"Swine Flu" vs "H1N1" terminology - tweets show that people do not adopt the new term

In a previous post I already gave an example of the kind of research we are doing using the Infovigil system - an infodemiology/infoveillance system, which we are developing in cooperation with public health agencies *.
As described in this article, one of the applications of infodemiology (aside from identifying emerging epidemics) is to measure and track knowledge translation and dissemination. When the WHO as well as public health officials around the world dropped the terminology "swine flu", replacing it with "H1N1", on April 29th 2009, I was interested to see whether and how quickly this terminology would be adopted by the public - and ideal test case for a question which could be answered by infodemiological methods.

Figure: H1N1 term (red) versus Swine Flu/Swineflu (blue) terms in tweets (yellow: both). (c) Gunther Eysenbach, licensed under Creative Commons Attribution license (CC-by V2)

The figure above depicts the terminology used in tweets during the first 3 weeks in May, as a proportion of all tweets containing either "swineflu", "swine flu", or "h1n1" (including hashtags). On May 1st, less than 10% of the tweets contained only H1N1 (but not swineflu or "swine flu") (red), and 7% contained both concepts (H1N1 and [swineflu or "swine flu"]) (yellow).
As could be expected, the adoption of the term H1N1 gradually increased during the first week of May, and peaked on May 8th, when 24% of all tweets contained only H1N1 (and not swineflu or swine flu), while the term "swine flu" (or swineflu) alone was used in "only" 63% of the tweets. However, since then, the "swine flu" term experienced a sudden "revival", with the "swine flu" (or swineflu) term alone again used in 74-79% of the tweets, and "H1N1" hovering around 16-20%, without any clear trend of the term's use increasing.

We are still analyzing what happend around May 8th/9th to cause the trend to reverse (suggestions welcome). One working hypothesis is that interest in the epidemic (and the number of tweets) has declined significantly since around that date, so the nature and content of tweets has changed as well.

Comment: As a public health researcher, I am sympathetic of the reason for the name change and disappointed (but not surprised) by the publics' (in particular journalists') reluctance to adopt the name change. Studies have shown over and over again, that terminology matters: one of my favorite studies shows that changing the diagnosis "chronic bronchitis" to "smokers lung" actually encourages people to stop smoking (Brandt et al., 1997). The WHO had good reasons for suggesting a terminology change: The inherent threat in using the term "swine flu" is that people translate this into the wrong preventive behaviour (such as avoiding pork). If opinion leaders (in particular the media) would have adopted the name change, then - I assume - the public would have had an easier time to follow.
But I guess it is true what people say about change: Only as babies in our wet diapers we embrace change. It is the same reluctance to change which prevents new medical and scientific evidence to penetrate into medical practice and into the consciousness of the public in a timely manner (it takes years to change medical practice even if new research findings are out). And it is this reluctance to change which we hope to visualize with our line of infodemiology work which focuses on knowledge translation questions.

---

* Infovigil is NOT funded by Google and has no relationships to Google Flutrends - they gladly took our ideas but didn't collaborate.

Reports on Twitter Fueling H1N1/Swine Flu Fear and Misinformation Are Vastly Overstated

Is Twitter fueling panic and misinformation? (Image Source: Posible caso de influenza, (C) Hello32020, licensed under CC-by license)

Recent media reports (e.g. on CNN and USA Today) suggest that microblogging sites such as Twitter are fueling an epidemic of misinformation, disseminating rumor and speculation about the H1N1 (Swine Flu) outbreak.
As infodemiology scholar these media reports struck me as largely anecdotal, not based on systematic evidence.
Using the infovigil system (which allows archiving and analysis of tweets and other sources on the Internet) we decided to address this issue more systematically.

Number of Tweets collected by the Infovigil system between May 1st and May 18th, 2009, containing the keywords or hashtags H1N1, swineflu, or Swine Flu (Image Source: Gunther Eysenbach, University of Toronto, licensed under Creative Commons Attribution license V2.0 CC-by)


My student Cynthia Chew and I just finished a preliminary content analysis of 400 randomly selected tweets (sent during the first weeks of the outbreak), and found that her data do not support the notion of twitter spreading misinformation. As she wrote in her abstract submitted to the Medicine 2.0'09 conference in September in Toronto, she found only 7/400 (1.75%) cases of misinformation in tweets.
News posts were the most common type of information shared (46%) followed by public health education (19.18%) and H1N1-related humour (18.25%). 36.75% of all posts quoted news articles verbatim and provided URLs to the source.
Take that, CNN and USA Today!


Gunther Eysenbach MD MPH
Senior Scientist, Centre for Global eHealth Innovation, Director, Consumer Health Informatics, Public Health Intelligence & Infoveillance Group

Patient empowerment and power distance, health care safety and plane crashes

I just finished reading a fascinating book chapter "The ethnic theory of plane crashes" in Malcolm Gladwells book Outliers. Malcolms argument is that

The single most important variable in determining whether a plane crashes is not the plane, it's not the maintenance, it's not the weather, it's the culture the pilot comes from. Planes are flown safely when the pilot and co-pilot are in open and honest communication. And in cultures where it is difficult for a junior person to speak openly to a superior, you have lots of plane crashes.
(Source: CNN Interview with Gladwell)

As summarized by Rob Verger:

Gladwell explores two plane crashes—one Colombian (Avianca Flight 52) and another, South Korean (Korean Air Flight 801)—and how the culture of the pilots perhaps contributed to each disaster. He focuses on how well the pilots communicated with each other and with air traffic control. Poor communication in these examples, he argues, has to do with something called a culture’s Power Distance Index (P.D.I.)—the term and concept come from psychologist Geert Hofstede—which is a measurement of “how much a particular culture values and respects authority,” as Gladwell defines it. Countries with a high P.D.I. generally value being more deferential towards authority, and thus not contradicting a superior (the U.S. and New Zealand both have a low P.D.I.). Gladwell argues that since both Colombia and South Korea rank towards the top of the P.D.I. list, the subordinate members of their cockpit crews were unable or unwilling to speak up as assertively as they should have about safety concerns.

Gladwell retells the story of Korean Air, which in the 1990s was plagued by a series of plane crashes. Investigators discovered that when Koreans spoke to each other in Korean, they "were trapped in roles dictated by the heavy weight of their country's cultural legacy." That meant that they were hesitant to challenge a superior directly. According to Gladwell, a number of plane crashes are the direct consequence of "power distance", resulting in miscommunication and eventually disaster.

As a health care researcher and patient advocate I was intrigued by the concept of power distance and the notion of measuring relationships between culture and safety.
And of course, being a professor in health policy and ehealth, I started thinking about the relationship between health care culture and quality/safety of care.

I would argue that in health care, the relationship between patient and health care professional is sometimes akin to the relationship between pilot and copilot. Patients need to speak up and need to be engaged to utter concerns or preferences to avoid disaster. Power distance, which is "the extent to which the less powerful members of organizations and institutions accept and expect that power is distributed unequally" is a barrier often impeding patient-doctor commmunication.

I also hypothesized that power distance is an inverse predictor for patient empowerment: The more consumers accept that there is and that there should be a "power distance" between patients and doctors, the less advocacy in terms of patient rights and access to information will take place, and public policy will be slower to adopt an empowered patient model.


To test this hypothesis, I correlated data (total scores) from the Euro Patient Empowerment Index, which scores patient rights, information access and financial incentives on a health policy level, with Hofstede's Power Distance Index - and found a surprisingly high inverse correlation between the two metrics.
Countries with a high Power Distance Index such as Poland also tend to have low Patient Empowerment scores, while countries such as Denmark, where there is less "fear of authority", also score high on the patient empowerment metric.


What does this all mean for health care policy? Well, if you believe Gladwell's argument that power distance is a predictor for aviation disasters, then it is very feasible to make the analogous argument in health care: that high power distance and low patient empowerment is often a cause for miscommunication with disastrous outcomes in health care.
And obviously there is a relationship to my other research area, ehealth, in that the Internet and the virtues of Medicine 2.0 (participation and openness, collaboration and disintermediation/apomediation) are powerful antidotes to reduce the power distance in health care between patients and doctors, but also between health care professionals (e.g. junior and senior doctors).

I am intrigued by these relationships between culture, policy, and health care quality, and I am trying to get my head around how to do further research in this area.

To avoid "ecological fallacies", such studies presumably must be undertaken at the individual level. So perhaps what we need to develop next is a patient empowerment index that can be administered at the individual rather than policy level, and which can be used to study relationships between empowerment and outcomes.