IJC International Journal of Cancer
Improving cervical cancer screening attendance in Finland Anni Virtanen1, Ahti Anttila1, Tapio Luostarinen2, Nea Malila2 and Pekka Nieminen3 1
Mass Screening Registry, Finnish Cancer Registry, Unioninkatu 22, Helsinki, Finland Finnish Cancer Registry, Unioninkatu 22, Helsinki, Finland 3 Department of Obstetrics and Gynecology, Helsinki University Central Hospital, Jorvi Hospital, PL 800, HUS, Finland 2
Good coverage amongst those at risk and high attendance within the cervical cancer screening program are of paramount importance to results. In Finland, a substantial proportion of cancer incidence and especially mortality among women in screening ages 30–60 years is seen among nonattendees.1,2 Attendance in the program is currently 70% with a slightly decreasing trend (www.cancerregistry.fi). In addition to personal invitations to screening, prebooked appointments in invitation letters and reminders sent to non-attendees increase screening attendance.3–12 Selfsampling for high risk human papillomavirus (hrHPV) DNA Key words: cervical cancer screening, attendance, self-sampling, HPV, reminder letters, socioeconomic factors Abbreviations: ASC-US: atypical squamous cells of undetermined significance; CI: confidence interval; CIN21/CIN2: cervical intraepithelial neoplasia grade 2 (moderate) or worse; CIN31/ CIN3: cervical intraepithelial neoplasia grade 3 (severe) or worse; HrHPV: high risk human papillomavirus; LSIL: low-grade squamous intraepithelial neoplasia; RR: relative risk Conflicts of interest statement: PN was a member of the GSK HPV-vaccine Endpoint Committee for HPV-vaccine trials. Other authors have no potential conflicts of interest. Grant sponsors: Academy of Finland and Finnish Cancer Society; Orion-Farmos Research Foundation DOI: 10.1002/ijc.29176 History: Received 21 May 2014; Accepted 22 Aug 2014; Online 1 Sep 2014 Correspondence to: Anni Virtanen, Mass Screening Registry/ Finnish Cancer Registry, Unioninkatu 22, FI-00130 Helsinki, Finland, Tel.: 1358-50-441-5099, E-mail: anni.virtanen@cancer.fi
C 2014 UICC Int. J. Cancer: 136, E677–E684 (2015) V
testing helps to further increase attendance among women who are not reached by the routine screening program.11–19 The ability (relative sensitivity) to detect moderate cervical intraepithelial neoplasias or worse (CIN21) of hrHPVtesting on self-taken samples has been shown to be somewhat lower than that of hrHPV-testing on clinician-based samples, but similar or even more sensitive to a Pap-smear, depending on the cytological threshold.20 The main aim of this study was to study the effect of reminder letters (1st reminder) and self-sampling tests (2nd reminder) as means to increase attendance within the routine cervical cancer screening programme in a wide setting of Finnish municipalities. Screening results among the attendees and original non-attendees were also explored. We further studied the socioeconomic characteristics of nonparticipants to routine screening to find groups that should be in the focus when planning interventions to increase attendance.
Material and Methods In the Finnish program, women aged 30–60 years are invited to screening by their home municipalities in 5-year intervals. Some municipalities also invite women aged 25 and/or 65 years. Women are identified for screening from the population register based on their age and home municipality, and all women with address information available are invited, irrespective of possible recent opportunistic screening or hysterectomy. The study was conducted as part of the routine screening program in 2011–2012 in 31 different municipalities; 11 municipalities took part in both years, whereas 11 only in 2011 and nine only in 2012. Overall the cohort consisted of 31,053 women. The municipalities were scattered from the
Epidemiology
High attendance is essential to cervical cancer screening results. Attendance in the Finnish program is currently at 70%, but extensive opportunistic screening occurs beside the organized. A shift from opportunistic to organized screening is imperative to optimize the costs and impact of screening and minimize potential harms. We evaluated the effect of reminder letters (1st reminder) and self-sampling test (2nd reminder) on program attendance. The study population consisted of 31,053 screening invitees in 31 Finnish municipalities. 8,284 non-attendees after one invitation received a reminder letter and 4,536 further non-attendees were offered a self-sampling option. Socioeconomic factors related to participation were clarified by combining screening data to data from Statistics Finland. Reminder letters increased participation from 72.6% (95% CI 72.1, 73.1) to 79.2% (95% CI 78.8, 79.7) and self-sampling further to 82.2% (95% CI 81.8, 82.7). Reminder letters with scheduled appointments resulted in higher increase than open invitations (10 vs. 6%). Screening of original non-attendees increased the yield of CIN31 lesions by 24%. Non-attendance was associated with young age, immigrant background, lower education level and having never been married. We showed that a total attendance of well over 80% can be achieved within an organized program when the invitational protocol is carefully arranged.
E678
Improving cervical cancer screening attendance
Epidemiology
What’s new? Self-sampling for high-risk human papillomavirus (hrHPV)-testing facilitates access to cervical screening and can thereby increase screening attendance. Attendance in screening programs is also influenced, however, by invitation protocols, such as the use of reminder letters for nonattendees or prefixed appointments noted in invitations. The present study explored attendance in a routine screening setting in Finland, where program attendance is about 70%. The authors found that the combined use of personal invitations and reminder letters, scheduled appointments in invitations and letters, and self-sampling tests sent to nonattendees can potentially raise total attendance to more than 80%.
southern to the most northern parts of Finland, involving both urban and rural areas. Screening visits were arranged locally, but all participating municipalities used the same screening laboratory of the Cancer Society of Finland for the analysis of the samples. The exact flow of women in the invitation protocol is showed in Figure 1. All women belonging to the screening cohort and with a valid address available from the Population Register Centre were invited to screening by personal letters. Non-attendees after the primary invitation received a second invitation (1st reminder) within the same year. However, in 2012 women were not sent a reminder letter if they declined from screening altogether when cancelling the given appointment (a feature added in 2012 to the program used for sending out the invitations). In all, 30 of the 31 municipalities used the same online booking system and primary invitation letter template with pre-booked appointment times. In 2011, 6 municipalities sent reminder letters without a pre-booked appointment, and 16 with a pre-booked appointment. In 2012, all reminder letters were sent with a pre-booked appointment. All invitation letters were sent by local health care personnel at the municipalities. As a second reminder, a self-sampling test was sent out to non-attendees after the reminder letter. Prior to mailing the sample-taking devices, the self-sampling possibility was introduced in an invitation letter with an opt-out option. The invitation letter was in three languages, Finnish, Swedish and English, and the self-sampling kits in Finnish or Swedish but available upon request also in English. The mailings of the self-sampling procedure were sent out centrally from the Mass Screening Registry. There were a few lapses in the invitation protocol; 273 women did not receive a reminder letter from their home municipality despite their non-attendance and 17 women did not receive a self-sampling test in error. The results presented here are based only on women who received invitations according to study protocol; primary invitation, reminder letter if no previous refusal and then self-sampling. As opportunistic samples are not registered in a joint database in Finland, the effects of self-sampling on overall screening coverage (including also Pap-smears taken outside the organized program) were studied using questionnaire data collected together with self-taken samples. Non-attending women were considered under-screened, if they reported their previous Pap-smear being taken 5 years ago or never.
The screening data from mass screening register was combined to data from Statistics Finland to clarify the socioeconomic factors related to participation; mother tongue, education level, type of home municipality and marital status. Statistics Finland divides municipalities into three types—urban, semi-rural and rural—according to the proportion of people living in urban settlements and the population of the largest urban settlement. Education level is recorded in Statistics Finland for those who have completed lower secondary education or higher. For the purposes of this study, we divided the education level into three categories; Primary (including only primary education, currently at 9 years in Finland; including also those with information missing as to the registration protocol in Statistics Finland), secondary (upper and lower secondary education) and tertiary (upper and lower tertiary education and doctoral degree or equivalent). Finnish and Swedish are the two official languages in Finland and mother tongue was thus divided into three groups; Finnish, Swedish or other. Municipalities were also divided to two geographical locations, north and south, using University Hospital areas as guidelines; the University Hospital area of Oulu representing the Northern part and all other areas the Southern part. With regard to age, the population was divided to seven 5-year age groups from 30 to 60, and 201 women aged 25–29 (from municipalities inviting also women aged 25) and 312 women aged 65–67 (from municipalities inviting also women aged 65) were combined to the youngest and oldest age groups, respectively. Self-sampling-test and HPV-analysis
The sample taking at home was done by the Delphi Screener (Delphi Bioscience BV, Scherpenzeel, The Netherlands) which produces a lavage-type sample by rinsing the cervix and upper vagina with saline. Samples were sent to the screening laboratory in a test-tube in regular mail and analyzed using the Hybrid Capture 2 (HC2)-assay. Only samples that produced a visible cell pellet after centrifugation at 1500 rpm were considered adequate. Of the 937 originally returned samples 30 (3.2%) did not fulfill this criterion. Women in question were offered a new device, and 13 returned a new sample with more material. In the end, 920 samples filled the criteria of an adequate sample and are included in the results of the study. Nineteen women chose to take part in screening with a Papsmear after receiving the self-sampling kit. They are included in self-sampling participants by intention-to-treat. Women with a hrHPV-positive result from the self-taken sample were either invited for a Pap-smear (women 80% attendance after primary invitation, reminder letters further increased attendance by 5–11%. The effect of self-sampling was also affected by the preceding attendance rate. When looking at the pooled effect in group of municipalities with lower than 75% attendance rate after reminder letters, self-sampling increased attendance by 6% on average (data not shown). A slightly lower 4% average increase was observed in both groups of municipalities with 75.0–79.9% or 80.0–84.9% attendance after reminder letters. Still, even in municipalities with high >85.0% attendance after reminder letters, self-sampling increased attendance by 3% on average. The effect of pre-booked appointments in reminder letters was also assessed based on 2011 results when 6 municipalities
Epidemiology
Figure 1. Invitation protocol and flow of the study population in 2011–2012. 1Including 651 women who attended screening before the invitation was mailed or who made an appointment to screening (e.g., by phone) and thus received no invitation letter. 2Option valid only in 2012 invitations. 3Women received a letter informing them about the upcoming self-sampling test and giving an opportunity to decline beforehand. 4Out of all women invited to self-sampling. 5Not included in the analysis.
E680
Improving cervical cancer screening attendance
Figure 2. Participation rate among all 30,053 women identified for screening in the study cohort in 2011–2012. 1Including 201 women aged 25–29 years. 2Including 312 women aged 65–67 years. Table 1. Effect of scheduled appointments in reminder letters Total participation rate % (95% CI)
1%
Both invitations with scheduled appointments (n 5 13,559) 1st invitation (with an appointment)
72.9 (72.2–73.7)
2nd invitation (with an appointment) 80.5 (79.8–81.2) 110.4
Characteristics of the non-attendees in routine screening
Self-sampling
Socioeconomic factors related to lower participation in routine screening after two invitations (primary and reminder) were young age, a mother tongue other than Finnish or Swedish (immigrants), a lower education level, living in a rural municipality and having never been married (Fig. 2). All these factors were statistically significant also in the adjusted model (Table 3).
83.8 (83.2–84.4) 14.1
Total increase
115.0
Only 1st invitation with a scheduled appointment (n 5 1,692) 1st invitation (with an appointment)
Epidemiology
opportunistic screening; Table 2). Nearly 29% reported a previous Pap-smear 5 or more years ago or never, and could thus be considered truly under-screened before they took part by self-sampling. Only these women thus demonstrably increased overall screening coverage.
71.1 (68.9–73.3)
2nd invitation (open invitation)
75.1 (73.1–77.2) 15.6
Self-sampling
79.5 (77.6–81.4) 15.9
Total increase
111.8
Achieved attendance rates after primary invitations, reminder letters and self-sampling tests (% of those identified for screening with CIs) and the rise in total participation rate. Based on 2011 results and grouped according to the use of pre-booked appointments in reminder letters (2nd invitations).
sent reminder letters without a pre-booked appointment (open invitations), and 16 with a pre-booked appointment. If reminder letters were sent with a pre-booked appointment, participation by the reminder letter was twofold compared to open invitation (28% versus 14% of those invited; data not shown) and the increase in total attendance was higher (10.3% versus 5.6% increase; Table 1). Effects on screening coverage
Of the 920 non-attendees who took part in screening with a self-taken sample, 57% reported a Pap-smear in the preceding screening interval, i.e. 0–4 years ago (including also
Characteristics of the attendees in self-sampling
There was slight variation in the crude participation rates to self-sampling within different age groups, but no trend with regard to decreasing or increasing age and differences were not significant in the adjusted model (Table 3). Participation by self-sampling was lower among women with a mother tongue other than Finnish or Swedish. A higher education level resulted in significantly higher participation by selfsampling, the difference being almost twofold between the lowest and higher levels of education. Differing from the results from routine screening, participation by self-sampling was highest in the semi-urban municipalities. There were no differences in participation with regard to geographical location or marital status. Self-sampling test results
Of the 920 returned self-taken samples, 109 (11.8%) were hrHPV-positive by HC2 (Table 4). 52 women were invited C 2014 UICC Int. J. Cancer: 136, E677–E684 (2015) V
Table 2. Effect of self-sampling on screening coverage Time from previous Pap-smear
n
% of participants
Improving cervical cancer screening attendance in Finland Anni Virtanen1, Ahti Anttila1, Tapio Luostarinen2, Nea Malila2 and Pekka Nieminen3 1
Mass Screening Registry, Finnish Cancer Registry, Unioninkatu 22, Helsinki, Finland Finnish Cancer Registry, Unioninkatu 22, Helsinki, Finland 3 Department of Obstetrics and Gynecology, Helsinki University Central Hospital, Jorvi Hospital, PL 800, HUS, Finland 2
Good coverage amongst those at risk and high attendance within the cervical cancer screening program are of paramount importance to results. In Finland, a substantial proportion of cancer incidence and especially mortality among women in screening ages 30–60 years is seen among nonattendees.1,2 Attendance in the program is currently 70% with a slightly decreasing trend (www.cancerregistry.fi). In addition to personal invitations to screening, prebooked appointments in invitation letters and reminders sent to non-attendees increase screening attendance.3–12 Selfsampling for high risk human papillomavirus (hrHPV) DNA Key words: cervical cancer screening, attendance, self-sampling, HPV, reminder letters, socioeconomic factors Abbreviations: ASC-US: atypical squamous cells of undetermined significance; CI: confidence interval; CIN21/CIN2: cervical intraepithelial neoplasia grade 2 (moderate) or worse; CIN31/ CIN3: cervical intraepithelial neoplasia grade 3 (severe) or worse; HrHPV: high risk human papillomavirus; LSIL: low-grade squamous intraepithelial neoplasia; RR: relative risk Conflicts of interest statement: PN was a member of the GSK HPV-vaccine Endpoint Committee for HPV-vaccine trials. Other authors have no potential conflicts of interest. Grant sponsors: Academy of Finland and Finnish Cancer Society; Orion-Farmos Research Foundation DOI: 10.1002/ijc.29176 History: Received 21 May 2014; Accepted 22 Aug 2014; Online 1 Sep 2014 Correspondence to: Anni Virtanen, Mass Screening Registry/ Finnish Cancer Registry, Unioninkatu 22, FI-00130 Helsinki, Finland, Tel.: 1358-50-441-5099, E-mail: anni.virtanen@cancer.fi
C 2014 UICC Int. J. Cancer: 136, E677–E684 (2015) V
testing helps to further increase attendance among women who are not reached by the routine screening program.11–19 The ability (relative sensitivity) to detect moderate cervical intraepithelial neoplasias or worse (CIN21) of hrHPVtesting on self-taken samples has been shown to be somewhat lower than that of hrHPV-testing on clinician-based samples, but similar or even more sensitive to a Pap-smear, depending on the cytological threshold.20 The main aim of this study was to study the effect of reminder letters (1st reminder) and self-sampling tests (2nd reminder) as means to increase attendance within the routine cervical cancer screening programme in a wide setting of Finnish municipalities. Screening results among the attendees and original non-attendees were also explored. We further studied the socioeconomic characteristics of nonparticipants to routine screening to find groups that should be in the focus when planning interventions to increase attendance.
Material and Methods In the Finnish program, women aged 30–60 years are invited to screening by their home municipalities in 5-year intervals. Some municipalities also invite women aged 25 and/or 65 years. Women are identified for screening from the population register based on their age and home municipality, and all women with address information available are invited, irrespective of possible recent opportunistic screening or hysterectomy. The study was conducted as part of the routine screening program in 2011–2012 in 31 different municipalities; 11 municipalities took part in both years, whereas 11 only in 2011 and nine only in 2012. Overall the cohort consisted of 31,053 women. The municipalities were scattered from the
Epidemiology
High attendance is essential to cervical cancer screening results. Attendance in the Finnish program is currently at 70%, but extensive opportunistic screening occurs beside the organized. A shift from opportunistic to organized screening is imperative to optimize the costs and impact of screening and minimize potential harms. We evaluated the effect of reminder letters (1st reminder) and self-sampling test (2nd reminder) on program attendance. The study population consisted of 31,053 screening invitees in 31 Finnish municipalities. 8,284 non-attendees after one invitation received a reminder letter and 4,536 further non-attendees were offered a self-sampling option. Socioeconomic factors related to participation were clarified by combining screening data to data from Statistics Finland. Reminder letters increased participation from 72.6% (95% CI 72.1, 73.1) to 79.2% (95% CI 78.8, 79.7) and self-sampling further to 82.2% (95% CI 81.8, 82.7). Reminder letters with scheduled appointments resulted in higher increase than open invitations (10 vs. 6%). Screening of original non-attendees increased the yield of CIN31 lesions by 24%. Non-attendance was associated with young age, immigrant background, lower education level and having never been married. We showed that a total attendance of well over 80% can be achieved within an organized program when the invitational protocol is carefully arranged.
E678
Improving cervical cancer screening attendance
Epidemiology
What’s new? Self-sampling for high-risk human papillomavirus (hrHPV)-testing facilitates access to cervical screening and can thereby increase screening attendance. Attendance in screening programs is also influenced, however, by invitation protocols, such as the use of reminder letters for nonattendees or prefixed appointments noted in invitations. The present study explored attendance in a routine screening setting in Finland, where program attendance is about 70%. The authors found that the combined use of personal invitations and reminder letters, scheduled appointments in invitations and letters, and self-sampling tests sent to nonattendees can potentially raise total attendance to more than 80%.
southern to the most northern parts of Finland, involving both urban and rural areas. Screening visits were arranged locally, but all participating municipalities used the same screening laboratory of the Cancer Society of Finland for the analysis of the samples. The exact flow of women in the invitation protocol is showed in Figure 1. All women belonging to the screening cohort and with a valid address available from the Population Register Centre were invited to screening by personal letters. Non-attendees after the primary invitation received a second invitation (1st reminder) within the same year. However, in 2012 women were not sent a reminder letter if they declined from screening altogether when cancelling the given appointment (a feature added in 2012 to the program used for sending out the invitations). In all, 30 of the 31 municipalities used the same online booking system and primary invitation letter template with pre-booked appointment times. In 2011, 6 municipalities sent reminder letters without a pre-booked appointment, and 16 with a pre-booked appointment. In 2012, all reminder letters were sent with a pre-booked appointment. All invitation letters were sent by local health care personnel at the municipalities. As a second reminder, a self-sampling test was sent out to non-attendees after the reminder letter. Prior to mailing the sample-taking devices, the self-sampling possibility was introduced in an invitation letter with an opt-out option. The invitation letter was in three languages, Finnish, Swedish and English, and the self-sampling kits in Finnish or Swedish but available upon request also in English. The mailings of the self-sampling procedure were sent out centrally from the Mass Screening Registry. There were a few lapses in the invitation protocol; 273 women did not receive a reminder letter from their home municipality despite their non-attendance and 17 women did not receive a self-sampling test in error. The results presented here are based only on women who received invitations according to study protocol; primary invitation, reminder letter if no previous refusal and then self-sampling. As opportunistic samples are not registered in a joint database in Finland, the effects of self-sampling on overall screening coverage (including also Pap-smears taken outside the organized program) were studied using questionnaire data collected together with self-taken samples. Non-attending women were considered under-screened, if they reported their previous Pap-smear being taken 5 years ago or never.
The screening data from mass screening register was combined to data from Statistics Finland to clarify the socioeconomic factors related to participation; mother tongue, education level, type of home municipality and marital status. Statistics Finland divides municipalities into three types—urban, semi-rural and rural—according to the proportion of people living in urban settlements and the population of the largest urban settlement. Education level is recorded in Statistics Finland for those who have completed lower secondary education or higher. For the purposes of this study, we divided the education level into three categories; Primary (including only primary education, currently at 9 years in Finland; including also those with information missing as to the registration protocol in Statistics Finland), secondary (upper and lower secondary education) and tertiary (upper and lower tertiary education and doctoral degree or equivalent). Finnish and Swedish are the two official languages in Finland and mother tongue was thus divided into three groups; Finnish, Swedish or other. Municipalities were also divided to two geographical locations, north and south, using University Hospital areas as guidelines; the University Hospital area of Oulu representing the Northern part and all other areas the Southern part. With regard to age, the population was divided to seven 5-year age groups from 30 to 60, and 201 women aged 25–29 (from municipalities inviting also women aged 25) and 312 women aged 65–67 (from municipalities inviting also women aged 65) were combined to the youngest and oldest age groups, respectively. Self-sampling-test and HPV-analysis
The sample taking at home was done by the Delphi Screener (Delphi Bioscience BV, Scherpenzeel, The Netherlands) which produces a lavage-type sample by rinsing the cervix and upper vagina with saline. Samples were sent to the screening laboratory in a test-tube in regular mail and analyzed using the Hybrid Capture 2 (HC2)-assay. Only samples that produced a visible cell pellet after centrifugation at 1500 rpm were considered adequate. Of the 937 originally returned samples 30 (3.2%) did not fulfill this criterion. Women in question were offered a new device, and 13 returned a new sample with more material. In the end, 920 samples filled the criteria of an adequate sample and are included in the results of the study. Nineteen women chose to take part in screening with a Papsmear after receiving the self-sampling kit. They are included in self-sampling participants by intention-to-treat. Women with a hrHPV-positive result from the self-taken sample were either invited for a Pap-smear (women 80% attendance after primary invitation, reminder letters further increased attendance by 5–11%. The effect of self-sampling was also affected by the preceding attendance rate. When looking at the pooled effect in group of municipalities with lower than 75% attendance rate after reminder letters, self-sampling increased attendance by 6% on average (data not shown). A slightly lower 4% average increase was observed in both groups of municipalities with 75.0–79.9% or 80.0–84.9% attendance after reminder letters. Still, even in municipalities with high >85.0% attendance after reminder letters, self-sampling increased attendance by 3% on average. The effect of pre-booked appointments in reminder letters was also assessed based on 2011 results when 6 municipalities
Epidemiology
Figure 1. Invitation protocol and flow of the study population in 2011–2012. 1Including 651 women who attended screening before the invitation was mailed or who made an appointment to screening (e.g., by phone) and thus received no invitation letter. 2Option valid only in 2012 invitations. 3Women received a letter informing them about the upcoming self-sampling test and giving an opportunity to decline beforehand. 4Out of all women invited to self-sampling. 5Not included in the analysis.
E680
Improving cervical cancer screening attendance
Figure 2. Participation rate among all 30,053 women identified for screening in the study cohort in 2011–2012. 1Including 201 women aged 25–29 years. 2Including 312 women aged 65–67 years. Table 1. Effect of scheduled appointments in reminder letters Total participation rate % (95% CI)
1%
Both invitations with scheduled appointments (n 5 13,559) 1st invitation (with an appointment)
72.9 (72.2–73.7)
2nd invitation (with an appointment) 80.5 (79.8–81.2) 110.4
Characteristics of the non-attendees in routine screening
Self-sampling
Socioeconomic factors related to lower participation in routine screening after two invitations (primary and reminder) were young age, a mother tongue other than Finnish or Swedish (immigrants), a lower education level, living in a rural municipality and having never been married (Fig. 2). All these factors were statistically significant also in the adjusted model (Table 3).
83.8 (83.2–84.4) 14.1
Total increase
115.0
Only 1st invitation with a scheduled appointment (n 5 1,692) 1st invitation (with an appointment)
Epidemiology
opportunistic screening; Table 2). Nearly 29% reported a previous Pap-smear 5 or more years ago or never, and could thus be considered truly under-screened before they took part by self-sampling. Only these women thus demonstrably increased overall screening coverage.
71.1 (68.9–73.3)
2nd invitation (open invitation)
75.1 (73.1–77.2) 15.6
Self-sampling
79.5 (77.6–81.4) 15.9
Total increase
111.8
Achieved attendance rates after primary invitations, reminder letters and self-sampling tests (% of those identified for screening with CIs) and the rise in total participation rate. Based on 2011 results and grouped according to the use of pre-booked appointments in reminder letters (2nd invitations).
sent reminder letters without a pre-booked appointment (open invitations), and 16 with a pre-booked appointment. If reminder letters were sent with a pre-booked appointment, participation by the reminder letter was twofold compared to open invitation (28% versus 14% of those invited; data not shown) and the increase in total attendance was higher (10.3% versus 5.6% increase; Table 1). Effects on screening coverage
Of the 920 non-attendees who took part in screening with a self-taken sample, 57% reported a Pap-smear in the preceding screening interval, i.e. 0–4 years ago (including also
Characteristics of the attendees in self-sampling
There was slight variation in the crude participation rates to self-sampling within different age groups, but no trend with regard to decreasing or increasing age and differences were not significant in the adjusted model (Table 3). Participation by self-sampling was lower among women with a mother tongue other than Finnish or Swedish. A higher education level resulted in significantly higher participation by selfsampling, the difference being almost twofold between the lowest and higher levels of education. Differing from the results from routine screening, participation by self-sampling was highest in the semi-urban municipalities. There were no differences in participation with regard to geographical location or marital status. Self-sampling test results
Of the 920 returned self-taken samples, 109 (11.8%) were hrHPV-positive by HC2 (Table 4). 52 women were invited C 2014 UICC Int. J. Cancer: 136, E677–E684 (2015) V
Table 2. Effect of self-sampling on screening coverage Time from previous Pap-smear
n
% of participants
