Evidence-based Medicine in Clinical Practice PDF

Evidence-based medicine in Clinical Practice (diagnosis/prognosis/therapy/harm) Chrysanthi Sardeli Obstetrician-Gynecologist Associate Professor of Pharmacology & Clinical Pharmacology School of Medicine, Faculty of Health Sciences,...

Evidence-based medicine in Clinical Practice (diagnosis/prognosis/therapy/harm) Chrysanthi Sardeli Obstetrician-Gynecologist Associate Professor of Pharmacology & Clinical Pharmacology School of Medicine, Faculty of Health Sciences, AUTh e-mail: [email protected] “A 21st century clinician who cannot critically read a study is as unprepared as one who cannot take a blood pressure or examine the cardiovascular system.” Evidence-based medicine (EBM), the practice of appraising the literature in a time-efficient manner to answer a clinical question about, and for, the patient, is a method to be so prepared In the practice of EBM, it is the physician’s duty to find the best and most current information and apply it judiciously for the benefit of the patient The goal is always to provide the best care possible to patients—using one’s clinical expertise to address patient values and expectations for treatment BMJ 2008:337:704-705 Evidence-Based Medicine: How to Practice and Teach EBM. 2005 Churchill Livingstone The 5 Steps Towards Evidence Based Practice 1. Ask the right clinical question: Formulate a searchable question 2. Collect the most relevant publications: Efficient Literature Searching Select the appropriate & relevant studies 3. Critically appraise and synthesize the evidence 4. Integrate best evidence with personal clinical expertise, patient preferences and values: Applying the result to your clinical practice and patient 5. Evaluate the practice decision or change: Evaluating the outcomes of the applied evidence in your practice or patient Clinical Process and knowledge Research evidence sought from literature requirements searches Etiology knowledge about causation Etiognostic Research Patient presentation knowledge about diagnosis Testing History Examination Diagnostic Research Investigations Diagnosis knowledge about prognosis Prognostic Research Therapy Changes prognosis Therapy Research knowledge about therapeutic effectiveness Clinical outcome Patient presenting with a specific NON-LIFE THREATENING condition EBM APPROACH 1. How common is the problem Prevalence 2. Is early detection worthwhile Screening 3. Is the diagnostic test accurate Diagnosis 4. What will happen if we do nothing Prognosis 5. Does this intervention help Treatment 6. What are the common harms of an Treatment intervention 7. What are the rare harms of an Treatment intervention Constructing a clinical question P I C O patient intervention comparison outcome Alternative Outcomes Who? What? Intervention? “How would I What is the “Compared to what describe a group ”Which patient oriented other treatment, of patients treatment, test outcome – better test, or perhaps similar to this or other prognosis? compared to doing particular intervention?” Higher rate of nothing” patient?” cure? Etc.?” Types of questions answered by EBM Diagnosis Question A question concerning the ability of a test to predict the likelihood of a disease Example: in geriatric patients with suspected carotid stenosis, is duplex ultrasound as good as magnetic resonance angiography in detecting significant carotid stenosis? Prognosis Question A question concerning outcome of a patient with a particular condition Example: in diabetic patients with foot ulcers, is the diagnosis of osteomyelitis with MRI as predictive of healing as an audible pulse on Doppler examination? Therapy Question A question concerning the effectiveness of a treatment or preventative measure Example: in patients with migraine headaches without auras, is Depakote more effective than Inderal for prophylaxis of headaches? Harm Question A question concerning the likelihood of an intervention to cause harm Example: for pregnant patients, does the consumption of large amounts of coffee (compared to non-coffee drinkers) increase the rate of spontaneous abortion? Searching the evidence De novo appropriate literature search Refer to critically appraised topics (CATs, summaries of a study and its results that a physician can create for later retrieval, review, and reuse) Article title, the clinical “bottom line,” the clinical question, a summary of the results, comments, the date the study was published, and any relevant citations  How should this patient be treated? Research relevant systematic reviews and meta-analyses Critically appraise them http://www.cochrane.org/resources /handbook/Handbook4.2.6Sep2006. pdf Research relevant individual articles (cohort studies, case control studies, case series, expert opinion) How to research relevant individual articles Ascertain the validity of an individual article, by determining if the study's results and conclusions were accurately deduced the methods used to arrive at the conclusions were free of error and bias How to research relevant individual articles Was there an independent and blind comparison to a reference standard (a method of defining the presence or absence of the disease or condition in question)? Criterion standard (the diagnostic standard for a particular disease or condition, used as a basis of comparison for other (usually noninvasive) tests. Ideally, the sensitivity and specificity of the criterion standard for the disease should be 100%) Those who perform tests and those who interpret the results should be independent of one another and blinded to the diagnostic and reference standard test results Was the diagnostic test evaluated in subjects similar to patients seen in practice? The study's subjects need to have similar baseline characteristics Was the reference standard obtained regardless of the diagnostic test's result? A study should be viewed with suspicion if it does not independently perform the reference standard test and diagnostic test on every participant, even if the reference standard was considered invasive or risky  Diagnostic tests Disease present Disease absent Test positive (+ve) (a) (b) Gold Standard True positive (TP) False positive (FP) Test negative (-ve) (c) (d) False negative (FN) True Negative (TN) Sensitivity: proportion of people with disease who have +ve test Sensitivity= TP = a. TP+FN a+c SNOUT-Highly Sensitive test, Negative result rule OUT disease Specificity: proportion of people free of disease who have -ve test Specificity= TN = d. TN+FP d+b SPIN-Highly Specific test, Positive result rules IN disease Sensitivity/Specificity for dummies High SENSITIVITY = low number of false negatives High SPECIFICITY = low number of false positives Best accuracy if both factors are close to 100% Can only deal with dichotomous tests (only two possible results) Likelihood ratios Positive likelihood ratio (+ve LR) = LR for +ve test defined as likelihood of +ve test among individuals with disease relative to likelihood of +ve test among those without disease Negative likelihood ratio (-ve LR) = LR for –ve test defined as likelihood of -ve test among individuals with disease compared to probability of -ve test among those without disease Cystoscopy (Reference Test) Result Bladder Cancer No Bladder Cancer Positive 44 179 Negative 35 1073 Total 79 1252 Likelihood ratio for positive test result = (44/79)/(179/1252) = 3.90 Likelihood ratio for negative test result = (35/79)/(1073/1252)] = 0.52 Advantages of the likelihood ratio approach The likelihood ratio form of Bayes Theorem is easy to remember: Posttest Odds = Pretest Odds x LR Likelihood ratios can deal with tests with more than two possible results (not just normal/abnormal) The magnitude of the likelihood ratio give intuitive meaning as to how strongly a given test result will raise (rule-in) or lower (rule-out) the likelihood of disease Computing posttest odds after a series of diagnostic tests is much easier than using the sensitivity/specificity method. Posttest Odds = Pretest Odds x LR1 x LR2 x LR3... x LRn Predictive value Disease present Disease absent Test positive (+ve) (a) (b) True positive (TP) False positive (FP) Test negative (-ve) (c) (d) False negative (FN) True Negative (TN) Positive predictive (+PV) value: proportion of people with +ve tests who have disease +PV= TP = a. TP+FP a+b +PV = post-test probability of disease given +ve result Negative predictive (-PV) value: proportion of people with -ve test who are free of disease -PV= TN = d. FN+TN c+d 100-(-PV)= post-test probability of disease given -ve test Nomogram for interpreting diagnostic tests Pretest probability Likelihood ratio Posttest probability Receiver Operating Characteristic (ROC) Curves Signal Detection Theory The sensitivity and specificity of a diagnostic test depends on more than just the "quality" of the test, they also depend on the definition of what constitutes an abnormal test Diagnosis (yes/no) Comparison of two diagnostic tests ROC curves In a ROC curve the true positive rate (Sensitivity) is plotted in function of the false positive rate (100-Specificity) for different cut-off points of a parameter Each point on the ROC curve represents a sensitivity/specificity pair corresponding to a particular decision threshold The area under the ROC curve (AUC) is a measure of how well a parameter can distinguish between two diagnostic groups (diseased/normal) Useful definitions Relative risk measures how much the risk is reduced in the experimental group compared to a control group Absolute risk reduction is the absolute difference in outcome rates between the control and treatment groups: CER – EER The number needed to treat is the number of patients that would need to be treated to prevent one additional bad outcome Diagnostic Test Results of Type 2 Diabetes Mellitus Compared With the Criterion Standard (N=1471) & Statistical Assessment Further questions that need to be answered Can these results benefit the patient? Yes -> apply / No -> start over Is the diagnostic test available and affordable in the physician's clinical setting? How can the physician determine a specific patient's pretest probability of having the target disorder? using the study's inherent disease prevalence the physician's clinical experience, regional and national statistics, and studies specifically developed to determine pretest probabilities for the target disorder Is the pre- to posttest probability shift valuable to the specific patient? Large LR+ values and small LR- values are indicative of significant shifts Prognostic Studies Cohort Studies In such studies, investigators identify a group of individuals with a previously specified characteristic of interest and follow up with them over a predetermined period of time and the results of the group with the disease of interest are then compared with normal subjects who do not have the disease Validity of prognostic studies Are the study subjects truly representative of the population of interest? studies analyze a representative sample of a patient population and generalize the results to a larger population Were the study subjects chosen at a common point in the disease course? the stage of the disease can impact whether the results are valid and applicable to a specific patient Was the follow-up period sufficiently long and complete? Short follow-up periods may allow too little time for an outcome of interest to occur and intention-to-treat analysis should be performed (5-20% rule) Were adjustments made for important prognostic factors? Studies on prognosis often stratify the study group into cohorts based on comorbid conditions or other factors (age, sex, environmental risk factors) that influence prognostic outcomes Were the study subjects and investigators blinded to the measures of interest? usually single-blinding, clinicians are blinded to the method by which the outcome is measured, but not to the characteristics of the patients involved in the study vs. double blinding (physician AND patient are blinded to the method and/or group allocation) Study outcome of prognostic studies How likely are the projected outcomes over time? Depending on endpoint What is the precision of the relationship between the endpoint and the outcome? The principle measure of precision is the 95% confidence interval (CI). The 95% CI quantifies the uncertainty of a measurement by reporting a range of values within which we can be 95% certain that the true value lies for the entire population Validity of Articles on Therapy To assess the validity of a study is to ask if its findings are true and accurate Were subjects randomly assigned to the treatment group? Were all the subjects accounted for and attributed at the end of the study? Was the study “blinded”? Were the study groups similar at the start of the investigation? Were the study groups treated equally? How big was the treatment effect? How precise is the estimate of the treatment effect?  Were subjects randomly assigned to the treatment group? Causation (“X causes Y”) cannot be established using observational studies (eg, case control study, cohort study), only an association between a therapeutic intervention and desired outcome Were all the subjects accounted for and attributed at the end of the study? All enrolled subjects must be accounted for at the end of the investigation All study participants should be analyzed in the groups to which they were originally assigned (‘’intention to treat’’ analysis), allowing for randomization to be preserved Was the study “blinded”? Study participants, clinicians, and investigators should not know which subjects are assigned to the intervention or control group Double-blinding refers to the processes of keeping group assignments concealed from study subjects and investigators, when possible, all measures of progress and improvement for such studies should be concealed from primary investigators through the use of independent evaluators  Were the study groups similar at the start of the investigation? There are always some established risk factors that may affect the study outcome, randomization does not always guarantee an equal balance of demographic factors and medical history between groups and sufficient power How big was the treatment effect? The proportion of those who suffer/exhibit a negative result in the control group determine what is called the control event rate (CER) The proportion of study subjects who suffer/exhibit a negative result in the intervention group determines the experimental event rate (EER) When the EER is subtracted from the CER and that total is then divided by the CER, the result is the RRR (relative risk reduction) The absolute risk reduction (ARR), which is calculated by subtracting the EER from the CER, takes the baseline risk into account The NNT (number needed to treat) is computed by taking the reciprocal of the AAR, or dividing 1 by the ARR  Were the study groups treated equally? More tests or follow-up visits, different clinicians can create bias How precise is the estimate of the treatment effect? A range of values is used to estimate where the true measure would lie, normally, this range of values is expressed by a 95% confidence interval (CI) and can be interpreted as: “We are 95% confident that the true value lies within the given interval” The P value is a statistical expression of significance (e.g. P