through S4BE This list will help build categories on the website, once the list is finished we will then ask people to give their top 10 resources for each point.
1. What is EBM Evidence Based Medicine
2. What is an adverse effect 3. What does the balance of benefits and harms mean 4. Whats the difference between odds & risks 5. What is a hazard ratio 6. What is an economic analysis 7. What is sensitivity 8. What is specificity 9. What is a positive predictive value 10. What is a negative predictive value 11. What is risk ratio 12. What is relative risk (RR) 13. What is absolute risk (AR) 14. How would you express the same data as RR and AR 15. What is an odds ratio 16. What is a two-by-two table 17. What does numbers needed to treat mean 18. What does numbers needed to harm mean 19. What is a systematic review 20. How does a systematic review differ from a traditional review 21. How do you critically appraise a paper about a treatment 22. How do you critically appraise a study about a diagnostic test and its accuracy 23. How do you critically appraise a systematic review 24. What is bias 25. What is attrition bias 26. What is detection bias 27. What is allocation bias 28. What is allocation concealment 29. What is performance bias 30. What is publication bias 31. What are levels of evidence or ranks 32. What does no evidence of effect/evidence of no effect mean and how do they differ 33. Prioritisation how are research topics prioritised 34. How might patients be involved in prioritising research (An example - James Lind Alliance 35. What is a meta-analysis 36. Distinction between systematic reviews & meta-analysis 37. Heterogeneity 38. Blinding what is it? What is effective blinding? What is not? 39. What databases are there? PubMed, EMBASE, LILACS, CENTRAL 40. Language bias positive results more likely to be published in English 41. What is a clinical trials register? 42. Placebos 43. What is confounding 44. What is confounding by indication 45. The I2 statistic 46. What are outcomes? What are relevant outcomes? What are surrogate outcomes? 47. What is a composite outcome? Why are they reported? What is the significance? 48. What are cut-off points 49. How do patients values and beliefs get taken into account when practising EBM 50. What is shared-decision making 51. What is unwarranted variation in practice 52. How do you search the literature 53. What is a ROC curve (receiver operator characteristics) 54. What is a guideline 55. How is evidence incorporated into guidelines 56. What is blinding? What is an example of effective and non- effective blinding 57. What is the CONSORT statement 58. What is randomisation? What is an example of effective and non-effective randomisation? 59. What is a case-control study 60. What is a cohort study 61. What is the Cochrane Collaboration 62. Who was Archie Cochrane? 63. What is a post hoc analysis? How do we interpret these results? 64. What is data dredging or a fishing trip 65. What is statistical power 66. How to you calculate power 67. What is a Type I error 68. What is a Type II error 69. What does prospective mean 70. What does retrospective mean 71. What is preference-sensitive care 72. What does fixed effects model mean 73. What does random effects model mean 74. What is the difference between descriptive and inferential statistics 75. What is the difference between standard deviation and standard error (and how are they related 76. What is a 95% confidence interval? 77. What is an intention to treat (ITT) analysis 78. What is an as treated analysis 79. What is a per protocol analysis 80. What do trialists do about participants who are lost to follow- up 81. What is a cross-over trial 82. What is a carry over effect
Harry also suggested having examples of assumptions which turned out to
be wrong proved by good research:
- Observational data suggesting HRT reduced heart attacks by 50%
but RCT showed no benefit (message observational trials can be misleading confounding factors) - People after heart attacks die from arrhythmias, anti-arrhythmias prescribed, CAST trial showed these drugs increase death (likely 200,000 people died) - SR- thrombolysis for heart attacks - SR- Steroids for premature babies