2EBM SKILLS - STATISTICS CHANCE - p = 1 in 20 (0.05).> 1 in 20 (0.051) = not significant< 1 in 20 (0.049) = statistically significantCONFIDENCE INTERVALSwhat is the range of values between which we could be 95% certain that this result would lie if this intervention was applied to the general population
3EBM SKILLS - A BASIC INTRODUCTION CHANCE, BIAS, CONFOUNDING VARIABLES
4TYPES OF STUDY - HYPOTHESIS FORMING CASE REPORTS / CASE SERIESCROSS SECTIONAL / PREVALENCE STUDIES measure personal factors & disease states - hypothesis FORMING - cannot indicate cause & effectCORRELATIONAL / ECOLOGICAL / GEOGRAPHIC STUDIES. prevalence &/or incidence measurement in one population c/w another pop.
5TYPES OF STUDY - HYPOTHESIS TESTING CASE CONTROL STUDIES
6CASE CONTROL EXAMPLE -SMOKING & LUNG CANCER DISEASECases ControlsEXPOSURE Yes a bEXPOSURE No c dOdds Ratio = ad/bc (1 = no association, > 1 = possible association, < 1 = protective effect)(lung cancer)EXPOSURE Yes(smoking) NoThe odds ratio would therefore be 56 x 246 = = 8.6.7 x
7TYPES OF STUDY - HYPOTHESIS TESTING COHORT STUDIES
8This is expressed as a divided by c . COHORT STUDIESOUTCOMEYes NoExposed a bNot exposed c dAttributable risk (absolute risk or risk difference)"What is the incidence of disease attributable to exposure"Answer = a - c.Relative risk "How many times are exposed persons more likely to develop the disease, relative to non-exposed persons?" i.e. the incidence in the exposed divided by the incidence in the non-exposed.This is expressed as a divided by ca+b c+d
9Exposed ( on oral contraceptive ) 41 9996 COHORT STUDY EXAMPLEDeep vein thromboses (DVT) in oral contraceptive users. (Hypothetical results).OUTCOME (DVT)Yes NoExposed ( on oral contraceptive )Not exposed (not on o.c.)These results would give an attributable risk of 34 and a relative risk of 6 - significantly large enough numbers to indicate the possibility of a real association between exposure and outcome. However, the possibility of biases very often arises.
11RANDOMISED CONTROLLED TRIALS OUTCOMEYes NoComparison intervention a bExperimental intervention c dRelative risk reduction: “ How many fewer patients will get the outcome measured if they get active treatment versus comparison intervention”a /a+b c/c+da/a+bAbsolute risk reduction: “What is the size of this effect in the population”a/a+b c/c+d
12RCT EXAMPLE - 4S STUDYSTABLE ANGINA OR MYOCARDIAL INFARCTION MORE THAN 6 MONTHS PREVIOUSLYSERUM CHOLESTEROL > 6.2mmol/lEXCLUDED PATIENTS WITH ARYHTHMIAS AND HEART FAILUREALL PATIENTS GIVEN 8 WEEKS OF DIETARY THERAPYIF CHOLESTEROL STILL RAISED (>5.5) RANDOMISED TO RECEIVE SIMVASTATIN (20mg > 40mg) OR PLACEBOOUTCOME DEATH OR MYOCARDIAL INFARCTION (LENGTH OF TREATMENT 5.4 YEARS ) WERE THE OUTCOMES
131/ARR = NUMBER NEEDED TO TREAT. RCT EXAMPLE - 4S STUDYOUTCOME (death)Yes NoComparison intervention (placebo)Experimental intervention (simvastatin)The ARR is (256/2223) - (182/2221) = =The RRR is 0.033/0.115 = 0.29 or expressed as a percentage 29%.1/ARR = NUMBER NEEDED TO TREAT.1/0.033 = 30.i.e. if we treat 30 patients with IHD with simvastatin as per 4S study, in 5.4 years we will have prevented 1 death.
15Why are RCTs the “gold standard” Breast cancer mortality in studies of screening with mammography; women aged 50 and over (55 in Malmo study, 45 in UK)
16SCREENING - WILSON & JUNGEN (WHO, 1968) IS THE DISORDER COMMON / IMPORTANTARE THERE TREATMENTS FOR THE DISORDERIS THERE A KNOWN NATURAL HISTORY & “WINDOW OF OPPORTUNITY” WHERE SCREENING CAN DETECT DISEASE EARLY WITH IMPROVED CHANCE OF CUREIS THE TEST ACCEPTABLE TO PATIENTSSENSITIVE AND SPECIFICGENERALISABLECHEAP / COST EFFECTIVEAPPLY TO GROUP AT HIGH RISK
17DISEASE PRESENT ABSENT SCREENING TEST POSITIVE A B NEGATIVE C D Sensitivity = a/a+c; Specificity = d/b+d;positive predicitive value = a/a+b;negative predicitve value = d/c+d.
18Value of exercise ECG in coronary artery stenosis DISEASEPRESENT ABSENTTEST POSITIVENEGATIVESensitivity = a/a+c = 60%; Specificity = d/b+d = 91%;positive predicitive value = a/a+b = 93%;negative predicitve value = d/c+d = 55%.
19Sensitivities and Specificities for different tests Alcohol dependency or abuse (as defined by extensive investigations in medical and orthopaedic in patients)SENS SPECGGT % %MCV % %LFTs % %“Yes” to 1 or > of CAGE ?s 85% %“Yes” to 3 or > of CAGE ?s 51% %
20MAKING SENSE OF THE EVIDENCE - ARE THESE RESULTS VALID - i. e MAKING SENSE OF THE EVIDENCE - ARE THESE RESULTS VALID - i.e. should I believe them?Randomised (where appropriate)?Drop outs and withdrawals?Followup complete?Analysed in the groups to which randomised?-“Intention to treat”.
21MAKING SENSE OF THE EVIDENCE- ARE THESE RESULTS USEFUL. - i. e MAKING SENSE OF THE EVIDENCE- ARE THESE RESULTS USEFUL?- i.e. should I be impressed by them, are they relevant to my patients (GENERALISABLE)How large was the treatment effect?How precise was the estimate of treatment effectWere all important clinical outcomes considered?Do benefits outweigh risks?