Using Clinical Epidemiology in Patient Care-Tom Ball 12/6/05 *Ball’s PDF of ‘Critical appraisal of journal articles’ was sent out 12/1 on listserv (same email with final exam assignments) Used in Obj.#5&6 *Powerpoint slides were sent out 12/6/05 on listserv. *Refs (below) are found in e-reserves. Search under “Moher”. Password is “med806” OBJECTIVES At the end of this session, you should be able to 1) List six (6) of the nine (9) epidemiologic concepts described. 1. “The triangle”: The expression of a disease is the effect of environment, host and agent. Look to modify each of these in treatment of any given disease 2. The iceberg. What we see presented to us in clinic or in the hospital represents the tip of the iceberg. Look for other influences to the disease process such as genetic factors, infectious agents, asymptomatic manifestations, etc. He included in this a picture showing icebergs with different percentages of their mass underwater, I suppose meaning we should not make assumptions about each individual disease presentation. 3. Causality is confusing. He shared with us his French hotel experience; who caused the glass to break? Diseases are multi-factorial in causation. It is up for debate whether they relate to one another like dominos (A causes B causes C) or stand as independent factors. He used refs 1,2&3 [below] to illustrate this point. Effect modification- Ref 2 shows that a gene was predictive only for 6-17mo olds. Effect modification indicates that there is a third unaccounted-for factor that modifies the outcome effect. In this case, effect modification was immune maturation (remember the window of vulnerability when a person is no longer protected by mother’s passive immunity and the patient has yet to fully develop a mature immune system), which in addition to the gene variation, led to the patient’s added vulnerability of contracting the disease. 4. Know your local epidemiology. Dr. Ball gave the story of seeing an 18-month-old with a scorpion bite and not knowing the course of treatment. Moral of the story = round up the resources: Health Dept, co-workers, colleagues, online resources. 5. What is normal? Dr. Ball gives 5 frameworks from which we may assess normal. See Obj#4. 6. Use of a test is determined by its accuracy and disease prevalence. See Obj#3. Dr. Ball used refs 4-6 to explain this. 7. The other “P”: Power. This basically boils down to (1) sample size and (2) effect size. This can be evaluated within the following frameworks: Disease Outcomes (risk), Treatment (chance of response vs. amount of response), Studies (sample size vs. effect size). Look beyond the p-value to effect size. The example here was using acyclovir to treat varicella. Results may be statistically significant, but not clinically significant. 8. Consider all outcomes. In class, we looked at a study of a drug that decreased coronary deaths, but at the expense of increasing suicide/trauma deaths. 2nd, duplicating study confirmed phenomenon. 9. Generalizable and Individualizable. Can we apply the results of a particular study to our patient? To all patients? Refs 7-9 We also used the example of a study of Fluticasone in 12-47 mo infants, wondering how applicable it would be to our 10-mo old patient. Dr. Ball mentioned that age and sex are two prominent effect modifiers and reiterated that future decisions will be based in pharmacogenetics. 2) Describe an example of how host, agent, and environmental factors all play roles in illness and can all be therapeutically manipulated. This is the triangle (#1 of 9 in Obj#1) Dr. Ball referenced Malawi (sp?), Tb testing and drug addiction among others. Important, modifiable factors with respect to Tb would include the underlying prevalence of Tb in the population with which you work, the immune competence of your particular patient, etc. 3) Describe why both test performance and prevalence of disease impact your use of a diagnostic test in your clinical decision-making. Test Performance Knowing disease states, accuracy, likelihood. Sensitivity and specificity Prevalence of Disease Knowing the test result, likelihood of disease. Positive predictive value and negative predictive value 4) Discuss three (3) different definitions of normal. 5 Frameworks from which to assess normal: Term Consequences Percentile All diseases with same prevalence. Patients normal until worked up. Culturally desirable Confusion over the role of medicine in society Diagnostic Need to know predictive values that pertain to your practice Therapeutic Need to keep up with therapies Self-reported Fuzzy & variable cut-offs for clinical decisions 5) List four (4) factors to consider when critically reviewing a study on therapy. 6) List three (3) factors to consider when critically reviewing a study on diagnosis. Dr. Ball indicated that these objectives were addressed by the questions present in the ‘Critical appraisals’ PDF under their respective titles ‘Therapy’ and ‘Diagnosis’. This page has columns of questions under eight headings. We specifically looked under ‘Therapy”, “Diagnosis”, and “Causation” in class. 7) Describe two (2) explanations for a negative result of a treatment trial. Given a study that indicates that there is not a ‘statistically significant’ difference between the outcome and the outcome happening by chance, (1) There truly is no difference (based on powerful evidence presented) (2) The study was insufficiently powered, usually by a low N (sample size) REQUIRED READING Read any three (3) of the nine articles (#9 is a package deal) listed below: Hint: Reading the abstract, scanning the methods for factors listed on the user’s guide sheet, reading the first paragraph of results, and reviewing the tables/figures in the results usually gets you what you need (and for #9 – look at the funding agencies). 1) Cohen S, et al. Psychological stress and susceptibility to the common cold. NEJM 1991;325:606-12. 2) Koch A, et al. Acute respiratory tract infections and mannose-binding lectin insufficiency during early childhood. JAMA 2001;285:1316-21. 3) Van den Hoogen BG, et al. A newly discovered human pneumovirus isolated from young children with respiratory tract disease. Nature Medicine 2001;7:719-24. 4) Wipf JE, et al. Diagnosing pneumonia by physical examination. Arch Intern Medicine 1999;159:1082-7. 5) Aetiology, outcome, and risk factors for mortality among adults with acute pneumonia in Kenya. Lancet 2000;355:1225-30. 6) Chaparas SD, et al. Tuberculin test, variability with the Mantoux procedure. Am Rev Respir Dis 1985;132:175-7. 7) Rose EA, et al. Long-term use of a left-ventricular assist device for end-stage heart failure. NEJM 2001;345:1435-43. 8) Honl M, et al. Comparison of robotic-assisted and manual implantation of a primary total hip replacement. J Bone Joint Surgery 2003;85-A:1470-8. 9) A) Meltzer EO, et al. Comparative outdoor study of the efficacy, onset and duration of action, and safety of cetirizine, loratadine, and placebo for seasonal allergic rhinitis. J Allergy Clin Immunol 1996;97:617-26. 9) B) Salmun LM, et al. Loratadine versus certizine: assessment of somnolence and motivation during the workday. Clin Therapeutics 2000;22:573-82.