Exam questions in Statistics and evidence-based medicine, spring 2012 2. sem. Medis/Medicin, Modul 2.4. Statistics : ESSAY-TYPE QUESTION 1. In order to estimate the effect of the drug Captopril on blood pressure the drug is administered to a random sample n = 49 subjects. Each subjects blood pressure was recorded before taking the drug and then 30 minutes after taking the drug. The appropriate test to show that Captopril results in a blood pressure decrease in patients 30 minutes after taking it is: (explain your choice) A B C D E the t-test for two independent samples the Mann-Whitney test the Kruskal-Wallis test paired-samples t-test none of the above Max point: 9 MULTIPLE-CHOICE QUESTION 2. Which of the following is not true? A B C D E The mean and the median are both measures of central tendency. The standard deviation and the variance are both measures of spread. Median is always equal to mean. For normal distributions, about 95% of observations lie within approximately two standard deviations of the mean. Confidence Interval is an interval of values that contains the true value of the population parameter with specified confidence. Max point: 5 1
MULTIPLE-CHOICE QUESTION 3. The histogram below indicates that the distribution of the data is: A B C Normal Skewed Symmetric D all of the above E none of the above Max point: 3 ESSAY-TYPE QUESTION 4. A researcher reported a study in which healthy men aged 45 to 65 received either a calcium supplement or a placebo for 12 weeks. He reported, The calcium group had lower blood pressure compared with the placebo group (p-value=0.008). (Note: blood pressure is measured in millimeters of mercury, abbreviated mm Hg) Which of the following sets of data is consistent with the researcher s statement? Why? A Difference = 10 mm Hg with 95% confidence interval 2.4 to 17.6. B Difference = 20 mm Hg with 95% confidence interval -5 to 45. Max point: 9 2
ESSAY-TYPE QUESTION 5. A study recruited 10,000 smokers and 10,000 non-smokers aged 30 to 39 and followed them for 20 years. 200 smokers and 50 non-smokers got throat cancer. (Abbreviation used: OR = odds ratio) Questions: a) What is the odds for smoking among those who got throat cancer? b) What is the odds for smoking among those who did not get throat cancer? c) What is the sample OR for smoking among those who got throat cancer compared to those who did not get throat cancer? d) What conclusion can you make based on this OR? e) What else would you need to calculate to estimate the population OR? (do not perform calculations) f) Explain what is the difference between the sample OR and population OR? Max point: 18 3
ESSAY-TYPE QUESTION 6. A study was conducted among former mine workers to assess the misclassification probabilities of a Chest X-Ray in the diagnosis of silicosis (a fibrotic chronic lung disease that is caused by prolonged occupational inhalation of silica dust). Chest X-Rays taken one month before the death of the subject were used and histological examination of the lung tissue (obtained at post mortem irrespective of the cause of death) was undertaken. Histology results were considered to be the reference or gold standard. The following results were obtained: Questions: a) Complete the missing figures in the remaining cells. b) What are the sensitivity and the specificity of Chest X-Ray? c) Interpret/explain the results (what do sensitivity and specificity show?). Max point: 13 4
ESSAY-TYPE QUESTION 7. A farmer wishes to compare two diets, to figure out which diet fattens his pigs up best. He has 10 pigs with which to compare the two diets, so 5 pigs are randomly assigned to each diet. He then weighs the pigs at the start of the experiment, and again 30 weeks later. Finally he calculates a gain in weight for each pig. (Weight gain = Weight after 30 weeks Weight at baseline) Assume that the weight gain for each diet is normally distributed. Questions: a) What test would you perform to find out, whether there is a difference between the diets? Do the data satisfy the assumptions for the test? b) What null hypothesis do you choose for your test? (Use mathematical terms to describe your hypothesis) c) Perform the test and outline the main results. d) Explain/interpret the results. (For example, answer following questions: What does a p-value (or confidence interval) show? What conclusions can you make based on this p-value (confidence-interval?) e) What diet should a farmer use to fatten his pigs up best? Why? Max point: 28 5
Evidens-based medicine : EXTENDED MATCHING 8. Evidens Begreber: A. De epidemiologiske metoderegler B. Evidensbaseret medicin og evidensbaseret praksis C. Evidensbaseret viden Opgave: Vælg fra listen af begreber det, der passer bedst til hvert af de nedenstående udsagn. Er viden af høj metodologisk kvalitet Er den samvittighedsfulde, eksplicitte og velovervejede brug af den tilgængelige bedste evidens Indeholder formål, design, metoder m. m. for en videnskabelig undersøgelse Udøves på baggrund af referenceprogrammer og evidensbaserede kliniske retningslinjer / vejledninger Skal følges for at en undersøgelse kan opnå høj evidens Opnås ved at integrere kliniske ekspertise med den bedst tilgængelige kliniske evidens fundet ved systematisk litteratursøgning Max point: 4 6
EXTENDED MATCHING 9. Evidens Begreber: A. Meta-analyser af RCT studier B. Kohorde-undersøgelser C. Systematiske reviews D. Case-control undersøgelser Opgave: Vælg fra listen af begreber det, der passer bedst til hvert af de nedenstående udsagn. Systematisk indsamling og analyse af al eksisterende evidensbaseret viden indenfor et område Sammenfattende statistisk analyse af resultater af tidligere udførte undersøgelser Prospektiv (fremadrettet) undersøgelse af en årsagssammenhæng Indeholder ofte også - meta-analyser og dermed statistiske beregninger Retrospektiv /bagudrettet) undersøgelse af en årsagssammenhæng Rangerer sammen med systematiske reviews højest i evidens-hierarkiet fra Oxford Center for Evidence-Based Medicine Max point: 5 7
MULTIPLE CHOICE OPGAVE 10. Hvilken type videnskabelig undersøgelse/studie design rangerer højest i evidens-hierarkiet fra Oxford Center for Evidence-Based Medicine? A Metaanalyser af RCT studier B Case-control studier C Systematiske reviews D Dobbelt blindede kohorte studier E Randomiserede kontrollerede forsøg (RCT) Max point 2 MULTIPLE CHOICE OPGAVE 11. Begrebet evidens betyder i folkesundhedsmæssig sammenhæng: A Ekspertviden B At behandlinger baseres på ekspertviden og klinisk evidens C Viden af høj metodologisk kvalitet D At behandlinger baseres på faglige erfaringer E Viden der er baseret på gentagne forsøg Max point 2 MULTIPLE CHOICE OPGAVE 12. Hvilken af følgende er en medicinsk database, kendt for at stille specielt høje krav til, at de epidemiologiske metoderegler er overholdt? A Dagens Medicin B Cochrane Library C Sundhed.dk D e-medicine E Medicin.dk Max point 2 8