The burden and costs of prevention and management of genital disease caused by HPV in women: A population-based registry study in Finland

Authors

  • Heini Salo,

    Corresponding author
    • Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), Helsinki, Finland
    Search for more papers by this author
  • Tuija Leino,

    1. Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), Helsinki, Finland
    Search for more papers by this author
  • Terhi Kilpi,

    1. Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), Helsinki, Finland
    Search for more papers by this author
  • Kari Auranen,

    1. Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), Helsinki, Finland
    Search for more papers by this author
  • Petri Tiihonen,

    1. Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), Helsinki, Finland
    Search for more papers by this author
  • Matti Lehtinen,

    1. Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), Helsinki, Finland
    Search for more papers by this author
  • Simopekka Vänskä,

    1. Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), Helsinki, Finland
    Search for more papers by this author
  • Miika Linna,

    1. Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), Helsinki, Finland
    Search for more papers by this author
  • Pekka Nieminen

    1. Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), Helsinki, Finland
    Search for more papers by this author

  • Conflict of interest: Heini Salo, Tuija Leino, Terhi Kilpi, Pekka Nieminen, were members of the National Expert Group on HPV Related Disease Prevention. Terhi Kilpi is principal investigator of a nationwide effectiveness study of the ten-valent pneumococcal conjugate vaccine mainly funded by GlaxoSmithKline. Matti Lehtinen has received grants for HPV vaccination studies through his employer University of Tampere, Finland from Merck&Co. Inc and GSK Biologicals. Pekka Nieminen has been a medical consultant for GlaxoSmithKline Ltd. and Sanofi-Pasteur MSD Ltd.

Correspondence to: Heini Salo, Department of Vaccination and Immune Protection/Vaccination Programme Unit, National Institute for Health and Welfare (THL), P.O. Box 30, FI-00271 Helsinki, Finland, E-mail: heini.salo@thl.fi

Abstract

The aim of this study was to evaluate the total burden and health care provider costs of prevention, management and treatment of HP-related genital disease outcomes including all organized and opportunistic screening tests. Information about HPV-related disease outcomes in the Finnish female population of 2.7 million was obtained from nationwide population-based registry data. We estimated the incidence, health care resource use, health provider costs and life years lost due to cervical, vaginal and vulvar cancer and intraepithelial neoplasia (CIN, VaIN, VIN), cervical adenocarcinoma in situ, and external genital warts. The average annual disease burden of HPV-related genital disease in the female population of Finland comprises altogether 241 cases of cervical, vaginal and vulvar cancer, 2,898 new cases of CIN, 34,432 cases of minor cytological abnormalities, and almost 4,000 cases of external genital warts. The total annual costs of screening, further diagnostics and treatment of HPV-related genital disease were €44.7 million of which the annual costs due to cervical cancer screening were €22.4 million and due to diagnostics, management and treatment of HPV-related genital disease outcomes were €22.3 million. The latter included €8.4 million due to minor cervical abnormalities detected by the current cervical screening practice. The extensive opportunistic Pap testing fails to keep the incidence of cervical cancer from increasing among women aged 30–34. In addition opportunistic screening among this and younger age group detects a significant number of cytological abnormalities, most of which are probably treated unnecessarily.

Human papillomavirus (HPV) is the most common sexually transmitted infectious agent, causing both subclinical and overt anogenital and oropharyngeal diseases. Up to 30–35% of young adults are HPV DNA positive, and 70–80% of both women and men acquire infection with at least one of the approximately 40 genital HPV types during the sexually active life.[1, 2] HPV infection is necessary but not sufficient for the development of cervical cancer and HPV DNA can be detected practically always in cervical cancer tissue.[3, 4] The so-called low risk HPV types 6 and 11 cause 90% of external genital warts whereas the high-risk HPV types (most notably HPV16, 18, 31, 33, 35, 45 and 52) cause also other anogenital and oropharyngeal neoplasia.[5, 6] Furthermore, HPV infection is associated with approximately 65% of vaginal and 40% of vulvar cancers.[7] Worldwide, cervical cancer is the third most common female cancer, with estimated 530,000 cases and 275,000 deaths in 2008.[8]

Prevention of HPV infection by using condoms is not particularly efficacious, partly because the virus infects and can be transmitted both from skin and mucosal surfaces.[9, 10] By contrast, primary prevention by prophylactic vaccines licenced in 2006 and 2007 has been proven to be efficacious against both the infection and the immediate precursors to cervical cancer.[11, 12]

Secondary prevention of cervical cancer is based on detection and treatment of cervical intraepithelial neoplasias (CIN) before HPV infection has progressed into invasive disease. In Finland, an organized cytology-based screening programme for cervical cancer was launched regionally in 1963 and became nationwide in the early 1970s.[13] Until the 1990s the age-adjusted incidence and mortality rates of cervical cancer decreased by 80%.[14] However, among the 20–39 year old women the incidence rate of cervical cancer has increased steadily during the last 15 years.[15]

Data of the total HPV disease burden and the associated costs are needed for evaluation of the effectiveness and cost-effectiveness of preventive interventions, both vaccination and screening, targeting HPV infection and disease. Both the epidemiology of cervical cancer and the performance of its organized screening are well-known in Finland. The total incidence of the less severe HPV disease manifestations, the health care utilization due to HPV diseases and the volume of opportunistic Pap testing have, however, remained unknown.

We estimated the total burden and health care provider costs of HPV-related genital disease, the organized cervical cancer screening and all opportunistic Pap testing in Finland. We utilized population-based registry data to estimate the incidence of HPV-related outcomes and the related health care resource use, health provider costs and life years lost. The estimates produced in this study will be used for evaluation of cost-effectiveness of cervical cancer screening and HPV vaccination programme in Finland.

Material and Methods

Study population

In 2008, the total population of Finland was 5.3 million, 2.7 million of which were women. The cervical screening and HPV-related disease burden were evaluated in all women aged 15 years or older and living in Finland in 1999–2008. Age- and gender-specific population sizes were obtained from the Finnish Population Information System, a computerized national database, which registers individual-based data including name, gender, personal identity code, address, date of birth, and death for all Finnish residents. The personal identity code remains unchanged throughout the lifetime and is used for patient identification in all health care registers in Finland.

Data sources used for cervical screening and HPV disease burden characterization

We established a nationwide dataset comprising all women who had at least one HPV-related outcome or Pap test recorded in any of the following registers during the specified periods: Finnish Cancer Registry (1990–2008), Mass Screening Registry (1999–2008), Hospital Discharge Register (1999–2008), Social Insurance Institution (SII) Register of Special Reimbursements for Medical Expenses (1999–2008) and Prescribed Medicines (2004–2008), Register of the Finnish Student Health Service (1999–2008), and HUSLAB Pathology Laboratory Register (HUSLAB, 2004–2008). All registers are population-based and nationwide, excluding the HUSLAB Register that covers one fifth of the Finnish female population. The dataset covers all health care providers in Finland.

The clinical HPV-related outcomes included in the analysis were cervical, vaginal or vulvar cancer and intraepithelial neoplasia (CIN, VaIN, VIN), cervical adenocarcinoma in situ (AIS), minor cytological abnormalities and external genital warts. Notifications of the cancer outcomes, AIS and CIN3 were obtained from the Finnish Cancer Registry. Hospitalizations and hospital outpatient visits due to any HPV outcomes were extracted from the Hospital Discharge Register with the International Classification of Diseases 10th revision (ICD-10) codes (Table 1, HPV-related outcome) and with the Finnish version of the Nordic Classification of Surgical Procedures (NCSP) codes for diagnostic and treatment procedures related to cervical, vaginal or vulvar cancer, AIS or intraepithelial neoplasia. Diagnostic and treatment procedures reimbursed in private health care services were extracted from the Register of Special Reimbursements, based on the same NCSP codes. In addition to the diagnoses of HPV outcomes, the final dataset included hospitalizations and outpatient visits with certain additional ICD-10 codes associated with the treatment of HPV outcomes (Table 1) and the cancer chemotherapy NCSP codes.

Table 1. ICD-10 codes for the selection and identification of the cases and health care resource use
 ICD-10 codes included
HPV-related disease outcome codes as first-listed diagnosis 
Cervical, vaginal and vulvar cancerC53, C52, C51
Carcinoma in situ of endocervix (AIS)D06.0
Cervical intraepithelial neoplasia (CIN1–3)N87, D06.1, D06.7, D06.9
Vaginal intraepithelial neoplasia (VaIN1–3)N89.0, N89.1, N89.2, N89.3, D07.2
Vulvar intraepithelial neoplasia (VIN1–3)N90.0, N90.1, N90.2, N90.3, D07.1
Abnormal cytological/histological findings in specimens from female genital organsR87.6, R87.7
Abnormal cytological finding on antenatal screening of motherO28.2
Anogenital wartsA63.0
Extra ICD-10 codes accepted to the final dataset of women with HPV related female genital disease outcome
Extra ICD-10 codes as first-listed diagnosis
Papillomavirus as the cause of diseases classified elsewhereB97.7
Malignant neoplasm of overlapping sites of female genital organsC57.8
Encounter for screening for malignant neoplasm of cervixZ12.4
Extra ICD-10 codes as first-listed diagnosis when the additional diagnosis was C53, C52, or C51
Secondary malignant neoplasmC77–79
Disorders in malignant neoplastic diseaseD63.0, M90.7
Post-radiation disordersK52.0, N30.4
Other noninfective disorders of lymphatic vessels and lymph nodesI89
Localized oedemaR60.0
Encounter for follow-up examination after completed treatment for malignant neoplasmZ08
Encounter for antineoplastic radiation therapy, chemotherapy, immunotherapyZ51.0, Z51.1, Z51.2
Encounter for palliative careZ51.5
Convalescence following radiotherapy or chemotherapyZ54.1, Z54.2
Extra ICD-10 codes as first-listed diagnosis when the additional diagnosis was C53, C52, C51, CIN, AIS, VaIN or VIN
Encounter for gynecological examinationZ01.4
Observation for suspected malignant neoplasmZ03.1

All Pap tests from both the organized screening and opportunistic testing were included in the dataset. The Pap tests from the organized screening are registered in the Mass Screening Registry, which includes information on the results of smears and diagnostic procedures and treatment of women following abnormal smears. The data on opportunistic Pap testing were collected from several sources: Hospital Discharge Register (secondary health care), SII register of Special Reimbursements (private health care services), Student Health Service Register (health care services for undergraduate university students), HUSLAB Register (smears from public primary health care of the Helsinki metropolitan area).

The incidence of external genital warts was estimated from the reimbursed prescriptions of podophyllotoxin or imiquimod in the Register of Prescribed Medicines. Superficial basal cell carcinoma and actinic keratosis are also treated with imiquimod. Therefore, the prescriptions for imiquimod for external genital warts were estimated by assuming a similar age-specific pattern of use to that of podophyllotoxin.

Screening and follow-up Pap tests

Pap tests are performed either for screening purposes or as a follow-up of previously detected cervical abnormalities. In our data, Pap test results were available from the organized screening and the HUSLAB Register data. Furthermore, the purpose of the test (screening vs. follow-up) was only available from the organized screening programme. Therefore, opportunistic Pap tests were classified as screening or follow-up tests according to the intervals between the consecutive tests of the case. Any opportunistic test taken within 30 months after a positive organized screening test was defined as a follow-up test. Similarly, an opportunistic test that was taken within 15 months after a positive test result or within 10 months after a test with unknown test result was regarded as a follow-up test. The same applied to a test obtained within 30 months after the first listed diagnosis code or treatment procedure code indicating intraepithelial neoplasia, or at any time after the first cervical, vaginal or vulvar cancer notification. In the student health service a Pap test was regarded as a follow-up test if it was taken within 15 months after a test with unknown test result.

Cases and incidence of HPV-related genital disease outcomes

For each HPV-related case (i.e., woman) the diagnosis indicating the most serious HPV outcome determined the outcome of the case in the following priority order: cancer, AIS, CIN3, VaIN3, VIN3, CIN2, VaIN2, VIN2, CIN1, VaIN1, VIN1, minor cytological abnormalities and external genital warts. Diagnoses indicating two dysplasias of the same grade at different sites (e.g., CIN2 and VaIN2), were categorized according to the cervical diagnosis.

All cases of minor cytological abnormalities had follow-up Pap testing without a diagnosis of CIN at any time during the follow-up. Some of these cases underwent a histopathological examination (biopsy) or even a procedure to remove a cervical lesion without having any other diagnosis than abnormal cytopathological or histopathological finding (R87.6, R87.7). About half of minor cytological abnormalities can be assumed to be due to persistent HPV infection producing lesions milder than CIN1 and the other half due to immature squamous metaplasia, that is, false-positive findings because of a cytology-based cervical screening performed to prevent HPV-related cancer.

For each HPV-related case, an episode of care was constructed by linking the health related events of a woman (outpatient visits, inpatient hospitalizations, gynaecological procedures by private providers, follow-up Pap tests and Cancer Registry notifications) retrieved from different registers using the unique personal identity code (Fig. 1). Each case manifests itself in exactly one episode of care, which includes the whole HPV-related disease history of an individual during the study period and which is categorized according to the severest detected outcome as described above.

Figure 1.

An example of events within three CIN3 episodes of care. The index event is the first outpatient visit, inpatient hospitalization, or Cancer Register notification with a diagnosis corresponding to the final classification of the episode of care (here CIN3). For Case 1 index event occurred in 2000 (i.e., in the beginning of the established dataset), for Case 2 in 2004 (i.e., in the middle of the established dataset), for Case 3 in 2007 (i.e., in the end of the established dataset). For each episode, the red line shows the follow-up time divided into 1-year periods with the time origin at the index event. The follow-up begins at the start of the first whole prediagnostic year and stops at the end of the last whole postdiagnostic year. The data covered years 1999–2008. NB: Pap tests taken solely for screening purposes (organized and opportunistic) are excluded from the episodes of care. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

The index event in the episode of care was the time point when a case was for the first time assigned the diagnosis that determined its outcome category. Any of the following could be an index event: an outpatient visit to secondary health care, inpatient hospitalization or a Cancer Registry notification. For each HPV outcome, the population-based incidence rate was calculated by age group from the number of cases in the Cancer Register data 1999–2008 (cervical, vaginal and vulvar cancer) or in the Hospital Discharge Register data 2004–2008 (other outcomes) and total person time.

The incidence rate of minor cytological abnormalities resulting in biopsy or treatment procedure was estimated from the nationwide dataset. Since only regional data were available on the opportunistic primary health care Pap testing, the national incidence of cases with mere follow-up Pap testing was extrapolated from the Helsinki metropolitan area data.

Diagnosis codes in the Hospital Discharge Register and prescriptions of pharmacotherapy usually used for external genital warts in the register of prescribed medicines were used to construct the episodes of care for genital wart cases. The secondary health care cases had been hospitalized or had visited an outpatient clinic with the first listed diagnosis code A63.0. The primary health care cases had received reimbursement for medicines generally prescribed for external genital warts.

Health care service use per outcome

Each episode of care was divided into 1-year pre- and postdiagnostic periods. Time before the index event (prediagnostic years) consisted of the diagnostic tests and procedures, excluding screening, that occurred before the ultimate diagnosis. Time from the index event onwards (postdiagnostic years) included the index event itself, the acute phase treatment and the follow-up of the HPV outcome. Only the period from 1999 to 2008 was considered when the health-care service use for whole pre- and postdiagnostic years was estimated (Fig. 1). Thus, if a woman had a CIN diagnosed in 1999 the prediagnostic years were left-censored from the dataset and her episode of care produced data only for the postdiagnostic years. Respectively, if a CIN was diagnosed in 2008 the postdiagnostic years were right-censored from the dataset. We calculated the average number of events (outpatient visits, inpatient hospitalizations, gynaecological procedures by private providers, follow-up Pap tests) per a prediagnostic or postdiagnostic year by dividing the total number of each event type by the total number of episodes of care in follow-up in that diagnostic year of interest (Fig. 1). The dataset from which the health care service use was estimated included all cases with index events in 1999–2008 and additionally those cancer, AIS, and CIN3 cases that had their index event recorded in the Cancer Registry during 1990–1998. To be regarded as a true case of cancer, AIS or CIN3, any such case notified to the Cancer Registry needed to have also a corresponding diagnosis in the Hospital Discharge Register.

Life years lost

The fatality data in the Cancer Registry based on the death notifications were used to estimate the annual mean number of deaths due to cervical, vaginal and vulvar cancer. Also the stage of the disease at diagnosis was taken into account by age group. The expected life years lost per cancer case were estimated from the age- and gender specific life expectancy of the population (Statistics Finland).

Health care costs and economic burden

The costs were evaluated from the health care provider perspective and were presented in 2010 prices. Table 2 summarizes the unit costs. The unit costs for hospitalizations and outpatient visits were estimated based on individual-level cost accounting data from one hospital district. Other unit cost estimates were mainly taken from the widely used national pricelist for unit costs of health care in Finland.[16] All unit costs included overhead costs.

Table 2. Unit costs of care
Pap-testScreening intention (€)Follow-upb (€)
  1. All estimates include overhead costs.

  2. a

    Includes unit price for the test and half the price of a visit to a physician.

  3. b

    Includes unit price for the test and the price of a visit to a physician.

  4. c

    Average unit cost per hospitalization or outpatient visit, estimates include diagnostic and treatment procedures (cost accounting data from Hospital District of Helsinki and Uusimaa).

Organized screening test (mean price for 5 university towns)28.028.0
Private providera76.7114.8
Student health servicea51.286.0
Secondary health care62.7125.4
Primary health carea50.784.9
Secondary health carecHospitalisation (€)Outpatient visit by appointment (€)
Cervical cancer as first listed diagnosis2,849.2194.8
Vaginal cancer as first listed diagnosis3,650.8196.9
Vulvar cancer as first listed diagnosis3,552.7159.4
CIN3/AIS as first listed diagnosis2,566.4244.5
CIN2 as first listed diagnosis2,604.6268.8
CIN1 as first listed diagnosis2,630.3233.2
VaIN3 as first listed diagnosis2,562.9267.1
VaIN2 as first listed diagnosis2,301.3247.8
VaIN1 as first listed diagnosis2,660.8207.8
VIN3 as first listed diagnosis1,693.3252.2
VIN2 as first listed diagnosis1,917.0276.9
VIN1 as first listed diagnosis3,008.0209.0
Genital warts (GW) as first listed diagnosis1,246.682.3
Other unit costsUnit cost (€) 
Outpatient visit by appointment, average125.4 
Primary health care visit68.5 
Private provider visit, gynaecology76.1 
GW treatment, imiquimod cream88.3 
GW treatment, podophyllotoxin solution/cream18.7 
Private provider diagnostic and treatment procedures  
CIN treatment511.4 
VIN treatment340.6 
VaIN treatment340.6 
CIN diagnostic tests450.0 
VIN diagnostic tests279.2 
VaIN diagnostic tests289.5 
VIN and VaIN diagnostic tests286.4 

The average cost per case for a given HPV-related outcome was derived by summing up the average costs of the consecutive prediagnostic and postdiagnostic years (Table 3). The cost per cancer and dysplasia cases was limited to three prediagnostic years and nine postdiagnostic years. The cost per a case of minor cytological abnormalities and external genital warts started at index event and was limited to four and five postdiagnostic years, respectively. Under the assumption of stationarity of the annual incidence of cases, the total annual health care costs due to a given HPV outcome were the product of the average annual number of cases and the average cost per case. For each outcome, we derived the variance of the cost per case (episode of care) from the joint distribution of the numbers of events per case over all pre- and postdiagnostic years (including their covariance between different years), and the distribution of costs per event.

Table 3. Average undiscounted cost per HPV related genital disease case and total annual costs
 Cost (€)/caseSDDiagnosis and treatment costs (€)/year
  1. Abbreviations: AIS, cervical adenocarcinoma in situ; CIN1/CIN2/CIN3, cervical intraepithelial neoplasia grades 1, 2 and 3; VaIN1/VaIN2/VaIN3, vaginal intraepithelial neoplasia grades 1, 2 and 3; VIN1/VIN2/VIN3, vulvar intraepithelial neoplasia grades 1, 2 and 3.

Cervical cancer22,71025,7523,580,983
AIS4,7302,375209,925
CIN32,7352,0282,749,011
CIN22,2901,5832,355,255
CIN11,7531,5991,494,193
Vaginal cancer24,42426,760404,007
VaIN32,6542,95793,759
VaIN22,3701,796115,334
VaIN11,8371,54982,576
Vulvar cancer15,86718,3461,234,915
VIN33,8443,389334,350
VIN22,4721,92266,385
VIN12,0282,88696,167
Minor cytological abnormalities without CIN diagnosis
With only follow-up Pap tests1171193,520,161
With cervical biopsy9855673,824,323
With surgical procedure1,2968751,044,835
With vaginal biopsy61927676,655
With vulvar biopsy574447196,275
Genital warts
Secondary health care386508410,118
Primary health care16575448,560
Total  22,337,787

Results

Annual number of cases, incidence rates and life years lost

In 1999–2008, there was an annual average of 153 cases of cervical cancer in Finland, corresponding to an incidence rate of 5.7 per 100,000 woman-years (Table 4). There were two peaks in the age-specific incidence rate. The first peak occurred in women aged 30–34 (10.1 per 100,000 women-years) and the second one in women aged 70 and older (11.8 per 100,000 women-years).

Table 4. Annual average numbers of HPV related female genital disease cases and corresponding incidence rates per 100,000 woman-years
 15–1920–2425–2930–3435–3940–4950–5960–6970–7980+Total
 NRateNRateNRateNRateNRateNRateNRateNRateNRateNRateNRate
  1. Annual averages in 1999–2008 (cervical, vaginal and vulvar cancer) or in 2004–2008 (all other outcomes).

  2. Abbreviations: AIS, cervical adenocarcinoma in situ; CIN1/CIN2/CIN3, cervical intraepithelial neoplasia grades 1, 2 and 3; VaIN1/VaIN2/VaIN3, vaginal intraepithelial neoplasia grades 1, 2 and 3; VIN1/VIN2/VIN3, vulvar intraepithelial neoplasia grades 1, 2 and 3

Cervical cancer0010.874.61610.1179.8287.4236.0176.12510.91913.31535.7
AIS00.331.963.984.863.671.930.810.300.100.1351.3
CIN32716.812074.5200122.9209135.315594.518048.46516.4217.1146.453.399837.1
CIN25735.5188116.5206126.8164106.012978.417246.17218.1258.393.821.5102438.0
CIN16741.8181112.112577.010165.58350.715541.78922.23110.573.310.684131.2
Vaginal cancer00000000.10010.420.520.852.264.0160.6
VaIN310.510.710.710.900.241.071.761.972.942.8321.2
VaIN221.342.632.032.232.172.0112.872.552.010.5471.8
VaIN121.053.231.821.621.072.0102.672.531.410.6441.6
Vulvar cancer0000.100.10010.651.482.0113.92310.32316.2722.7
VIN310.831.721.021.031.9123.2194.9155.2167.2138.1873.2
VIN210.610.921.421.421.241.061.531.121.110.8261.0
VIN121.131.631.642.321.171.892.472.483.531.8471.8
Cases of minor cytological abnormalities (without CIN diagnosis) with
Follow-up Pap tests874548.12811173933712071322420833149191472241940544813652790936.0680302.2188121.1297641209
Cervical biopsy278174.1546338.1414254.4350226.0364221.0731196.7617154.4314105.516372.56038.43868143.6
Surgical procedure4024.87646.88250.68454.58350.420254.316240.55016.7188.032.280029.7
Vaginal biopsy10.952.874.264.185.0184.9194.9248.1209.195.81194.4
Vulvar biopsy64.01811.32012.31610.61911.34913.16616.65819.35323.42214.433012.3
Genital wart cases
Secondary care212132.7341211.4169103.77649.16438.710027.06215.53411.594.210.5107640.0
Primary care755473.5970601.0460282.5161103.910865.617045.89323.33411.4104.521.02774103.0

In 2004–2008, an average of 2,898 new cases of CIN were treated and altogether 34,432 cases of minor cytological abnormalities were diagnosed annually (Table 4). The corresponding incidence rates per 100,000 woman-years were 1.3 for cervical AIS, 37.1 for CIN3, 38.0 for CIN2 and 31.2 for CIN1. The age-specific incidence rates for both AIS and CIN3 peaked in women aged 30–34 (4.8 and 135.3 per 100,000 women-years, respectively). The age-specific incidence rate for CIN2 peaked in women aged 25–29 (126.8 per 100,000 women-years) and for CIN1 in women aged 20–24 (112.1 per 100,000 women-years). The majority (86%) of cases of minor cytological abnormalities were those with follow-up Pap testing without any further procedures or treatment. The average annual age-specific incidence rate of these cases increased sharply already in women aged 20–24 (1,739 per 100,000 women), remained high in women aged 25–49 (1,987 per 100 000 women), and finally decreased in women aged 50 and older.

There were 16 vaginal and 72 vulvar cancer cases on average each year, corresponding to incidence rates of 0.6 and 2.7 (per 100,000 women-years), respectively (Table 4). For both cancer types, the age-specific incidence started to increase after 50 years of age. There were almost 4,000 women consulting health care due to external genital warts annually. Approximately 70% of these were treated in primary health care. The age-specific incidence peaked in women aged 20–24 both in cases treated in primary health care (601.0 per 100,000 women) and secondary health care (211.4 per 100,000 women).

There were on average 56, 12 and 30 deaths due to cervical, vaginal and vulvar cancer per year, with the estimated life years lost as 993, 127 and 346, respectively. Women aged 70 and older composed 60% of all deaths due to cervical cancer and accounted for 30% of all life years lost.

Utilization of health care services and cost per case

In all HPV outcomes the majority of the health care resource use was outpatient care and occurred within the first year after the diagnosis (index event). In this first postdiagnostic year, there were on average 13.6 outpatient visits per case for cervical cancer, 2.0–2.7 for AIS and CIN, 1.2–2.4 for minor cytological abnormalities, and 2.1 for external genital warts. Inpatient care was rare in all other HPV outcomes but cancer. Most hospitalizations occurred in the first postdiagnostic year, during which there were on average 4.1 hospital admissions per cervical cancer case. Genital wart cases treated in primary health care had in the first postdiagnostic year on average 1.3 prescriptions of pharmacotherapy.

Both utilization of health care services and costs were spread over several prediagnostic and postdiagnostic years. The estimated cost per case varied with the HPV outcome, prediagnostic and postdiagnostic year, and age (Fig. 2 and Table 3). The average cost per case was highest for cancer cases, €22,710. The corresponding cost for CIN varied from €1,750 to €2,740. The average cost per case was higher for older women in all HPV outcomes. The costs associated with diagnostics (prediagnostic years) accounted for an average 4% of the total costs per cervical cancer case and 23–31% per AIS and CIN cases. The costs associated with treatment (first post diagnostic year) accounted for an average 68% of the total costs per cervical cancer case, 45% per AIS case, and 36–39% per CIN case. The costs associated with the follow-up (second to ninth postdiagnostic years) accounted for an average of 29–35% of the total costs per cervical cancer case, AIS, and CIN. The undiscounted costs of prediagnostic and postdiagnostic years (Fig. 2) added up to the total cost per case (Table 3).

Figure 2.

Average undiscounted cost per case per prediagnostic and postdiagnostic year and by age-group for a given HPV related outcome (AIS, cervical adenocarcinoma in situ; CIN1/CIN2/CIN3, cervical intraepithelial neoplasia grades 1, 2 and 3). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Costs and frequency of Pap testing

The average annual number of Pap tests obtained for screening purposes was 446,000 during 2004–2008. Of these tests 40% were organized and 60% opportunistic screening. Furthermore, 18% of opportunistic screening tests were taken from women under the age of 25. In fact, 65% of women aged 20–24 had an opportunistic Pap test taken at least once over the 5-year period. The total cost of all Pap tests obtained for screening purposes was €22.4 million, of which only €5.0 million incurred in organized screening. In addition to screening Pap tests, there were 84,000 follow-up Pap tests, the costs of which are included in the episodes of care.

Annual health care costs due to diagnostics and treatment of HPV outcomes

The estimated annual costs of female HPV-related genital disease to health care was €22.3 million, of which cervical cancer, AIS, and CIN accounted for €10.4 million, minor cytological abnormalities of cervix €8.4 million, and external genital warts €0.9 million (Table 3). The costs of cervical cancer in women aged less than 25 years, aged 25–64 and aged 65 and older accounted for 1, 66 and 33% of the total costs to health care, respectively. These three age groups accounted for 19, 75 and 6% of the costs of AIS, CIN and minor cytological abnormalities of cervix, respectively. With the €22.3 million spent on diagnostics and management of HPV-related genital disease in women and the €22.4 million spent on screening Pap tests, the total annual costs of the whole HPV-related disease burden add up to €44.7 million, that is, €19.90 per a female citizen aged 15 and older.

Discussion

The average annual disease burden of HPV-related genital disease in the 2.7 million female population of Finland comprises altogether 241 cases of cervical, vaginal or vulvar cancer, almost 3,000 new cases of CIN, 34,432 cases of minor cytological abnormalities, and almost 4,000 cases of external genital warts. The first peak in the incidence of both cervical cancer and the most severe grade of CIN was seen in the relatively young women aged 30–34 years. The total annual costs of HPV-related genital disease were €44.7 million, of which the annual costs of cervical cancer screening were €22.4 million and the costs of further diagnostics, management and treatment of HPV-related genital disease outcomes were €22.3 million. The management of HPV-related cervical diseases cost €18.8 million, of which €3.6 million were due to cervical cancer and €6.8 million to AIS and CIN, and €8.4 million due to minor cytological abnormalities. The vast majority of the cervical disease findings are initially detected in organized and opportunistic cervical cancer screening. Of all Pap tests obtained for screening purposes, opportunistic screening accounted for 60% of the 446,000 annual tests and 78% of the €22.4 million total costs.

An obvious strength of our study is the reliable nationwide, individually linkable register data. The accuracy of the diagnoses and the registration coverage of outpatient care have proven to be good in the Hospital Discharge Register.[17] The data included Finnish Cancer Registry notifications, Mass Screening Registry data, and all out- and inpatient treatment provided in Finnish hospitals for HPV-related genital disease (Hospital Discharge Register). The data were complemented with social insurance data from private health care providers, prescribed medicines and opportunistic Pap testing. By linking all these data items through the unique person identifier we managed to characterise an episode of care with screening, diagnostic procedures, treatment, and the follow-up for each individual and thereby the total health care resource utilization in the whole population.

The estimate of the number of opportunistic Pap tests covered all data sources and the whole country. With the exception of the number of public primary health care Pap tests that were extrapolated from the data on Helsinki metropolitan area, representing one-fifth of the female population, all other data sources were nationwide. The relatively large standard deviation of the cost per episode of care reveals the heterogeneity across cases in the health care resource use and in unit costs in secondary health care. However, our results rather underestimate than overestimate health care resource utilization in Finland. For instance, the data on private providers include only the reimbursed Pap tests and surgical procedures.

The incidence of external genital warts was 143 per 100,000 women-years (all ages) in this study, which is practically the same as in a recent study from England (142 per 100,000 women-years).[18] However, the annual average cost of external genital warts per female citizen aged 15 and older was almost one-third higher in England. The difference is due to the higher estimate of recurrence of the disease in England. The disease burden of vulvar and vaginal cancer was relatively low. However, together with HPV-associated anal and oropharyngeal cancers these, to a large extent HPV vaccine preventable cancers and their precursors[19-22] will further increase the potential impact of HPV vaccination.

The results of this study reflect the health care system in Finland, a country which has been successful in reducing the incidence and mortality due to cervical cancer. Cervical cancer is now the 19th most common cancer in women in Finland, although it is the third most common cancer worldwide. The number of detected cases of cervical cancer or CIN depend largely on how screening is implemented. Notably, the detected disease burden depends not only on the organized programme but also on opportunistic screening practice. The incidence rate of CIN and minor cytological abnormalities in our study was consistent with the coverage of Pap testing.

There was a significant disease burden among young women. The cervical cancer incidence peaked in the 30–34 age-group. Despite the current screening programme and the extensive opportunistic screening, the cancer incidence is not at an acceptable level in this age group. This finding is supported by a recent study on the effectiveness of organized screening in Finland showing that screening at ages below 40 and especially below 30 was associated with a clearly smaller reduction in cancer risk than screening at older ages.[23]

This is the first time the high coverage of opportunistic screening in young women not yet eligible to organized screening has been quantified in Finland. The voluminous opportunistic screening among women aged less than 25 years produced 20% of the total costs due to CIN1, CIN2 and minor cytological abnormalities. In particular, younger women, at the age when the HPV infection is known to be most prevalent and mostly transient,[24-26] were intensively tested outside the organized programme. Too early HPV detection poses a risk of an excess number of diagnosed lesions, likely to regress without treatment.[24-26] Active management and treatment of CIN findings among younger women causes unnecessary costs and, even more importantly, increases the risk of reproductive health problems, such as preterm delivery.[27] Reducing opportunistic screening, especially among women aged less than 25, would likely reduce the over management and treatment of reversible lesions and thereby bring considerable savings without significant health impairment. However, it is also reasonable to assume that some of cytological abnormalities detected through the extensive opportunistic screening among young women would have developed into cancer unless treated at an earlier stage. The interaction between opportunistic and organized screening calls for further research.

One-third of the costs of cervical cancer occurred in women aged 65 and older. In this age group, not currently targeted by the national screening programme, the coverage of opportunistic Pap testing started also to decrease rapidly. This is reflected as a second peak in the age-specific incidence of cervical cancer in women aged 70 and older. The considerable burden of cervical cancer among women aged 65 and older raises the question whether it is appropriate to stop organized screening at 60. Considering a continuation of organized screening beyond this age gains further support from a recent Finnish study in which the risk reduction due to screening seemed to continue in those aged 65 and older.[23]

Vaccine efficacy estimates against all cervical intraepithelial neoplasia grade 3 or greater (CIN3+) have varied from 43 to 93%.[19, 28] A fair part of the disease burden, revealed in this study, could thus be prevented by HPV vaccinations. However, the burden of disease could be reduced sooner by redirecting resources within the screening activity from the arbitrary and costly opportunistic testing to a more comprehensive and more effectively targeted organized screening. The current study demonstrates that the extensive opportunistic Pap testing in Finland has failed to keep the incidence of cervical cancer from increasing among women aged 30–34. In addition, opportunistic screening among this and younger age group keeps detecting a considerable number of cytological abnormalities that are subsequently treated even though they would likely regress without any treatment.

In addition to the appropriate information about the carefully estimated disease burden presented here, policy making requires proper analyses on the effectiveness and cost-effectiveness of the HPV vaccination programme and cervical cancer screening as well as the optimal combination of these. This necessitates modelling of both the transmission of the underlying HPV infection and progression of the disease. The estimates produced in this study are vital for the models that are used for the evidence-based decision making for prevention of HPV disease.

Acknowledgement

The authors thank the National Expert Group on Papillomavirus Disease Prevention as a steering group of the study.

Ancillary