Cervical cancer is the fourth most common cancer among women globally. The world-age standardized incidence rate (ASIR) was approximately 5 per 100,000 women-years, with 604,127 new cases and 341,831 deaths in 2020 [1]. Cervical cancer had a very high disease burden and was a leading cause of death in Africa, Sub-Saharan Africa, Melanesia, South America, and South-East Asia. In Thailand, the cervical cancer ASIR and mortality rate are 16.4 and 7.4 per 100,000 women-years, respectively [1,2].

The target goal to conquer cervical cancer as planned by the World Health Organization (WHO) by 2030 is to increase the access and acceptability rates of cervical cancer vaccination, screening, and treatment to 90%, 70%, and 90%, respectively [3]. Focusing on cervical cancer screening, which is an essential step in achieving this goal [4] and detecting high-risk human papillomavirus (HPV), a vital etiological factor for cervical cancer, play important roles [5]. HPV testing is recommended as part of a comprehensive program for the prevention and control of cervical cancer [6,7].

The barrier to successful cancer screening lies not only in an effective screening technique but also in the screening coverage of the target population. The Thai National Cancer Institute reported 69% screening coverage among the target groups, and only 28% among women who should undergo screening based on recommendation [8]. Previous studies from other countries have reported many reasons for not undergoing cervical cancer screening, such as self-perception of good health, unavailability, shyness, problems of health access (the distance or personnel), and reluctance to undergo pelvic examination and/or to reveal personal or sexual history [9-11]. In Thailand, few studies reported the causes of ‘no’ or ‘inadequate screening’ as shyness or embarrassment, fear of pain, lack of knowledge about the cause of cervical cancer and importance of screening, anxiety about the abnormal results, low patient priority, and self-perception of good health or absence of any risk that the test deemed unnecessary [12-14]. Hence, self-sampling to collect specimens for HPV testing in patients who are not comfortable undergoing cervicovaginal examinations in health units may solve these problems [1,15,16].

Many studies have reported a high concordance rate of HPV detection between self-sampling and physician-sampling [17-19]. This finding has led to the implementation of self-sampling for HPV testing in some regions, including Europe, Africa, America, and some Asian countries such as Malaysia and Myanmar [20]. In Thailand, HPV-based testing has been included in the National Cervical Cancer Screening Policy since 2020. In 2022, HPV self-sampling was offered by the Thai government to women in some areas as a pilot study and is planned to be launched nationwide in the future.

As this a screening technique performed by the women themselves, the women’s knowledge of cervical cancer screening and attitude toward the self-sampling procedure are important. This information should help solicit and construct educational programs for target women.

This study aimed to compare attitudes toward self-sampling for HPV testing before and after specimen collection in women undergoing colposcopy. Possible factors associated with the outcomes were also evaluated. Knowledge of cervical cancer screening and HPV was also assessed, and is presented in detail elsewhere.

Of the 500 participating women, one did not complete the attitude questionnaire and was excluded from the analysis. Among the remaining 499 women, two did not provide clinical or demographic data. Table 1 summarizes the clinical and socioeconomic characteristics of the 497 women included in this study. The mean age was 39.0±11.4 years. Most patients were premenopause (85.3%) and had no HPV vaccination (87.7%). Approximately 72.0% of participants had been sexually active in the past year. Nearly half of the women (45.0%) had a history of abnormal cervical testing; 22.7% had prior abnormal cervical cytology, 15.3% had prior high-risk HPV infection, and 6.0% had a history of >high-grade cervical lesions (HSIL).

Before self-sampling, the mean sum attitude score toward HPV self-sampling was 37.8±5.7. On examining each item of the attitude questions, the pre-self-sampling scores for each attitude ranged from 4.2 to 4.4, except for two items. The two items with lower scores compared to the others were ‘confidence to perform self-sampling’ (score 3.9) and ‘reliable results compared to physician sampling’ (score 3.7).

After the self-sampling process, the mean sum attitude score increased to 39.7±5.2 (1.9 higher), which was significantly higher compared to the pre-self-sampling (P<0.001). The score for each attitude after sampling ranged from 4.2 to 4.6, showing a 0.2-0.3 higher score compared to pre-sampling. Despite the overall improvement, one item, namely confidence in the reliability of self-sampling results compared to physician sampling, still had a lower score (3.9) than the other items. Table 2 provides a detailed overview of attitude scores before and after self-sampling.

When we assessed the association between attitude before the use of HPV self-sampling and the women’s characteristic features using univariate analyses, a significant negative correlation of attitude with age (B=−0.079; P=0.077) and postmenopause status (B=−0.097; P=0.031) was revealed. In contrast, a significant positive correlation of attitude with active sexual activity within the past year (B=0.095; P=0.035), education level of bachelor’s degree or higher (B=0.098; P=0.029), and family monthly income greater than 672 United States Dollar (B=0.089; P=0.047) were found. Although knowledge of HPV (B=0.097; P=0.055) and prior HSIL pathology (B=0.078; P=0.082) tended to have positive correlation, they were not significant. On multivariate analysis, the only remaining significant factor associated with a positive attitude was sexual activity within the past year (B=0.105; P=0.048) (Table 3).

The present study was a multicenter study in Thailand that assessed women’s attitudes toward HPV self-sampling. We found rather high overall and attitude scores for each item even before the self-sampling (37.8±5.7). The attitude score was higher after the actual use of self-sampling (39.7±5.2). This result was similar to that of a previous single-center study in Thailand by Phoolcharoen et al. [19], who reported a mean attitude score of 4.2 out of 5 after use among 250 women attending colposcopy clinics. Another survey by Kittisiam et al. [21] in Thailand also reported a 40.3% acceptance rate of 2,810 healthy Bangkok women toward self-sampling. In several other studies, 57-100% preference rates toward HPV self-sampling have been reported [16-18,21,22]. This wide range of acceptance may lie not only in individuals’ characteristics but also in religious and cultural differences.

We demonstrated significantly improved attitudes after actual self-sampling for all items. This was consistent with previous studies that reported that women’s perceptions improved after an actual experience of self-sampling in terms of privacy, comfort, ease of use, feeling more relaxed during sampling, and time saving [16-18,22-25]. Privacy and comfort are the major benefits, especially for women who have never been or are under-screened due to embarrassment [18,25]. Although there was a statistically significant increase in attitude scores post-sampling (with a 1.9-point increase), further studies involving a larger sample size and healthy women are needed to confirm the clinical significance of this improvement.

The two issues that had the lowest attitude scores in this study were confidence in performing self-sampling and the reliability of self-sampling. These were consistent with findings from previous studies [16,17,19,22,25]. Nevertheless, our study demonstrated an improvement in confidence about performing self-sampling after a single actual performance. Poorer acceptability might be because the women did not know about HPV self-sampling or had scant opportunities to try the self-method, especially older women who were more familiar with traditional Pap smears [26]. Furthermore, due to the conservative culture, Thai women usually avoid touching or inserting any device into the vagina, for example, using a tampon or performing a pelvic examination [19,25,27].

Age and postmenopause were negatively associated with attitude. Previous studies also found that older age, lower income, level of knowledge, religious beliefs, and Pap smear experience are associated with low acceptance of HPV self-sampling [16,21,26]. In contrast, we found that the features showing a positive association were sexually active status, high education level, and previous HSIL cytology. The only independent positive factor was sexual activity within the past year. These women may have less fear or be more comfortable inserting devices into the vagina than non-sexually active women. Other studies have revealed that women with better HPV knowledge tend to have greater acceptance of and confidence in self-sampling [21,26,28].

Previous studies in Thailand reported a low level of knowledge about HPV and cervical cancer screening [21,29]. A Canadian study demonstrated a strategy for the expanded coverage of HPV self-sampling by using healthcare providers to provide more details and information on the accuracy of HPV self-sampling [26]. The national policy should promote the population’s knowledge of cervical cancer prevention and HPV infection as important causes.

With our finding that the women’s attitude regarding the reliability of the testing results did not improve much after actual self-sampling, comparable performance of self-compared to physician sampling should be reassured to the women directly or as a public educational program. These factors should enhance the acceptance of, or attitudes toward, and self-sampling for HPV testing. Direct communication and assertion by regular or local healthcare providers may reassure women about the clinical application and reliability of self-sampling. Illustrating the basic lower genital tract anatomy before self-sampling using a model or animation may increase the confidence and trustworthiness in the HPV self-sampling results.

This study has some strengths, as it was a prospective multicenter study. Therefore, the heterogeneity among women from different institutions may represent a variety of female groups. Furthermore, data were collected immediately before and after self-sampling to avoid recall bias. However, this study had some limitations. First, a self-sampling survey of attitudes was conducted after colposcopy. This may have affected our results. Second, the study participants had specific characteristics; for example, all participating institutes were located in the metropolitan urban area of Thailand, only 1.2% of the women in this study had never had any sexual activity (intercourse), and all study participants had undergone abnormal cervical cancer screening. Hence, the clinical features of women that influence their attitudes may differ from those of women with different characteristics or backgrounds, and the study findings may not be applicable to all groups of women. Further studies on women from different backgrounds are required to determine the overall national situation. The data will serve as basic information for planning a comprehensive solicitation program to achieve the WHO target of 70% cervical cancer screening.

Attitudes toward HPV self-sampling were accepted, especially among sexually active women. Attitude scores improved after the actual self-sampling. Confidence in performing the procedure and the reliability of the results were issues to be resolved.