Cervical Cancer Screening Cascade for women living with HIV: a cohort study from Zimbabwe

Abstract

Countries with high HIV prevalence, predominantly in sub-Sahahran Africa, have the highest cervical cancer rates globally. HIV care cascades successfully facilitated the scale-up of antiretroviral therapy. A cascade approach could similarly succeed to scale-up cervical cancer screening, supporting WHO's goal to eliminate cervical cancer. We defined a Cervical Cancer Screening Cascade for women living with HIV (WLHIV), evaluating the continuum of cervical cancer screening integrated into an HIV clinic in Zimbabwe. We included WLHIV aged ≥18 years enrolled at Newlands Clinic in Harare from June 2012-2017 and followed them until June 2018. We used a cascade approach to evaluate the full continuum of secondary prevention from screening to treatment of pre-cancer and follow-up. We report percentages, median time to reach cascade stages, and cumulative incidence at two years with 95% confidence intervals (CI). We used univariable Cox proportional hazard regressions to calculate cause-specific hazard ratios with 95% CIs for factors associated with completing the cascade stages. We included 1624 WLHIV in the study. The cumulative incidence of cervical screening was 85.4% (95% CI 83.5-87.1) at two years. Among the 396 WLHIV who received screen-positive tests in the study, the cumulative incidence of treatment after a positive screening test was 79.5% (95% CI 75.1-83.2) at two years. The cumulative incidence of testing negative at re-screening after treatment was 36.1% (95% CI 31.2-40.7) at two years. Using a cascade approach to evaluate the full continuum of cervical cancer screening, we found less-than 80% of WLHIV received treatment after screen-positive tests and less-than 40% were screen-negative at follow-up. Interventions to improve linkage to treatment for screen-positive WLHIV and studies to understand the clinical significance of screen-positive tests at follow-up among WLHIV are needed. These gaps in the continuum of care must be addressed in order to prevent cervical cancer.

Fig 1. Conceptual model for a Cervical Cancer Screening Cascade nested in the three pillars of cervical cancer prevention: Primary, secondary and tertiary.

The conceptual model considers the three pillars of cervical cancer prevention: primary, secondary and tertiary. A cascade for secondary cervical cancer prevention is detailed and is the focus of this study. The screening arm is represented by the first line and box colors white to dark blue. The preventative treatment arm is represented by the second line and box colors orange to red. The definitions for stages of the cascade can be found in S1 Table. HPV: human papilloma virus.

Fig 2. Flow diagram of women included in the analysis.

ART; antiretroviral therapy, n; number of women.

Fig 3. Number and percentage of women living with HIV (WLHIV) for each stage of the Cervical Cancer Screening Cascade: Screening and preventative treatment arms.

Panel A: screening arm; left: percentages using a fixed denominator (denominator-denominator approach), referring to the denominator “women in care”; right: using a changing denominator (numerator-denominator approach), referring to all WLHIV reaching the previous stage; panel B: preventative treatment arm; left: percentages using a fixed denominator (denominator-denominator approach), referring to the denominator “screen-positive at first screening”; right: using a changing denominator (numerator-denominator approach), referring to all WLHIV reaching the previous stage. The guideline indicated timeframes are described in the Methods and S1 Table. Grey areas represent the percentage of eligible WLHIV who did not achieve a respective cascade stage (retention gaps).

Fig 4. Analysis of cumulative incidence–percentage of women living with HIV (WLHIV) transitioning through the Cervical Cancer Screening Cascade by year.

The two-year cumulative incidence of WLHIV transitioning to each stage of the Cervical Cancer Screening Cascade—Panel A: Cervical Cancer Screening Cascade, time zero: date of first ART clinic visit; Panel B: screening arm, time zero: date of first ART clinic visit; Panel C: preventative treatment arm, time zero: date of initial screen-positivity.