FC 08-03INCREASING SCREENING ATTENDANCE AMONG LONG-TERM SCREENING NON-ATTENDERS: RANDOMIZED HEALTHCARE POLICY

10. Self-sampling
K.M. Elfström 1, K. Sundström 2, D. Öhman 3, D. Bzhalava 2, Z. Bzhalava 2, A. Carlsten Thor 3, C. Eklund 2, Z. Gzoul 3, H. Lamin 4, J. Dillner 5, S. Törnberg 3.
1Department of Laboratory Medicine, Karolinska Institutet, Screening Unit, Regional Cancer Center Stockholm Gotland (Sweden), 2Department of Laboratory Medicine, Karolinska Institutet (Sweden), 3Screening Unit, Regional Cancer Center Stockholm Gotland (Sweden), 4Department of Pathology, Karolinska University Hospital (Sweden), 5Department of Laboratory Medicine, Karolinska Institutet, Department of Pathology, Karolinska University Hospital (Sweden)

Background / Objectives

The organized cervical screening program of Stockholm, Sweden reaches a 10-year population coverage of 96%, with the remaining 4% of the population constituting a high risk group for cervical cancer.  The organized coordination and quality assurance allows for pilot implementation of novel screening strategies designed to increase coverage among long-term non-attenders. We performed a randomized health services study within the real-life organized screening program. 


Methods

A comparison of the population registry with the regional screening registry identified that 16,437 out of the 413,487 resident women between 23 and 60 years of age had not taken a cervical screening test in at least 10 years, despite annual renewed invitations. Among these long-term non-attenders, 8000 women were randomized to either a) ordering a self-sampling kit using an open source e-Health application b) mailing a HPV self-sampling kit directly to the woman c) an invitation to call the coordinating midwife of the screening program with questions and concerns regarding screening; and d) standard annual renewed invitation letter (routine practice). HPV positive women were referred directly to colposcopy. Participation rates by study arm and outcome of screening tests were identified by registry linkages. 


Results

Overall participation, defined as returning a self-sampling kit (or other screening participation) by arm was as follows: a) 10.7%; b) 18.7%; c) 1.9%; and d) 1.7%. The relative risk of participating in study arm a was 6.3 (4.4-8.9), 11.0 (7.8-15.5) in arm b, and 1.1 (0.7-1.7) in arm c, compared to routine practice (repeat renewed invitation with new appointment) in study arm d. HPV prevalence among women who returned kits in study arms a and b was 12.2%. In total, 63 women were referred to colposcopy, out of which 44 women attended. 


Conclusion

Offering self-sampling increased attendance, even among women who had not responded to more than 10 invitations with appointments in an organized program. Attendance was higher when kits were sent directly but offering women to order a kit did increase attendance at lower costs. 


References