1 Bonnie Spring, Ph.D., ABPP Northwestern University Evidence-Based Practice in Clinical Psychology: What It Is, Why It Matters, What You Need to KnowBonnie Spring, Ph.D., ABPPNorthwestern University
2 Why it matters: EBBP Rationale improve quality and accountability for health care practice (IOM, 2001, Crossing the Quality Chasm)shared vocabulary and concepts for transdisciplinary, biopsychosocial research, practice, health care policystimulate development of evidence base for behavioral treatments2005 APA Policy
3 Why it matters: Potentionally Useful Infrastructure Clinical Practice Guidelines:Increasingly based on ongoing systematic review of research (esp. RCTs) (e.g., USPTF, Cochrane, CDC/AHRQ)Research reporting guidelines (CONSORT, TREND, QUOROM)Evidence grading & knowledge synthesis systems (e.g., GRADE, AHRQ)Policy, often coverage/reimbursement implications (VA/DOD, CMS, NICE) (P4P?)Evidence-Based Practice: (life-long learning)Question formulation, search strategies, critical appraisalSUMSEARCHClinical Evidence, First Consult, BMJ updates, Best Evidence Topics, CATCRAWLER, CATBANK – clinical scenario & bottom line
4 Overview History of evidence-based practice (EBP) Core elements of EBP EBP pedagogy in psychologyEBP pedagogy in other health disciplinesUseful infrastructure and potential opportunities for synergy
5 Origins of Evidence-Based Practice Missionary zeal.
6 Emergence of Evidence-Based Medicine Flexner report :155(31!) (1915) 76(1930)Archie Cochrane – epidemiology, health services research - Effectiveness and Efficiency: Random Reflections on Health Services1973 – John Wennberg – widespread practice variationclinical epidemiology determinants and consequences of health care decisions (McMaster U – David Sackett, Gordon Guyatt)1985 – IOM: 15% medical practices evidence-based [2001 Crossing the Quality Chasm]Evidence-based medicine, Brian Haynes & Ann McKibbon – search strategiesCochrane CollaborationSackett - How to Practice and Teach EBMGreen – US; Yellow – Europe; White – Canada. Flexner – quality control; teaching methods & what txs; science. Cochrane – epidemiology, but how do you know what treats – yeast vs vit C in POW. RCT. Wennberg – nomothetic – diff diseases get diff treatments in diff places. Medicare patients in high-cost areas get more expensive care (more ICU, test, hospitalizations) but no better outcome. Because more academic med centers there. McMaster - idiographicAbraham Flexner – proprietary med schools, German educator and not MD. With the approval of the American Medical Association, Abraham Flexner, a college graduate of Johns Hopkins and an educator, undertook a muckraking investigation for the Carnegie Foundation for the Advancement of Teaching of 155 medical schools in the United States and Canada. Medical reformers were convinced that there were too many schools of low quality producing an overabundance of doctors. Flexner's radical conclusion was that only 31 medical schools were fit to survive. His model of the medical school of the future was Johns Hopkins School of Medicine. By 1915 the nation's medical schools had been reduced to 96 and by 1930 there were only 76 schools training the nation's physicians Cochrane – Scottish epidemiologist, screened entire populations - reflections on health services research; RCT as way to determine which treatments most effective & cost-effective to benefit public health. Methods used to determine best evidence introduced by them
8 Alternative Definitions of Evidence-Based Practice NomotheticGuidelines: (public health, medicine) – focus on problem/disorder & level of evidence for practices (based on systematic review) (e.g., NICE, VA, apa)ESTs: (psychology) focus on intervention (& disorder)EBP: (psychology, medicine, nursing, social work) focus on decision-making about individual patientsGuidelines – Not standards, aspirational, but do tend to influence coverage policies – rationing care, often w/ good reason. Nomothetic; eliminate practice variation; RCTs of treatments for disorders; way to disseminate education; implement best practices.Somewhat different evidence criteria for ESTIdiographic……….. Lifelong Learning
9 APA Policy Statement adopted August 2005 “Evidence-based practice in psychology is the integration of the best available research with clinical expertise in the context of patient characteristics, culture, and preferences.”-adapted fromIOM, 2001 & Sackett, 2000
10 Clinical Decision-Making Best available research evidenceClinical Decision-MakingPatient’s values, characteristics, and circumstancesClinical Expertise
11 Syllabus ProjectPrompt: Does anyone on the list teach a course on evidence-based practice (EBP)? Specifically, I am searching for syllabi that cover one or more "legs" of the three-legged EBP stool: a) research evidence, b) clinical expertise, c) patient values, preferences, characteristics. November, 2006
12 Listservs Sampled ABCT APA Division 12 SSCPNET (Section III, Div 12) CUDCPAPA Division 38ABMRSBM EBBM, MRBC, Obesity, CA SIGs
13 Outcome 39 syllabi 17 additional recommended articles and books 273 page documentDiscipline:30 psychology 3 public health3 medicine 1 nursing1 PE/health/sport studies140 requestsNovember, 2006
14 Evidence-Based Practice ModalCourse Title: CBT, EST, EVT, Psychological Interventions, Psychotherapy ResearchTexts: Barlow, Handbook Psychologic Disorders, Bergen & Garfield Handbook of Psychotherapy and Behavior ChangeContent: ESTsAdditionalAdditional Texts:-Persons, Case Conceptualization-Dawes, House of CardsAdditional Content:-Assessment-Case formulation, functional analysis-Clinical judgment-Diversity-Iatrogenic effects-Research methodsIatrogenic – critical incident debriefing
17 courtesy of Barbara Walker, Indiana University, 2006
18 Best available research evidence SynthesizerLocateCritically appraiseMeta-analysisConsumerAppraise quality & relevanceIntegrateResearcherDesignConductAnalysisReportingBest available research evidenceClinical Decision-MakingPatient’s values, characteristics, and circumstancesClinical ExpertiseClinicianCommunicateAssess patientDeliver EBPPatientUnderstandingPreferencesAccess
19 Researcher Training in Psychology versus Medicine DesignCorrelational (convenience classes)Experimental (from animal studies)ConductBrief, tight controlLittle missing data; replace casesAnalysis - completerReportingClinical MedicineDesignObservational (population)Clinical Trial –test of policy applied to populationConductLong, intercurrent eventsMissing data;Analysis – ITTReporting – CONSORTBased on animal study – noise exposure; if didn’t experience full course of noise, can’t tell, so replace. In clinical medicine, test if vitamin C or eating citrus. If people drop out of citrus group or shift to A, keep them citrus. Meaninful if they drop out or shift. Generalizing to what would be policy effect.
20 Researcher, Synthesizer, Consumer Training in Analysis PsychologyANOVA/regressionClinical MedicineOdds RatiosEpidemiology TerminologyAbsolute risk (p[disease] in a particular population)Relative risk (p[disease/exposed]/p[disease/unexposed)Attributable risk (p[disease/exposed] -p[disease/unexposed)Number needed to harm (1/attributable risk)Odds ratio (odds[disease/exposed]/odds[disease/unexposed])Why categorical – diagnosis; yes/no decision
21 Clinical Significance NNH = 5. If 5 patients treated with TX1, 1 would be more likely to have AE than if all had received TX0NNT = patients would need to be treated with TX1 to see one success not seen with TX0
23 Excerpt from CONSORT checklist METHODS Participants3Eligibility criteria for participants and the settings and locations where the data were collected.Interventions4Precise details of the interventions intended for each group and how and when they were actually administered.Objectives5Specific objectives and hypotheses.Outcomes6Clearly defined primary and secondary outcome measures and, when applicable, any methods used to enhance the quality of measurements (e.g., multiple observations, training of assessors).Sample size7How sample size was determined and, when applicable, explanation of any interim analyses and stopping rules.Randomization -- Sequence generation8Method used to generate the random allocation sequence, including details of any restrictions (e.g., blocking, stratification)Randomization -- Allocation concealment9Method used to implement the random allocation sequence (e.g., numbered containers or central telephone), clarifying whether the sequence was concealed until interventions were assigned.Randomization -- Implementation10Who generated the allocation sequence, who enrolled participants, and who assigned participants to their groups.Blinding (masking)11Whether or not participants, those administering the interventions, and those assessing the outcomes were blinded to group assignment. When relevant, how the success of blinding was evaluated.
25 Best available research evidence SynthesizerLocateCritically appraiseMeta-analysisEvidence User LocateAppraise quality & relevanceIntegrateResearcherDesignConductAnalysisReportingBest available research evidenceClinical Decision-MakingPatient’s values, characteristics, and circumstancesClinical ExpertiseClinicianCommunicateAssess patientDeliver EBPPatientUnderstandingPreferencesAccess
26 Synthesizer: Systematic Reviewer- explicit, systematic, transparent to avoid bias Specific research question (PICO)Search protocol to select papers – key wordssystematic search of the literature (EMBASE, CINAHL, Cochrane Controlled Trial register, DARE)explicit inclusion and exclusion criteriaExplicit, transparent rating of methodological qualityData extractionAnalysis: qualitative or quantitativeConclusionDiscussion of strengths and limitations
27 The 5 Step EBM Model for Evidence Users (Consumers) Ask: formulate the questionAcquire: evidence - search for answersAppraise: the evidence for quality and relevanceApply the resultsAssess the outcomeThink of these as competenciesAppraise – both quality of the evidence and relevance to the pt at handThe question must have 4 parts – pt or problem being addressed, the intervention being considered, the comparison intervention, the clinical outcomes of interestRemember that clinical experience, pt preference, social context also play a role!Individual clinical expertise: the proficiency and judgment of the clinician gained from clinical experience and practice. Best available external clinical evidence - clinically relevant research. Accuracy/precision of diagnostic tests Power of prognostic markers Efficacy/safety of therapeutic, rehabilitative, and preventative regimens Basic medical sciences – genetics, immunology
28 Asking: Well-Built Clinical Questions Background: What are effective treatments for bulimia nervosa?Foreground: In patients withPatient: binge eating disorderIntervention: does interpersonal therapyComparison: compared to CBT reduceOutcome: frequency of binge episodes
29 Critically appraising the evidence Use of standardized a priori appraisal methods to answer:Is the evidence valid?Internal validityIs the evidence applicable/relevant?External validityIs the evidence clinically significant?Is the evidence valid?Internal validity: how true are the study results for the people in the study? Varies by the validity of the chosen study design to answer the study question, the potential biases of the chosen study design and how well they were controlled in the study execution. Sources of bias: selection, performance, detection, attrition.External validity: How well do the study’s results apply to the person or clinical population you are interested in? Less well defined criteria for all studies.Historic approach has been hierarchical study designs (Handout—study designs). Some of the “bias” towards RCT has come from this.More recent approaches are sophisticated by considering trial design in combination with study design (how well does this type of study answer the question at hand?), how well study was executed (adequate randomization and followup), and questions related to external validity (spectrum of patients considered)—along with looking at various summary reports (SR is highest level of evidence). (Handout levels of evidence and grades of recommendation).How strong is the evidence? The chart also gets at this attribute—especially consistency and bias. Coherence is addressed by looking across evidence at contradicting as well as consistent results, and conisdering the evidence in light of what is known in other sciences (basic sciences/biology, systems research, sociology/psychology, etc.).What does the summarized evidence suggest about harms and benefits? Putting it all together.Common sense dictate:When there is evidence of benefit and value—do it.When there is evidence of no benefit, harm or poor value, don’t do it.When there is insufficient evidence to know for sure—be conservative.Use of the balance sheet. Balance sheets include:The characteristics of the underlying population that affect the questionThe most important health (and economic outcomes)—positive and negativeThe most important options (choices)The probability or magnitude of each outcomes with each optionThe absolute differences of outcomes with each optionLimitations--Can have strong but contradictory evidence—HRTWeighing benefits and harms—whose perspective?None of these methods is perfect—and coming up with hard numbers can be a lot of work—but focusing on exact estimates and complete options and outcomes forces precision in our thinking, accurate communication to patients, and consistency of care with patient preferences weighed in.A patient can’t really weigh in for a decision involving real tradeoffs with qualitative words:This treatment has benefits—but is also has some risks…What do you want to do?
30 Clinical Decision-Making Clinical epidemiology disciplinestudy of determinants and consequences of clinical decisionsapply EBP/5A’s/critical appraisal at clinical encounter to overcome automatic, unconscious decision-making biases (aka bad clinical intuition)
31 barriers between research and practice 30 kg of guidelines per family doctor per year25000 biomedical journals in print8000 articles published per day95% of studies cannot reliably guide clinical decisions2001 Bazian Ltd
32 Clinical Decision-Making Health Informatics disciplineinfrastructure, resources, devices, structures (e.g., algorithms, guidelines) needed to store, retrieve, manage and use health information and the time and place that a decision needs to be made.-Decision support.
33 Secondary Synthesized Evidence (AKA “evidence-based capitulation”) Research proliferates rapidly. Clinical performance demands increase. Practicing clinicians too busy to use all EBM steps will all patients.Increased focus on pithy clinical practice guidelines, synopses, and structured abstractsMD ConsultACP Journal ClubCochrane Database of Systematic Reviews“Up-to-date”InfoPOEMS (Patient Oriented Evidence that Matters)
34 Best available research evidence SynthesizerLocateCritically appraiseMeta-analysisConsumerAppraise quality & relevanceIntegrateResearcherDesignConductAnalysisReportingBest available research evidenceClinical Decision-MakingPatient’s values, characteristics, and circumstancesClinical ExpertiseClinicianCommunicateAssess patientDeliver EBPPatientUnderstandingPreferencesAccess
35 Clinically Supervised Training in Evidence-Based Treatment Needs work: papers by Woody and by Weissman
36 Best available research evidence SynthesizerLocateCritically appraiseMeta-analysisConsumerAppraise quality & relevanceIntegrateResearcherDesignConductAnalysisReportingBest available research evidenceClinical Decision-MakingPatient’s values, characteristics, and circumstancesClinical ExpertiseClinicianCommunicateAssess patientDeliver EBPPatientUnderstandingPreferencesAccess
37 Patient PreferencesShared decision-making requires information only available to patient (e.g., valuation of harms/hassles, alternative outcomes & treatments)Utility assessment: All possible outcomes assigned a value between 0 (death) and 1 (perfect health).Time trade-off approachThe proportion of life in a particular health state (e.g., severe depression) that you would give up to attain perfect health (e.g., 30%). Utility of that health state is 1-(30%) = .70Standard gamble approachThe point where you are indifferent to the choice between spending the rest of your life in the health state in question and a gamble between perfect health and instant death where the probability of perfect health represents the utility of the health state.
38 Teaching evidence-based practice = teaching a process DidacticsSmall groups, problem-based learningPreceptorships/clinical supervisionStandardized patients and evidence stationsEmbedded throughout curriculumTeaching EBP is about teaching a process – not just what tx have empiricalRole models
39 Medical Decision Making in the NU-FSM curriculum MDM-I (first week of medical school)Sensitivity, specificity, pre- and post-test probabilities, innumeracy, uncertainty in medicineMDM-II (last two weeks of M1 year)EpidemiologyStatisticsMDM-III (beginning of M2 Spring Quarter)Decision analysisMeta-analysisCost-effectiveness analysisClinical guidelinesM3 MDM (once a month in M3 year)Review papers pertaining to clinical casesUse of CAT
40 Evidence-Based Behavioral Practice (EBBP) NIH Office of Behavioral and Social Sciences Research contract N01-LM :Resources for Training in Evidence-Based Behavioral Practice,Invite SMDM to the Scientific Advisory Board
41 OBSSR 5-Year PlanYear 1: develop training website, Council, Scientific Advisory Board, white paper on training, skills, competencies reflecting education in evidence-based behavioral practice (EBBP)Year 2: develop, implement a web-based, research-focused training module(s) on EBBP; field test in graduate curriculaYear 3: launch interactive web-based training courses; establish practice network, develop first EBBP clinical practice training module
42 OBSSR 5-Year PlanYear 4: With practice network, develop modules on application of evidence-based clinical decision-making to intervention with specific cases. Field test in internship/residency/post-doctoral training programs and practice network.Year 5: Link website to systematic reviews of behavioral interventions, treatment manuals, outcome assessments. Develop and field test clinical decision-making modules that integrate patient preference and clinical competency assessments.
43 SuggestionsTo enhance the evidence base for psychological treatments and support lifelong learning, clinical psychology training might benefit from enhanced coverage of:Researcher skills in methods: clinical trial design, analysis, reporting, synthesisClinician training in 5-step (5A’s) EBP model – cover 2 A’s
44 SuggestionsPsychology informatics could use infrastructure development (PSYCinfo & Cochrane; library access; coverage in secondary synthesized sources like Up-to-Date; practice-based research networks)Psychology could use appropriate patient preference measures that support shared decision-makingA discipline of clinical psychology decision-making needs to develop to systematize integration of research evidence, clinical expertise, and patient clinical data and preferences
45 Concluding QuestionsWhat training modules and materials would be helpful?Will you partner with us to help develop and try these out?
48 The Evidence Pyramid for Treatment Effectiveness Questions ***USE THE BEST EVIDENCE AVAILABLE***
49 Alternatives to evidence-based medicine Eminence based medicineEloquence based medicineVehemence based medicineNervousness based medicineTo understand why guidelines, consider where coming from. Eminence- more senior expert, making the same mistakes with increasing confidence over impressive number of years, an art or craft done w/ clinical expertise. Vehemence-based: brow-beating your colleagues into incorporating your treatment into practice guidelines. Nervousness-based – in US fear of litigation – only bad test or treatment is one you didn’t think of orderingConfidence-based – restricted to surgeons(Isaacs and Fitzgerald, 1999, BMJ)
50 Levels of Clinical Evidence in the Primary Literature (psycINFO, MEDLINE) RCT, Practice GuidelineConsensus Development ConferenceRandomized Controlled TrialQuality ImprovementCohort Studies, RiskCohort StudiesEtiologyCohort Studies, Prognosis, Survival AnalysisCohort Studies, Case Control, Case SeriesPrognosisRandomized Controlled Trial, Double Blind, Clinical TrialsDouble-Blind Randomized Controlled TrialTherapySearch FiltersMethodologyType of Question
51 EBM ResourcesPocket guides with web-linked updates (Sackett; Guyatt & Rennie)Cochrane LibraryBMJ:Centre for EBM:Centre for Evidence-based mental health: