Although preventable, Cervical Cancer (CC) is the leading cause of cancer deaths among women in Sub-Saharan Africa with the highest incidence in East Africa. Kenyan guidelines recommend an immediate screen and treat approach using either Pap smear or visual screening methods. However, system (e.g., inadequate infrastructure, weak treatment, referral and tracking systems) and patient (e.g., stigma, limited accessibility, finance) barriers to comprehensive country wide screening continue to exist creating gaps in the pathways of care. These gaps result in low rates of eligible women being screened for CC and a high loss to follow up rate for treatment. The long-term goal of 70% CC screening and treatment coverage can partly be achieved by leveraging electronic health (eHealth, defined here as systems using Internet, computer, or mobile applications to support the provision of health services) to support service efficiency and client retention. To help address system level barriers to CC screening treatment and follow up, our team developed an eHealth tool—the Cancer Tracking System (CATSystem), to support CC screening, treatment, and on-site and external referrals for reproductive age women in Kenya. Preliminary data showed a higher proportion of women enrolled in the CATSystem receiving clinically adequate (patients tested positive were treated or rescreened to confirm negative within 3 months) follow up after a positive/suspicious screening, compared to women in the retrospective arm.
This study presents the findings from a pilot study utilizing an eHealth tool—the Cancer Tracking System (CATSystem), an adaptation of a web-based eHealth intervention called the HIV Infant Tracking System (HITSystem) (21). Designed for use in low to middle income settings, the primary goals of the CATSystem are to (a) increase rates of CC screening, (b) improve the treatment, referral (internal and external), and follow up rates of women screened positive with precancerous and cancerous lesions and c) identify missed re-screening and treatment opportunities. Using algorithm driven alerts for providers and SMS to patients, the CATSystem is designed to support CC screening, treatment, and referrals for reproductive age women (HIV + and HIV–) in Kenya. The system accesses satellite broadband via modems, generates a provider dashboard linking to patients who are overdue for a service or in need of patient outreach, and sends automated customized text messages to women to support screening and treatment follow-up per national Kenyan Ministry of Health guidelines (14). Decentralized for data entry, authorized providers (mentor mothers, data clerks, or clinicians) enter data in real-time at implementing hospitals, allowing the generation of timely alerts and provider follow up (patient tracing, phone calls, or SMS); however, as a web-based intervention, centralized updates to the programming of the system are automatically applied across sites as they become available. This study was an observational study with historical controls to evaluate an 11-month pilot of the CATSystem at one provincial level hospital in Rift Valley, Kenya. The standard of care at the facility includes paper-based record keeping for CC screening (VILLI/VIA, pap smear and colposcopy) and on-site treatment with cryotherapy and loop electrosurgical excision procedure (LEEP). Patients needing chemotherapy or radiation treatment are referred to Kenyatta National Hospital, in Nairobi, which is approximately 160 km (3–4 h drive one-way) from the study hospital. Adapted from the HITSystem, which is an eHealth intervention that has proven effective in improving maternal and infant HIV care in Kenya (21, 22), the CATSystem is an eHealth intervention that aims to improve follow up after an abnormal CC screen and increase rescreening rates per Kenya Ministry of health Guidelines (see Figure 1). Women are enrolled in the CATSystem through the comprehensive HIV care centers (CCC) and maternal and child health/Family Planning (MCH/FP) department. Demographic and contact information are captured at the time of enrollment. The patient is assessed for risk factors of CC including HIV status, age of onset of sexual activity, number of sexual partners, history of sexually transmitted infections, prior positive screenings, or history of in situ carcinoma of the vulvar or vaginal epithelium. Women are screened and those who have an abnormal CC screen are ideally treated on the same day. If the patient is unable to undergo same day treatment after a positive screen, the system sends an automated SMS alerts prior to the scheduled appointment and alerts in case of a missed appointment. They are tracked until they complete clinically indicated care based on their unique clinical presentation. After completing appropriate management for suspicious malignancy/in situ carcinoma or invasive malignancy, the patient continues to get automated alerts for follow up care based on national guidelines. Women who have a normal screen are prompted for rescreening at the indicated interval. Algorithm-driven electronic alerts notify clinical providers when patients are overdue or missing key services (treatment, labs, follow up care, missed appointments). The system also keeps a record of clinical findings, visual images of the cervix and lab results from each encounter. Kenya ministry of health national cervical cancer screening guidelines. The CATSystem was piloted in the Comprehensive Care Center (CCC, where HIV services are provided) and Maternal and Child Health Departments (MCH) from October 22nd, 2019, to January 26th, 2021. These departments are the two main points where CC screening occurs. The inclusion of the CCC also allowed us recruit women living with HIV who are vulnerable to CC due to the associated risk between HIV and contracting the Human Papilloma Virus (HPV) (26, 27). Each department had one or two nurses or clinical officers to conduct screenings and cryotherapy treatments. The facility had two gyno-oncologists to perform complicated procedures [biopsies, loop electrosurgical excision procedure (LEEP)]. Treatments for invasive CC (chemotherapy, radiation, radical hysterectomy) were referred to the highest tier referral hospitals. During this time, hospital operations across the country were limited or shut down due to multiple healthcare worker strikes (2019, Dec 2020–Feb 2021) and COVID-19 mitigation strategies (May–June 2020) affecting daily hospital operations and pilot study data collection. Data from enrolled participants were compared to data from historical controls who were screened for CC in the 6 months prior to CATSystem implementation April 2019–October 2019. All women ages 18–50 years who received CC screening in CCC or MCH during the study period were eligible. Due to the CATSystem’s use of SMS text messages, women also needed cell phone access to participate in the study. All participants provided written informed consent prior to study participation. During the historical control period, paper CC screening registries from CCC and MCH were reviewed. Data were entered in an Excel spreadsheet by study staff. Although the registries followed the required ministry of health format, the data quality was highly variable regarding consistency, completeness, and evidence of follow up. Where paper records were incomplete, study staff followed up with providers to fill in gaps as much as was possible and ensured that duplicate records were reconciled; however, in many cases retrospective data remained incomplete. Clinic providers or study staff informed women presenting for CC screening about the purpose of the CATSystem and asked eligible women if they would like to participate in the study or would prefer to receive standard of care (paper-based record keeping without action alerts or communication). Participants’ demographic and place/s of residence information were entered into the CATSystem upon enrollment. All subsequent counseling and clinical care data (including appointments, laboratory tests and results, treatment, rescreening) were entered into the CATSystem by a study research assistant. The CATSystem then used dates of services and other clinical criteria (i.e., screening results, appointment dates) to trigger electronic alerts to prompt providers when time-sensitive actions along the CC cascade of care were needed (see Figure 2 for an image of the dashboard). Clicking on each alert name would bring up a list of patients requiring that service, allowing providers to easily identify and initiate follow up among patients with incomplete services. Alerts were only resolved once the indicated action had taken place and had been recorded in the CATSystem. Finally, an informal “lunch and learn” style group discussion was held with providers (CCC Nurse, MCH Nurse, MCH Clinical Officer, two Gyno-oncologists and Laboratory technician) where facilitators and challenges to utilization and identified modifications to the CATSystem that could improve its utility and implementation. Cancer tracking system (CATSystem) dashboard. The primary outcome of our study was “clinically appropriate care” after an abnormal CC screen defined as completion of any of the following actions: (1) Onsite treatment for precancerous lesions, (2) onsite or referred LEEP treatment for more severe precancerous lesions, (3) referral to a treatment center if suspected of invasive cancer, or (4) treatment of coinfections (e.g., cervicitis or STI), followed by a re-screen with appropriate follow-up within 3 months. We compared the pilot data to a 6-month retrospective record review of all female patients seen in settings where CC screening should have been conducted prior to CATSystem implementation. Descriptive statistics were calculated for demographic variables and risk factors for CC. Continuous variables were expressed as mean ± SD where applicable. Due to limitation in availability of historical control data, output is only available for intervention arm demographic and risk factor variables. The primary outcome was proportion of patients with a positive screening obtaining clinically appropriate care. A chi-square analysis was done on the categorical outcome to determine significance. Data was analyzed using SPSS v27.
N/A