Send to

Choose Destination
Cancer Epidemiol Biomarkers Prev. 2018 Nov;27(11):1342-1351. doi: 10.1158/1055-9965.EPI-18-0480. Epub 2018 Aug 14.

Determinants for Participation in Human Papillomavirus Self-Sampling among Nonattenders to Cervical Cancer Screening in Denmark.

Author information

Unit of Virus, Lifestyle, and Genes, Danish Cancer Society Research Centre, Copenhagen, Denmark.
Unit of Statistics and Pharmacoepidemiology, Danish Cancer Society Research Centre, Copenhagen, Denmark.
Department of Pathology, Copenhagen University Hospital Hvidovre, Hvidovre, Denmark.
Clinical Research Centre, Copenhagen University Hospital Hvidovre, Hvidovre, Denmark.
Unit of Virus, Lifestyle, and Genes, Danish Cancer Society Research Centre, Copenhagen, Denmark.
Department of Gynecology, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark.


Background: Offering human papillomavirus-based self-sampling to nonparticipants in routine cervical cancer screening can increase screening participation. However, little is known about characteristics of women who accept self-sampling. In this population-based study, we investigated determinants for participation in self-sampling among Danish nonattenders to routine cervical cancer screening.Methods: During 2014 to 2015, a random sample of screening nonparticipants ages 27 to 65 years living in the Capital Region of Denmark were invited for self-sampling. Of 21,314 eligible women, 4,743 participated in self-sampling. Information on sociodemographic characteristics and mental and physical health of all the women was obtained from nationwide registries, and 3,707 women completed a questionnaire on lifestyle, sexual behavior, and reasons for nonparticipation in routine screening. We used logistic regression to estimate ORs for participation in self-sampling, crude, and adjusted for sociodemographic characteristics.Results: Basic education [ORadjusted = 0.79; 95% confidence interval (CI), 0.72-0.88], low income (ORadjusted = 0.66; 95% CI, 0.59-0.73), origin from a nonwestern country (ORadjusted = 0.43; 95% CI, 0.38-0.48), and being unmarried (ORadjusted = 0.66; 95% CI, 0.61-0.72) were associated with lower self-sampling participation. Long-term unscreened women (ORadjusted = 0.49; 95% CI, 0.45-0.53), women with prior schizophrenia or other psychoses (ORadjusted = 0.62; 95% CI, 0.48-0.80), women with poor self-perceived health (ORadjusted = 0.42; 95% CI, 0.25-0.69), and women who perceived screening as unnecessary (ORadjusted = 0.54; 95% CI, 0.37-0.80) or irrelevant (ORadjusted = 0.81; 95% CI, 0.78-0.96) were less likely to self-sample.Conclusions: Certain population groups, including women with low socioeconomic position or of nonwestern origin, were less likely to participate in self-sampling.Impact: Targeted approaches may be needed to increase screening participation in these groups. Cancer Epidemiol Biomarkers Prev; 27(11); 1342-51. ©2018 AACR.

Supplemental Content

Full text links

Icon for HighWire
Loading ...
Support Center