Clinical performance of human papillomavirus based cervical cancer screening algorithm: The result of a large Danish implementation study

Abstract

Introduction

In Denmark, where human papillomavirus (HPV) ‐based cervical cancer screening is being implemented, the aim of this pilot implementation study was to test a specific screening algorithm, assess follow‐up examination attendance, and measure the proportion of precancer lesions found in relation to the number of women referred for colposcopy.

Material and Methods

From May 2017 to December 2020, 36 417 women in the uptake area of the Department of Pathology, Vejle Hospital, Region of Southern Denmark, were included in the HPV group. Women positive for HPV16/18 irrespective of cytology and women positive for other high‐risk HPV (hrHPV) types having concomitant abnormal cytology were referred directly to colposcopy. Women positive for other hrHPV types and normal cytology were referred to repeat screening after 12 months, and hrHPV negative to routine screening after three years. We obtained information on screening results and subsequent histological diagnosis from the Danish Pathology Databank through September 2022.

Results

3.6% of the women were referred to colposcopy after primary screening, 5% to repeat screening after 12 months, and 91.4% back to routine screening. High follow‐up rates were observed: 96% attended colposcopy after primary screening, with 91% attending colposcopy after repeat screening. CIN3+ was detected at colposcopy following the primary screening in 28.1% of HPV16/18‐positive women and 18.2% of those positive for other hrHPV types with concomitant abnormal cytology. Of the women with other hrHPV and simultaneous ASCUS/LSIL, 8% had CIN3+. At the repeat screening, 43% had become hrHPV negative, 55% were persistently positive for other hrHPV, and 2% had turned positive for HPV16/18. At the colposcopy following repeat screening, 10.1% of the women positive for other hrHPV were diagnosed with CIN3+, in comparison with 11.1% of the HPV16/18‐positive women.

Conclusions

In this pilot implementation study, an algorithm for HPV‐based screening was evaluated in a Danish setting. The results demonstrated high attendance at follow‐up examinations and provided insights into the number of colposcopy referrals and the detection of CIN2 and CIN3+ cases. The results suggest that women testing positive for other hrHPV in combination with ASCUS/LSIL at primary screening could potentially be referred to repeat screening instead of an immediate colposcopy.

This Danish pilot study assessed an HPV‐based screening algorithm, revealing a high follow‐up attendance. Positive hrHPV with ASCUS/LSIL may motivate repeat screening after 12 months, potentially reducing immediate colposcopy referrals while maintaining effective detection of CIN2 and CIN3+ cases.

Abbreviations

ASCUS

atypical squamous cells of undetermined significance

CIN

cervical intraepithelial neoplasia

CIN2

cervical intraepithelial neoplasia grade 2

CIN3+

cervical intraepithelial neoplasia grade 3 or worse

hrHPV

high‐risk human papillomavirus

HSIL

high‐grade squamous intraepithelial lesion

IQR

interquartile range

LSIL

low‐grade squamous intraepithelial lesion

PPV

positive predictive value

Key message.

This Danish pilot study assessed an HPV‐based screening algorithm, revealing a high follow‐up attendance. Positive hrHPV with ASCUS/LSIL may motivate repeat screening after 12 months, potentially reducing immediate colposcopy referrals while maintaining effective detection of CIN2 and CIN3+ cases.

1. INTRODUCTION

In recent years, a considerable number of countries have either transitioned to or are in the process of transitioning to human papillomavirus (HPV)‐based cervical cancer screening, marking a global shift in the primary screening method. HPV tests used for cervical cancer screening purposes detect high‐risk HPV (hrHPV) infections, which are responsible for virtually all cervical cancer cases and a considerable proportion of cervical precancer lesions. ¹ As a result, HPV‐based screening has a higher sensitivity than conventional cytology, which has contributed to it becoming the primary method for cervical cancer screening. However, in comparison with cytology screening, HPV testing has reduced specificity, which might lead to increased colposcopy referral rates. ² , ³ Therefore, different triage strategies have emerged worldwide, aiming to find a balance between optimizing colposcopy referrals and identifying cases of severe cervical neoplasia.

As nations increasingly transition toward primary HPV screening, the demand for real‐world evidence to complement controlled research settings has grown. Randomized control trials are limited by selection bias and stringent follow‐up protocols. ⁴ These factors hinder the direct applicability of trial outcomes to routine clinical contexts. Assessing the efficacy of HPV‐based screening in actual clinical settings is necessary. Furthermore, given the diversity in HPV prevalence and cervical (pre)cancer among countries, policymakers require national‐level data to build evidence‐based screening algorithms.

Denmark introduced a nationwide cervical cancer screening initiative in the mid‐1990s based on cytology‐based screening. The program has gradually evolved, and since 2007, all women aged 23–65 have been invited for routine screening every 3–5 years. In 2018, the National Board of Health suggested the implementation of HPV‐based screening for women between the ages of 30–59. The implementation of HPV‐based screening began on January 1, 2021. However, national data are scarce, with only results from the first year of a large pilot implementation study taking place in one region of Denmark (HPV SCREEN DENMARK) available. ² , ⁵

Using information from this pilot implementation study, which started in 2017, the aim of the present study was to evaluate a specific HPV‐based screening algorithm regarding the proportion of women referred to colposcopy, the follow‐up examination attendance, and the proportion of high‐grade precancerous lesions detected in relation to the number of women referred for colposcopy.

2. MATERIAL AND METHODS

2.1. Design and study population

The study design of HPV SCREEN DENMARK has previously been described. ² , ⁵ Briefly, HPV SCREEN DENMARK is an implementation study embedded into the routine cervical screening program in part of the Region of Southern Denmark. From May 2017 to December 2020, women aged 30 to 59, attending the routine screening and residing in the uptake area of the Department of Pathology at Vejle Hospital, Lillebaelt Hospital, Region of Southern Denmark, were assigned to HPV‐ or cytology‐based cervical screening depending on their municipality of residence. Women living in the municipalities Vejle, Fredericia, Kolding, and Middelfart received HPV‐based screening as the primary screening method, and women living in one of the other nine municipalities (Assens, Faaborg‐Midtfyn, Kerteminde, Nyborg, Odense, Svendborg, Nordfyns Langeland, and Ærø) were screened with cervical cytology. General practitioners or gynecologists collected liquid‐based cervical cytology samples (ThinPrep, Hologic) and sent them to the Department of Pathology at Vejle Hospital for further processing. The women continued with the assigned screening method for the following routine screenings. In this study, we will focus on the group of women who were offered HPV testing.

2.2. HPV testing

HPV testing was performed using the Cobas 4800 HPV test (Roche). Cobas, which is a PCR‐based assay, detects 14 hrHPV types (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68), of which HPV16 and HPV18 are reported individually, and the other 12 hrHPV types are pooled and analyzed as one group and referred to as other hrHPV types. ⁶

2.3. Clinical management

An overview of the screening triage for women in the HPV group is shown in Figure 1. Women who tested negative for hrHPV returned to routine screening and were invited after 3 years to an HPV test. Women with HPV16/18 were referred directly to colposcopy (the cytology results were also available but did not influence the colposcopy referral). For women with other hrHPV types, the result of the cervical sample collected at the primary screening was evaluated and used for further triage; the women with concomitant atypical squamous cells of undetermined significance or worse (ASCUS+) were referred directly to colposcopy, whereas those with normal cytology were referred to repeat screening after 12 months. The repeat screening consisted of a hrHPV test and cytology. At the repeat screening, hrHPV‐negative women with normal cytology returned to routine screening, whereas all other women were referred to colposcopy. All hrHPV‐positive women were referred to colposcopy regardless of the cytology status.

FIGURE 1.

Flowchart over the HPV group in HPV SCREEN DENMARK.

2.4. Histological follow‐up data

The Danish Pathology Databank stores the results of all cervical cytology tests and biopsies obtained from colposcopies and potential subsequent conizations and hysterectomies performed by hospitals or private gynecologists. The cervical intraepithelial neoplasia (CIN) nomenclature was used for the classification of histological diagnoses. The results are registered under each woman's unique personal identification number, which is allocated to all residents in Denmark at birth or immigration. For this study, we retrieved data from the Pathology Databank until September 30, 2022.

The outcomes were defined as cervical intraepithelial neoplasia grade 2 (CIN2) and cervical intraepithelial neoplasia grade 3 or worse (CIN3+). In Denmark, there is a consensus to treat CIN2 conservatively in women of reproductive age, whereas CIN2 in postmenopausal women and all CIN3 are treated with a conization.

3. RESULTS

From May 2017 to December 2020, 91 903 women were screened and included in the implementation study HPV SCREEN DENMARK. Of these, 36 436 women (39.6%) were allocated to the HPV screening group. We excluded 17 women due to lack of valid HPV test at baseline and two women due to total hysterectomy before baseline. The median age among the remaining 36 417 women in the HPV group was 44 years with an interquartile range (IQR) of 37–51 years, and the median time from baseline until the end of follow‐up (30 September 2022) was 3.5 years (IQR 3–4).

3.1. Primary screening

At the primary screening, 91.4% (n = 33 278) of the 36 417 women were hrHPV negative, and 8.6% (n = 3139) were hrHPV positive (Figure 1). The median age was 44 years (IQR 37–50) in the hrHPV‐negative women and 41 years (IQR 35–47) among hrHPV‐positive women.

Of the hrHPV‐positive women, 23.9% (n = 751) tested positive for HPV16/18 and were referred directly to colposcopy. Among the remaining 76.1% (n = 2388) who were positive for other hrHPV types than HPV16/18, 23.6% (n = 563) had abnormal cytology and were referred directly to colposcopy and 76.4% (n = 1825) had normal cytology and were referred to repeat screening after 12 months (Figure 1).

3.2. Colposcopy after primary screening

Among the 1314 women (3.6%) referred to colposcopy after primary screening, 1267 women attended the colposcopy, corresponding to 96.4% attendance (Table 1). The median time from primary screening to colposcopy was one month (IQR 1–2). CIN2 and CIN3+ were detected in 13.4% (n = 170) and 23.8% (n = 302) of the women, respectively. Four women had an invalid result at the colposcopy, and the remaining 63.8% of the women had ≤CIN1.

TABLE 1.

Attendance and occurrence of CIN2, CIN3+, and cervical cancer at colposcopy among hrHPV‐positive women referred to immediate colposcopy following primary screening.

HPV group	Colposcopy			Outcome
	Referred	Attended		CIN2			CIN3+ a			Cervical cancer
	N	N	%	N	%	(95% CI)	N	%	(95% CI)	N	%	(95% CI)
HPV16/18 Positive
All	751	718	95.6	84	11.7	(9.3–14.1)	202	28.1	(24.8–31.4)	13	1.8	(0.8–2.8)
Normal	472	441	93.4	46	10.4	(7.6–13.3)	65	14.7	(11.4–18.0)	5	1.1	(0.1–2.1)
ASCUS/LSIL	142	141	99.3	27	19.1	(12.7–25.6)	37	26.2	(19.0–33.5)	0	0.0	(0.0–0.0)
HSIL+ b	135	134	99.3	11	8.2	(3.6–12.9)	100	74.6	(67.3–82.0)	7	5.2	(1.5–9.0)
Invalid	2	2	100	0	0.0	(0.0–0.0)	1	50.0	(0.0–100)	1	50.0	(0.0–100)
Other hrHPV Positive
All	563	549	97.5	86	15.7	(12.6–18.7)	100	18.2	(15.0–21.4)	3	0.5	(0.0–1.2)
ASCUS/LSIL	420	411	97.9	62	15.1	(11.6–18.5)	33	8.0	(5.4–10.7)	0	0.0	(0.0–0.0)
HSIL+ b	143	138	96.5	24	17.4	(11.1–23.7)	67	48.6	(40.2–56.9)	3	2.2	(0.0–4.6)
All	1314	1267	96.4	170	13.4	(11.5–15.3)	302	23.8	(21.5–26.2)	16	1.3	(0.6–1.9)

^a Adenocarcinoma in situ (AIS) is included in the group of CIN3+.

^b Atypical squamous cells cannot exclude high‐grade squamous intraepithelial lesion (ASC‐H) is included in HSIL+.

The colposcopy was attended by 95.6% (n = 718) of the 751 HPV16/18‐positive women. CIN2 was detected in 11.7% (n = 84) and CIN3+ in 28.1% (n = 202) (Table 1). Even though the algorithm referred HPV16/18‐positive women directly to colposcopy, the cytology result at baseline was available in our study. The colposcopy attendance and detection of CIN3+ were higher in the groups with abnormal cytology (Table 1). The CIN3+ group included 13 cervical cancer cases, five of which were in women with concurrent normal cytology, seven in women with a high‐grade squamous intraepithelial lesion or worse (HSIL+), and one in which the cytology was invalid. The mean age at cervical cancer diagnosis was 38.7 years, with a range of 31–50 years. Four of the five cancer cases found in women with normal cytology were adenocarcinomas, with the fifth being squamous cell carcinoma. Six of the women with concomitant HSIL+ had squamous cell carcinomas, and one had adenocarcinoma. The last cancer case was a squamous cell carcinoma discovered in a woman with invalid cytology at primary screening.

In the group of women positive for other hrHPV and with abnormal cytology, 97.5% (n = 549) of the 563 referred women attended the colposcopy (Table 1). At the colposcopy, 15.7% were diagnosed with CIN2 and 18.2% with CIN3+. The highest proportions of CIN2 and CIN3+ were seen in the group with HSIL+ at baseline; 17.4% had CIN2, and 48.6% had CIN3+. Meanwhile, 8% of the women with ASCUS or low‐grade squamous intraepithelial lesion (LSIL) were diagnosed with CIN3+ (Table 1). There were three cases of cervical cancer in the group of CIN3+, and all were detected in the group of women with HSIL at primary screening. Two of the cancers were squamous cell carcinomas, and one was an adenocarcinoma. The mean age at the time of diagnosis was 48 years (range 41–57 years).

3.3. Repeat screening

The repeat screening aimed at 12 months was attended by 96.1% (n = 1753) of the 1825 women referred to repeat screening. The median interval between baseline and repeat screening was 13 months (IQR 12–15 months). At the repeat screening, 42.7% (n = 748) of the women had cleared the infection and were hrHPV negative, whereas 55.0% (n = 965) were still positive for other hrHPV types, 1.7% (n = 30) tested positive for HPV16/18, and 0.6% (n = 10) had an invalid HPV test (Figure 1).

Of the hrHPV‐negative women, 94.5% (n = 707) had normal cytology and returned to routine screening, 1.9% had ASCUS/LSIL, 0.3% had HSIL+, and 3.3% had no cytology at the repeat screening. The women with ASCUS+ were referred to colposcopy.

Most women who tested positive for other hrHPV at repeat screening had normal cytology (80.6%), and the same applied to women testing HPV16/18 positive (66.7%) (Figure 1). However, all women who were hrHPV positive, regardless of cytology were referred to colposcopy.

3.4. Colposcopy after repeat screening

After repeat screening, 1011 women were referred to colposcopy: 30 after testing positive for HPV16/18, 965 after being positive for other hrHPV, and 16 women testing hrHPV negative but with concomitant ASCUS+ (Figure 1). Out of them, 922 women (91.2%) attended the colposcopy. The median time interval between repeat screening and colposcopy was two months (IQR 1–3 months). Thirteen of the women with no record of a colposcopy had their repeat screening three months before the end of follow‐up (September 30, 2022), resulting in a limited time for them to attend a colposcopy. Of the 922 women who attended the colposcopy, 10.1% were diagnosed with CIN2 and 10.1% with CIN3 + (Table 2). CIN2 and CIN3+ were most common in women with HPV16/18 and other hrHPV with contemporaneous abnormal cytology. Women positive for HPV16/18 with HSIL+ had a CIN3+ prevalence of 33.3% and a CIN2 prevalence of 66.7%, whereas women positive for other hrHPV types with HSIL+ had a CIN3+ prevalence of 45.8% and a CIN2 prevalence of 18.8%. Two cases of cervical cancer (of squamous cell type) were detected at the colposcopy after repeat screening in the group of women positive for other hrHPV; one had normal cytology at repeat screening and was 59 years old at the time of diagnosis, and the other had HSIL and was 38 years old. Out of the 25 women who only had a negative hrHPV test and no cytology at the repeat screening, 4 had a colposcopy during the follow‐up, and no one was diagnosed with CIN2 or CIN3+ (data not shown).

TABLE 2.

Attendance and prevalence of CIN2, CIN3+, and cervical cancer at colposcopy among women referred to colposcopy following repeat screening after 12 months.

HPV group at 12‐month repeat screening	Colposcopy			Outcome
	Referred	Attended		CIN2			CIN3+ a			Cervical cancer
	N	N	%	N	%	(95% CI)	N	%	(95% CI)	N	%	(95% CI)
HPV16/18 Positive
All	30	27	90.0	5	18.5	(3.9–33.2)	3	11.1	(0.0–23.0)	0	0.0	(0.0–0.0)
Normal	20	18	90.0	1	5.6	(0.0–16.1)	2	11.1	(0.0–25.6)	0	0.0	(0.0–0.0)
ASCUS/LSIL	7	6	85.7	2	33.3	(0.0–71.1)	0	0.0	(0.0–0.0)	0	0.0	(0.0–0.0)
HSIL+ b	3	3	100.0	2	66.7	(13.3–100)	1	33.3	(0.0–86.7)	0	0.0	(0.0–0.0)
Other hrHPV Positive
All	965	890	92.2	87	9.8	(7.8–11.7)	90	10.1	(8.1–12.1)	2	0.2	(0.0–0.5)
Normal	776	710	91.5	52	7.3	(5.4–9.2)	53	7.5	(5.5–9.4)	1	0.1	(0.0–0.4)
ASCUS/LSIL	136	128	94.1	26	20.3	(13.3–27.3)	15	11.7	(6.1–17.3)	0	0.0	(0.0–0.0)
HSIL+ b	48	48	100.0	9	18.8	(7.7–29.8)	22	45.8	(31.7–59.9)	1	2.1	(0.0–6.1)
Invalid	5	4	80.0	0	0.0	(0.0–0.0)	0	0.0	(0.0–0.0)	0	0.0	(0.0–0.0)
hrHPV Negative
ASCUS+	16	5	31.3	1	20.0	(0.0–55.1)	0	0.0	(0.0–0.0)	0	0.0	(0.0–0.0)
All	1011	922	91.2	93	10.1	(8.1–12.0)	93	10.1	(8.1–12.0)	2	0.2	(0.0–0.5)

^a Adenocarcinoma in situ (AIS) is included in the group of CIN3+.

^b Atypical squamous cells cannot exclude high‐grade squamous intraepithelial lesion (ASC‐H) is included in HSIL+.

4. DISCUSSION

This paper reports the results of the regional Danish implementation study of HPV‐based cervical cancer screening, including results from the primary screening, first round of repeat screening, and colposcopy. Following the primary screening, 3.6% of the women were referred to immediate colposcopy, 5% to repeat screening after 12 months, and the remaining 91.4% were scheduled for routine screening after three years. We observed a high follow‐up attendance, with 96% attendance at the colposcopy after primary screening, 96% attending repeat screening, and 91% attending the colposcopy after repeat screening. CIN3+ and CIN2 were detected in overall 28.1% and 11.7% of the HPV16/18‐positive women, 18.2% and 15.7% of the other hrHPV‐positive women who attended colposcopy after primary screening, and 10.1% and 10.1% of the women who attended colposcopy after repeat screening. Cervical cancer was found in 16 women after primary screening (13 cases in the group of HPV16/18 positive and 3 cases in women positive for other hrHPV with concurrent ASCUS+), and 2 cases were found at colposcopy following repeat screening.

The screening algorithm utilized in this implementation study is sensitive and cautious. During the primary screening, we referred all HPV16/18‐positive women, regardless of cytology status, directly to colposcopy. This strategy was chosen since HPV16/18 is known as the most carcinogenic type to cause CIN3 + . ⁷ Other studies/countries have chosen strategies where hrHPV‐positive women only with accompanying abnormal cytology were referred to colposcopy, and hrHPV‐positive women with concomitant normal cytology to repeat screening. ³ , ⁸ , ⁹ , ¹⁰ However, we observe that nearly 63% of the women who were HPV16/18 positive had normal cytology at primary screening, and still, respectively, 10.4% and 14.7% of them were diagnosed with CIN2 or CIN3+ at the colposcopy, which is slightly higher than in a similar study. ¹¹ In the group of CIN3+, five cases were cervical cancer. Referral to immediate colposcopy appears appropriate in our hands. The high proportion of women with CIN2/3 among HPV16/18‐positive women with concomitant normal cytology highlights the potentially low sensitivity of cytology and the importance of HPV testing in cervical screening.

The 12 hrHPV types that were pooled in the group of other hrHPV are known to be less carcinogenic than HPV16/18, and therefore, immediate colposcopy following the primary screening is unlikely to be required for all women in this group. We, therefore, used cytology as a supplementary triage and only referred women with simultaneous ASCUS+ to colposcopy. Women with normal cytology were directed to repeat screening after 12 months. The overall positive predictive value (PPV) of CIN3+ was 18% in women positive for other hrHPV. Depending on the level of risk one accepts, our findings may suggest that women with only concomitant ASCUS/LSIL could be referred to repeat screening rather than colposcopy. Australia was one of the first countries to implement HPV‐based cervical cancer screening, and they implemented in 2017 a similar screening algorithm as used in the present study. ¹² However, they only referred women positive for other hrHPV with concomitant HSIL+ to colposcopy, whereas women positive for other hrHPV types with ASCUS/LSIL on the cytology were referred to repeat screening after 12 months. In the future, a similar triage approach in a Danish setting could be a potential option.

At the repeat screening, all hrHPV‐positive women, independent of genotype or cytology status, and hrHPV‐negative women with ASCUS+ on cytology were referred to colposcopy. The PPV for CIN3+ for women becoming HPV16/18 positive at the repeat 12‐month screening was lower compared to primary screening (11% vs. 28%). This is most likely because women who tested positive for HPV16/18 at repeat screening had acquired the infection within the last 12 months, giving them less time to develop cervical abnormalities than the HPV16/18 group at primary screening, which included both recently acquired and persistent infections. The CIN2 prevalence at repeat screening was 18.5%, which was greater than at primary screening (11.7%), demonstrating that the HPV16/18 infection had begun to produce cell alterations even though it had been acquired within the previous year. The PPV for CIN3+ for women positive for other hrHPV was also lower after repeat screening compared to primary screening (10% vs. 18%). The reduced PPV at repeat screening is most likely due to the fact that all women who participated in repeat screening had normal cytology one year before, and the group referred to colposcopy included all women positive for other hrHPV, including those with normal cytology. At the primary screening, only women with ASCUS+ were referred to colposcopy. On the other hand, the fact that women tested positive for other hrHPV types for the second time may indicate difficulties in clearing the infection or higher susceptibility to new infections, and this could increase their risk of developing CIN3+. Because other hrHPV types are a pooled response for 12 hrHPV types, we were unable to determine whether it was a new HPV type acquired or it was a persistent HPV infection.

Most of the cervical cancer cases detected in this cohort were detected in the HPV16/18‐positive women at colposcopy following primary screening (13 of 18 cervical cancer cases). Squamous cell carcinoma was the most frequent histological type for HPV 16/18 positive women with concurrent HSIL+ (6 of 7), whereas adenocarcinoma was more common in women with normal cytology (4 of 5). Adenocarcinomas develop from glandular cells in the endocervix, the inner region of the cervix, where cell sampling is more difficult, which may explain why adenocarcinomas were more common in cytology normal women. For women positive for other hrHPV types, either at primary or repeat screening, the most common histological type was squamous cell carcinomas (4 of 5 cases) and most women had concurrent HSIL (4 of 5). The mean age at the time of diagnosis was higher in women positive for other hrHPV types (48.2 years) than in HPV16/18‐positive women (38.7 years).

Based on data from the first two years, Australia's screening program has been adjusted so that women who test positive for other hrHPV and have ≤ASCUS/LSIL on cytology at primary or repeat screening are referred to a second repeat screening after 12 months rather than colposcopy. ¹³ The decision was based on the observation that only 4.6% and 2.7% of women positive for other hrHPV with ASCUS/LSIL or normal cytology, respectively, had CIN3+ detected at colposcopy following repeat screening. ¹² We detected more CIN3+ following repeat screening in both women positive for other hrHPV and ASCUS/LSIL (11.7%) and normal cytology (7.6%) than reported by Smith et al. ¹² even though our repeat screening population only consisted of women positive for other hrHPV with concomitant normal cytology at primary screening. In Denmark, national guidelines recommend a biopsy from each quadrant of the cervix as the clinical practice during colposcopy despite an apparently normal examination, as even trained colposcopists may miss areas of dysplasia if biopsies are only taken in case of visible changes. Taking four randomly selected biopsies can improve the detection of dysplasia requiring treatment. ¹⁴ , ¹⁵ , ¹⁶ However, there is no agreement in Australia on biopsies at normal colposcopy. ¹²

Denmark has no universally agreed specific risk threshold for when women should be referred for colposcopy or repeat screening. A higher colposcopy referral rate and detection of CIN3+ were expected when utilizing HPV‐based screening since a cervical HPV test is known to be more sensitive but less specific than cytology. ² , ⁵ However, after identifying and treating the most prevalent cases in the first round of screening, the probability of finding CIN3+ should be lower in subsequent rounds. As a result, it is important to remember that this is only a snapshot of the current situation and that further monitoring and adaptation of screening algorithms are required to reduce the proportion of unnecessary colposcopies, as studies have shown that colposcopies can cause psychological and physical distress. ¹⁷ , ¹⁸

The strength of this study is the population‐based design and its incorporation into the Danish routine screening program, which gives us the real‐life performance of the HPV‐based screening. Furthermore, all cervical cytology and histology results collected in Denmark are stored in the Danish Pathology Databank, allowing us to include all subsequent cervical diagnoses throughout the country without loss to follow‐up. Our study also had some limitations. We could not establish whether women who tested positive for other hrHPV types at both primary and repeat screening had the same HPV type on both occasions because we used a pooled test for the 12 other hrHPV types.

5. CONCLUSION

In this pilot implementation study, an algorithm for HPV‐based screening was implemented and tested in a Danish context. The results indicate high follow‐up examination attendance, along with insights into the number of colposcopy referrals and the detection of CIN2 and CIN3+ cases. The study results may suggest that in the future, women testing positive for other high‐risk HPV along with ASCUS/LSIL at primary screening could be referred to a repeat screening after 12 months rather than to immediate colposcopy. The results of this study can help policymakers develop a cervical cancer screening program in which referral levels are balanced with detection levels.

AUTHOR CONTRIBUTIONS

Marianne Waldstrøm, Susanne K. Kjær, Christian Munk, and Dorthe Ørnskov designed and initiated the study. Marianne Waldstrøm and Susanne K. Kjær obtained funding. Marianne Waldstrøm organized the implementation of HPV‐based screening. Dorthe Ørnskov was responsible for HPV testing and other laboratory procedures. Sofie Lindquist performed registry linkages, data management, and statistical analyzes. Sofie Lindquist and Susanne K. Kjær wrote the first draft of the manuscript. All authors contributed to the interpretation of results and revising the article gave final approval of the version to be published and agreed to be accountable for all aspects of the work.

FUNDING INFORMATION

This study was funded by Lillebaelt Hospital, Region of Southern Denmark, and the Mermaid project (Mermaid 2). HPV tests for this study were provided at a reduced cost by Roche.

CONFLICT OF INTEREST STATEMENT

HPV tests used in this study were provided at a reduced cost by Roche. Susanne K. Kjær has received research grants outside the study through her institution from Merck. All other authors declare no potential conflict of interest.

ETHICS STATEMENT

The Health Research Ethics Committee in the Region of Southern Denmark assessed the study to be an implementation study exempt from informed consent (S‐20160146). The Danish Data Protection Agency (18–21 475) and the National Board of Patient Safety (3–3013‐2597) approved data collection and the use of registry data. Datasets were pseudonymized before analysis.

Lindquist S, Kjær SK, Frederiksen K, et al. Clinical performance of human papillomavirus based cervical cancer screening algorithm: The result of a large Danish implementation study. Acta Obstet Gynecol Scand. 2024;103:1781‐1788. doi: 10.1111/aogs.14915