Evaluation of a mHealth Intervention to Increase Adherence to Triage of Self-collected HPV+ Women (ATICA Project)

Cervical cancer is a disease of health and gender inequalities, and is primarily a cancer of poor, socially vulnerable women. Even though it is almost entirely preventable, in Latin America it is the leading cause of cancer death among women.1 In Argentina, it is the second cause of female cancer death: every year 5,000 women are diagnosed with cervical cancer and more than 1,800 die from the disease.1 In this country, and generally across Latin America, high mortality is related to problems in the continuity of the prevention process, including low screening coverage and loss to follow-up, diagnosis, and treatment.2,3 New screening tests detect the presence of high-risk types of HPV which cause cervical cancer. HPV tests have important advantages over the Pap as a primary screening test, including higher sensitivity (over 90%)4 and a high negative predictive value,5 which allow reduced screening frequencies. Very importantly, through self-collection strategies, HPV testing can reduce barriers to screening and increase coverage.6 HPV self-collection is highly accurate,7 and is acceptable among women in different countries.8-11 However, HPV screening identifies women who are HPV+ (around 10%). Triage tests for HPV+ women determine those who need additional diagnostic procedures. While several triage methods are available for detecting precancerous lesions, including cytology and colposcopy,12 to date most such triage methods require HPV+ women to attend health centers to determine appropriate follow-up. Although adherence to follow-up and treatment is a widespread problem for cervical cancer programs in Latin America,2 evidence for interventions to improve adherence to follow-up is lacking.

The introduction of self-administered HPV testing in Jujuy has been successful in increasing primary cervical cancer screening coverage. Yet the proportion of self-collected HPV+ cases who currently undertake cytological triage within 120 days of screening is low (about 30%), which hinders the impact of the increased HPV screening coverage attained.

Aims of the study

1. To evaluate the effectiveness of a mHealth intervention targeted to women and health providers to increase adherence to cytological triage among self-tested HPV+ women compared to usual practice.
2. To identify and understand the processes and factors which contribute to the success of the mHealth intervention strategy (or lack thereof).

Setting Jujuy province is located in northwest Argentina and 85% of its population live in urban areas.19 The primary health care (PHC) system has 270 health centres and employs about 700 full-time CHWs who visit approximately 110,000 households twice a year for health-related tasks.

Since 2012, HPV-testing has been the primary CC screening test, available for women aged 30 years or older attending public health centres. Women are screened with HPV-testing every five years. Since 2014, HPV self-collection is offered during the CHW routine home visits. Women self-collect samples with a cervical sampler kit (Qiagen, Gaithersburg, MD, USA), which is comprised of a cervical brush, specimen container, and transport medium. Community health workers transport specimens at room temperature to health centres; from where they are sent to the provincial HPV laboratory to be analysed for 13 high-risk HPV types using hybrid-capture 2, following the manufacturer's instructions. According to the national guidelines,21 HPV-positive women with self-collected tests must go through cytology (triage Pap), and those HPV-positive women whose triage Paps are classified as atypical squamous cells of undetermined significance or worse (ASCUS+) are referred to colposcopy, then to biopsy if colposcopy images are classified as abnormal following the International Federation of Cervical Pathology and Colposcopy (IFCPC) classification.22 Women with histologically confirmed cervical intraepithelial neoplasia (CIN) of grade 2 or worse are referred to treatment.

All HPV-testing/diagnoses/treatments of women screened in the public health system are registered on the national screening information system (SITAM, for its initials in Spanish). Results of HPV tests and triage Paps are instantly available online to providers at public health establishments.

Setting The research will take place in the province of Jujuy, located in Northwest Argentina. Eighty-five percent of Jujuy's population lives in urban areas and the rate of mobile phone penetration was 82% of urban households in 201113 Jujuy has a high cervical cancer mortality rate (10.6 per 100.000 women during 2012-2014).

The Jujuy public health system includes a main hospital and 300 primary health care centers. For the uninsured, health services are provided free of cost. The Primary Health Care System (PHC) integrates approximately 700 paid full-time community health workers (CHWs) who twice yearly visit approximately 110,000 households for health-related tasks such as immunization and promoting maternal and child health.

Study design The study design follows the structure of an effectiveness-implementation hybrid type I trial14 and uses a mix-methods approach.

It combines a cluster randomized trial to evaluate the effectiveness of a multi-component mHealth intervention with a mixed methods approach involving quantitative and qualitative evaluations of the implementation using the RE-AIM15 and Consolidated Framework for Implementation Research (CFIR)16 frameworks.

The study will integrate quantitative and qualitative methods and follow Greene's typology of mixed-methods for convergence and complementarity.17 A formative research will be carried out to develop the framing and content of SMS messages (qualitative secondary method).

Then a pragmatic cluster randomized trial will be carried out to measure the effectiveness of the multi-component mHealth intervention in improving triage adherence among HPV+ women (quantitative primary method for Specific Aim 1).

Finally, during the post-intervention phase (Implementation evaluation for Specific Aim 2), a quantitative survey will be carried out to measure acceptability of the strategy by CHWs (quantitative secondary method), and semi-structured interviews (qualitative secondary methods) will be conducted among health authorities and health professionals,18 to receive feedback on their own perspectives and opinions about the intervention.19 HPV+ women will also be interviewed via a structured questionnaire (quantitative secondary method) to collect information about acceptability of the intervention, reasons for triage adherence/non-adherence, and their views about the strategy.

Formative research. Focus group discussions (FGD) will be carried out to explore women's opinions about how to frame and phrase SMS messages. Barriers women face in accessing triage will be explored, and the resulting information will be used to develop strategies CHWs can use to address them during the intervention. Drawing from FGD results, SMS messages will be designed and pilot tested with a validation protocol that has already been used in Argentina.20 Development of Automated Messaging system: an Automated System (AS) will be developed, which will then send SMS messages and e-mails based on data from the national Screening Information System (SITAM). This AS will will work in an asynchronous manner. In order to avoid sending a SMS message to the wrong person, an identity verification process will be initiated. Only matching data will be classified as "valid" and receive subsequent SMS messages. E-mails sent to CHWs will not include personal information about women, but a secure link to registries of women to be viewed in SITAM. SMS messages and e-mails will comply with security procedures.

Intervention: cluster-randomized trial. Approximately 200 CHWs from Jujuy's PHC will be randomized into two groups: 1) mHealth Intervention (MH) Group: Women with positive self-collected tests will receive a multi-component intervention. HPV+ women will receive periodic SMS messages during the first month after registration of self-collection test result in SITAM, notifying them that test results are available and they should go to the health center. In addition, CHWs will receive an e-mail and SMS message prompting them to visit those HPV+ women who, at 60 days since availability of the HPV result in SITAM, have not attended triage. CHWs will be asked to confirm receipt of the e-mail for monitoring project processes. Within 15 days of being notified, CHWs will visit these HPV+ women, for an in-person reminder and counseling about the importance of attending health centers and have triage performed. 2) Usual Care (UC) Group: Women with positive self-collected tests will receive usual care, i.e., women opting for self-collection will be instructed to go the health center in 30 days to pick up results.

Hypothesis: Adherence to triage will be higher in the MH intervention group compared to women receiving UC.

Study procedures. CHWs from both groups will identify eligible women and invite them to participate in the study during their routine home visits. Once women have performed HPV-self collection, CHWs will screen for eligibility criteria, including use of mobile phones. Those women who use a mobile phone will be explained the study objectives and will be invited to participate in the study. Women accepting to participate will be asked to provide a telephone number. Based on previous studies,21,22 the investigators expect approximately 85% of women will provide a telephone number; with 70% providing their own personal mobile phone.

MH Group: CHWs will carefully explain participants the sequence of SMS messages to be received depending on the Test results, and how they should proceed after reception of the SMS messages. If the woman does not have a personal phone, she will be asked if she is willing to be contacted and receive the proposed SMS messages on a shared phone. CHWs will inform women about the nature and content of the SMS messages to be received. They will also explain that if they have not performed triage 60 days after a positive test results they will come back for an in-person visit. Therefore, consent to participate also means women agreement to be contacted and to receiving SMS messages, even if phones are shared by several family members.

Upon SITAM registration of the HPV result, a first SMS message will be sent requiring women a confirmatory reply with their id number. This confirmatory approach to assure a patient's identity has been used and proved effective in other studies.22,23 After confirmation, women will receive a second SMS message with different information depending on their HPV status: if HPV-, they will be informed about the result and date of next screening; if HPV+, they will be informed that test results are available at the health center. Women will know in advance that this SMS message will mean they are HPV+. During the first four weeks after registration of a HPV result in SITAM, a SMS message will be sent weekly unless a Pap result is registered in SITAM. Sixty days after the date of registration of the HPV test result in SITAM, CHWs will receive an e-mail and SMS message informing them that there are HPV+ women who have not had Pap triage (if this is the case). The e-mail will contain a secure link to the list of HPV+ women to be visited. CHWs will be asked to confirm receipt of the e-mail for monitoring project processes. CHWs will visit these HPV+ women within 15 days of being notified, for an in-person reminder and counseling about the importance of attending health centers and have triage performed. CHW visit information will be recorded, including visit attempts, date of visit and issues addressed during counseling.

UC care group: Once women have agreed to participate in the project, CHWs will provide UC counseling using specific24 standard provincial protocol and materials. They will inform women that they should go to the health center in a month to get their test results.

Randomization and sample size: All CHWs (clusters) (approximately 700) of the Jujuy province offer self-collection and will be eligible to participate in the study. After classifying them into four groups according to gender and urban/rural setting, a stratified sample of 240 CHWs will be randomly selected with allocation proportional to strata. Selected CHWs will be invited to participate in the study following the order defined by the random list until 200 are enrolled. Enrolled CHWs will be randomly allocated to the MH or the UC group (3:2 ratio). Blinding of intervention and outcome assessments is not feasible due to the characteristics of the study. Primary outcome: The target sample size (120 CHWs in the MH group and 80 CHWs in the UC group, each one providing 2 HPV+ women) will have 97% power to detect a 20% triage difference between the two groups when the control group has a 30% triage in a 120-days period (two-sided test, alpha=0.05). Correlation induced by the CHWs was included in power calculations assuming an intra-class correlation coefficient of 0.10. Secondary outcome: Under the same assumptions, the proposed sample size will have 90% power to detect a difference of 10% in triage between groups when the control group has a 15% triage by day 60.

Analysis plan: All analysis will be conducted in an intention-to-treat basis, i.e. an enrolled woman with a positive HPV result will be considered in the study group assigned to her CHW, regardless of adherence of the CHW to the intervention or the number of SMS messages delivered to the woman. The effect of the multi-component mHealth intervention on the primary outcome will be assessed using a generalized estimating equations approach. The model, logit link, and binomial distribution will include trial group as fixed effect and CHWs as the clustering variable. Potential modification of the effect of the intervention due to rural/urban location of the woman and CHW gender will be evaluated, additionally including these factors and their interactions with study group in the former model. Baseline women factors will be compared between groups. Factors showing clear unbalance between arms will be included in the model proposed above to adjust for potential confounding. The same approach will be used to estimate the intervention effect on the secondary outcome. An interrupted time series approach will be used to explore the differential impact of each component of the intervention (SMS messages to women from day 1-59, SMS message and e-mail to CHWs by day 60) on the total number of women with triage at each week. Baseline for each woman will be defined as the day the HPV tests results are recorded in SITAM. Segmented regression will be used to assess the trend of the outcome in each period (1-59; 60-120 days) and any potential discontinuity (change in level or profile) in the time series of the intervention group compared to the control group. A linear model accounting for the autocorrelation of the data and including period (1-59; 60-120 days), intervention indicator (yes/no), and functions of time before and after the breakpoint (date when the second component starts) will be fitted. Given that the number of HPV+ women with triage is not expected to follow a linear trend with time, smooth functions (spline) of time will be used to model the relation between outcome and time in each study group. Under the assumption that the trend in the outcome would remain unchanged in the absence of the second component, this approach will allow investigators to describe the counterfactual curve expected in the intervention group had the second component not been implemented. Analyses will be conducted using STATA and/or SAS. The trial will follow the CONSORT extension for cluster randomized trials reporting guidelines 25.