Seasonal Variation in the Internet Searches for Cancer Recurrence: An Infodemiological Study

Abstract

Background While few clinical and epidemiological studies have assessed how seasonality affects cancer recurrence, it has not been studied with the utility of the internet data. In this study, we aim to test whether cancer recurrence presents seasonality on a population level, utilizing internet search query data.Methods This infodemiological study used Google Trends to find query data for the term "cancer recurrence" from January 01, 2004, to December 31, 2018 in the USA, the UK, Canada, and Australia. Time series seasonal decomposition and the cosinor analysis were used to analyze and describe the seasonal trends for cancer recurrence.Results A general upward trend in UK and northern hemisphere were observed. Statistically significant seasonal trends on "cancer recurrence" in the USA (p=1.33×10 -5 ), the UK (p=0.012), and northern hemisphere (p=5.67×10 -7 ) were revealed by cosinor analysis, with a peak in early summer and nadir in early winter.Besides, a seasonal variation was also found in Australia (p=2.3×10 -4 ), with a peak in late summer and nadir in late winter. ConclusionsThe evidence from internet search query data showed a seasonal variation in cancer recurrence, with a peak in early summer(northern hemisphere)/late summer(southern hemisphere). Besides, the relative search volume of "cancer recurrence" appeared a general upward trend in UK and northern hemisphere in recent years.