Evidence-Based Clinical Practice: Clinical Practice Guidelines Cy h iia a Jac us C y nt nth J ac kev iicc iiu BScPhm, PharmD Q C a rd i o l o g y, F C S H P, FA H A , F C C P, FAC C D,, M S c , B C P S - A AQ P ro fe s s o r o f P h a r m a c y P ra c t i c e ©2019 Cynthia Jackevicius All Rights Reserved Reproduction and/or broadcast of this material without the explicit written permission of Dr. Cynthia Jackevicius is prohibited. Objectives Define and describe clinical practice guidelines (CPGs). Identify the major evidence sources for locating CPGs. Describe the three elements of the framework analyzed in CPGs. Evaluate the validity of a CPG, according to the framework described in the JAMA Users’ Guide on Summarizing the Evidence. Evaluate the clinical importance of the results from a CPG. Determine whether a CPG has an explicit and sensible process for identifying, selecting and combining evidence to create the guideline. Identify whether a CPGs is current and up to date. Assess whether a CPG has been subject to peer review and how bias may be present in developing CPGs. Describe and interpret the grades and levels of evidence for commonly used CPGs. Describe the “killer Bs” when determining if CPGs are applicable to your practice setting. When given a patient scenario, evaluate the applicability of the CPGs to the specific patient. Case SP is a 22-month-old, 24.4 lb female who is being seen by her pediatrician. You are the clinical pharmacist at the medical practice, and working with the pediatrician that day. According to her mother, SP has been crying since last night with constant ear tugging while complaining about something stuck in her ear. Her fever last night was 100.5 F. Her mother also tells you that since she started her new daycare program a few months ago, SP has gotten sick twice and is just getting over her last cold. Her temperature in the office is 100.7 F. The pediatrician suspects acute otitis media. Is an antibiotic appropriate for this patient? P : the child with acute otitis media , I : antibiotic , C : no antibiotic , C: infection Steps for EBP Ask and frame the question Acquire the evidence Act Ask Appraise the evidence Apply and integrate the evidence with our clinical expertise and patient values and circumstances Act on your findings Apply Acquire Appraise Ask: PICO P I C O What is Our Clinical Question? In [population], does [intervention] [versus comparator] affect [outcome]? Types of Foreground Questions Therapy (antibiotic) Harm Diagnosis Prognosis Search: Pyramid of EBP Resources Examples Evidence based textbooks Guidelines Evidence based abstraction journals Systematic Reviews Original journal articles Summaries and Guidelines PreAppraised Research NonPreAppraised Research 8 What Are Guidelines? Clinical Practice Guidelines (CPGs) Systematically developed statements to help practitioners and patients make decisions about appropriate care in specific clinical situations Attempt to bridge between producers and consumers of health care research Aim to influence what clinicians to in order to reduce practice variation, lower costs, and improve health outcomes of patients Guideline Characteristics Often address multiple clinical questions Usually informed by existing evidence May include economic or patient preference considerations Often attempt to grade evidence and recommendations Require strategies for dissemination Why Do We Need CPGs? Evidence Recommendations Values Impact of Guidelines Two big questions: ◦ Do guidelines change what doctors do? ◦ Do guidelines change patient-important outcomes? One systematic review of studies answers these questions ◦ 11 studies looked at the effect on patient-important outcomes ◦ 9 of those studies showed “improvements” Grimshaw, et al. Lancet. 1993;342:1317-22. Performance Matters! Lower Death Rate if Guidelines are Followed In-Hospital Mortality (%) 7 6 5.9 5.0 5 4.6 4 3.6 3 2 1 0 <65% 65%-75% 75%-80% >80% Adherence to Guidelines by the Hospital Peterson ED. CRUSADE. Annual Scientific Sessions of the AHA; November 17, 2002; Chicago, Ill. Physician Attitudes Survey of 201 internal medicine physicians: ◦ 74.7% felt that the guidelines were applicable to over half their patients ◦ 76.5% agreed that guideline-based performance measures encourage evidence-based medicine ◦ Over 70% perceive increased bias when committee members have industry-sponsored research or presentations Shea, et al. Am J Med Qual. 2007;22:170-6. Barriers to Implementation Knowledge Lack of Familiarity Lack of Awareness Attitudes Behavior Lack of Agreement with Specific Guidelines Lack of Outcome Expectancy Lack of Agreement with Guidelines in General Lack of Self-Efficacy External Barriers Guideline Factors Environmental Factors Lack of Motivation/Clinical Inertia Cabana, et al. JAMA. 1999;282:1458-65. Finding Guidelines National Guideline Clearinghouse www.guidelines.gov ◦ Was a reputable source to find multiple guidelines on a topic ◦ Defunded by the government ◦ Picked up by a commercial entity ECRI Guidelines Trust ECRI Guidelines Trust https://guidelines.ecri.org/ ◦ Set up to replace www.guidelines.gov ◦ Must sign-in to access, making it more difficult to use than guidelines.gov ◦ Still a work in progress – watch for future developments Finding Guidelines Professional Association Websites ◦ ACC/AHA – Cardiology ◦ JNC8, AHA/ACC – Hypertension ◦ NCEPIV, AHA/ACC, NLA – Lipids ◦ ADA – Diabetes ◦ IDSA – Infectious Diseases Secondary Resources ◦ PubMed with guidelines filter ◦ EMBASE with guidelines filter Sometimes, you just have to Google! Activity – Find the Guideline Treatment of deep vein thrombosis ◦ PubMed ◦ “Chest Guidelines” ◦ Compare the two ACC/AHA heart failure Community-acquired pneumonia Critical Appraisal Critical Appraisal of CPGs Is the evidence from the CPG valid? If valid, is the evidence from the CPG important? If valid and important, can you apply to the patient? Appraising Validity Steps Involved Is the evidence valid? • Do the recommendations consider all relevant patient groups, management options and possible outcomes? • Are there systematic reviews of evidence that estimate the relative effect of management options on relevant outcomes? • Is there an appropriate specification of values and preferences associated with outcomes? • Are the guidelines timely and current? Are There Systematic Reviews? The systematic review is the ideal method to evaluate the effect of the different options on relevant outcomes for the condition in the guideline. How was the evidence acquired? Did they do a systematic review? Is there a summary of the evidence used to make recommendations? HOW DO THEY GET THE EVIDENCE FOR RECOMMENDATION ? Appropriate Specification of Values and Preferences? Who was on the panel? Did they receive money for the review ???? Any specialist ? Multiple viewpoints ? ◦ They might have conflicts of interest or other biases. What were their priorities? ◦ Did they value one outcome over another? ◦ Should state underlying value judgments Do recommendations consider relevant patient groups, options, outcomes? Think of their PICO ◦ Subgroups of patients? ◦ Management options/comparators? ◦ Patient-important outcomes? Patient already had DVT : looking for treatment not prevention Check for therapeutic option Cover the outcome or not Example of Values Statement: CCS Atrial Fibrillation Guidelines CCS AFib Guidelines 2012 Timeliness Consensus panels take time to deliberate Check the dates of the: ◦Guideline itself ◦Evidence search ◦References list Validity Summary SPOT S • Systematic Review was conducted to obtain evidence P • Priorities and values were described O • Outcomes, Options, and Patients were sufficiently broad T • Timely and current Appraising Importance of Results Steps Involved Is the evidence important? • What are the key recommendations? • How strong are the recommendations? Key Recommendations : HOW PRACTIAL THEY ARE What is the clinical importance? ◦ Risks versus benefits of recommendations ◦ Absolute risk versus relative risk reductions Are they practical? (Ex: not available in US, too expensive) ◦ Are the interventions described in sufficient detail? ◦ Ease of implementation Strength of Recommendations Should consider: ◦ Quality of evidence ◦ Magnitude of benefits ◦ Magnitude of harms ◦ Burden to the patient and the health care system ◦ Costs Hierarchy of Evidence High Systematic reviews of RCTs Single RCT Systematic review of observational studies Single observational study Unsystematic clinical observations Low Example Systems GRADE American College of Chest Physicians (ACCP) American College of Cardiology/American Heart Association (ACC/AHA) American Diabetes Association (ADA) American Academy of Pediatrics/American Academy of Family Physicians GRADE Approach Grades of Recommendation, Assessment, Development and Evaluation (GRADE) Working Group ◦ Developed explicit criteria for evaluating evidence and formulating recommendations ◦ Many organizations and countries involved Two Components of GRADE Quality of Evidence ◦ High ◦ Moderate ◦ Low ◦ Very low Strength of Recommendation ◦ Strong ◦ Weak (conditional) GRADE: Quality of Evidence Quality Rating Definition High High confidence that the evidence reflects the true effect. Future research unlikely to change confidence in estimate of effect. Moderate Moderate confidence that the evidence reflects the true effect. Future research my change confidence in estimate of effect. Low Low confidence that the evidence reflects true effect. Future research will likely change both confidence in and the estimated effect. Very Low Very low confidence that the evidence reflects the true effect. Any estimates are very uncertain. Hierarchy of Evidence is Too Simplistic High Systematic reviews of RCTs Single RCT Systematic review of observational studies Single observational study Unsystematic clinical observations Low Reasons to Downgrade Quality Reason What to Assess Limitation of methodology (bias) FRISBE of individual studies Inconsistency of results Heterogeneity Indirectness of evidence PICO Applicability Imprecision of results Wide confidence intervals Low number of events Publication bias Funnel plots Egger’s, Begg’s GRADE: Rating Quality of Evidence *Quality of study moves down 1 or 2 grades. †Quality of study moves up 1 or 2 grades. Go to kahoot.it on your browser Study #1: Rate the Quality Study #1: Clinical trial Randomized Concealed Open label : not blinding Per protocol analysis : bad Surrogate endpoint : Bad Wide CI : Bad, imprecise What is the study quality? Because High Moderate Low Very low Study #2: Rate the Quality Study #2: Cohort study Hazard ratio = 3.5 – big magnitude Dose response gradient – upgrade to high Exposure before outcome Confounding adjusted using propensity scores Narrow CI – high What is the study quality? High Moderate Low Very low Two Components of GRADE Quality of Evidence ◦ High ◦ Moderate ◦ Low ◦ Very low Strength of Recommendation ◦ Strong ◦ Weak (conditional) Scenario 1: Recommend or Not?? Benefits Harms Scenario 1: GRADE: Strong or Weak? Benefits Harms GRADE: Clinical Implications Strong Recommendations ◦ ◦ ◦ ◦ ◦ For clinicians: Beneficial outcome: Most individuals should receive intervention Harmful outcome: Most should not receive intervention For patients: Beneficial outcome: Most individuals would want the recommended intervention; only a small proportion would not ◦ Harmful outcome: Most would not want to receive intervention Weak Recommendations ◦ For clinicians: Should offer the intervention and include discussion to make decision (shared decision-making); examine the evidence and be prepared to weigh individualized pros and cons ◦ For patients: May be split on whether or not they want the intervention Factors that Determine Strength of Recommendation Quality of the evidence Balance between desirable & undesirable effects Values and preferences Costs (resource allocation) Scenario 2: Recommend or Not? Benefits Harms Live Longer Decreased QOL Prevent Heart Attack Increase Stroke Scenario 2: GRADE: Strong or Weak?? Benefits Harms Live Longer Decreased QOL Prevent Heart Attack Increase Stroke Scenario 3: Recommend or Not? Benefits Harms Scenario 3: GRADE: Strong or Weak?? Benefits Harms ACC/AHA Rating System (OLD) SIZE OF TREATMENT EFFECT PRECISION OF TREATMENT EFFECT Class I: Benefit >>> Risk Level A ◦ Procedure SHOULD be performed Class IIa: Benefit >> Risk ◦ Additional studies needed ◦ It is REASONABLE to perform Class IIb: Benefit ≥ Risk ◦ Procedure MAY BE CONSIDERED Class III: Risk ≥ Benefit ◦ Procedure NOT HELPFUL and MAY BE HARMFUL ◦ Multiple populations evaluated ◦ Data derived from multiple RCTs or metaanalyses Level B ◦ Limited populations evaluated ◦ Data derived from a single RCT of nonrandomized studies Level C ◦ Very limited populations evaluated ◦ Only consensus opinion of experts, case studies or standards of care JACC 2009;54:2205-41. ACC/AHA Rating System (NEW) ADA Levels of Evidence Diabetes Care 2019;42:S1-2. AAP/AAFP Recommendations Statement Definition Strong recommendation Anticipated benefits clearly exceed harms and the quality of evidence is excellent Recommendation Anticipated benefits exceed harms, but the quality of evidence is not as strong Option Either the quality of evidence is suspect or carefully performed studies have shown little clear advantage to one approach over another No recommendation There is a lack of pertinent published evidence and that the anticipated balance of benefits and harms is unclear Pediatrics 2004;113:1451-65. Applying to the Patient Steps Involved Can the results be applied to my patient? • Is the patient the intended target of the guideline? • Did the guideline make exceptions for your patient population? • Do the recommendations fit the current clinical scenario for your patient? • Do the recommendations match your patient’s preferences? “Killer Bs” Burden of illness: Is it too low in the community or is our patient’s PEER too low? Beliefs of individuals and the community: Does it go against the guideline? Bad Bargain: How high is the opportunity cost to implement? Is it a bad deal? Barriers: Geographical, organizational, traditional, authoritarian, legal, behavioral… too high? Questions?? Resources Chapter 26. How to use a patient management recommendation. Guyatt, ed. User’s Guide to the Medical Literature. American Medical Association, 2015. Chapter 28.1. Assessing the strength of recommendations: The GRADE approach. Guyatt, ed. User’s Guide to the Medical Literature. American Medical Association, 2015.