Evaluation of an Mhealth Intervention to Increase Adherence to Triage of HPV+ Women

Description of the problem and study rationale Cervical cancer remains a serious health problem, particularly in developing countries. In Argentina, every year 5,000 women are diagnosed with cervical cancer and more than 1,800 die from the disease 1. In several Latin American countries, it is the leading cause of cancer death among women and mainly affects women of low socioeconomic status. This is due in part to the difficulties of health policies to promote women's participation in prevention programs, namely accessing screening, follow-up and treatment of precancerous lesions. Proof of the causal link between HPV and cervical cancer has allowed for the recent development of HPV-based screening tests, which detect HPV types considered high-risk. These tests have important advantages over the Pap as a screening test: high sensitivity (over 90%)2 and high negative predictive value3, allowing for a reduction in screening frequency as well as for the identification of HPV positive women who, even without presenting a precancerous lesion, constitute a high-risk group that requires special follow-up within the health system. Very importantly, through self-collection, HPV-testing can reduce barriers to screening and increase coverage.4 The method is highly accurate5,6, and is acceptable for women in different countries.7-10 Argentina has become a pioneer in the introduction of new technologies for HPV-based cervical cancer prevention. The National Ministry of Health has approved the incorporation of the test as a nation-wide screening method. In this context the PI conducted the Jujuy Demonstration Project (JDP) to introduce HPV-testing for primary screening.11 This project was part of a national strategy to strengthen cervical cancer screening re-launched in 2008,12 which also included introduction of HPV vaccination in 2011.13 The JDP is a four-year implementation project (2011-2014) to develop, implement and evaluate the programmatic components of an HPV-based screening strategy11. In addition, in 2012 the PI conducted a population-based cluster randomized trial (Self-collection Modality Trial, initials EMA in Spanish). The study was the first carried out in a programmatic, real-world context of a low-resource region with the main objective to assess the effect on women´s screening uptake of HPV self-collection offered by CHWs during home visits and to assess the self-test acceptability by women and CHWs. Our intervention resulted in a four-fold increase in screening uptake (85.9% vs 20.2%), showing that offering HPV self-collection at home visits through CHWs is an effective strategy to improve cervical screening coverage10. However, in screening programs, high coverage will not result in a decrease of disease burden if women are not diagnosed/treated. Several triage methods are available including cytology, visual inspection methods, and colposcopy. In Argentina, cytology is used as triage method; therefore, all HPV-positive women who have performed self-collection at home need to attend health centers to undergo cytology. However, the proportion of women who have completed triage is low; therefore new innovative strategies are needed to increase attendance to cytology of these women. The aim of the study is to evaluate the effectiveness of a mHealth intervention versus usual care to increase adherence to cytological triage in women with HPV-positive self-collected test. Setting of the study The province of Jujuy was selected to develop this study due to the fact that screening based on HPV testing as primary test was introduced in 2011. The province of Jujuy has a high mortality rate (11.8 x 100.000 women over the 2008-2011 three-year period) and it is one of the provinces set as a priority in the National Program of Cervical Cancer Prevention. Since 2011, HPV-testing is the primary screening test, available for all women aged ≥30 attending public health centers and since 2014 HPV- self-collection has been introduced as programmatic strategy for screening under-users. The Primary Health Care System (PHC) integrates approximately 700 paid full-time CHWs who twice yearly visit approximately 110.000 households for health-related tasks as immunization and promotion of maternal/child health. CHW performance is evaluated annually and scored as good or sub-optimal by PHC supervisors according to achievement of established goals (e.g. number of home visits).10 One hundred percent of screening, follow up and treatment in public health system are registered in SITAM. SITAM is the national online information system that registers screening/diagnosis/treatment events from women attending the public health system.14 Goals of the study A cluster randomized trial will be conducted; it will include a total number of 240 Community Health Workers (CHWs) and 7200 women. The aim of this study is: 1. To evaluate the effectiveness of a mhealth intervention versus usual care to increase adherence to cytological triage among women with HPV-positive self-collected tests. 2. To assess the intervention acceptability by women and CHWs. Cluster randomized trial An overall number of 240 CHWs from the Primary Health Care System (PHCS) of the Province of Jujuy will be randomized into two groups that will be named mHealth-Intervention Group (MH) and Usual Care Group (UC) 1\. mHealth Intervention Group: Women with HPV self-collected tests will receive a mixed intervention which includes counseling through an interactive Apps specifically devised to increase adherence to triage which will be run on a tablet, and SMS text messages (SMS message) to remind them to attend triage. In addition, Heads of CHWs, chiefs of gynecology services and CHWs will receive reminders via e-mails to contact women if after 60 days from the HPV-results HPV+ they have not performed triage. 2\. Usual Care Group: Women with HPV self-collected tests receive usual care. 1. Procedures to carry out in each group: • mHealth intervention Group: During the home visits that CHWs carry during health rounds, selected CHWs will visit every household they are in charge of and will contact women aged 30+ to offer them HPV self-collection according to programmatic guidelines. Those women who accept self-collection will be invited to participate in the study, providing them the necessary information about the goals and activities involved. Eligible women willing to participate in the study will be invited to sign an informed consent. The CHW will update the F883 template, the standard primary health care form used by CHWs to collect socio-demographic and health data of the household and its members. CHWs will use a specific questionnaire to collect basic data about inclusion/exclusion criteria. Then CHWs will use an APPs specifically designed for the project, run on a tablet, to provide women with information about the meaning of the test results, next steps to be followed both for positive and negative results, and the importance of adherence to triage and diagnostic/treatment procedures if needed. An automated system to delivery SMS messages will be developed and linked to SITAM. Once the test sample is processed at the HPV laboratory and results registered in SITAM, women will receive a SMS message asking them to send a confirmatory reply by text messaging which should include its national ID number. Once this confirmatory reply is received by the system, women will receive a second SMS message informing them that the results are available at the health center, and dates and times to receive them. Seven days after the first text message, HPV+ women will receive a reminder to collect the test results at the health center. The message will ask women to disregard the message in case they have already gone to the health center to get results. Sixty-days after the first SMS message, Heads of CHWs, chiefs of gynecology services and CHWs will receive an e-mail and SMS message informing them about HPV+ women who have not been triaged with cytology, asking them to contact those specific women, with information about their name and addresses. • Usual care group: During the home visits that CHWs carry during health rounds, selected CHWs will visit every household they are in charge of and will contact women aged 30+ to offer them HPV self-collection according to programmatic guidelines. Those women who accept self-collection will be invited to participate in the study, providing them the necessary information about the goals and activities involved. Eligible women willing to participate in the study will be invited to sign an informed consent. The CHW will update the F883 template, the standard primary health care form used by CHWs to collect socio-demographic and health data of the household and its members. CHWs will use a specific questionnaire to collect basic data about inclusion/exclusion criteria. HPV-positive women will receive the visit of the CHW to inform them that they should go to the health center to receive results, following standard practices in the provincial protocols and guidelines. 2. Training for CHWs: All the CHWs of the province of Jujuy will be informed about the study and a special training will be provided for all those who were selected to participate in the study. During the training sessions, the goals of the study, the main activities to be carried out and the specific tasks that each one of them will have to perform will be explained. 1. Sample selection and size A total of 240 CHWs, who work in the province, will be stratified in four groups according to gender and urban/rural setting. A stratified sample of 240 CHWs was randomly selected with proportional allocation to strata. Within strata, CHWs were randomly assigned (1:1) to study groups. All selected CHWs were informed about the study by the PHC heads. We will produce computer-generated random number lists for the CHW random selection and the intervention allocation. Blinding of intervention and outcome assessments was not feasible. The study is designed to have power of more than 0·9 to detect a 20·0% increase in triage for the intervention group compared with a 30·0% triage in the control group (two-sided alpha=0·05). Based on PHC records13 we estimated that CHWs would enroll an average of 30 women in six months, of which 4 will be positive . Correlation induced by CHWs is included in sample size calculations assuming an intra-class correlation coefficient of 0·10, resulting in a sample of 120 CHWs per arm, and a total of approximately 7200 women enrolled. 2. Database: A database built specifically for the study will be used; it will include data about randomization Group, agreement to participate in the study and socio-demographic data. Data on HPV-testing and triage will be uploaded importing the data from SITAM, using the Unique Identification Number (DNI). Data entry will be done through specific software that will include range and inter-item consistency checks. Entries out of the expected range will not be allowed. Data entry will be done by 3 data enters, with re-entry of 20% of questionnaires for quality control. Data will be stored in files (TXT, DBF, SQL) compatible with their analysis with statistical packages (Stata). The data cleaning process will actively search for errors in a planned way. Data screening will include searching for four types of oddities: outliers (including inconsistencies), strange patterns in distributions, and unexpected analysis results. Screening methods will include browsing of data tables after sorting, printouts of variables not passing consistency checks, graphical exploration of distribution, frequency distributions and cross-tabulations and summary statistics. Once errors have been identified, we will go to previous stages of the data flow to see whether the value is consistently the same. If no error is found in the data entry phase, we will question the interviewer about what may have happened. If possible, we will contact the interviewee for repeated measurement. Accuracy of previous HPV test reported by the woman will be crossed with information from the laboratory database (SITAM). Those women having reported not having a previous HPV test, but with a test in the SITAM data base will be excluded from the study. Plan for missing data Data are exhaustive and complete for all samples arriving to the HPV laboratory. Because we define the outcomes as number of Pap smear effectively recorded in the laboratory database (SITAM)/ number of women in the study arm, we do not expect to have missing data in the main outcome. 3. Data analysis: Data analysis will be done through specific commercial softwares (STATA 10.0 or SPSS 15.0). Effectiveness to enhance women adherence to cytological triage Adherence to triage will be considered for each CHW. This will be defined as the number of women with triage smears within 30, 60 and 90 days. There will be a comparison of the percentage of HPV-positive women that did the Pap test within those time intervals in the mHealth intervention group and the Usual Care group. The effect of the mHealth intervention against usual care will be estimated using a means difference test or a non-parametric test for independent samples.