High Cohort Retention in a Phase III Trial of the Diaphragm and Lubricant Gel in Southern Africa Presented by Constancia Watadzaushe University of Zimbabwe – University of California, San Francisco Collaborative Research Programme IAS 2007 Conference GOODMORNING , I AM GOING TO PRESENT ON THE HIGH COHORT RETENTION IN A CLINICAL TRIAL OF THE DIAPHRAGM AND LUBRICANT GEL THAT WAS CONDUCTED IN SOUTHERN AFRICA

Methods for Improving Reproductive Health in Africa (MIRA) Phase III randomized controlled trial to examine the effectiveness of the diaphragm and Replens gel in preventing HIV acquisition in women THE METHODS OF IMPROVING REPRODUCTIVE HEALTH IN AFRICA WAS A PHASE III RANDOMISED CONTROLLED TRIAL THAT EXAMINED THE EFFECTIVENESS OF THE DIAPHRAGM AND REPLENS GEL IN PREVENTING HIV ACQUISITION AMONG WOMEN

Background: The MIRA Trial UZ-UCSF = 2502 MRC = 1515 PHRU = 1028 Five study clinics in Zimbabwe (Harare) and South Africa (Durban & Soweto) Sites were a mixture of rural, peri-urban and urban; with formal and informal settlements A TOTAL OF 5045 WOMEN PARTICPATED IN THE STUDY ACROSS 3 SITES WITH 5 STUDY CLINICS. THE CLINIC SITES WERE BASED IN HARARE (ZIM), DURBAN AND SOWETO IN SOUTH AFRICA. THE SITES WERE A MIXTURE OF RURAL, PERI URBAN AND URBAN WITH FORMAL AND INFORMAL SETTLEMENTS

Baseline participant characteristics (n= 5045) Average age = 28 (range 18 – 50) 44% high school educated 23% employed 56% earned income in the past year 59% married 68% living with partner 94% Christian Average age at sexual debut = 18 (10-31) 22% had circumcised partner, 20% didn’t know THE STUDY PARTICIPANTS WERE ON AVERAGE 28 YEARS OF AGE RANGING BETWEEN 18 AND 50 YEARS, 445 HAD HIGH SCHOOL EDUCATION AND 235 WERE EMPLOYED. 595 OF THESE WOMEN WERE MARRIED AND 685 WERE LIVING WITH A PARTNER. THE AVERAGE AGE OF SEXUAK DEBUT WAS 18 YEARS AND IT RANGED BETWEEN 10 AND 31 YEARS. 22% HAD CIRCUMCISED PARTNERS AND 20% DID NOT KNOW.

Retention and LTFU considerations Protocol assumed 4-7% loss to follow-up (LTFU) per year over 3 years given an overall expected HIV incidence in the trial of 4% Minimizing LTFU is essential in clinical trials to ensure the validity of trial results Important not only to limit LTFU, but to ensure that LTFU is not different by study group “LTFU” defined as no closing visit with HIV endpoint data Sample size considerations - HIV incidence estimated as 4% across sites, therefore we aimed to limit LTFU to approximately this level THE TRIAL DESIGN ASSUMED A 4.7% LOSS TO FOLLOW UP OVER A PERIOD OF 3 YEARS OVER AN ESTIMATED HIV INCIDENCE RATE OF 4 % ACROSS ALL SITES. LTFU WAS DEFINED AS NO CLOSING VISIT WITH HIV END POINT DATA IT IS ESSENTIAL TO LIMIT LTFU SO AS TO ENSURE VALIDITY OF RESULTS AND IT WAS IMPORTANT TO ENSURE LTFU WAS NOT DIFFERENT BY STUDY GROUP. WE

Follow-up schedule Screening visit Enrollment visit (2 weeks after screening visit; repeat screening visit conducted if > 2 weeks since first screening visit) Week 2 visit Quarterly follow up visits during between 12 and 24 months, depending on date of enrollment Closing visit at 12 – 24 months, depending on date of enrollment A PARTICIPANT WAS ENROLLED INTO THE STUDY WITHIN TWO WEEKS AFTER THE SCREENING VISIT AND A REPAET SCREENING WAS DONE SHE IF SHE CAME BACK FOR ENROLLMENT TWO WEEKS AFTER THE FIRST SCREENING VISIT. PARTICPANTS WERE THEN FOLLOWED UP QUARTELY FOR A PERIOD OF 12 TO 24 MONTHS DEPENDING ON THEIR DATE OF ENROLLMENT. CLOSING VISITS WERE DONE BETWEEN 12 TO 24 MONTHS DEPENDING ON THEIR DATE OF ENROLLMENT

Concerns about LTFU prior to implementation Migration Long follow-up period (24 months) Lack of perceived incentive to stay in study for control group? Before the trial was implemented there were concerns about migration, long follow up periods which was a maximum of 24 months There was also concern that those in the control group may perceive no benefit of staying in the study that long

Retention strategies prior to study implementation Initiation training Every staff member’s responsibility Resources allocated and sites encouraged to hire many outreach workers Outreach workers given space in clinic Community launch activities VERY DETAILED LOCATOR FORM THE FOLLOWING STRATEGIES WERE THEREFORE IMPLEMENTED PRIOR TO IMPLEMENTATION OF THE STUDY STAFF WAS TRAINED ON THE IMPORTANCE OF RETAINING PARTICIPANTS THROUGHOUT THE TRIAL AND ON HOW THIS COULD BE DONE ADEQUATE RESOURCES FOR FOLLOW UP WERE ALLOCATED AND SITES WERE ENCOURAGED TO RECRUIT MANY OUTREACH WORKERS SO THAT THE OUTREACH:PARTICPANT RATIO WAS MANAGEABLE OUTREACH WORKERS WERE ALLOCATED SPACE IN THE CLINICS , THIS ENABLE THEM TO HAVE ONE TO ONE CONTACT WITH PARTICPANT WHILE IN THE CLINIC AND ESTABLISH RAPPORT COMMUNITY LAUNCH ACTIVITIES WERE DONE TO SENSITISE THE COMMUNITY AND MOST IMPORTANTLTY STAFF WAS TRAINED TO OBTAIN VERY DETAILED LOCATOR INFORMATION FROM EACH PPT

Locator form information Home address and phone number, space to draw map Husband’s name, work address and phone number School-going child’s name, school name and address Parent or nearest relative’s name, address and phone number Second contact after parent or relative Rural address of the participant For each, participant indicated and initialed whether it was okay to reach that contact person, by phone or by post, and whether it was okay to mention the study by name VERY DETALIED LOCATOR INFORMATION WAS COLLECTED FROM EACH PPT FOR EFFICIENT TRACKING PURPOSES INFORMATION OBTAINED INCLUDED THE PPT’S HOME ADDRESS, PHONE AND DIRECTION MAPS WERE DRAWN ON THE LOCATOR WHERE ADDRESSES WERE NON-EXISTENT WE ALSO OBTAINED THE PPT’S SCHOOL GOING CHILD’S NAME, NAME OF THE SCHOOL AND ADDRESS.THIS WAS DONE TO ENSURE THAT IN THE EVENT THE PPT MOVED FROM HER CURRENT KNOWN ADDRESS THE CHILD WAS UNLIKELY TO CHANGE SCHOOLS THEREFORE WOULD REMAIN A RELAIBLE CONTACT A SECOND CONTACT DETAILS WAS ALSO OBTAINED AFTER THE PARENT AND RELATIVE FOR EACH OF THESE DETAILS THE PARTICIPANT WAS ASKED TO INDICATE AND INITIAL THAT IT WAS OKAY TO REACH THE CONTACT BY PHONE OR POST AND WHETHER IT WAS OKAY TO MENTION THE STUDY BY NAME

Retention strategies during study implementation: 1 Retention strategies during study implementation: 1. extensive follow-up Reminder letters sent prior to visits Phoning or home visits 24-hours after missed visits Peer educators used for outreach (MRC site) THE FOLLOWING STRATEGIES WERE USED DURING THE IMPLEMENTATION OF THE TRIAL REMINDER LETTERS WERE SENT TO THE PPT’S BEFORE THE NEXT SCHEDULED VISIT WAS DUE IF PPT MISSED HER SCHEDULED VISIT SHE WAS EITHER CONTACTED BY PHONE OR MAKING A HOME VISIT WITHIN 24 HOURS PEER EDUCATORS WERE ALSO USED FOR OUTREACH

2. improving customer service Offered Saturday Clinics and evening hours Created participant suggestion boxes Provided transport Appointed Client Liaison Officer to interface with participants regarding expected waiting periods Played family movies and served refreshments in waiting areas Provided childcare while women were in their visits AS THE TRIAL PROGRESSED WE IMPROVED OUR CUSTOMER SERVICE TO TRY AND MEET THE PPT’S NEEDS AS FAR AS THEIR PARTICIPATION IN THE TRIAL WAS CONCERNED WE OFFERED WEEKEND AND AFTER HOURS CLINICS FOR THOSE PPTS WHO WERE FINDING IT DIFFICULT TO GET TIME OFF FROM WORK WE CREATED A SUGGESTION BOX FOR THE PPTS PROVIDED TRANSPORT FOR THOSE WHO WERE LIVING IN AREAS WERE TRANSPORT WAS DIFFICULT TO ACCESS A CLIENT LIASION OFFICER WAS APPOINTED AT EACH SITE TO INTERFACE WITH PARTICPANTS ABOUT THEIR WAITING PERIODS PLAYED FAMILY MOVIES AND SERVED REFRESHMENTS IN WAITING AREAS AND FOR THOSE PPTS WHO BROUGHT CHILDREN TO THE CLINIC CHILDCARE WAS PROVIDED WHILST THE PPT WAS HAVINGWERE IN THEIR VISITS

3. small gifts and competitions All IRB approved Token gifts at visit milestones (all < R15 each) given at 6 month intervals: Purse/ clock Umbrella Sewing Kit Tupperware set Participant lottery – one voucher for every visit attended and quarterly draw for winners. Prize = grocery voucher Incremental reimbursement scale for every follow-up visit completed TO ENCOURAGE OUR PPTS TO KEEP COMING FOR THEIR VISITS WE GAVE THEM TOKEN GIFTS AT VISIT MILESTONES E.G AT MONTH 6 VISITS( HALFWAY ) WE ALSO HELD A PARTICIPANT LOTTERY ,, THE DRAW WAS DONE QUARTELY FOR PREVIOUS VISITS THAT HAD ALREADY BEEN ATTENDED AN INCREMENTAL REIMBURSEMENT WAS ALSO DONE AS THE VISTS ACCRUED THESE SMALL GIFTS AND COMPETITIONS WERE ALL IRB APPROVED

4. Focusing on hard to reach participants Rural outreach trips Closing visits done at homes Coordinators/ PIs visiting participants who repeatedly were “difficult to bring to the clinic” Fast-track provision for hard-to-reach or working participants “Final push” for 2 months after last scheduled visit We also focused on participants that we termed as Hard to reach. These participants were hard to reach for one reason or the other. They either relocated or kept promising to come for visit. For those who migrated to rural areas or other towns Rural Outreach trips were done to give these participants busfare to come to the clinics. For those who said they did not have time to come to the clinic a MIRA team visited them at home and did a closing visit in their home. And for those participants who repeatedly were difficult to bring back to the clinic, a coordinator or PI visited them at home. And a final push for participants who had not closed after the last scheduled visit was done for 2 months to try and bring them all back to the clinic for a closing visit

Results – Visit attendance # attended/ #expected

Results – final retention and LTFU Intervention Arm Control Arm Followed 2521† 2518 † Completed scheduled closing visit 2322 (92%) 2349 (93%) Early withdrawals †† 57 (2.3%) 42 (1.7%) Lost to follow up 142 (5.6%) 127 (5.0%) - At last scheduled follow-up visit, on October 4, 2006, 89% of enrolled participants had a closing visit - After another 2 months of intensive follow-up and tracking, 92.5% of participants had come for their scheduled visit - Overall 5.3% of participants were classified as LTFU, meaning that they never returned to the clinic for a closing visit and a final endpoint (HIV) test. Some of these participants had some follow-up data and contributed some woman-months to the analysis, others were never seen after enrollment † Baseline sample †† Withdrawals include participants who were discontinued due to death (23) or who came back to the study clinic and withdrew or were withdrawn early because of relocation (22), partner-related reasons (16), personal reasons (13), no time to participate (13) or other reasons (12)

Conclusions Pre-emptive planning and continuous monitoring of retention rates, combined with additional tracking time ensured high return of participants Achieving high retention is resource intensive and requires proper allocation of financial, administrative and human resources Many of our successful strategies were staff-driven, therefore it’s important to involve all members of staff in retention discussions, and to share ideas across sites In conclusion, pre-emptive planning and continuous monitoring of retention rates , combined with additional tracking time ensured high return of participants Achieving high retention is resource intensive and requires proper allocation of financial, administrative and human resources Many of our successful strategies were staff-driven, therefore it is important to involve all members of staff in retention discussions and to share ideas across all sites

Thank you!

Co-authors and Acknowledgements This study was funded by the Bill and Melinda Gates foundation UZ-UCSF Agnes Chidanyika UCSF, Women’s Global Health Imperative Liz Montgomery Mi Lam Ariane van der Straten Nancy Padian Medical Research Council (MRC), Durban Sharika Gappoo Perinatal HIV Research Unit (PHRU), Soweto Sinazo Pato Ibis Reproductive Health Naomi Lince The MIRA team