A new paradigm for determining diagnosis and prognosis is evolving that is set to revolutionise the way patients are managed, thereby increasing effectiveness without compromising productivity or quality of care provided. It is called ‘practising evidence-based medicine’.
Clinical practice is changing due to rapid advances in technology and the clinicians need to change their practice methodology in order to continue to deliver high levels of specialised care. An average clinician today is faced with a multitude of problems. Ever increasing medical knowledge in the form of 27kg of guidelines, more than 3000 new papers per day, 1000 new Medline articles, 46 randomised clinical trials, and the number of biomedical journals alone doubling since 1970. Couple it with the average workload for a clinician of anything between 100 to 200 consultations a week resulting in 5000 to 10000 per year (UK figures).
Every encounter with a patient identifies appear more often than not to reveal some gaps in the understanding of the aetiology, diagnosis, prognosis, or therapy of their illness on the part of most clinicians. Recent research reveals that even as seasoned clinicians generate about five knowledge ‘needs’ for every in-patient encounter, and two ‘needs’ for every three outpatient’s encounters. To bridge these gaps and fulfill the ‘needs’, it has become imperative to practice evidence-based medicine. To evaluate the best evidence and make the necessary corrections, it has become necessary to perform outcomes analysis.
Evidence-based Medicine
The most commonly accepted definition of ‘Evidence-based Medicine’ is that it is the conscientious, explicit and judicious use of current best evidence in making clinical decisions about the care of individual patients.
Dr.DavidEddy of Kaiser Permanente is widely credited for having coined the term ‘Evidence-based Medicine’. He explains that when there is evidence that something works and is good and benefits the patient, one should do it. When one has evidence that there is no benefit and that it is ineffective, where it is going to harm the patient either directly or indirectly by stealing resources, one should not do it. When there is insufficient evidence to determine either way, one must be conservative, relying on individual clinician discretion. This means the integration of individual clinical experience with the best available external clinical evidence from systematic research.
Individual clinical expertise means the proficiency and judgment that individual clinicians acquire through clinical experience and clinical practice.
Increased expertise is reflected in many ways, but especially in more effective and efficient diagnosis and in the more thoughtful identification and compassionate use of individual patients’ predicaments, rights, and preferences in making clinical decisions about their care.
Best available external clinical evidence means clinically relevant research, often from the basic sciences of medicine, but especially from patient centered clinical research into the accuracy and precision of diagnostic tests (including the clinical examination), the power of prognostic markers, and the efficacy and safety of therapeutic, rehabilitative, and preventive regimens. Such evidence validates previously accepted diagnostic tests and treatments, and where necessary, replaces them with new ones that are more powerful, more accurate, more efficacious, and safer.
Any practice that uses best evidence, while delivering healthcare, is called ‘evidence-based practice’. Such practice generally is the application of up-to-date information from relevant, valid research about the effects of different diagnostic tests and the predictive power of prognostic factors across the broad field of healthcare, including education, practice management and health economics.
Outcomes Analysis
Outcomes analysis is a non-prejudiced analysis of the outcome of an event, episode or encounter. It is used by the nurses as a measure for patient acuity, the hospital administrators and the health economists to perform cost-benefit analysis, and care providers to justify their clinical decisions regarding treatment plans as well as to validate the clinical protocols when followed. Not only are the outcomes of an event, but also the variances between different treatment methodologies are measured. In a clinical setting, it also allows one to find out how well a particular treatment method is faring.
The aim of outcomes analysis is not to find ways to reduce, but to put a cost on a given treatment regimen. It is entirely possible that the measure ends up recommending a treatment regimen that is costlier because it is more beneficial. It can also be used as a costing method to perform cost-benefit analysis and effort estimations.
The various parameters that are evaluated are as follows:
- Quality Adjusted Life Year (QALY)
• Patient preferences
• Resource use including time, money, equipment, hospital stay, etc.
• Nursing Acuity
• Cost effectiveness.
The Process
Evidence-based medicine converts reading and appraising the information into using it to benefit individual patients while concurrently adding to the clinician’s knowledge base. Instead of reading all the articles in a journal, it is better to focus on the ones that are related to specific problems. It is critical that one follows a constructive method of framing the pertinent question related to the problems on hand, and then searching for evidence related to that question. The aim is to keep one’s knowledge at a more usefully productive level.
Evidence-based medicine requires some knowledge regarding the calculations and interpretations of relative risks, absolute and relative risk reductions, odds ratio, numbers needed to treat/harm, sensitivity, specificity, likelihood ratio, pre-test probabilities, etc.
Outcomes analysis is an inherent requirement for the total adoption of evidence-based medicine. Without the results arrived at from analysis of outcomes being added to the knowledge repository for future reference, the internal expertise is not enriched, i.e. there is no value-add of the process for future patient with similar clinical picture demanding the answers to similar questions.
Calculations in Evidence-based Medicine
There two broad types of calculations that go into effective practice of evidence-based medicine. These are evaluating the following:
- Evidence regarding the efficacy of a certain treatment as opposed to another, including no treatment. Mostly results from randomised clinical trials are used. It is a type of prognostic assessment
2. Evidence regarding a particular diagnostic test or patient finding. It is a type of diagnostic assessment.
For Assessment of Treatment Protocols
Absolute Risk Reduction (ARR): This is the difference in the risk of the outcome between patients who have undergone a particular method of treatment (called experimental) and those who have not undergone that method (called control). This measure tells us the percentage of patients who were spared the adverse outcome as a result of having received the experimental rather than the control therapy. It is calculated as |EER – CER|
Relative Risk (RR): This is the ratio of the risks in the experimental to the control groups and is represented as a percentage of the original risk. It is calculated as |EER – CER| / CER
Relative Risk Reduction (RRR): This is the extent to which an experimental treatment reduces a risk, in comparison with the control, and assesses the effectiveness of a treatment. This is calculated by subtracting the RR from 1. If the RRR is 0, then the experimental treatment is no different from the control. The relative risk reduction is fundamentally an estimate of the percentage of baseline risk that is removed as a result of the experimental therapy. It is calculated as |CER-EER| /CER. A risk is considered to be baseline if the risk of an adverse event among patient either in the control group or who are receiving the standard or inferior therapy.
Numbers Needed to Treat (NNT): This is the most recently introduced measure of treatment efficacy, and is defined as the number of patients who need to be treated to achieve 1 additional good outcome. It the reciprocal of the ARR, and is measured if the outcome of the experimental treatment is positive. When the outcome is negative, numbers needed to treat (NNH) is measured. This is the number of patients who need to be treated with the experimental method to cause 1 additional patient being harmed as compared to those who are treated with the control method. The thumb rule is that if EER > CER, then calculate NNT else calculate NNH. The numbers needed changes inversely in relation to the baseline risk. If the risk of an event doubles, one needs to treat only half as many patients to achieve the same results, and if the risk decreases by a factor of four, one needs to treat four times as many. It is calculated as 1 / ARR
Odds Ratio (OR): These are the odds of an event (usually adverse) occurring and are usually the measure of choice in the analysis of case-control studies. Generally, the odds ratio has certain optimal statistical properties that make it the fundamental measure of association in many types of studies. The statistical advantages become particularly important when data from several studies are combined, as in meta-analysis. Among such advantages, the comparison of risk represented by the odds ratio does not depend on whether the investigator chose to determine the risk of an event occurring (e.g. fatal) or not occurring (e.g. improvement). This is not true for relative risk where the definitions of experiment and control can alter the figures. In some situations the odds ratio and the relative risk will be close like in case control studies of a rare disease. The odds ratio is calculated by dividing the odds in the experimental group by the odds in the control group. It follows that efficacious treatments generate odds ratios that are less than 1, which is analogous to the relative risk for the adverse event (EER/CER) being less than 1.
Meta-analysis: It is a statistical procedure that integrates the results of several independent studies considered to be ‘combinable’ and should be viewed as an observational study of the evidence. Well-conducted meta-analyses allow a more objective appraisal of the evidence than traditional narrative reviews, provide a more precise estimate of a treatment effect, and may explain heterogeneity between the results of individual studies. Ill conducted meta-analyses, on the other hand, may be biased owing to exclusion of relevant studies or inclusion of inadequate studies. Methods used for meta-analysis use a weighted average of the results, in which the larger trials have more influence than the smaller ones. Results from each trial are graphically displayed, together with their confidence intervals. A black square and a horizontal line represent each study, which correspond to the point estimate and the
95 per cent confidence intervals of the odds ratio. The 95 per cent confidence intervals would contain the true underlying effect in 95 per cent of the occasions if the study were repeated again and again. A solid vertical line is drawn that corresponds to no effect of treatment (odds ratio 1.0). When the confidence interval of any study includes 1 the difference in the effect of experimental and control treatment is not significant at conventional levels (p>0.05). An area made of black squares reflects the weight of the study in the meta-analysis. A diamond shape represents the combined odds ratio, calculated using a fixed effects model, with its 95 per cent confidence interval. It should be noted here that a result that is meta-analytical in origin should be viewed with a higher degree of confidence than one that is not.
For Assessment of Findings / Diagnostic Tests
Sensitivity: Representing the accuracy of a positive test result, this is the proportion of patients who have the target disorder and also test positive for the diagnostic test.
Specificity: Representing the accuracy of negative test result, this is the proportion of patients who do not have the target disorder and also test negative for the diagnostic test.
Likelihood Ratio (LR): This measures how likely the presence (or absence) of a finding (or diagnostic test) would result in ruling in (or out) a diagnosis. The ratio is used to assess how good a diagnostic test (or finding) is, to help in selecting an appropriate diagnostic test or a sequence thereof. It is better than sensitivity and specificity numbers because it is less likely to change with the prevalence of the disorder, can be calculated for several levels of signs and symptoms, can be used to combine the results of multiple diagnostic tests, and can be used to calculate the post-test probability for a target disorder. A likelihood ratio greater than 1 produces a post-test probability that is higher than the pre-test probability, while an LR lesser than 1 accomplishes the reverse, thereby altering the chances of finding the target disorder. A diagnostic test result with a very high LR (e.g. >10) would virtually rule in a disease when found positive, while one with a very low LR (e.g. <0.1) that would virtually rule out the chance that the patient has the disease.
Pre-test Probability (Priori of Bayes Rule): It is defined as the probability of the target disorder before a diagnostic test result is known. It is especially useful for (1) interpreting the results of a diagnostic test, (2) selecting one or more diagnostic tests, (3) choosing whether to start therapy without further testing (treatment threshold) or while awaiting further testing, (4) deciding whether it’s worth testing at all (test threshold). The probability of the target disorder can be calculated as the proportion of patients with the target disorder, out of all the patients with the symptoms, both those with and without the disorder.
Post-test Probability (Posteriori of Bayes Rule): It is defined, as the probability of the target disorder being present after the diagnostic test result is known.
Bayes Rule – This is based on the theorem proposed in mid-nineteenth century by Rev. Thomas Bayes and is on probability inference. It is a means of calculating the probability that it will occur in future trials from the number of times an event has occurred.
