JMIR Publications

ABSTRACT

Background:

Background：Cervical cancer poses a significant threat to women's health worldwide, and increasing screening coverage is a crucial intervention to reduce its incidence. However, in most countries, particularly low- and middle-income countries (LMICs), multiple barriers hinder the improvement of screening rates, making it difficult to achieve the World Health Organization’s (WHO) cervical cancer elimination target of 70% screening coverage. AI-assisted screening presents a promising approach to rapidly achieving universal cervical cancer screening, as demonstrated by practices in Hubei Province, China.

Objective:

Objective:

This study explores the influencing factors in AI-assisted large-scale screening through interviews with stakeholders in Hubei Province, providing insights for LMICs to accelerate progress toward cervical cancer elimination.

Methods:

Methods:

A semi-structured interview guide was developed under the guidance of the Macro-Model of Health Systems (MMHS), employing a multi-stage stratified sampling approach to recruit stakeholders from 14 county/district-level screening institutions for semi-structured interviews. The interviews were conducted between January and August 2024. Thematic analysis was applied to analyze the collected data.

Results:

Results:

Twelve factors influencing the comprehensive coverage of AI-assisted cervical cancer screening were identified. Government leadership and screening technology were identified as critical factors for achieving full coverage, while screening funding and public health awareness were important factors. Additional factors encompassed structural dimensions including multi-sectoral coordination, quality control measures, screening personnel qualifications and information system integration; process dimensions such as institutional service delivery capacity and community health promotion efforts; along with outcome measurements comprising population coverage rates, positive case detection rates, treatment initiation rates and follow-up completion rates.

Conclusions:

Conclusions:

Achieving large-scale cervical cancer screening requires coordinated efforts across three dimensions: governance, financing, and service delivery. At the governance level, governments should incorporate screening into public welfare initiatives, establish it as a performance indicator and enhance interdepartmental collaboration. Regarding financing, it is essential to secure sufficient funding with matching provisions and ensure rational allocation of resources. For service delivery, active utilization of AI technologies should be implemented to enhance screening efficiency, while concurrently elevating population awareness of self-health management and cultivating routine screening behaviors. This study provides an evidence base for developing innovative screening strategies and offers actionable insights for low- and middle-income countries to achieve universal screening coverage, thereby accelerating global progress toward cervical cancer elimination.