JMIR Publications

ABSTRACT

Background:

Screening is important for early detection of cancer of the cervix. In Low- and Middle-Income Countries (LMICs) Visual Inspection with Acetic acid (VIA) is usually the method of screening, However, interpretation of VIA results is subject to inter- and intra-observer variability.

Objective:

The goal of this study was to analyze the process, barriers and facilitators of introducing under field conditions an Artificial Intelligence-Decision Support System (AI-DSS) device, to increase consistency of interpretation of VIA results in Bangladesh and Uganda. We used the Consolidated Framework for Implementation Research (CFIR) to structure the findings.

Methods:

We operationalized the CFIR (innovation, process, outer setting, inner setting, individuals) for this specific study and defined the constructs for analysis. We performed the study in Uganda and Bangladesh. We extracted relevant information from routine data, from patient surveys and from facility surveys. We interviewed health workers who used the AI-DSS devices to evaluate their experiences. A panel of experts performed an analysis of the quality of pictures and the performance of health workers. We monitored implementation through quarterly meetings and documented the process.

Results:

Innovation AI-DSS: the applied AI-DSS was evidence-based with a relative advance of improving accuracy of diagnosis. The hardware design using an adapted mobile phone was adequate. The user interface was user-friendly and intuitive. However, operating the device while performing VIA, was challenging Outer setting: Since the device was not registered as medical device, it was only used for research purposes. In both countries there was no official government policy concerning use of AI in healthcare. All facilities had gynecological examination rooms, but some facilities in the research areas did not have permanent electricity or internet connection. Inner Setting: The normal procedure for VIA was followed. The AI-DSS did not need additional materials or supplies. AI-DSS did not overburden health workers. Individuals: The AI-DSS was well accepted by women, when privacy was guaranteed, but there seemed to be more trust in the judgement of health workers. Health workers used the feedback from AI to evaluate their own assessment. The pictures helped supervisors to give a second opinion and were used for training purposes. The technical support team was able to help through remotely and improved the performance. Implementation process: Training health workers in taking good pictures was very important. A permanent monitoring process during the implementation was established, which led to early detection and correction of shortfalls.

Conclusions:

This study showed that the CFIR is a proper tool for categorizing barriers and facilitators when introducing an AI-DSS device in practice. The device has the potential for improving quality of the assessment by health workers who perform VIA. However competent staff trained in performing VIA will be indispensable for capturing high quality pictures. Clinical Trial: 2022300195