Impact of the cervical cancer awareness months on public interest in Japan: A Google Trends analysis, 2012–2021

The immunization and screening rates for human papillomavirus in Japan are lower than those in other countries. We aimed to evaluate the impact of cervical cancer awareness months on public attention using Google Trends analysis. Between 2012 and 2021, we analyzed the trends in relative search volumes (RSVs) for “Shikyuu-keigan” (cervical cancer in English) in Japan, during the cervical cancer awareness month (CCAM) in January and cervical cancer prevention awareness enhancement month (CCPAEM) in November. We performed a joinpoint regression analysis to identify a statistically significant trend change point. Additionally, we compared the mean RSVs of each awareness month with the rest of the year. Significant trend change points were observed, but none were found in CCAM and CCPAEM periods. Comparison of mean RSVs among CCAM, CCPAEM, and the rest of the months did not suggest any significant increases in RSVs during these awareness periods. In conclusion, CAM and CCPAEM did not raise public interest in cervical cancer in Japan. Although the results are based on internet users, the findings might suggest a need to develop a more effective and attractive approach to achieve the 90-70-90 targets of cervical cancer prevention by 2030.


Scientific Reports
| (2022) 12:15391 | https://doi.org/10.1038/s41598-022-19798-x www.nature.com/scientificreports/ awareness to alleviate the burdens of malignant diseases, cancer awareness months have been globally established, in which January is set as the cervical cancer awareness month (CCAM). Additionally, since 2010, November is the cervical cancer prevention awareness enhancement month (CCPAEM) in Japan, which has been endorsed by CancerNet Japan 18 . Recently, several studies have reported the usefulness of measuring public attention toward malignant diseases during the Cancer Awareness Months by the method of Google Trend analysis 19,20 . However, impacts of those awareness campaigns on public interests has not been fully evaluated in Japan. Therefore, this study aimed to evaluate the increase in public awareness of cervical cancer during CCAM and CCPAEM.

Results
Trends in the relative search volumes for the search term. The monthly trend of the relative search volume (RSV) for the search term "Shikyuu-keigan" from 2012 to 2021 is depicted in Fig. 1. Data from the "All" and "Health" categories showed almost identical fluctuations over time. The largest surge occurred in June 2013 (RSV at 100%), followed by September 2015 (RSVs at 73% in the "All" and 64% in the "Health" category), and July 2016 (RSV at 75% in the "All" and 66% in the "Health" category). Subsequently, there was no increase in the RSVs recorded at > 60%.
Joinpoint analysis for the 10-year period. The graphics and monthly percent changes (MPCs) in the RSV obtained by the joinpoint analysis are shown in Fig. 2  in RSV for the search time in the "All" category obtained by the joinpoint analysis are shown in Fig. 3 and Table 2, respectively. During the decade, we did not observe a clear surge in RSVs by means of the joinpoint analysis in either the CCAM or CCPAEM periods. During CCAM periods, there was only a temporal increase in RSVs in 2018, which did not form a cluster of increased RSVs. We found an aggregation of increased RSVs around the CCAM period in 2019, but its peak was at week 6 (middle of February) and the joinpoint analysis did not find a significant trend change point during the awareness month. During the CCPAEM periods, we detected temporal increases in RSVs in 2012 and 2020 without forming a monthly cluster of RSVs. In addition, we observed no The graphics and WPCs in RSV for the search time in the "Health" category obtained by the joinpoint analysis are provided in Fig. 4 and Table 3. Similar to the "All" category analyses, there was no uptrend of RSVs in these awareness months. During CCAM periods, there was also a one-point increase in RSVs in 2018, which did not form a significant RSV mass. During the CCPAEM period, we detected a cluster of increased RSVs in 2020. However, this was not determined to be significant by the joinpoint analysis. Outside these awareness months, we found significant surges of RSVs around week 24 in 2013, week 37 in 2015, week 26 in 2016, and week 36 in 2021. All these surges corresponded to those in the "All" category.  Table 4. In the "All" category, the Kruskal-Wallis test showed a p-value of 0.03; however, a significant difference was not observed among any mean RSVs after Bonferroni adjustment. Otherwise, there were no significant differences among the three categories. In the "Health" category in 2012, the mean RSV in CCAM was significantly lower than that in the other weeks (p = 0.028). In 2020, the mean RSV in CCPAEM was significantly higher than that in the rest of the week (p = 0.044).

Discussion
We demonstrated the public awareness of cervical cancer in Japan by employing Google Trends data for the last decade. In the long run, the highest RSV was observed in June 2013, followed by an increasing trend from April 2020 by the end of 2021. Although the data collected by the internet search does not necessarily reflect the entire  Table 1. www.nature.com/scientificreports/ awareness situation in the country, the year-by-year data suggest that public interests did not rise during these awareness months in Japan, implying that the awareness-raising activities in CCAM and CCPAEM were not powerful or influential enough to promote public attention in Japan. Japan is one of the advanced countries with high income, high literacy, and good accessibility to the internet, and an infodemiology study like our approach is relatively applicable. Thus, we consider the results of this study relatively accurately reflect the current state of public interest in cervical cancer in Japan; simply telling, there is a lack of awareness of cervical cancer prevention   Table 2. www.nature.com/scientificreports/ in Japanese society as a whole. However, beginning in April 2022, the public expense assistance system was resumed in Japan, and we assume that these awareness-raising campaigns will attract more attention than ever. Internet access to online searching engines, such as Google, has won the people's favor to gain knowledge on various topics. This greatly affected their attitudes, behaviors, intentions, and decisions at both the individual and national levels 21 . Even 10 years ago in 2012, more than half of the people in the United States answered that they had access to the internet for health-related issues 22 . Currently, data from the Google Trends platform have been   Table 3. Table 3. Weekly percentage changes (WPC) in relative search volumes (RSV) of "Shikyuu-keigan (cervical cancer)" in "Health category", by each year, 2012-2021. SE, standard error; CI, confidence interval. *Significantly different from zero (p < 0.05).  27 . A review of social events that could have evoked increases in RSVs over the last decade should be discussed. In June 2013, when the highest RSV was achieved, the Japanese Ministry of Health, Labor, and Welfare decided to discontinue the proactive recommendation for HPV vaccination 28 , despite being already included in the lineup of the national immunization program. This government announcement was launched following the increased reports of vaccine-related adverse reactions and social concerns. In September 2015, when we observed a second momentary surge, the government joint meeting announced that it would continue to refrain from recommending HPV vaccination 29 . In July 2016, the Japanese government and two pharmaceutical companies were accused of the health hazards that occurred with HPV vaccination. According to the petition, 63 women aged 15 to 22 years filed a lawsuit with the courts in four cities in Japan. These instances could be largely associated with the rise in RSVs in each term.

Year
Impact of the COVID-19 pandemic on public interest in cervical cancer is of interest to be considered. The World Health Organization declared the COVID-19 pandemic as the Public Health Emergency of International Concern at the end of January 2020. The present data indicated that MPC in RSV of cervical cancer in both All and Health categories showed significantly increasing trends since April 2020 ( Fig. 2 and Table 1). Although the COVID-19 pandemic appears to have promoted public interest in cervical cancer, the causal relationship between these events is unexplainable. Generally considering, the unprecedented pandemic must have limited educational opportunities for preventive medicine for the public. In fact, an infodemiology study using the Google Trends reported a sharp drop in global online searching for cervical cancer after the declaration of the COVID-19 pandemic 30 .
To the best of our knowledge, this is the first attempt to investigate the impact of monthly awareness campaigns for cervical cancer on public awareness by utilizing Google trend data. The power of internet search data has become greater to estimate public interest and test a hypothesis in various fields. However, the limitations of the present study should be mentioned. First, Google Trends data is based on internet access, and thus, the present study included only those who had access to the internet and used Google to search 31 . However, the percentages of internet penetration between 2012 and 2019 were sufficiently high at 79.5% and 92.7%, respectively 32 . The market shares of Google searches in Japan remained over 70% during half of the study period 33 . This data suggests the robustness of Google Trends as a surrogate for public awareness. Second, the validity of searching and analysis methods used in this study should be evaluated. Although we consider the Japanese term "Shikyuu-keigan" to Table 4. Comparison of mean relative search volumes (RSV) of "Shikyuu-keigan (cervical cancer)", 2012-2021. *A significant difference was not observed among any mean RSVs after bonferroni adjustment. **Mean RSV in CCAM was significantly lower than that in other weeks (p = 0.028). ***Mean RSV in CCPAEM was significantly higher than that in other weeks (p = 0.044). P values were calculated by Kruskal-Wallis test. CCAM, Cervical Cancer Awareness Month (January); CCPAEM, cervical cancer prevention awareness enhancement month (November). www.nature.com/scientificreports/ be the most familiar word for cervical cancer among the general population, whether it could comprehensively encompass the internet searching activity for the disease was uncertain. Also, collaborating Google trend data and joinpoint analysis to estimate public interest should be evaluated. However, in addition to our efforts 34,35 , other researchers have applied this combination method in recent studies [23][24][25][26] . Third, as indicated in a previous study 36 , Google Trends lacks full transparency and reproducibility. This is because RSVs are calculated based on non-published assumptions. Fourth, due to the nature of RSV data, non-scientific news, such as the death of the celebrity 37 , remarkably influences the results. Fifth, multiple comparisons may be associated with a statistical difference in the tested parameters. Despite these limitations, we believe our research could evaluate the public interest during the awareness months for cervical cancer in Japan, where vaccination and screening programs for preventable diseases need to be further promoted.
In conclusion, we highlighted that both global and domestic awareness months for cervical cancer failed to raise public interest during the last decade in Japan. Multifaceted and repeated approaches from another perspective are warranted to increase this trend, to ensure we meet the 90-70-90 targets for cervical cancer prevention by 2030.

Methods
Data source. Google Trends is a publicly available free data source based on Google search data. The open tool allows users to customize and obtain search term volumes that are entered into the Google search engine. It does not provide an absolute count of Google searches; rather, it provides an RSV with a scale of 0-100 (0 is the lowest and 100 is the highest popularity) 26,[38][39][40] . This can indicate the degree of interest of a particular word at a certain time point. To date, the open database has beenwidely utilized to measure public attention in various fields, including healthcare research 34,35,[41][42][43] . This analysis allows us to estimate the relative popularity of a specific search terms of interest in each particular category (for example, "health, " "sports, " "business, " and so on), place (by country), and time range (various adjustments is available). In this study, we accessed the Google Trends platform and downloaded data in .csv format on February 1, 2022.
Search input and variables. Our search protocol for Google Trends is provided in Supplementary Fig. 1, which was utilized in previous studies 26,34,35,40,44 . We searched in the Google Trends on 31st January 2022. The location of the search was selected for Japan, and we extracted data from 2012 and 2020, as well as each year of the study period. We put a Japanese word "Shikyuu-keigan" (cervical cancer in English) into the search term and chose the total category to extract the popularity of the term over a long period. In this way, we obtained monthly RSVs for the decade and weekly RSVs for each year and subjected them to statistical analysis to estimate whether the term was searched during the monthly awareness campaigns.