The impact of Germany's human papillomavirus immunization program on HPV-related anogenital diseases: a retrospective analysis of claims data from statutory health insurances.

The German HPV immunization program has led to significant declines in female anogenital disease among young women in Germany, highlighting the importance of the vaccination. Moreover, the data suggest that increasing vaccination coverage in Germany could further strengthen the public health impact of its HPV immunization program.

We conducted a retrospective analysis of claims data from the “Institut für angewandte Gesundheitsforschung Berlin GmbH” (InGef) research database, which had data available from a nine year period (January 2013 to December 2021). The InGef Research Database comprises anonymized claims data of about 4 million individuals (4.8% of the German population and 5.5% of the German SHI population) and includes the majority of SHI in Germany. The database represents the German population in terms of age and gender according to the Federal Office of Statistics database. The claims data of the participating SHIs are collected in a specialized trust center, anonymized, and transferred to InGef. The analysis of German SHI claims data are permitted by social law and no review by an independent ethics committee was required to conduct this study. To assess the impact of the HPV-immunization program, we compared prevalences of HPV-related anogenital diseases among women born in 1980, who were not eligible for vaccination under the vaccination recommendation from STIKO (pre-VR cohort) to women born in 1990–1991, who were eligible for the STIKO recommended vaccination (post-VR cohort). For the time span available in the InGef database, the youngest age range available for comparison was 32–33 years for women in the pre-VR cohort and 28–30 years for women in the post-VR cohort. Because HPV-related disease prevalence changes with age, we also developed a peri-VR cohort of women born in 1983–84, who were only eligible for catch-up vaccination, for a sensitivity analysis. The peri-VR cohort had data available in the InGef database at ages 28–33. Figure  1 illustrates the overall study design along with the timeline for the cohorts to pass through the significant dates of the HPV-immunization program. Fig. 1 Study design with observation periods and vaccination program related timelines. Pre-VR cohort is shown in green and post-VR cohort in blue. The peri-VR cohort used for sensitivity analyses, is shown in pink, ages for vaccination per the STIKO recommendation (12–17 years) are shown in a solid darker bar. Ages for catch-up vaccination (18–26 years) are show in lighter bars. Periods for which cohorts are available in the Ingef database and 28–30 years old are shown in yellow. Periods for which cohorts are available in the InGef database and 32–33 years old are shown in orange Study design with observation periods and vaccination program related timelines. Pre-VR cohort is shown in green and post-VR cohort in blue. The peri-VR cohort used for sensitivity analyses, is shown in pink, ages for vaccination per the STIKO recommendation (12–17 years) are shown in a solid darker bar. Ages for catch-up vaccination (18–26 years) are show in lighter bars. Periods for which cohorts are available in the Ingef database and 28–30 years old are shown in yellow. Periods for which cohorts are available in the InGef database and 32–33 years old are shown in orange For the construction of the VR cohorts, women in the InGef research database born in the relevant birth years who were continuously observable in the SHI in the respective observation period were identified. Observation periods are defined as the time when the individuals had the relevant age for comparison, i.e., 32–33 years of age for the pre-VR cohort and 28–30 years of age for the post-VR cohort. Age was determined by the quarter of birth in the year of birth. Women switching the SHI in the respective observation period were excluded. However, women who died in the respective observation period were maintained in the cohort. HPV-related anogenital diseases assessed included CIN2+, anogenital warts and vulvar, vaginal, and anal precancer/cancers. Women were considered as diagnosed with the respective disease if they had at least one record of the corresponding ICD-10-GM diagnosis (see Table 1 ) in the outpatient setting (verified diagnosis) or in the inpatient setting (primary or secondary discharge diagnosis) during their individual observation period. Since the detection of CIN2+ cases within the population depends on the frequency of pap-smear tests and their frequency of use may vary over time, cervical cancer screening test data were also determined. Cervical cancer screening tests were identified using uniform assessment standard (Einheitlicher Bewertungsmaßstab (EBM)) codes 01730, 01733, 01760, 01761, and 01762. Table 1 ICD-10-GM codes used for identification of HPV-related anogenital diseases ICD-10-GM codes Cervical precancer/cancer (CIN2+) N87.1, N87.2, C53.-, D06.- Anogenital warts A63.0 Vaginal precancer/cancer N89.0, N89.1, N89.2, D07.2 Vulvar precancer/cancer N90.0, N90.1, N90.2, N90.3, D07.1 Anal precancer/cancer D01.3, C21.- ICD-10-GM codes used for identification of HPV-related anogenital diseases To demonstrate the population-level impact of HPV vaccination on HPV-related anogenital diseases among women in Germany, prevalence was calculated separately for CIN2+, anogenital warts, and vaginal, vulvar, and anal precancer/cancer and for each of the VR cohorts. Only women who underwent at least one cervical cancer screening test during their individual observation period were considered for analysis of CIN2+ prevalence. Prevalence of CIN2+ was calculated as the percentage of women within a respective VR cohort with a relevant ICD-10-GM diagnosis during their individual observation period who received at least one cervical cancer screening compared to all women in the respective cohort who received cervical cancer screening. Prevalence of anogenital warts, vaginal precancer/cancer, vulvar precancer/cancer, and anal precancer/cancer were calculated separately as the percentage of women within a respective VR cohort with a relevant ICD-10-GM diagnosis during their individual observation period compared to all women in the respective cohort. Due to structure of the data source, with data only available from 2013 to 2021, persons in the different cohorts had differing lengths of observation within the study period. To account for this limitation, prevalence rates for each of the anogenital diseases were annualized. The annualized prevalence rates were calculated by relating the number of women affected to the average observation period within the respective cohort. The number ( n ) and percentage (%) of women with at least one diagnosis for respective HPV-related diseases as well as 95%-Clopper–Pearson confidence intervals are presented below. Odds ratios were calculated to quantify the difference in prevalence pre-VR compared to post-VR and Pearson’s χ 2 test was used to assess the significance of differences between pre-VR and post-VR cohorts.

Demographic characteristics of the study population are presented in Table  2 . The number of individuals living in eastern German rural areas is halved from pre-VR to post-VR, while at the number of women in family insurance in pre-VR is more than twice that of the post-VR. In the pre-VR cohort, 22,533 women aged 32 to 33 years were identified. Of these, 16,823 (74.7%) had at least one pap-smear test during the relevant period. In the post-VR cohort, 38,987 women aged 28 to 30 years were identified. Of these, 33,510 (86.0%) had at least one pap-smear test. Table 2 Characteristics of study population by cohort Pre-VR cohort Post-VR cohort N % N % 22,533 100 38,987 100 Birth year  1980 22,533 100.0 – –  1990 – – 18,969 48.7  1991 – – 20,009 51.3 Place of residence  East-rural 1935 8.6 1594 4.1  East-urban 2371 10.5 3756 9.6  West-rural 4541 20.2 7892 20.2  West-urban 13,412 59.5 24,890 63.9 Insurance status  Regular 20,012 88.8 37,194 95.4  Family 2185 9.7 1700 4.4  Retired 334 1.5 84 0.2  Unknown 2  < 0.1 0 0.0  Cervical cancer screening 16,823 74.7 33,510 86.0 Characteristics of study population by cohort Prevalence for CIN2+ showed a decline of 51.1% from the pre-VR cohort (0.92%, 95%-CI[0.78%, 1.08%]) to the post-VR cohort (0.45%, 95%-CI[0.38%,0.53%]). This difference between the pre-VR and post-VR cohort was significant (OR = 0.49, χ 2  = 40.30, p  < 0.001). For the individual ICD-10-GM codes, a similar pattern as for the overall CIN2+ diagnosis was observed. The post-VR cohort showed lower prevalence rates for moderate cervical dysplasia (CIN II, N87.1), severe cervical dysplasia (N87.2), malignant neoplasm of cervix uteri (C53) and carcinoma in situ of cervix uteri (CIN III, D06). Only for N87.1 no significant differences between the VR cohorts were observed. The other three relevant ICD-10-GM codes showed significant differences in the comparison between pre-VR and post-VR cohort (N87.2: OR = 0.401, χ 2  = 28.81, p  < 0.001; C53: OR = 0.30, χ 2  = 9.31, p  = 0.002; D06: OR = 0.44, χ 2  = 10.31, p  = 0.001). Table 3 and Fig.  2 A provide a comprehensive overview and visualization of the CIN2+ prevalence results. Table 3 Annualized prevalence of cervical precancer/cancer (CIN2+) in women with cervical cancer screening and annualized prevalence of anogenital warts, vaginal precancer/cancer, and vulvar precancer/cancer Pre-VR cohort Post-VR cohort At the age of 32–33 years At the age of 28–30 years Count % 95% CI Count % 95% CI Prevalence of cervical precancer/cancer (CIN2+) in women with cervical cancer screening n  = 16,823 n  = 33,510  ≥ 1 CIN2+ diagnosis 155 0.92 0.78–1.08 151 0.45 0.38–0.53 Moderate cervical dysplasia (CIN II) 30 0.18 0.12–0.25 51 0.15 0.11–0.20 Severe cervical dysplasia 75 0.44 0.35–0.56 60 0.18 0.14–0.23 Malignant neoplasm of cervix uteri 17 0.10 0.06–0.16 10 0.03 0.01–0.05 Carcinoma in situ of cervix uteri (CIN III) 34 0.20 0.14–0.28 30 0.09 0.06–0.13 Prevalence of anogenital warts, vulvar precancer/cancer, and vaginal precancer/cancer n  = 22,533 n  = 38,987  ≥ One diagnosis of anogenital warts 100 0.44 0.36–0.54 104 0.27 0.22–0.32  ≥ One diagnosis of vaginal precancer/cancer 8 0.04 0.02–0.07 3 0.01 0.00–0.02  ≥ One diagnosis of vulvar precancer/cancer 9 0.04 0.02–0.08 9 0.02 0.01–0.04 Fig. 2 Annualized prevalence of CIN2 +, anogenital warts, vaginal precancer/cancer, and vulvar precancer/cancer: Prevalence estimates, associated 95% CIs and p-values for relevant group comparisons are presented for the different anogenital diseases Annualized prevalence of cervical precancer/cancer (CIN2+) in women with cervical cancer screening and annualized prevalence of anogenital warts, vaginal precancer/cancer, and vulvar precancer/cancer Annualized prevalence of CIN2 +, anogenital warts, vaginal precancer/cancer, and vulvar precancer/cancer: Prevalence estimates, associated 95% CIs and p-values for relevant group comparisons are presented for the different anogenital diseases Comparisons of prevalences for anogenital warts (OR = 0.60, χ 2  = 13.00, p  < 0.001) and vaginal precancer/cancer (OR = 0.22, χ 2  = 4.72, p  = 0.030) between the pre-VR and post-VR cohort were significant. From pre-VR to post-VR, this reflects a significant prevalence reduction of 38.64% and 75.00%, respectively. Only a very small number of anal precancer/cancer cases were found for both cohorts (< 5). Presentation of the results and evaluation with inferential statistical methods was not possible due to data protection regulation. Table 3 and Fig.  2 B–D provide a comprehensive overview and visualization of the prevalence of the included anogenital diseases. To explore whether the age difference at time of observation between the pre-VR and post-VR cohorts influenced findings, additional peri-VR cohorts were identified at 28–30 and 32–33 years of age. No significant differences in prevalence were found for CIN2+, vaginal precancer/cancer, nor vulvar precancer/cancer between the two age-specific peri-VR cohorts. Anogenital warts were more prevalent in the younger age group (OR = 0.79, χ 2  = 4.69, p  = 0.030).

Human papillomavirus (HPV) is the most common sexually transmitted infection [ 1 ]. Although most HPV infections are transient, persistent infection increases the risk of developing anogenital as well as head and neck cancers; virtually all cervical cancers globally are attributable to high-risk HPV types as well as approximately 30, 50, 70, and 90% of vulvar, oropharyngeal, vaginal and anal cancers, respectively [ 2 – 4 ]. Low-risk HPV types cause anogenital warts as well as recurrent respiratory papillomatosis (RRP), both diseases characterized by benign papillomas [ 5 – 10 ]. HPV vaccination for prevention of HPV infection and related diseases has been available in Germany since 2006, when a quadrivalent HPV vaccine (4vHPV) targeting high-risk HPV types 16 and 18, responsible for 70% of cervical cancers, and low risk HPV types 6 and 11, responsible for 90% of genital warts, was approved by the European Medicines Agency [ 11 ]. In 2007, the German Standing Commission on Vaccination (STIKO) recommended HPV vaccination of girls in the age group 12 to 17 years [ 12 ] with some health insurance funds additionally offering catch-up vaccinations with full reimbursement up to the age of 26 years. In 2018, STIKO recommended HPV vaccination be extended to boys. Current vaccination recommendations include vaccination for both boys and girls aged 9 to 14 years and catch-up vaccinations for missed immunizations preferably until the age of 17 years [ 13 , 14 ]. Vaccine uptake in Germany has been persistently suboptimal. In 2008, one year after the STIKO vaccine recommendation (VR), only 32.2% of 12–17-year-old females and 12.3% of 18–26-year-old females had received at least one HPV vaccination [ 15 ]. In 2020, 68.5% of 18-year-old girls in Germany had been vaccinated at least once and 54.1% had been fully vaccinated, putting Germany’s vaccination-coverage rate behind that of other countries [ 16 ]. A large body of literature demonstrates the real-world impact of HPV vaccines globally [ 17 , 18 ]. Studies investigating the impact of the implementation of HPV-vaccination programs on cervical diseases show particularly large reductions for high-grade abnormalities such as cervical intraepithelial neoplasia 2+ (CIN2+) and for individuals who received vaccinations at younger ages prior to sexual debut [ 17 ]. For example, a study from Denmark, a country with high vaccination coverage, showed that the rate of CIN2+ decreased by up to 73% in birth cohorts eligible for HPV vaccination as compared to older birth cohorts [ 19 ]. Reductions have also been observed for other anogenital diseases; Depending on the age at vaccination, reductions of up to 93% have been observed for anogenital warts [ 20 ]. In Germany, there are few real-world studies of the impact of HPV vaccination. In 2010, Thöne et al. using statutory health insurances (SHI) data compared populations in 2005, pre-VR, and 2010 (post-VR) and showed a 60% reduction in anogenital warts incidence in 16–20-year-old females [ 21 ]. Similarly, the prevalence of anogenital warts was observed to decreased from 4.7% to 1.7% among 24–27-year-old women between 2010 and 2015 [ 22 ]. In a previous study we conducted based on SHI claims, the prevalence of CIN2+ in 20- to 26-year-old females significantly declined from 0.64% in 2013 to 0.35% in 2018 [ 23 ]. These findings suggested that HPV vaccination may have been influencing CIN2+, but since these time periods were both post-VR, the impact of HPV vaccination could not be directly assessed. The present study assesses the population-level impact of Germany’s HPV-immunization program on HPV-related anogenital diseases in women. Within impact studies, the prevalence of vaccine-preventable diseases before introduction of an NIP is compared with the prevalence of the diseases after the introduction of the program. In this study we compare prevalences of CIN2+, genital warts, as well as vulvar, vaginal, and anal precancer/cancer among women in Germany born in 1980 and so were not eligible for the vaccine (pre-VR) to those born in 1990–1991 who were vaccine eligible (post-VR). We hypothesized that the prevalence of HPV-related anogenital disease is significantly lower in females in the post-VR than pre-VR cohort.

The population-level reductions in prevalence of HPV-related anogenital disease among young women post introduction of HPV vaccination demonstrate the important public-health impact of the German HPV-immunization program. Further, the results, especially when compared with global impact studies, suggest that increasing HPV-vaccination coverage in Germany would strengthen that public-health impact and increase the likelihood of HPV elimination in Germany with time.

The aim of the present study was to investigate the population-level impact of the German HPV-vaccination program on HPV-related anogenital diseases in young women. We found a significant reduction in prevalence for CIN2+, anogenital warts, and vaginal precancer/cancer when comparing the post-VR to pre-VR cohorts. The more than 50% reduction in CIN2+, the most concerning albeit uncommon HPV-related disease for young women from a public-health standpoint, is striking. These findings highlight the important, positive public-health impact of the German HPV-immunization program on HPV-related disease. They align with findings of existing literature on the real-world impact of HPV immunization and extend them by adding a German perspective [ 17 ]. In comparison to the present findings, studies in other countries reported even larger prevalence reductions in CIN2+ and anogenital warts after the introduction of HPV vaccination [ 17 , 18 ]. An impact study from Canada found a 69% decrease in CIN2+ in teenage females [ 24 ], while, as mentioned above, a study from Denmark found a 73% decrease in CIN2+ in the youngest eligible birth cohort [ 19 ]. Anogenital warts were found to decrease by 83% in the UK [ 25 ] and 88% in Australia [ 26 ] after the introduction of national immunization programs. One potential explanation for the lower prevalence reduction in Germany compared to other countries is the lower vaccine coverage rates in Germany. Based on HPV-vaccination coverage estimates from WHO and UNICEF, Germany is one of the lowest performers among high-income countries. Only 48.5% of German females born between 1990 to 1992, the years of the post-VR cohorts in this study, had received full HPV vaccination with three doses by the age of 18 to 20 years [ 27 ]. Countries such as Canada and Australia, for which studies showed greater prevalence reduction in anogenital diseases, have coverage rates of more than 80% [ 28 ]. Thus, while the current results demonstrate success of the vaccine in reducing the prevalence of anogenital diseases, the comparison with reductions in other high-income countries should encourage further efforts to increase vaccination coverage in Germany. Some in the post-VR cohort were already 17 years old when the immunization program was initiated and thus would only have been within the recommended age range for vaccination for less than a year and would have been at the older end of that age range. Previous studies have shown that vaccination coverage is directly correlated with years of eligibility based on the primary recommendation; birth cohorts eligible throughout the recommended age range have the highest vaccination coverage [ 23 , 29 ]. Furthermore, real-world evidence demonstrates HPV immunization is most effective prior to an individual’s sexual debut [ 30 ]. As median age of sexual debut in Germany is 15 years [ 31 ], the impact of HPV vaccination on the prevalence of anogenital disease found in the present study may underestimate the impact among subsequent birth cohorts that had more years of eligibility and were younger at the time of HPV vaccination. The women studied herein were among the youngest age groups regarding onset of HPV-related cancers. While the age-specific rates of new cases of cervical cancer increase significantly from the age of 25–29 and are highest at the age of about 40–44, the number of cancer diagnoses for the other investigated anogenital diseases only increases in the later decades. The young age of the present cohorts may explain why anal precancer/cancer could not be assessed in this study. Future studies comparing pre-VR cohorts to older post-VR cohorts will be needed to assess this outcome. Such studies may also show higher impact of vaccination due to HPV-natural history and the increased burden of HPV-related disease in older women. Despite its strengths, this study has several limitations. Due to the structure of the InGef research database, the pre-VR cohort was slightly older (ages 32–33) than the post-VR cohort (ages 28–30) and post-1991 birth cohorts, which would have had more eligible years for vaccination were not available in the InGef database at the relevant ages. Our sensitivity analyses did not suggest that the age difference between pre- and post_VR cohorts influenced these findings. In addition, claims data are primarily collected for reimbursement purposes, meaning only patients who visited a physician and received a diagnosis for an anogenital disease, thereby triggering reimbursement were recorded. Undiagnosed patients were not included. Similarly, diagnoses were only considered if ICD-10-GM codes were added. Finally, the post-VR cohort years of observation included the start of the COVID-19 pandemic. To assess the potential impact of COVID-19 on these findings, we performed a sensitivity analysis which included the 1989 birth cohort in the Post-VR cohort and censored all 2020 data. This sensitivity analysis did not suggest COVID-19 meaningfully impacted these results. Rates of cervical cancer screening increased by 0.2% to 86.2% in the sensitivity analysis. For CIN2+, the prevalence reduction pre- to post-VR decreased by 2.1% (48.9% instead of 51.1%). For anogenital warts the prevalence reduction pre- to post-VR increased 2.2% (38.6% sensitivity post-VR instead of 36.4%) and for vaginal precancer/cancer the decrease remained at 75.0%.