Original ResearchTracking search engine queries for suicide in the United Kingdom, 2004–2013
Introduction
Suicide rates have long been viewed as an early warning sign of wider social, political, or economic distress.1, 2, 3 Yet delays inherent in collection and analysis of suicide surveillance often take several years before such data become available. For example, suicide rates for 2013 in England and Wales were not published until February 20154 while, in other countries, delays are even longer. The most recent data for the entire Belgian population are from 2008.5
Recently efforts have been made to take advantage of new data created by people's engagement with social media and search engines. Infectious disease epidemiologists have drawn on internet search data to provide early warning of potential outbreaks,6,7 such as by tracking search engine queries relevant to infectious diseases, including influenza,8 HIV,9 and gastroenteritis.10 These applications have now gone beyond infectious disease, from monitoring interest in abortion11 to electronic cigarettes.12
Prior academic evidence suggests that this approach may also be fruitful for mental health and suicide epidemiology.13,14 Using web search engines for premeditating suicide is not unheard of in the general public. Even in the case of the tragic Germanwings plane crash in the French Alps this spring, German prosecutors revealed that the pilot, Andreas Lubitz, had conducted internet research on ‘cockpit doors’ and ‘suicides’ days before the crash.15
However, for search data to be useful in population health research, there are remaining questions about their validity. In particular, there are concerns about representativeness of data, especially if they fail to capture the intentions of vulnerable groups that may not have routine access to search engines. A recent American study by McCarthy14 matched Google internet search engine activity to US suicide rates reported by the Centers of Disease Control (CDC), finding a strong negative correlation with suicide rates in the general US population but a strong positive correlation with rates among youth. McCarthy14 also found a cyclical trend in search activity for suicide and searches for suicide risk factors (e.g. depression) that correlated with seasonal patterns in suicides, but only in individuals aged 15–24 years of age. In addition, there was no acknowledgement that search activity may vary by gender. Lastly, no tests were employed to determine the significanceof association between search activity and suicide rates.
Thus far, McCarthy's results have not been replicated in other countries and specifically, Western industrialised countries. In this paper we have two specific aims. First, we sought to determine if the cyclical trend seen in the USA was also evident for search activity related to suicide and three common suicide risk factors in the United Kingdom: depression,16, 17 unemployment,2, 18, 19, 20, 21 and marital strain.22, 23, 24 Second, we test the validity of suicide search data as a potential marker of suicide risk by evaluating whether web searches for suicide associate with suicide rates among those of different ages and genders in the United Kingdom.
Section snippets
Methods
We obtained search engine data from Google Trends, a publicly available repository of information of trends and patterns of user searches on Google. Data can be disaggregated by geographic region and time period although, clearly, they do not provide information on the demographic characteristics of those undertaking the searches. While Google Trends does not show the absolute number of searches, it calculates a query share for a search term. In practice, this means that Google calculates the
Results
The search activity for suicide, depression, job loss, and divorce were determined separately using Google Trends and averaged across years to determine any cyclical patterns in weekly search activity. Only monthly data, not weekly data, were available for searches of suicide in 2004 and 2005 and for unemployment-related searches in 2004. Therefore, we excluded those years when determining weekly trends in search activity. Both suicide (Fig. 1) and depression (Fig. 2) exhibited a recurring
Discussion
Google Trends is a publicly available and novel epidemiological tool that has been used to study population health and behaviour in a variety of fields, such as infectious and chronic disease. In this paper, we set out to apply this dataset to mental health and suicide epidemiology in the United Kingdom working age population. More specifically, the primary goals were to determine how well searches for suicide correlated with suicide rates (per 100,000 population) in the UK, and whether this
Limitations
Our findings have limitations and caveats. The most obvious limitation is that Google Trends is collecting anonymised search information on a regional or country level. Unlike a traditional survey, we are not able to utilise methods that allow us to determine, for example, how these relationships may differ according to household characteristics or socio-economic groups. Because these data are anonymised, we are not able to analyse the relationship between high suicide risk individuals and the
Acknowledgements
We would like to thank Google and the Office of National Statistics (ONS) for releasing access to search data and suicide data, respectively, in the public domain.
Ethical approval
None sought.
Funding
DS is funded by a Wellcome Trust Investigator Award and ERC Research Grant 313590-HRES.
Competing interests
None.
Declarations
None.
References (50)
- et al.
Tracking the rise in popularity of electronic nicotine delivery systems (electronic cigarettes) using search query surveillance
Am J Prev Med
(2011 Apr) Insights on unemployment, unemployment insurance, and mental health
J Health Econ
(2011 Mar)Internet monitoring of suicide risk in the population
J Affect Disord
(2010 May)- et al.
Increase in state suicide rates in the USA during economic recession
Lancet
(2012 Nov 24) - et al.
The public health effect of economic crises and alternative policy responses in Europe: an empirical analysis
Lancet
(2009 Jul) - et al.
Austerity: a failed experiment on the people of Europe
Clin Med Lond Engl
(2012 Aug) - et al.
Seasonality in seeking mental health information on Google
Am J Prev Med
(2013 May) Vitamin D, light and mental health
J Photochem Photobiol B
(2010 Nov 3)- et al.
Distribution of the vitamin D receptor and 1 alpha-hydroxylase in human brain
J Chem Neuroanat
(2005 Jan) - et al.
Vitamin D and the brain
Best Pract Res Clin Endocrinol Metab
(2011 Aug)
Vitamin D, effects on brain development, adult brain function and the links between low levels of vitamin D and neuropsychiatric disease
Front Neuroendocrinol
Developmental vitamin D deficiency causes abnormal brain development
Psychoneuroendocrinology
Seasonality and its distinct clinical correlates in bipolar II disorder
Psychiatry Res
Seasonality and bipolar disorder: a systematic review, from admission rates to seasonality of symptoms
J Affect Disord
Prevalence and correlates of binge eating in seasonal affective disorder
Psychiatry Res
Exposure to websites that encourage self-harm and suicide: prevalence rates and association with actual thoughts of self-harm and thoughts of suicide in the United States
J Adolesc
Economic suicides in the great recession in Europe and North America
Br J Psychiatry J Ment Sci
Suicides associated with the 2008-10 economic recession in England: time trend analysis
BMJ
The effect of the late 2000s financial crisis on suicides in Spain: an interrupted time-series analysis
Eur J Public Health
Suicides in the United Kingdom, 2013 registrations
Digital disease detection — harnessing the web for public health surveillance
N Engl J Med
The use of google trends in health care research: a systematic review
PLoS ONE
Monitoring influenza activity in the United States: a comparison of traditional surveillance systems with google Flu trends
PLoS ONE
Predicting new diagnoses of HIV infection using internet search engine data
Clin Infect Dis
Cited by (32)
Internet Search Engine Queries of Common Causes of Blindness and Low Vision in the United States
2021, American Journal of OphthalmologyGoogle Trends: Opportunities and limitations in health and health policy research
2019, Health PolicyCitation Excerpt :Over time, Google search engine data has increasingly been utilized to understand health behaviours. For example, it has been shown that changes in the volume of suicide-related searches may provide an early warning of changing mental health risk [11–13], although the association is strongest for suicides among younger people and middle-aged women, both groups more likely to take overdoses (and thus require information on how to do it) than among older men, among whom hanging is more common [14]. A related study examined the commonly held view that media coverage of celebrity suicides can either increase or decrease suicidal ideation, finding limited evidence for both, but only with the most prominent celebrities [15].
A Novel Framework to Forecast COVID-19 Incidence Based on Google Trends Search Data
2024, IEEE Transactions on Computational Social SystemsThe association of internet searches and actual suicide in Spain
2023, Journal of General PsychologyDeath Attitudes and Life Goals of Adolescents Interested in Death Content on Social Networks
2023, Lecture Notes in Networks and SystemsKorean adolescent suicide and search volume for “self-injury” on internet search engines
2023, Frontiers in Psychiatry