Detecting the impact of diagnostic procedures in Pap-positive women on anxiety using artificial neural networks.

Introduction: Women who receive a result of an abnormal Papanicolaou (Pap) smear can fail to participate in follow up procedures, and this is often due to anxiety. This study aimed to apply artificial neural networks (ANN) in prediction of anxiety in women with an abnormal Pap smear test, prior to and following diagnostic procedures.
Methods: One hundred-seventy two women who received an abnormal Pap screening result took part in this study, completing a questionnaire about socio-demographic characteristics and Hospital Anxiety and Depression Scale (HADS), right before and two to four weeks after diagnostics (i.e. colposcopy/biopsy/endocervical curettage). A feedforward back-propagation multilayer perceptron model was applied in analysis.
Results: Prior to diagnostic procedures 50.0% of women experienced anxiety, while after diagnostics anxiety was present in 61.6% of women. The correlation-based feature selection showed that anxiety prior to diagnostic procedures was associated with the use of sedatives, worry score, depression score, and score for concern about health consequences. For anxiety following diagnostics, predictors included rural place of residence, depression score, history of spontaneous abortion, and score for tension and discomfort during colposcopy. The ANN models yielded highly accurate anxiety prediction both prior and after diagnostics, 76.47% and 85.30%, respectively.
Conclusion: The presented findings can aid in identification of those women with a positive Pap screening test who could develop anxiety and thus represent the target group for psychological support, which would consequently improve adherence to follow-up diagnostics and enable timely treatment, finally reducing complications and fatal outcome.
Competing Interests: The authors have declared that no competing interests exist.
(Copyright: © 2024 Ilic et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.)