Impact of Preventive Strategies on HPV-Related Diseases: Ten-Year Data from the Italian Hospital Admission Registry

Simple Summary HPV-related diseases are mainly represented by cancers. Furthermore, real world data with respect to the effects of primary and secondary preventive strategies are lacking. Therefore, the aim of this study is to assess the effectiveness of preventive strategies in accordance with Italian HPV-related hospital admissions. From 2008 to 2018, there was a decrease (APC = −3.8%) in all HPV-related diseases. The increase in cervical cancer screening adherence was related to a decrease in invasive cervical cancer and an increase in HPV vaccine coverage, which was found to arise from a decrease in “in situ” cervical cancer. In this study, the need to improve the acceptance of preventive strategies for HPV-related diseases, as well as the homogenous information furnished by all healthcare workers involved in their promotion (e.g., gynecologists, general practitioners, pediatricians) is highlighted. Abstract Human papillomavirus (HPV)-related diseases are still a challenge for public health. Some studies have shown the effects of preventive strategies on them, but studies at the national level are few in number. Therefore, a descriptive study through hospital discharge records (HDRs) was conducted in Italy between 2008 and 2018. Overall, 670,367 hospitalizations due to HPV-related diseases occurred among Italian subjects. In addition, a significant decrease in hospitalization rates for cervical cancer (average annual percentage change (AAPC) = −3.8%, 95% CI = −4.2, −3.5); vulval and vaginal cancer (AAPC = −1.4%, 95% CI = −2.2, −0.6); oropharyngeal cancer; and genital warts (AAPC = −4.0%, 95% CI = −4.5, −3.5) was observed during the study period. Furthermore, strong inverse correlations were found between screening adherence and invasive cervical cancer (r = −0.9, p < 0.001), as well as between HPV vaccination coverage and in situ cervical cancer (r = −0.8, p = 0.005). These results indicate the positive impact of HPV vaccination coverage and cervical cancer screening on hospitalizations due to cervical cancer. Indeed, HPV vaccination also resulted in a positive impact on the decrease in hospitalization rates due to other HPV-related diseases.


Introduction
Human papillomavirus (HPV) is still the main infection that occurs via sexual activities [1][2][3][4][5][6]. HPV is also the etiological factor that is the most relevant for cervical cancer [2,3,7,8]. Worldwide, the number of HPV genotypes that have, until now, been identified is almost 200 [1, 2,4,9,10]. These are classified into low-grade (the most widespread HPV 6 and 11) and high-grade (such as HPV 16 and 18) risk with respect to malignant oncological diseases [10,11]. Due to the high clearance capacity of the human immune system [3], only 10% of HPV infections progress to precancerous and cancerous lesions [8,12]. In addition, certain age classes, determined according to sex, were found to be more affected

Study Design
In this descriptive study, the hospitalization trends due to HPV-related diseases in Italy between 2008 and 2018, excluding the administrative region of Sicily, were evaluated. The source of the data was the Italian hospital discharge records (HDRs) database, which is a medical recording system for each admission to Italian hospitals since 1991 [23]. The main objective of HDR is for healthcare performance payments, followed by disease classification in fixed fee categories that are named diagnoses related groups (DRGs), in which homogeneous diagnoses were included [24][25][26]. However, after several implementations regarding disease classification and the clinical information of discharged patients [23,25,[27][28][29][30][31][32][33], the variables contained in HDRs can be used both for epidemiological and healthcare planning purposes. The variables contained in the HDRs are socio-demographic (such as sex, date of birth, birthplace, educational level, citizen, residence, marital status, etc.) and medical (healthcare setting; admission and discharge dates; diagnoses and interventions encoded following the International Classification of Disease 9th edition; Clinical Modification [ICD9-CM]; and outcome of discharge) [23,25,31,32,34].

Target Population
According to the Italian Statistics Institutes, an annual average of 54,992,245 inhabitants have lived in Italy (excluding Sicily) in the study period with a percentage of 51.5% females [35]. Among them, it has been estimated that 79.6% (from a total of 15,281,090) of the female target population, aged 25-64 years old, had a cervical cancer screening from 2008 to 2017 [36][37][38]

Inclusion Criteria
The identification of HPV-related hospital admissions was in accordance with the following codes of ICD9-CM, as reported in other studies [7,34] In order to increase the accuracy of cervical cancer identification, as per the inclusion criteria, the following interventions codes were also considered: cervix conization (67.2), cervical lesion cauterization (67.32), and cervical lesion cryosurgery [7].
Both primary and secondary diagnoses or the intervention codes of the HDRs were considered. The quality of HDRs was checked in order to remove those with missing, duplicated, or inaccurate data.

Statistical Analysis
Absolute and relative frequencies were calculated for qualitative variables (sex, marital status, citizenship, administrative Italian region of residence, and HPV-related diseases). The skewness and kurtosis tests were used to analyze the continuous variables (i.e., the age and length of hospital stay in days) in order to choose the mean and standard deviation (SD) for normal distribution. Otherwise, the median and interquartile range (IQR) were used. A nonparametric K-sample test on the equality of medians was performed for the annual median of the patients' age, as well as the median in relation to the length of stay, in days, per year. Annual and total hospitalization rates for each HPV-related disease (cervical cancer; vulval and vaginal cancer; penile cancer; oropharyngeal cancer; anal cancer; and genital warts) were calculated per 100,000 inhabitants using the related attributable fractions. They were previously reported in another manuscript: cervical cancer 100%; genital warts 100%; anal cancer 88%; vulval and vaginal cancer 77%; penile cancer 50%; oropharyngeal cancer 26% [7]. The Breusch-Pagan/Cook-Weisberg tests were utilized for the heteroskedasticity evaluated variances of annual hospitalization rates.
The annual percent change (APC) and the average APC (AAPC) were evaluated by the Joinpoint Regression Program, Version 4.9.1.0. (April, 2022; Statistical Research and Applications Branch, National Cancer Institute), following the methodology described by Kim et al. [51]. In the Joinpoint Regression Program, a constant variance was chosen in the case of homoscedasticity among the annual hospitalization rates, the adherence to cervical cancer screening, and the HPV coverages. Otherwise, a calculation of standard error (SE) was performed and included before the Joinpoint analysis [51]. The best model was chosen in accordance with the permutation test. Confidence intervals at 95% (95% IC) and p-values were reported. The p-value was considered statistically significant when α ≤ 0.05. Stata/SE 14.2 (Copyright 1985-2015, StataCorp LLC, 4905 Lakeway Drive, College Station, TX 77845, USA. Revision 29 January 2018) was used in order to perform statistical analysis.
HDRs, due to cervical cancer, had the highest percentage of admissions (43.5%, n = 291,368). Conversely, instances of penile cancer had the lowest percentage (2.4%, n = 15,804). The median length of a hospital stay was 5 days (IQR = 3-11).  Table 2 details the hospitalization rates due to HPV-related diseases, per year, between 2008 and 2018. In addition, the trends of these declined for all of them. Among the cancers, the highest reduction was found for cervical cancer with a value of 31.2% (from 115.1 to 79.2 per 100.000). This was followed by oropharynx cancer with a reduction of 30.9% (from 12.3 to 8.5 per 100.000), and then anus cancer with a reduction of 17% (from 7.1 to 5.9 per 100.000). Overall, the condition with the highest reduction was GW with a value of 33.4% (from 11.7 to 7.8 per 100.000).  trend from 2008 to 2014 (APC = −3.7%, 95% CI = −4.7, −2.8; p < 0.001) and a slightly increasing trend in the period 2014-2018 (APC = 1.3% 95% CI = −0.6, 3.3; p = 0.136). The other diseases saw a decreasing trend without Joinpoint. Additionally, the highest value in AAPC was for genital warts (AAPC = −4.02%), followed by oropharyngeal cancer (AAPC = −3.99%).  Table 3). Instances of female death due to cervical cancer during hospitalization were 2085 (0.72%); their median age was 65 years old (IQR = 52-77) (data not reported in the table).  Table 3). Instances of female death due to cervical cancer during hospitalization were 2085 (0.72%); their median age was 65 years old (IQR = 52-77) (data not reported in the table).

Discussion
This study showed the hospitalization trends of HPV-related diseases from 2008 to 2018 in Italy. Despite its significant decrease along the whole period with an AAPC = −3.8 (95% CI = −4.2, −3.5; p < 0.001), cervical cancer was the HPV-related disease with the high-

Discussion
This study showed the hospitalization trends of HPV-related diseases from 2008 to 2018 in Italy. Despite its significant decrease along the whole period with an AAPC = −3.8 (95% CI = −4.2, −3.5; p < 0.001), cervical cancer was the HPV-related disease with the highest hospitalization rate (93.5/100,000). The Italian AAPC for cervical cancer was milder in comparison to the Sicilian data, in which the hospitalization rate for cervical cancer had a statistically significant AAPC = −9.9%, [7]. Furthermore, the Italian reduction in HPV cervical cancer from 2008 to 2018 was slightly higher than that observed in the period 2001-2012 (APC = −2.9%, 95% CI = −3.8, −2.1) [13]. This can be understood as a further demonstration of the impact of preventive strategies on HDRs.
The females' median age and length of hospital stay slightly increased during the study period (p < 0.001). This was similar to the Sicilian data and may be an indirect effect of the preventive strategies. Indeed, vaccine availability was available to people born after 1996, and the older females did not have the benefit of HPV vaccination. In addition, these older females could have received the diagnosis of cervical cancer later, when it began to be symptomatic, and could have had other diseases or complications, thus resulting in extending the length of their hospital stay [7].
The rate for cervical cancer was similar to that reported by a Swedish study, which was based on registry data among females who were not vaccinated against HPV (93.5/100,000 vs. 94/100,000) [52]. Another study in England has recently shown that the incidence rate ratio for cervical cancer was inversely correlated to the HPV vaccination recipients' age, which was lower among those vaccinated at the age of 8 years old than in other age cohorts (0.13 vs. 0.38 in the 10-11 years old cohort and 0.66 in 12-13 years old cohort) [49].
A globally observed reduction in hospitalization rates due to in situ and invasive cervical cancer was detected in most of the age classes, even if the rate was higher among those older than 75 years old for invasive cancer, and those aged 15-24 for in situ cancer. In addition, the significant increase in HPV vaccination coverage among females aged 12-21 years old and adherence to cervical cancer screening among females aged 25-64 years old seem to reduce the hospitalization rates due to in situ and invasive cervical cancer, respectively. This could suggest an impact of HPV vaccination and cervical cancer screening on cervical cancer reduction [7,8,17]. Furthermore, it could be related to the availability of preventive healthcare services for either HPV vaccination centers, or for cervical cancer screening throughout Italian regions. As well as the spatial-temporal variability of the HPV vaccination coverage [19, [39][40][41][42][43][44][45][46][47], most of the Southern Italian regions had access to the most updated, high valent HPV vaccines before the Northern regions [19]. Additionally, the cervix screening uptake had variable percentages from one Italian region to another, and this could influence the cervical cancer trends [36][37][38]. On the other hand, lifestyles, individual risks, or protective factors could also affect the onset of cervical cancer, such as smoking, sexual promiscuity, or genetic polymorphisms [53].
According to the newest meta-analysis data, whatever vaccine valence was undertaken against HPV, nearly 100% of cervical lesions and infections due to some HPV serotypes (such as 16 and 18) can be avoided, thus showing an almost decennial antibody protection in several trial studies [54]. This supports the results of this study, whereby the long-lasting effectiveness of more than 10 years of vaccine introduction in Italy is confirmed once again [55]. For these reasons, HPV vaccination should be promoted by all healthcare workers and should reach all eligible HPV vaccine recipients [56,57]. This will improve coverage in the at-risk population, thereby decreasing the burden of HPV-related malignancies such as cervical cancer.
Hospitalization rates due to oropharyngeal cancer in Italy decreased constantly during the period 2008-2018 (APC = −4.0%, p < 0.001). Nevertheless, these low trends may be attributed to HPV vaccination [58]. Indeed, sexual practices can transmit the virus, especially HPV16, and its survival is possible for the anatomy of the oropharynx district. Therefore, the link to cancer in this area cannot be excluded [5]. On the other hand, oropharyngeal cancer could be an example of an oncological disease that is caused by either HPV or damaging substances (such as alcoholic drinking and smoking) [5,58]. Therefore, other studies need to be conducted, where the impact of alcoholic drinks and smoking habits among people who are either vaccinated or unvaccinated against HPV, should be evaluated.
Penile cancer had the lowest hospitalization rate in Italy during the years 2008-2018 because of its rarity, as is shown in the European data [59]. However, the penile cancer rate was higher than what was reported in European and Northern American countries (2.7/100,000 vs. 1.0/100,000) [59]. Despite its severity, there is still no prevention strategy for this oncological disease and HPV-vaccination may only have an effectiveness of 33.3% [59]. The low effectiveness can be due to the fact that there are other risk factors for penile cancer, which can be either genetic, inflammatory, or environmental [59]. Moreover, the HPV-vaccination coverage in Italy for adolescent males is too low and too recent in order to correlate to penile cancer trends or to consider as an effective prevention measure.
This study had some limitations. HDRs could underestimate the real trend of these HPV-related diseases because some of them, especially the less severe ones such as cervical cancer, may have been treated in an outpatient setting, whose clinical and administrative sources of data were not available and not included in this study. In detail, precancerous cervical lesions are more frequently treated in outpatient settings during colposcopy, which is an outpatient procedure, according to national laws [60][61][62]. Furthermore, the inappropriate use of codes for HPV-related diseases could not be excluded. Moreover, the HPV attributable fraction of HDRs is not equivalent to an exact measure of HPV status. Although the previously reported limits are different, this study suggests that hospital discharge records can represent a very efficient way for evaluating the trends of HPV-related diseases, as well as the direct/indirect benefits of primary (HPV vaccination) and secondary (cervical cancer screening) prevention.

Conclusions
Cervical cancer is the HPV-related disease with the highest impact on women's health. HPV vaccination and cervical cancer screening in Italy are associated with a reduction in hospitalization rates due to in situ and invasive cervical cancer. Additionally, other HPV-related diseases reported in this study decreased globally. Further studies should confirm our findings investigating, at an individual level, the effectiveness of preventive strategies. While waiting for these results, our analyses provide evidence-based data to support adherence to HPV preventive strategies and thus achieve the elimination of HPV.  Institutional Review Board Statement: An ethical review and approval were waived for this study due to the fact that the hospital discharge records used were extracted in an anonymous manner with a coding number in order to exclude multiple admissions.
Informed Consent Statement: Patient consent was waived due to the hospital discharge records being extracted in an anonymous manner with a coding number in order to exclude multiple admissions. Data Availability Statement: Data will be available after sending a motivated request to corresponding author.