RECRUITMENT, ADHERENCE, AND RETENTION STRATEGIES TALES FROM CLINICAL TRIALS Kelley R. Branch, MD, MSc, FACC Associate Director, Clinical Trials Services Unit Medical Director, CCU/5NE A Clinical Trial Story… Once upon a time, there were 4 hypertension trials… COMPARISON OF SHEP, STOP-H, MRC-92 AND SYST-EUR CHARACTERISTICS SHEP STOP-H MRC-92 Syst-Eur Sample Size 4736 1627 4396 4695 Mean Age 71.5 75.6 70.3 70.3 160-219 180-230 160-209 160-219 BP SBP Criteria DBP Primary Outcome Design <90 90+ <114 <95 Total Stroke Total mortality Total stroke Total stroke P R DB P R DB P R SB P R DB Sys-Eur Trial: Accrual Trial Adherence Mean F/U Time SHEP 54 mo STOP-H 25 mo MRC-92 70 mo Syst-Eur 30 mo Mean Baseline BP 170/77 190/104 185/91 174/86 14/6 11/5 BP Differential 41247.00 19.5/8 Adherence Lost to F/U 6% 0% 25% % Crossovers 33% 23% 31% 52%D 37%B 47%P % Adherence 90%/67% 84%/77% P 5% (116) A 5% (121) 23% 85%/72% SYST - EUR A WORST CASE ANALYSIS? 600 500 STROKE 507 CHD TOTAL EVENTS Number 400 Late Recruitment = Fewer Events LOST TO FOLLOW-UP 300 262 237 High Lost to Follow Up = No 245 Definitive 200 Conclusions from the Trial 131 255 124 100 10 0 SYST-EUR SHEP 6 Major Trial Issues • Recruitment - Timely Enrollment • Adherence • Complete Follow Up RECRUITMENT MANTRA: “Get Sufficient Population In a Reasonable Time” RECRUITMENT FUNDAMENTAL POINT Successful recruitment depends on developing a careful plan with multiple strategies, maintaining flexibility, establishing interim goals and preparing to devote the necessary effort. Friedman, Furberg and DeMets RECRUITMENT • Successful recruitment has been documented in many trials • Clinical Sites: Past performance predicts future • Centers carefully selected by past performance (http://www.fhcrc.org/science/phs/swog/recrcct/) RECRUITMENT:BASIC ISSUES • Planning – Sources and support – Strategies • • • • • Conduct - Implementation Monitoring - Short and long term goals Problems - Expect them to happen Solutions - Make them occur Have reasons for participation RECRUITMENT: CAREFUL PLANNING • • • • BE CONSERVATIVE IN YOUR ESTIMATES Design easy recruitment Establish interim goals Have contingency plans • 3 TO 6 MONTH PERIOD TO SEE RESULTS TRIAL PLANNING • • • • • • • • Increase likelihood of getting sufficient participants Staff – Organized, experienced Institutional support - proper facilities Publicity - start before trial Multiple recruitment strategies - at least 3 Pilot test strategies Contingency plans Statistical power - assumes constant enrollment ADVANTAGES: WIDE ENTRY CRITERIA • Easier screening and recruitment • More feasible and affordable • Broader range of variables and larger study size • Reliable overall result • Greater public health impact • Testing subgroup hypotheses 14 RECRUITMENT DATA: Variable Success in 13 NHLBI Studies Study AMIS BHAT CAPS CARDIA CDP CSSCD HDFP LRC MILIS MRFIT POSCH SHEP Pilot TIMI-1* Number of Participants Recruited Actual Actual/Planned 4,524 3,837 502 5,182 8,345 3,241 10,940 3,843 985 12,886 838 551 316 1.06 0.95 1.00 1.02 1.00 1.01 1.04 1.08 0.82 1.07 0.84 1.10 0.93 Person-Years in Recruitment Planned Recruitment Time Period Actual/Planned Actual/Planned (R) 1.00 1.21 1.08 1.00 1.22 1.13† 1.50 1.54 2.71 1.13† 1.58† 1.17 0.96 *TIM-1 was stopped on the recommendation of Data and Safety Monitoring Committee; the treatment showed strong evidence of efficacy. † The ‘projected’ time was revised after additional clinics joined the study. 0.83 0.82 0.82 0.86 0.55 1.16 1.02 0.34 0.35 0.81 0.25 0.71 0.98 16 SELECT Trial Accrual Projected and Actual SELECT Accrual Actual and Projected Accrual vs. Estimated Accrual 40000 35000 30000 Actual accrual as of March 31, 2004: 24,166 Reach accrual goal of 32,400 at end of April 2004 25000 Actua Proje 20000 Estim 15000 10000 5000 0 Projected ACCORD Initial Trial ACCORD Main Trial Accrual 10000 80 9000 70 Number Randomized 60 7000 50 6000 5000 40 4000 30 3000 20 2000 1000 ACTUAL GOAL 0 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Weeks 0 Number of Sites Active 8000 RECRUITMENT STRATEGIES (N=3) How to Get Patients Chart Review Websites Media Efforts Registries Direct Mail Blood Bank Donors Mass Screening Screening Occupational Laboratory Lists Medical Referrals 20 Checklist: OVERALL RECRUITMENT PROGRAM Start recruitment on target date Choose physically accessible location Use at least three recruitment strategies Recruitment Coordinator - overall responsibility Trial-wide recruitment coordinator network Accurate tracking system Match staff and screenees OVERALL RECRUITMENT PROGRAM Provide staff back-up Be aware and anticipate staff burnout Inform medical and lay communities Recruits - Solicit in simple language Medical associations and hospital staffs contacted by the Principal Investigator OVERALL RECRUITMENT PROGRAM Identify excellent, experienced staff Calendar for ENTIRE recruitment period Pretest your recruitment strategies Regular review and evaluation of program Develop contingency plans Flexible clinic hours PATIENT REASONS FOR PARTICIPATION • Answer scientific question accurately • Altrusim: Benefit other patients - current and future • Benefit to themselves – additional monitoring – second opinion of their condition – reassurance regarding diagnosis 24 RECRUITMENT OF STUDY POPULATION MANTRA: “Get Sufficient Population In a Reasonable Time” RECRUITMENT FAILURE CAUSES: • Late start • Inadequate planning • Insufficient effort • Overly optimistic expectations POTENTIAL PROBLEMS Expect them-they will occur • Inadequate funding for screening process • Unwillingness to refer or allow participation • Overestimation of prevalence • Overly rigorous entry criteria POSSIBLE RECRUITMENT SOLUTIONS • EXTEND THE TIME FOR ENROLLMENTX? • RELAX INCLUSION/EXCLUSION CRITERIA-X • ACCEPT A SMALLER SAMPLE SIZE-X • RECYCLE PREVIOUS INELIGIBLES-O • CHANGE THE DESIGN-XXX Recruitment: Summary • • • • Plan, plan, plan Design for success with recruitment program At least 3 recruitment strategies Problems happen • Sufficient population in reasonable time 28 Clinical Trials ADHERENCE 29 ADHERENCE DEFINITION Adherence is the extent to which a person’s behavior coincides with medical or health advice in terms of taking medications, following diets, using devices, or executing life-style changes. TERMINOLOGY: ADHERENCE VS. COMPLIANCE • Adherence is preferred term • Adherence: Active, choice, interactive • Compliance: Passive, non-selective NHLBI Workshop, Bethesda, MD 1987 OVERALL ADHERENCE PLAN • Develop a bottom line - cannot be transgressed – Minimum amount of data which is essential • Set adherence goals depending on protocol – “Acceptability” trial – “Alteration of natural history” trial • Teach adherence techniques, plan for poor adherence – Run-in and test dosing procedures – Have a maintenance plan for everyone BOTTOM LINE: MINIMUM ACCEPTABLE ADHERENCE • Know primary outcome status on every randomized participant. • Human behavior will allow few to purposely harm a worthy scientific project. 33 Adherence is bad in clinical trials. Get over it. 34 SAMPLE SIZE ADJUSTMENT FOR REDUCED ADHERENCE • Key Point - Adherence correction p term-sample size formula, a squared function. 2N = 2(z + z)2 (1 - 2)2(1-p)2 p = Reduction in Adherence .01 .05 .10 .20 .30 .50 SS Increase 1.02 1.11 1.23 1.56 2.04 4.00 MRC 35 “ALTERATION OF NATURAL HISTORY” TRIAL • Enrolled group must do the intervention • Looking for efficacy on clinical outcomes • Adherence is crucial • e.g., Phase IV trials LRC: ANALYSIS FOR PREDICTORS OF ADHERENCE Adherence after first month associated with: • Adherence in first month- most powerful predictor (r=.59 or r²=.34) – r²=.36 with smoking and other factors added • • • • Smoking status Age Extent of Psychological Distress No statistical association with: – Exercise -Overall risk status – Weight -Motivational level – Vitamin consumption 37 FACTORS AFFECTING ADHERENCE TO INTERVENTIONS Effect on Adherence Positive Negative Patient Education Low social class “Blue collar: occupation Social isolation Regimen Supervised Therapy duration Parental Number of drugs administration Dosing frequency Cardiac drugs Symptomatic drugs Respiratory drugs Diabetic drugs Illness Disability Severity of symptoms Severity of illness Psychiatric illness No Effect Age Sex Race Adverse effect Disease duration Clinical improvement Concurrent illness “RUN-IN” PERIOD • Pre-randomization procedure • Single blind • Placebo used • Stress test for "pill-taking behavior” 39 CONCLUSIONS ABOUT “PLACEBO RUN-IN PERIOD” What does it do • Identifies a group of individuals who don’t adhere well during designated run-in • Successful repeat run-in performers (6.9%) adhere less well during trial • Those identified representative of those enrolled What doesn’t it do • Identify all who will adhere poorly to intervention Uncertainties • If those who “fail” would all be poor adherers • Cost/Benefit-advantageous “TEST-DOSING” PERIOD • Pre-randomization procedure • Single blind • Active drug used • Identify those with severe adverse effects Non-Adherence 42 SIGNS OF POTENTIAL NONADHERENCE: “RED FLAGS” 1. Missed visits 2. Difficulty in reaching by phone or failure to return calls 3. Rescheduling appointment twice (change in behavior) 4. Complaints about office visits 5. Impatience during clinic visit 6. Length of time (mandatory) at each visit 7. “Distance” during interview 8. Length of time since participation in study was discussed between physician and participant 9. Humor dealing with negative aspects of trial medication SIGNS OF POTENTIAL NONADHERENCE: “RED FLAGS” 10. Sarcasm about trial or study medication 11. Any expression by participant that he/she may discontinue study medication 12. Unusual or unexplained change in adherence to study medication 13. Unconcern by participant about adherence rate 14. Reassignment to new primary-care manager 15. Reassignment to other new clinic personnel 16. Illness with increased attention to “trial related disease” 17. Hospitalization for any reason 18. Any major change in life style which is imminent DISTRIBUTION OF ADHERENCE PROBLEMS IN A CADRE OF DROPOUTS AND OTHERS IN AN RCT Type of Problem Adverse effects Percent Dropouts Others* 19 22 Medical problems 11 20 Psychosocial problems 69 58 * Those who experienced either a 10% drop in medication adherence or a 10 day delay from their clinic visit window MECHANISMS INVOLVED IN PARTICIPANT NON-ADHERENCE • • • • • • Lack motivation Lack of knowledge (disease, intervention) Rejects medical diagnosis Denies significance of disease process Self-debate over intervention regimen Rejects intervention regimen MEDICAL THERAPEUTICS TEAM Psychologist-Behaviorist Nurse-Clinician Therapeutic Plan Physician Participant (Patient) Intervention Schedule Dietitian-Nutritionist Physician Assistant PRINCIPLES AND GOALS: PARTICIPANT COUNSELING IN DROPOUT RECOVERY Principles for Counseling Corresponding Goals 1. To establish contact with participants 1. To maintain contact with participants. 2. To “undercut” participant’s resistance for reinstitution of some aspect of the trial protocol. 2. To complete as much of the trial protocol as possible. 3. To convey a caring attitude to the participant about his overall health status and the importance of health care to these participants. 3. To resolve any somatic, adverse drug effects, or behavioral problems preventing protocol adherence. 4. To maximize the participant’s 4. To reinstate the protocol in small opportunities for success of increments using informal protocol completion using contracts and shaping. standardized behavioral techniques. 5. To give positive reinforcement for fulfillment of protocol activities 5. To emphasize the positive contribution at any level of protocol adherence. 6. To resume study drug at a low but definite priority for the participant. 6. To restore and maintain study power. 50 RECOVERY OF DROPOUTS BAYLOR-METHODIST CLINIC OF CPPT • 94 % were recovered for some regular visit with clinic personnel (90% within 6 months ) • Remaining participant was contacted regularly by telephone • 3% recidivism • 70% reinstituted study medication • Average adherence: study medication 35 % 51 Adherence: Key Points • Adherence is key to knowing the magnitude of effect – Withdrawal may be an outcome – Good trial = Good Adherence • Plan, set goals, make contingencies • Enhancing and actively monitoring participant adherence essential through trial • KNOW FINAL OUTCOME FOR EACH SUBJECT Adherence: Contingency Plans Contingency Plans • • • • Identify cause for non-adherence Motivation Negotiation Withdrawl of consent MOTIVATION • Waning motivation is a common element for trial participants with adherence difficulties, e.g. clinical trial fatigue. • Strong resolve is critical, if one is to cope with problems of life and continue trial participation. PARTICIPANT MOTIVATION How staff can contribute to it • Must know continuing importance of the trial. • Information from other studies. – Be proactive-don’t wait for them to ask/tell you. • Remind them that the DSMB meets regularly. – Considers potential benefit and harm. – Last meeting ended-vote for continuation. • Reassure participant of your position. NEGOTIATION “YOU DON’T GET IN LIFE WHAT YOU DESERVE-YOU GET WHAT YOU NEGOTIATE!” Ronald Karass-in Flight Add 58 NEGOTIATED ADHERENCE REGIMENS (Informal Contracts) • Reduced Dose • “Drug Holiday” • Follow-up only • Final assessment at trial end 59 RECHALLANGE: RESTARTING STUDY MEDICATION • INFORMAL CONTRACT -BE CAUTIOUS. • What was the reason for stopping? – Has that reason gone away? • Can you make small steps to your goal? • Part of a “Win, Win” is participant success 60 WITHDRAWAL OF CONSENT: HOW TO DEAL WITH IT • Use your “Pause Button” immediately. • Few will want to harm what is worthwhile. • You get what you negotiate. – “Seek first to understand, then be understood.” • • • • Know EXACTLY what your participant means. Make it clear you understand their position. Make clear your goal of minimum adherence. “Is there a way both can achieve goals?” 61 DROPOUTS: HOW TO DEAL WITH THEM • “Sense it coming”- use the “red flags” • A lesson in using your “Pause Button” • “Seek first to understand, then be understood.” • Issues frequently complex. • May not be solvable at the first interaction. 62 DROPOUTS: HOW TO DEAL WITH THEM • You are playing for- “Win, Win!” – Forcing resolution-may lead to “No.” • Get agreement to talk again. • Maintaining contact is your first principle. 63 BOTTOM LINE: MINIMUM ACCEPTABLE ADHERENCE Know primary outcome status on every randomized participant. Human behavior will allow few to purposely harm a worthy scientific project. 64 Summary • Recruitment – Plan, design for success – Timely Enrollment – 3 Recruitment Strategies • Adherence – Develop a bottom line – Set adherence goals depending on protocol – Teach adherence techniques, plan for poor adherence Summary • Complete Follow Up – Know primary outcome status on every randomized participant – Dropouts can drop back in