A, News and search trends for all reports or queries each day that included the term HIV (human immunodeficiency virus) since 2004 (news trends in blue and search trends in black). B, The same trend for 1 week around Charlie Sheen’s disclosure of his HIV-positive status (news trends increased 265% [95% CI, 24%-504%] and search trends increased 417% [95% CI, 384%-449%]). The dotted vertical line indicates the day of Sheen’s disclosure. C, Hourly trends for queries regarding condoms around the time of Sheen’s disclosure. The dotted vertical line indicates the hour of Sheen’s disclosure (7:00 AM Eastern Standard Time). D, Hourly trends for queries regarding HIV symptoms around the time of Sheen’s disclosure. The dotted vertical line indicates the hour of Sheen’s disclosure (7:00 AM Eastern Standard Time). E, Hourly trends for queries regarding HIV testing around the time of Sheen’s disclosure (all times Greenwich Mean Time). The dotted vertical line indicates the hour of Sheen’s disclosure (7:00 AM Eastern Standard Time). F, Effect estimates comparing observed vs counterfactual volumes for separate 24-hour periods up to 96 hours following Sheen’s disclosure. G, Effect estimates comparing observed vs counterfactual volumes for separate 24-hour periods up to 96 hours following Sheen’s disclosure. H, Effect estimates comparing observed vs counterfactual volumes for separate 24-hour periods up to 96 hours following Sheen’s disclosure.
Ayers JW, Althouse BM, Dredze M, Leas EC, Noar SM. News and Internet Searches About Human Immunodeficiency Virus After Charlie Sheen’s Disclosure. JAMA Intern Med. 2016;176(4):552-554. doi:10.1001/jamainternmed.2016.0003
Celebrity Charlie Sheen publicly disclosed his human immunodeficiency virus (HIV)–positive status on November 17, 2015. Could Sheen’s disclosure, like similar announcements from celebrities,1 generate renewed attention to HIV? We provide an early answer by examining news trends to reveal discussion of HIV in the mass media and Internet searches to reveal engagement with HIV-related topics2 around the time of Sheen’s disclosure.
News trends, gathered via the Bloomberg Terminal (http://www.bloomberg.com/professional/), included counts of global English-language reports with the term HIV. Internet searches, gathered via Google Trends (https://google.com/trends/), included counts of searches originating from the United States for 4 categories: HIV (all searches with “HIV”), condoms (“condom” or “condoms”), HIV symptoms (“symptom,” “symptoms,” or “signs of” and “HIV”), and HIV testing (“test,” “tests,” or “testing” and “HIV”). News about HIV was measured per 1000 reports. To compare relative search popularity, queries about HIV were divided by the total number of searches for each day or hour and then scaled to range from 0 to 100. Approximate raw search counts were from Google Adwords (http://www.google.com/adwords/). All measures from January 1, 2004 (when Google began data collection), through November 24, 2015, were analyzed. As these analyses are based on public meta-data, institutional board review was not required.
Our approach was quasi-experimental, comparing outcomes around Sheen’s disclosure with counterfactual estimates for expected news about and search volumes for HIV if Sheen’s disclosure had not occurred. Data analysis was conducted from November 17 to December 8, 2015. Expected volumes were calculated using an autoregressive integrated moving average algorithm by Hyndman and Khandakar3 and data from all days since 2004 for daily analyses or all hours in the 4 days preceding Sheen’s disclosure for hourly analyses (both after excluding 1-day or 24-hour washout when Sheen’s HIV status was rumored). The ratio of observed and counterfactual volumes with t tests was computed using R, version 3.2.1 (R Foundation for Statistical Computing).
Since 2004, news reports about HIV decreased from 67 (95% CI, 57-77) stories per 1000 to 12 (95% CI, 11-13) stories per 1000 in 2015. The day of Sheen’s disclosure, however, coincided with a 265% (95% CI, 17%-500%; P < .01) increase in news reports mentioning HIV (97% of which also mentioned Sheen), with more than 6500 stories on Google News alone, which placed Sheen’s disclosure among the top 1% of HIV-related media days in the past 7 years (Figure).
Sheen’s disclosure also corresponded with the greatest number of HIV-related Google searches ever recorded in the United States. About 2.75 million more searches than expected included the term HIV, and 1.25 million searches were directly relevant to public health outcomes because they included search terms for condoms, HIV symptoms, or HIV testing.
All searches regarding HIV were 417% (95% CI, 383%-450%; P < .001) higher than expected the day of Sheen’s disclosure. Hourly trends suggested that searches related to condoms were 72% (95% CI, 10%-171%; P < .048) higher than expected 24 hours after Sheen’s disclosure, first increasing the evening following the disclosure and remaining higher for 48 hours. Searches related to HIV symptoms were 540% (95% CI, 332%-808%; P < .001) higher than expected and searches related to HIV testing were 214% (95% CI, 151%-358%; P < .001) higher, both of which spiked the hour of Sheen’s disclosure and remained higher 72 hours later.
Just as with celebrities Rock Hudson’s and Magic Johnson’s disclosures of their HIV-positive status,4 Sheen’s disclosure may be similarly reinvigorating awareness and prevention of HIV. News and search trends are only proxies for awareness and healthy behaviors, respectively, and sentinel surveillance (such as rates of HIV testing) is needed to clarify these early findings. Yet, our findings can still inform practice, building on theories where availability of and seeking information foreshadow action.5 Efforts to leverage the “Charlie Sheen effect,” such as coordination of prevention-focused press with coverage related to Sheen as with celebrity Angelina Jolie’s prophylactic mastectomy,6 may be prudent.
While no one should be forced to reveal HIV status, Sheen’s disclosure may benefit public health by helping many people learn more about HIV infection and prevention. More must be done to make this benefit larger and lasting.
Corresponding Author: John W. Ayers, PhD, MA, Graduate School of Public Health, San Diego State University, 9245 Sky Park Ct, Ste 230, San Diego, CA 92123 (firstname.lastname@example.org).
Published Online: February 22, 2016. doi:10.1001/jamainternmed.2016.0003.
Author Contributions: Dr Ayers had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Ayers, Althouse, Dredze, Leas.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: All authors.
Conflict of Interest Disclosures: Drs Ayers and Althouse reported sharing an equity stake in Directing Medicine LLC, which advises how to implement some of the methods embodied in this work. Dr Dredze reported receiving consulting fees from Directing Medicine LLC and Sickweather LLC, who use social media for public health surveillance. The methods described in this article are not proprietary. No other disclosures were reported.
Additional Contributions: Kurt Ribisl, PhD, and Diana Francis, BA, University of North Carolina, provided feedback but were not compensated for their contribution.