confounding. Some relevant critique of methodological limitations of the three articles presented. Critique is analytical (rather than descriptive), and incorporates some elements of bias, confounding, and chance variation. Detailed and relevant critique of methodological limitations of the three articles, incorporating elements of bias, confounding, and chance variation. Extensive and relevant critique of methodological limitations of the three articles, incorporating elements of bias, confounding, and chance variation. Cohesively written, and some use of references to support critique. Cohesively written, and comprehensive use of references to support critique. 20% Assessment of bias and confounding 25% Few references to support critique. No references to support critique. Accordance with other evidence 20% Summary of the evidence Little relevant evidence presented.

Some relevant evidence presented. Discussion of evidence is descriptive. There is some attempt to identify strengths and weaknesses, based on interpreted information. Some relevant evidence presented. Interpretation of evidence is analytical (rather than descriptive). Appropriate integration and interpretation of strengths and weaknesses. Detailed and relevant evidence presented. Analytical and accurate synthesis and analysis of strengths and weaknesses of the literature. Difference between screening and diagnosis

Extensive and relevant evidence presented. Strong synthesis and critical analysis of strengths and weaknesses of the literature demonstrated. Poor summary of the three articles’ findings. Findings over looked or ignored. No synthesis of results. Some of the three articles’ findings are presented and synthesized, but information is partly confused or inconsistent. Summary of the three articles’ findings presented and synthesis provided with some inconsistencies. Comprehensive summary and of the three articles’ findings and welldeveloped synthesis. Comprehensive summary of the three articles’ findings and highly developed synthesis Little relevant critique of the source population in the article, or consideration of applicability to other relevant population. Critique is descriptive. Difference between screening and diagnosis

Some relevant critique of the source population in the article, and consideration of applicability to other relevant population. Critique is descriptive. Some relevant critique of the source population in the article, and consideration of applicability to other relevant populations. Critique is analytical (rather than descriptive). Detailed and relevant critique of the source population in the article, and consideration of applicability to other relevant populations. Extensive and relevant critique of the source population in the article, and consideration of applicability to other relevant populations. Discussion of evidence is descriptive; there is little attempt, if any to integrate and interpret information. 25% External validity (generalisability) 10% Use of references to support critique. Note: Every additional 50 words over the essay limit will incur a 0.5 mark penalty 3 4 MD Population Health 1 Assignment 2 Exemplar Question 1: Describe the characteristics of the three papers you selected. The intervention in my three selected studies was the use of cannabis based products among children with epilepsy (Elliott et al., 2019; Stockings et al., 2018; Szaflarski et al., 2018). The cannabis based products varied between the three selected studies and included cannabidiol, which can be extracted from cannabis plants but does not include tetrahydrocannabinol (THC) as well as cannabis and cannabinoids. Dosages also varied between studies. The main outcome I was interested in was a reduction in seizure frequency, as per my original PICO question.

All three studies had several additional primary and secondary outcomes, however, including quality of life, sleep, death, gastro intestinal events, emergency department visits, adverse events, seizure severity and complete seizure freedom. My selected papers comprised of two systematic reviews (Elliott et al., 2019; Stockings et al., 2018) and a single arm open label add on prospective trial (Szaflarski et al., 2018). All three of my selected papers included paediatric patients, my target population, though the Stockings and Szaflarski papers also included some adults. The systematic review samples were drawn from multiple settings including the UK, USA, the Netherlands and Spain. I note there was some crossover on the studies included in both of the systematic reviews. The first systematic review (Elliott et al., 2019) included four randomised controlled trial (RCTs) with n = 550 children/adolescents with an age range of 7-14 years as well as 19 nonrandomised studies (NSRs) (n=1115). The second systematic review (Stockings et al., 2018), included 6 RCTs with n=555 participants and 30 observational studies with n=2865 participants and a mean age of 16.1 (range 0.5-55 years). The open label trial (Szaflarski et al., 2018) included 72 children (mean age 10.1 ±4.9 years) and 70 adults drawn from a North American setting. Difference between screening and diagnosis

Question 2: To what extent can the observed association between the exposure and outcome be attributed to bias and confounding? Both of systematic reviews were conducted using Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines (Moher et al., 2015), an evidence-based set of items for reporting in systematic reviews and is therefore a strength of both of the reviews. MD Population Health 1 Assignment 2 Exemplar They both had robust search strategies, conducted through multiple electronic databases and clinical trials registries using transparent search processes and appropriate keywords/MeSH headings. Stockings et al. also examined reference lists of other systematic reviews, a strength, though did not search for grey literature, a weakness versus Elliott et al.’s review which did this additional searching. A particular strength in both reviews is they also included non-English language papers. Both reviews had strong screening protocols with two independent reviewers examining articles for inclusion/exclusion. An added strength in Stocking et al.’s paper was use of a third reviewer to reach consensus if there was disagreement on article inclusion.

Risk of bias assessment (RoB) was done well in both reviews with two reviewers assessing methodological quality ratings for RoB in included RCTs using the Cochrane Collaboration RoB tool (Higgins & Green, 2008). Non-randomised studies were assessed using several various validated checklists including the RoB in non-randomised studies (NSRs) of interventions (ROBINS-I) tool (Sterne et al., 2016). The extraction and analysis of results appeared reasonable in both reviews, though Elliot et al. review did not provide great detail on data extraction processes which is a weakness in that review. Stockings et al. also graded their included studies, using the Grades of Recommendation, Assessment, Development and Evaluation tool (Higgins & Green, 2008). This, as well as other extra measures undertaken make the Stockings et al. a slightly better well-conducted review. Irrespective of the quality of the systematic review process in both cases, multiple NSRs included in each review which were mostly judged to be at serious to critical risk of bias, by both reviews.

Methodological quality for these included studies was therefore typically graded as low or very low. As one would anticipate with systematic reviews including high numbers of patients (Jorge & Lilian Martins, 2014), the results of both of these reviews are unlikely to have occurred due to chance with relatively narrow confidence intervals for the primary outcome of seizure frequency reduction; Stockings et al. 95% CI 39.0% to 58.1% and Elliott et al 95% CI = − 27.0% to − 12.6%). MD Population Health 1 Assignment 2 Exemplar Similar to the NSRs included in both reviews, Szaflarski et al.’s study was at high risk of bias. One of the key methodological flaws was absence of a control group, therefore any results arising from this trial could have arisen due to multiple biases. For example, study participants and investigators were not blinded to treatment opening up the possibility for measurement error and reporting bias. Further, there may have been unique characteristics present in the study participants that could have helped an improvement in their condition. In the absence of a control group, it is impossible to judge this. Seizure frequency in this study is unlikely to have been affected by chance variation (p=0.01). Difference between screening and diagnosis

Question 3: Do the findings of the three articles accord with each other? The findings of my three chosen articles were fairly consistent with each other, despite differences in research design. All three articles found a short term reduction in seizure frequency though how this was reported across the studies varied. Elliott et al. found a significant reduction in the median frequency of monthly seizures with cannabidiol compared with placebo, over 14 weeks (− 19.8%, 95% CI = − 27.0% to − 12.6%; moderate certainty due to RoB). It is important to note however that this was the pooled results of just three RCTs included in the review that had this specific outcome. When they assessed seizure frequency across six of their included NSRs that had seizure frequency as an outcome, they found wide variance in the reported reduction in total seizures with use of a cannabis- based product with reductions between 30% and 90% (duration = 8 weeks to >16 months; low certainty due to RoB).

Stockwell also pooled results from included NSRs and RCTs separately. Across the 17 observational studies that had a 50+% reduction in seizure freedom as on outcome they found that 48.5% of patients in those studies patients reported 50%+ reductions in seizures (95% CI 39.0% to 58.1%). From the two RCTs that also had a 50+% reduction in seizure freedom as on outcome, they found cannabidiol was more likely to produce >50% reduction in seizures than placebo (RR 1.74, 95% CI 1.24 to 2.43; low certainty due to RoB). Szaflarski et al. found that in their paediatric sample bi-weekly seizure frequency decreased from a mean of 231.8 ± 535.0 at entry into the study to 71.5 ± 177.25 at 48 weeks (p = 0.0112). MD Population Health 1 Assignment 2 Exemplar High variability is seen in these results as observed by the large standard deviations, however, with the results of low certainty due to the high RoB, including a significant drop out rate in the paediatric group in this study (only 37% of the original sample remained in the trial at 48 weeks). Question 4: What is the summary of the evidence? Overall, even though these studies report a reduction in seizure frequency, this evidence surrounding the use of cannabis based products is weak. First, there is variability in the type of cannabis product as well as dosing level, which leads to a lack of clarity around the efficacy of any particular treatment regime.

Second, Szaflarski et al.’s study lacks a control group and is therefore highly flawed. Third, while the two systematic reviews appear to be robust, well conducted reviews, the underlying evidence contributing to their conclusions around seizure frequency are at best of moderate certainty due to the inclusion of multiple observational studies at high risk of bias. In both cases the authors have concluded that the evidence is of low quality and that there is only low to moderate certainty in the results, even with the pooled results of included randomised controlled trials. Therefore, even if these reviews are given more weight – and perhaps Stockwell et al.’s additional weight due their slightly better methodological approach that Elliot et al.’s – when synthesising this evidence one can conclude that the evidence is at best weak due to the underlying biases. Based on my selected literature, I would therefore not recommend adoption of this therapeutical option to reduce seizure frequency among paediatric patients with epilepsy at this time, particularly in light of the adverse events reported across each of the studies and the variance in dosages reported. Nonetheless, there is clearly some suggestion of efficacy of this treatment approach and therefore a need for high quality, randomised clinical trials to inform any change in clinical practice. Question 5: Are the findings externally valid, that is generalisable?

If these studies were free of bias, it may be possible to generalise these results given the age range of the participants and the inclusion of participants from multiple settings such as the United Kingdom and the United States of America, who are unlikely to systematically differ MD Population Health 1 Assignment 2 Exemplar from the source or local population in the nature of their condition. In reality, however, the results of these three studies are not internally valid, that is free from bias, therefore their findings cannot be generalised to other paediatric populations, even those with the same type of epilepsies that were included in these studies (i.e. Lennox–Gastaut or Dravet Syndromes). Difference between screening and diagnosis

This is in part is due to high risk of selection bias in both the included NSRs in both systematic reviews as well as the single arm open label trial, leading to uncertainty as to the representativeness of the included sample. As a consequence, therefore, these results cannot be generalised to either the source population from which the study samples were derived, or the Greater Western Sydney population. Word Count: 1628 (including in text citations and question headings). References Elliott, J., Dejean, D., Clifford, T., Coyle, D., Potter, B. K., Skidmore, B., . . . Wells, G. A. (2019). Cannabis-based products for pediatric epilepsy: A systematic review. Epilepsia, 60(1), 6-19. doi:10.1111/epi.14608 Higgins, J. P. T., & Green, S. (2008). Cochrane Handbook for Systematic Reviews of Interventions. Hoboken: Hoboken: John Wiley & Sons, Incorporated. Jorge, F., & Lilian Martins, F. (2014). How sample size influences research outcomes. Dental Press Journal of Orthodontics, 19(4), 27-29. doi:10.1590/2176-9451.19.4.02 7029.ebo Moher, D., Shamseer, L., Clarke, M., Ghersi, D., Liberati, A., Petticrew, M., . . . Group, P.-P. (2015). Difference between screening and diagnosis

Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev, 4, 1. doi:10.1186/2046-4053-4-1 Sterne, J. A., Hernán, M. A., Reeves, B. C., Savović, J., Berkman, N. D., Viswanathan, M., . . . Higgins, J. P. (2016). ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ, 355. doi:10.1136/bmj.i4919 Stockings, E., Zagic, D., Campbell, G., Weier, M., Hall, W. D., Nielsen, S., . . . Degenhardt, L. (2018). Evidence for cannabis and cannabinoids for epilepsy: a systematic review of controlled and observational evidence. Journal of Neurology, Neurosurgery & Psychiatry, 89(7). doi:10.1136/jnnp-2017-317168 Szaflarski, J. P., Bebin, E. M., Cutter, G., Dewolfe, J., Dure, L. S., Gaston, T. E., . . . Ver Hoef, L. W. (2018). Cannabidiol improves frequency and severity of seizures and reduces adverse events in an open-label add-on prospective study. Epilepsy & Behavior, 87, 131-136. doi:10.1016/j.yebeh.2018.07.020 Over the time mobile phone usage have enormously increased throughout the world. As mobile phones emit electro … Difference between screening and diagnosis