Introducing a Telemedicine Strategy to Deliver Cervical Cancer Screening Results in El Salvador

Telemedicine, the use of telecommunication technologies to provide healthcare services remotely, has grown exponentially in the last decade. While some telemedicine applications are only feasible in high-resource settings, other types have been shown to be effective in low and middle-income countries (LMICs). Interventions that utilize mobile devices such as cellular phones and digital tablets, collectively known as mobile health or mHealth, may enable greater patient access to some services and reduce the strain on resource-limited health systems. Here we describe the use of an mHealth strategy to facilitate delivery of cervical cancer screening results in El Salvador.

This was a cross-sectional comparative study in primary health clinics of the Paracentral region of El Salvador. Eight clinics matched for size and location (rural vs. urban) were assigned to either the intervention (n=4) or the control condition (n=4). Recruitment continued in the clinics' catchment areas until the target of 550 women in each group was reached for a total of 1,100 participants. Given previous research, we expected that 12% of all women (n=66 in each group or 132 total) would be HPV+, while the remainder would be HPV- (n=484 in each group or 968 total). Using a Fisher's exact test with one-sided 0.10 type 1 error, a sample of 126 HPV+ women (63 in each group) would have 80% power to detect a 14% increase (80% vs 94%) in the proportion of women who obtained their test results. For those with HPV- results, using a Fisher's exact test with two-sided 0.05 type 1 error, a sample of 880 women (440 in each group) would have 80% power to detect a 10% increase (60% vs 70%) in the proportion who obtained their results.

Women were recruited to the study during screening visits at the participating clinics. Eligibility criteria were similar to those of the HPV screening program. Women aged 30 to 59, with no history of cervical cancer, no HPV testing in the last 5 years, and willing and able to provide informed consent were included. Those with no access to a working mobile phone were excluded. Informed consent procedures and an intake questionnaire were administered by experienced research assistants before participants underwent HPV sampling.

In the intervention group, all screened women received a SMS text message when their test results became available. Texts were sent automatically through Twilio (Twilio Inc., San Francisco, CA), a commercially available digital platform embedded into the study's REDCap database. The text messages were developed by native Spanish speakers in consultation with MoH personnel to ensure they were brief, clear, and understandable at a basic reading level. For HPV+ women, the text stated that the test had been processed and that the individual needed to return to the local clinic to receive their results. This ensured that this group of individuals received counseling and indications for follow-up treatment. Texts also included a telephone number to call in case of questions. For HPV- women in the intervention group, the message stated that the test result was negative and recommended that the recipient be re-screened in 5 years. A phone number was also included in case of questions. In the control group, all women regardless of HPV status followed the routine procedure of scheduling an appointment after 30 days to receive their results in-person at their primary care clinic. A survey was administered to ascertain delivery of HPV test results and to evaluate acceptability of the result delivery method. In the intervention group, HPV- women were administered the survey via phone call 4 weeks after screening. After three unanswered phone call attempts, a community health worker visited the home to conduct the survey. For participants who were required to pick up their test results in person (i.e., HPV+ positive women in the intervention group, HPV+ and HPV- women in the control group), the survey was conducted when they returned to the clinic. Women who did not show up after 30 days or missed their previously scheduled appointment were contacted at least three times via phone call or received a home visit by a community health worker to remind them to pick up their results.

Fisher's exact test was used to compare the proportion of women who received their results in each group within 45 days. Descriptive statistics were used to compare sociodemographic variables and results of the acceptability survey.