On the whole, where STIs are concerned, social media have tended to be considered as a potential force for the good in public health, offering a new resource for the management of HIV patients, or opportunities for disseminating health messages via peer education (Swanton & Mullan (STIs); Peer group education (STIs/blog)).  Recently, however, there have been a number of studies that have drawn attention to the negative implications of social media.  Last June a study by Beymer & Morisky (STIs), based on data on MSM attendees at the Los Angeles Gay and Lesbian Centre, concluded that, among the 7,000 participants, those who had used geo-sexual networking apps to meet up with a partner had greater odds for testing positive for gonorrhoea (OR 1.25) or chlamydia (OR 1.37) than those who employed in-person methods.

Recently, this more negative side has been receiving ever more attention in the US, especially in connection with HIV transmission.  A yet unpublished but widely publicized study, Agarwal and Greenwood (A&G), investigates hospital attendances for asymptomatic HIV (including acute and silent phases of the infection)  in Florida over the period 2002-2006 when the piece-meal introduction of the digital commerce platform, Craigslist, appears to have greatly facilitated on-line social transactions through its “casual encounters” forum.  It has also offered researchers the chance to record what they describe as a “natural experiment”, as successive counties have experienced the effects of entry into the platform.  A&G estimate the health “penalty” of entry into Craigslist at a 13.5% increase in attributable HIV infections – equivalent in financial terms to an additional burden of $592 million on the State of Florida.   This finding has recently been cited in connection with the precipitous rise in STIs in Rhode Island recently reported in an official Rhode Island Goverment press release and in the press coverage (Huffington Post) – 79% in syphilis; 30% in gonorrhoea; 33% in HIV over the year 2013-2014.

But A&G are concerned with more than estimating the magnitude of the effect.   The recent paper also claims to be the first study to attempt to determine exactly where that penalty of increased HIV infection due to social networking is actually falling – a question that is evidently of great interest to public health specialists who need to be able to target their interventions.  On the face of it, this is something of a puzzle.  HIV appears to be most heavily concentrated amongst the very sectors of the population who are most digitally disadvantaged.  So what could be going on?  To answer this question, A&G seek to disaggregate the Craigslist effect by ethnicity, income-level (as determined by enrolment in Medicaid) and gender.  What emerges from their analysis is that the effect of Craigslist entry is contained almost exclusively within the Afro-Caribbean (as opposed to Latino or “Caucasian”) population.  A&G seek to explain this apparently disproportionate penalty accruing to the digitally disadvantaged.  They argue that the “digital divide” is probably not “binary”, but more like a continuum.  We should not, in other words, necessarily think of “digital disadvantage” – at least for an important proportion of the disadvantaged – in terms of the total absence of access or skill.  It is therefore conceivable that it should be associated with a negative effect, i.e. the increased HIV incidence following Craigslist entry.  “Digital disadvantage”, they argue, is likely to be a matter of the limited capacity to utilize on-line resources for “welfare-enhancing activities” rather than a total unavailability of those resources.