Patient vs. clinician-collected cervical cytology for screening for CIN: a systematic review and meta-analysis

TL;DR

This study compares self-collected and clinician-collected cervical cytology for detecting cervical cancer, finding that self-collection could help more women get screened, though it's slightly less accurate.

Contribution

The study provides a meta-analysis of self-collected cervical cytology's diagnostic accuracy compared to clinician-collected methods, highlighting its potential for increasing screening participation.

Findings

Self-collected cytology had a pooled sensitivity of 0.698 and specificity of 0.805.

Clinician-collected cytology showed higher sensitivity (0.765) but lower specificity (0.613).

Variation in self-collection methods affects diagnostic performance, suggesting a need for standardization.

Abstract

This systematic review and meta-analysis aimed to compare the diagnostic accuracy of self-collected versus clinician-collected cervical cytology for screening for cervical intraepithelial neoplasia (CIN), using cervical biopsy histopathology as the reference standard. The study sought to evaluate self-collection’s potential to enhance cervical cancer screening participation, particularly among under-screened women, by addressing barriers such as discomfort and logistical challenges.

We searched PubMed (MEDLINE), Web of Science, Scopus, and Cochrane databases from their inception to March 1, 2024, to identify relevant studies comparing self-collected and clinician-collected cervical cytology.

Eligible studies included adult women (≥18 years) undergoing both self-collected and clinician-collected cervical cytology (conventional Pap smear or liquid-based cytology), with cervical biopsy as the reference standard. Studies were required to report diagnostic outcomes (e.g., sensitivity, specificity) and include a clinician-collected comparator. Non-English studies, those lacking biopsy as the reference standard, or single-arm studies were excluded.

Two independent reviewers screened studies, extracted data, and assessed methodological quality using the QUADAS-2 tool. A diagnostic meta-analysis was conducted with Review Manager 5.4 to calculate pooled sensitivity and specificity with 95% confidence intervals, using both unweighted and weighted methods. Summary receiver operating characteristic (SROC) curves were generated to visualize diagnostic performance.

Five studies were included, revealing that self-collected cytology had a pooled sensitivity of 0.698 and specificity of 0.805, while clinician-collected cytology showed a higher sensitivity of 0.765 but lower specificity of 0.613, with variation in sensitivity and specificity observed across different self-collection methods (e.g., brushes, tampons, lavages). QUADAS-2 assessment identified methodological concerns, particularly in patient selection and flow and timing, suggesting risks of bias and limited generalizability to primary screening settings. Methodological limitations in studies, including wide variation in self-collection methods, highlight the need for further research.

Self-collected cervical cytology demonstrates reasonable diagnostic accuracy, offering a promising approach to increase screening uptake among under-screened populations by overcoming barriers to traditional clinician-based methods. However, its lower sensitivity compared to clinician-collected cytology underscores the need for careful implementation, including optimized assay selection and targeted population strategies. Heterogeneity in self-collection methods, which varied widely (e.g., Mermaid rinse, tampons, brushes), may contribute to differences in diagnostic performance and underscores the need for standardized approaches. Further high-quality research is essential to refine self-collection techniques and strengthen its role in global cervical cancer prevention efforts.

Figures