The present study revealed a strong correlation between the sentinel surveillance data and relative search volume on Google Trends. The strength of the Internet search engine query data has significant implications for real-world public health interests. Additionally, our findings suggested that Google Trends may have the potential to enable early detection of a RSV epidemic even if a national surveillance system is unavailable.
The analysis of Google Trends data pertaining to infectious diseases was initially used to determine its utility in detecting the influenza A(H1N1)2009 pandemic, in which it demonstrated an ability to forecast influenza disease activity [12]. The Google Trends data for estimating the activity of the influenza virus, dubbed “Google Flu Trends”, showed a favorable correlation in the United States and Europe [13, 15, 18, 19]. Recently, the utility of Google Trends for describing RSV activity in the United States was also evaluated [20, 21]. Although this has been done only in the United Sates thus far, our findings indicated that it may be applicable to other nations as well. Furthermore, our analysis suggested that Google Trends might be capable of detecting an increase in the RSV infection trend simultaneously with or even before the national surveillance system. Further evaluation in other countries or regions is needed. Google Trends is a readily available tool that can be used with great effect to advise the public health sector of infection risks. For countries that do not have a nationwide surveillance or alert system, Google Trends may serve as a useful, alternative warning system provided that the Internet penetration rate is at level comparable with that of the nations discussed.
Early recognition of a RSV epidemic is important because it enables timely palivizumab administration to prevent infections in high-risk patients. The RSV epidemic was observed earlier in 2021 than in 2018 or 2019. This anomaly may be explained by the fact that no epidemic surge occurred in the preceding year. In situations such as that seen in 2021, early prophylaxis with palivizumab should be considered. In Japan, the Infectious Disease Surveillance Center issued an alert concerning a RSV epidemic in week 18 in 2021 [22]. Our trend analysis using Joinpoint regression indicated that the RSV epidemic started earlier, at around week 11. If the Google Trends database had been used to monitor the RSV trend, a timelier warning might have been issued to the public health sector.
Although Google Trends analysis has important implications for early epidemic detection, the peak of the epidemic curve in Google Trends was higher than in the surveillance data for 2021. We suspected that this discrepancy was affected by the public’s interest in the RSV epidemic, which the search volume on Google Trends reflects. Additionally, in 2020, a small peak was observed in the Google Trends data at week 9 while no peak was observed in the surveillance data. A small peak of this sort, which was apparently unrelated to any disease trend, may create the false impression that an epidemic is imminent. No epidemic surge in the sentinel surveillance was observed in 2020 owing to the governmental strategy for dealing with the COVID-19 pandemic (declaration of emergency status), including sheltering at home and closing schools. [23,24,25]. Moreover, topics of public interest, such as the announcement of a new drug or vaccine for RSV, will likely affect the search volume on Google Trends, undermining the reliability of the findings. Thus, Google Trend analysis is not an infallible method of predicting an infectious disease epidemic. Further studies are needed to evaluate the advantages and disadvantages of Internet search engine query data pertaining to other diseases and in other countries.
The present study had some limitations. First, the generalizability of the findings to other countries and regions was not evaluated. However, previous studies of Google Flu Trends demonstrated the service’s utility in the United States and Europe [10, 11, 13, 16, 17]; we may therefore expect a similar utility in predicting RSV trends. Second, we were able to obtain only a “relative” search volume because the “actual” search volume of Google Trends data is not available to the public. If the total number of Internet searches was very small, the results of an analysis of Google Trends data might become susceptible to over- or underestimation.
In conclusion, our study found a strong correlation between the relative search volume on Google Trends and sentinel surveillance data on RSV infections. Additionally, the Google Trends database was found to be able to detect an increasing trend in RSV infections simultaneously with or even before the national surveillance system. With its wide availability and user-friendly interface, Google Trends will likely gain more attention for its utility as a surveillance system for infectious diseases even among patients and their guardians.