Pocket Colposcopy sharpens cervical triage for women living with HIV in Kenya
A randomized trial in Kisumu, Kenya, suggests that Pocket Colposcopy can strengthen cervical cancer triage for women living with HIV when paired with acetic acid inspection. In a setting where HPV screening is already part of routine care, the central problem was not screening alone but what happened next: which women truly needed treatment, which needed biopsy confirmation, and which could be safely spared unnecessary intervention.
The trial compared naked-eye visual inspection with acetic acid, or VIA, against an enhanced approach using the Pocket Colposcope, a portable device linked to a mobile application for image capture and storage. Across six HIV care clinics between November 2022 and June 2023, investigators enrolled HPV-positive women living with HIV aged 25 to 65 years who were planning to undergo treatment at the clinic. Women with pregnancy, prior cervical cancer, or suspected invasive disease were excluded. Randomization was stratified by clinic, and the study was designed to reflect a “screen, triage, treat” pathway recommended by the World Health Organization for this higher-risk population.
The result was a direct comparison between a familiar low-cost method and a device-assisted version designed to bring more structure to the exam room.
Biopsy-confirmed disease and exam performance diverged in meaningful ways
Across the cohort, the positive exam rate was 17.3% in the VIA arm and 14.3% in the Pocket arm. Histology found CIN2/3 in 15.4% of participants overall, with prevalence of 12.2% in the VIA arm and 18.8% in the Pocket arm. Sensitivity, specificity, and negative predictive value were similar between the two approaches, but the Pocket Colposcope produced a numerically higher positive predictive value, 37.5% versus 20.6%, although that difference did not reach statistical significance.
One result stands out plainly: the device-assisted arm found more treatable disease while also reducing overtreatment.
That pattern matters in an environment where overtreatment drains scarce resources and exposes women to procedures they may not need. It also matters because women living with HIV face a higher burden of cervical precancer and cancer, making missed disease particularly costly. The study’s design used biopsy confirmation to anchor the visual diagnosis, with biopsies taken from aceto-white epithelium when present and from random sites when the cervix appeared negative. In women with lesions too large for standard treatment or with suspicion of invasive carcinoma, referral rather than enrollment was the correct path.
Why the Pocket approach fits the screening gap in resource-limited clinics
Cervical cancer prevention is often described as a screening problem, but in practice it is an infrastructure problem. High-income settings rely on colposcopy, pathology, and referral networks that are difficult to reproduce in many low-resource clinics. VIA remains inexpensive, but its performance depends heavily on examiner skill, lighting, and the ability to review or share images afterward. Those constraints can produce variation from one clinic to another and from one clinician to the next.
The Pocket Colposcope was built to address those gaps. The device is inserted through a speculum, positioned close to the cervix, and paired with a phone-based application for data capture and image storage. It provides white and green light imaging, although only white light images were used for diagnosis in this trial; pre-acetic and green light images were collected for future algorithm development. The concept is simple: make the exam more visible, more documentable, and easier to teach.
In the Kisumu trial, that portability did not come at the expense of acceptability. Both providers and patients found the Pocket approach acceptable for triage. In a service line where patient comfort, workflow, and trust all influence whether screening is repeated or completed, that detail carries operational weight. A device that performs well but disrupts clinic flow would have limited reach. Here, the workflow was embedded into routine HIV care, and providers completed the exam as part of their regular duties after structured training.
Training and workflow were part of the intervention, not an afterthought
The trial did not treat technology as a stand-alone fix. Kenyan providers received training through a “train the trainer” model that paired an experienced clinical officer with Duke University staff, followed by further instruction from KEMRI study staff. The training covered cervical cancer basics, HPV, VIA technique, Pocket Colposcope imaging, software use, and clinic workflow optimization. Weekly calls helped resolve issues during implementation.
That design matters because visual triage is not simply about the lens in front of the cervix. It also depends on the person holding it, the confidence of the team using it, and the way results move through the clinic. The Calla Health App supported patient data collection and image storage across facilities, which adds an important layer for quality control and remote review. For clinics trying to build capacity without building a full colposcopy service, image storage and transfer can help bridge the gap between bedside assessment and expert oversight.
What the trial suggests for HPV-positive women living with HIV
The study focused on a group with clear clinical urgency: women living with HIV who tested HPV-positive in routine care. In that population, the WHO recommends a triage step rather than immediate treatment for everyone, because screen-and-treat alone can drive overtreatment and still miss disease. The Pocket Colposcope appears to fit that triage role better than naked-eye VIA alone, at least in this Kenyan implementation study.
The device identified more biopsy-proven treatable disease, had comparable diagnostic accuracy measures to standard VIA, and showed a higher positive predictive value without a clear penalty in acceptability. It also offered a practical benefit that may be as important as any single accuracy metric: the possibility of image-based confirmation and remote review. For a cervical cancer prevention program operating under resource constraints, that can mean better supervision, better training, and better consistency across sites.
A technical error in the randomization scheme created imbalance in participant numbers between arms, and the study was conducted at six clinics in one region of western Kenya, so broader generalization will need caution. Still, the trial directly addresses a challenge that has hampered cervical cancer prevention in many low- and middle-income settings: how to make triage accurate enough to guide treatment without building a high-cost colposcopy service from scratch.
The remaining question is not whether cervical cancer prevention is needed. It is whether low-cost image-assisted triage can be scaled reliably enough to change everyday practice in clinics that already carry a heavy HIV burden. This trial moves that question from theory into the exam room.