Nathan Lawrentschuk, Division of Urology, Department of Surgical Oncology, Princess Margaret Hospital, 610 University Avenue 3-130, Toronto, Ontario M5G 2M9, Canada. e-mail: email@example.com
What’s known on the subject? and What does the study add?
The media is a powerful tool and the policy makers as well as the public are highly susceptible to media reporting of events. The interpretation of the two large randomised prostate cancer trials (ERSPC and PLCO) published in the NEJM in 2009 were conflicted in their results regarding the benefits of screening. What is agreed upon is that the methodology for the PLCO trial was somewhat flawed but also that both trials were reported on too early – prostate cancer has an extraordinarily long leadtime and time to mortality from diagnosis.
The paper adds valuable data for the first time a comprehensive prospective analysis of the established print media (online version) reporting across the UK versus north America (USA and Canada) and also Australasia as well as online-only reporting observed the reporting of the trials and how they interpreted the conflicting study results from March 2009. Interestingly geographical bias as anticipated was reported-North America media against screening predominantly and europe (UK) for it whilst Australasia rested between. Urological and related bodies need to be aware of such geographical media bias in when and how they report and publish articles and the imapact they may ultimately have on public awareness and health policy.
• The publication of two large screening studies for prostate cancer (CaP), the Prostate Lung Colorectal Ovarian Cancer (PLCO) and the European Randomized Study of Screening for Prostate Cancer (ERSPC), has generated intense interest in medical and lay press not only as a result of their robust size, but also their opposing outcomes and differing methodologies, making interpretation controversial.
• To characterize the world online media response to the studies by assessing reports for quality and message, as well as noting geographical differences.
MATERIALS AND METHODS
• Major newspapers in North America, UK and Australia reporting online and Internet-only news organizations were analyzed for their reporting of CaP screening in response to the trials for a period of 6 months post-release.
• Content, positive or negative projection regarding screening, and use of expert commentary were recorded.
• Statistical analysis of the results was then undertaken.
• In total, 48 newspapers reported the CaP screening studies with a median (range) publication time for newsprint online of 1.5 (0–175) days and same day appearance for online news sources in the range 0–110 days.
• Only 23% of newsprint articles indicated that screening was a positive endeavour, whereas 31% were negative and the remainder were neutral (46%).
• Some 78% of UK articles indicated insufficient screening, whereas 57% in the USA and 80% in Canada reported screening as being excessive. Online media reflected USA reporting.
• World newsprint media in general portrayed screening in a negative light after publication of the ERSPC and PLCO studies.
• North American media concluded that prostate-specific antigen (PSA) screening was excessive, whereas the UK media indicated that an inadequate level of PSA screening is occurring.
• The media influences public opinion and government policy and it is important that urological organizations are aware of the true impact.
European Randomized Study of Screening for Prostate Cancer
Prostate Lung Colorectal Ovarian Cancer
The publication of the two large population-based screening studies for prostate cancer (CaP) in 2009, the Prostate Lung Colorectal Ovarian Cancer (PLCO)  and the European Randomized Study of Screening for Prostate Cancer (ERSPC)  screening trials, has generated intense interest from both the medical and lay press. Attention to these studies stemmed not only from their robust size and randomization, but also from their opposing outcomes and different methodologies. Therefore, the interpretation of the results has been variable and somewhat controversial.
Even before the creation of the Internet, most individuals, including physicians and scientists, first learnt about developments in medicine through the media . Press coverage of medical stories, however, has often been characterized as inaccurate, superficial or sensationalized . In the field of urology, previous studies have indicated that misrepresentations and inaccuracies occur in media reports on CaP screening  and, alternately, that media hype has even drifted into the medical literature . Misinterpretations are not surprising and often are compounded by urologic oncology websites (often created by urologists) that lack WHO standards and quality control . Over the last few years, the general public has turned to the Internet rather than hard copy news media . It is thus imperative that urologists and their peak professional organizations remain at the forefront of urologic oncology and the debates that ensue within the public arena.
The present study aimed to characterize the response of world newsprint delivered through online media and also from online only news agencies to the ERSPC and PLCO screening studies by assessing articles for their quality and underlying message. Geographical differences are also identified given that the two screening studies originate from different continents.
MATERIALS AND METHODS
ESTABLISHING QUALITY NEWS SOURCES ON THE INTERNET
Initial research was conducted using the search engines Google (http://www.google.com), Yahoo (http://www.yahoo.com) and Wikipedia (http://www.wikipedia.com) to establish the most commonly accessed news sources from the top twenty ranked newspapers in the USA, UK, Canada, Australia and New Zealand, which all appear in an online version on the Internet. The top ten worldwide Internet-only (i.e. never printed newspapers) news sources were determined by the volume of visits to the websites. The search engines, Google and Yahoo, were chosen because they are used by 85–95% of the population who conduct searches on the Internet . Wikipedia was chosen for its accuracy in ranking news outcomes and recent event analyses such as newspaper circulation and visits to websites .
MONITORING OF NEWS REPORTING ON THE INTERNET
Online information gathering was undertaken daily for the first month from the release of the trials (19 March 2009) and then weekly for 6 months. Specifically, the search terms used for the analysis were : ‘newspaper’, ‘news’, ‘media’, ‘prostate’, ‘cancer’, ‘screening’, ‘prostate-specific antigen’, ‘European’, ‘American’, ‘ERSPC’, ‘PLCO’, ‘study’, ‘USA’, ‘United States’, ‘Australia’, ‘New Zealand’, ‘Canada’, ‘Scotland’, ‘Wales’ and ‘Northern Ireland’. These terms were entered into Google and Yahoo as independent searches so that no news articles were missed. The major newspapers (ranked in the top twenty by circulation as outlined above) in the USA, Canada, UK, Australia and New Zealand as well as Internet-only news organizations were also analyzed in the same searches for their reporting of CaP screening in response to the two major screening trials (ERSPC and PLCO) released 18 March 2009. The reports had to be related to CaP screening in the context of initial reporting of one or both of the large randomized controlled trials for CaP screening. News stories that resulted from the initial reports of the two studies (i.e. the cascade of news articles in response to the initial reports of the studies) were ignored because they were either political, specialist group or individually motivated, and did not focus on the CaP screening studies.
Articles were analyzed for basic demographics: date of publication on the Internet (from GMT 12.00, 18 March 2009); content (author gender, health or general reporter, use of expert commentary and profession if supplied). Use of terminology related to CaP such as, ‘PSA test’, ‘screening’, ‘mortality’ and if the article quoted or mentioned the ERSPC or PLCO trial (or both) were all quantified variables. Whether the overall message was positive or negative toward screening was also recorded.
News stories that were gathered were reviewed by two urologic oncologists, a medical student and two lay people for content and, importantly, for the perceived tone and general message of the byline as well as the article itself. The articles were rated as either ‘positive’, ‘negative’ or ‘neutral’ with respect to Cap screening from the byline and article content itself. Specific reference to PSA ‘screening’ or ‘testing’ as excessive, insufficient or not commented upon were also recorded. Concordance was reached in 88% cases, with 12% being decided by consensus after re-reading of the stories by a fifth lay person. Other data were factual (e.g. date of publication) or the number of times words such as ‘screening’, mortality’ or ‘test’ were utilized and required no interpretation.
Data were entered into a spreadsheet and analyzed using Graphpad Prism 4 (GraphPad Software Inc., La Jolla, CA, USA). Articles from different continents were compared using the Mann–Whitney U-test as a non-parametric test for assessing whether two independent samples of observations come from the same distribution. Where an item was being tested, proportions were used to correct for article length.
In total, 48 online news organizations carried a report on the CaP screening studies, with 37 of these originating in print press and then being published online. The remaining 11 derived from online-only news organizations (Tables 1 and 2). The median time for a print press article to appear on the Internet was 1.5 days after the studies were published (range 0–175 days). By contrast, for online news organizations, the median appearance on the Internet was the same day as the release of the trials (range 0–110 days). The median number of words devoted to articles overall was 577. Online news organizations printed longer articles (671 words) compared to newsprint online (534 words), with no significant difference between the two modalities (P= 0.59). Two-thirds were confirmed as specialist health reporters, with status being unclear in one-fifth of reports. The gender of journalists was in favour of males. Neither the gender of the reporter, nor whether the reporter specialized in health had any impact on the overall message (P > 0.05).
Table 1. Results of the analysis of the online media articles reporting in response to the release of the two major randomized prostate cancer screening trials
Includes highest ranked internet news service in each region.
One article mentioned urologist and family doctor. P+, positive; N-, negative. ERSPC, European Randomized Study of Screening for Prostate Cancer; PLCO, Prostate Lung Colorectal Ovarian Cancer.
Mortality rate for prostate cancer (yes given/total articles)
Experts quoted, median (range)
Table 2. Univariate analysis of article results by geographical region
Article length (words)
Percentage population reached
Overall message screening
View on PSA screening
Use of word ‘test’
Use of word mortality
Use of word screen
All data are P-values.
North America vs UK–Australia/New Zealand
North America vs UK
USA vs Canada
USA vs UK
USA vs UK–Australia/New Zealand
Canada vs UK
Canada vs Rest of World
Internet vs North America
Internet vs UK–Australia/New Zealand
Internet vs USA
Internet vs Canada
Internet vs UK
The overall message regarding screening as taken from the article byline was generally negative (22/48) and less often positive (10/48), with the remainder being neutral (16/48). Article bylines reflected these messages in almost every case compared to the overall article message (i.e. if the byline appeared positive, so was the actual article; P= 0.91). More specifically, PSA screening was represented as excessive in 26/48 compared to 12/48 news organizations, whereas a minority (10/48) remained neutral regarding this topic. The use of key words and phrases such as ‘death/mortality’, ‘PSA screening’ and/or ‘PSA testing’ were largely consistent across the different regions. However, the use of overall CaP mortality figures to place screening in context was reported in only one-third of articles. Many articles quoted ‘experts’ in CaP with the median number of experts quoted being 3 (range 1–4). Most were identified as representing various cancer societies or patient advocacy groups, with only 15% of articles quoting an urologist.
Google news noted an insignificant peak in stories related to CaP in the month of March 2009 compared to months before and after (Fig. 1).
DIFFERENCES BASED ON GEOGRAPHY
Differences in news reporting based on geography are displayed in Table 2. Focusing on North America vs the UK/Australia/New Zealand, there were obvious differences. The UK reported only on the ERSPC in seven of nine articles, whereas North American agencies reported on both the ERSCP and PLCO trials in 22 of 24 publications. The overall message regarding screening also depended on geography, with the UK/Australia/New Zealand being in favour in seven of 24 reports and negative in four of 13 compared to North America where 12 of 24 reports were negative and two of 24 were positive (P= 0.03; again driven primarily by the UK articles). The bylines of the articles also reflected these differences (P < 0.001). This result was echoed by the perception of screening being considered excessive more often in North America, although this was not significant (P= 0.15). The use of words, phrases and mortality data was similar between different countries, as was the use of experts and reports quoting urologists.
Internet-only news agencies tended to view screening similar to the newsprint stories as they appeared online from North America and their overall message was negative regarding screening, with almost all significant differences between the Internet and other regions being between Internet news agencies and the UK only (Table 2).
A more detailed analysis of geographical differences is presented in Table 2 detailing comparisons between USA, Canada, UK and UK/Australia/New Zealand news sources. Most results reflected the differences based on North America and Non-North America as already outlined (i.e. Canada and US were very similar, whereas the UK was different and Australia/New Zealand was inbetween).
The PLCO  and ERSPC  screening trials have delivered a message of concern to urologists with respect to CaP screening-related suggestions of an over diagnosis and overtreatment of CaP. However, caution regarding PSA screening was recognized well before these trials accrued . The present study did not aim to resolve such issues but rather aimed to determine how the news media relayed the message from these trials to the public and whether there was any geographical differences in reporting. Against this background, the reporting of the large CaP screening trials presented a somewhat polarized message depending on geography. The UK articles were more in favour of screening compared to North American articles (USA and Canada) where screening was deemed excessive. Articles from Australia and New Zealand were balanced inbetween. The driving force may have been that with the UK being more connected to the European study (ERSPC), it placed less emphasis on the North American PLCO study.
In 2002, Yamey and Wilkes commented upon the ‘widespread belief in America; that every man should know his PSA status, yet they went on to assert that this was driven by politics and not evidence . Therefore, one possible explanation of different geographical reporting with respect to the screening trials relates to differing health systems. For example, regarding PSA testing, Cohen suggested that, ‘In the US, the angle for the press may well be, ‘Are you paying for something you don’t need?’ Whereas in the UK they will be questioning whether there’s a good thing that the UK National Health System (NHS) aren’t giving you’. Nevertheless, the PSA screening debate will remain and, despite its controversy, PSA testing remains widespread in both regions [14,15].
Differences in CaP medical reports in the media are well known with ≈10% of articles related to CaP being factually inaccurate or having claims not supported by the scientific literature . From this research, six key issues have been identified in media articles in the past as misrepresenting CaP screening: (i) mortality from CaP; (ii) expert agency support for screening; (iii) the efficacy of screening in preventing death from CaP; (iv) the importance of early detection; (v) the accuracy of the prostate-specific antigen test; and (vi) prevalence and severity of adverse effects from treatment . All of these issues were referred to in many of the news articles that were surveyed. The heterogeneity of reporting in reference to these issues did not allow for meaningful comparisons between news sources.
In the majority of case, the general adult population now receives their print news from the Internet . Therefore, this was considered to comprise the best forum for tracking the media’s reporting of the two large CaP screening studies. However, as Albert Einstein stated, ‘Information is not knowledge’ and this is a problem in our ‘information age’. This applies not only to the general public, but also to the media whom report to the general public. As urologists, we must be aware of the information that is being disseminated and, importantly, peak professional bodies representing urologists, such as the American Urology Association, should maintain a role in public education as they have advocated . Hence, it was a disturbing to find so few urologists or affiliated organizations being quoted in the articles reviewed. Public opinion helps shape health policy by influencing politicians through patient advocacy groups and the media. Clearly, we must continue our vigilance in the area if we are to remain relevant to the ongoing debate regarding CaP screening, as aptly stated in the editorial accompanying the two large trials: CaP screening is a ‘controversy that refuses to die’. We must also take a greater lead in ensuring quality and honest reporting on the Internet, which has thus eluded urologic oncology .
In medicine, inaccurate reporting can generate false hopes and unwarranted fears. The clinical aftermath of such inaccuracies emphasizes the need for a solution to the problem of misleading journalism . Certainly, it is easy for the press to be blamed by the medical profession and vice versa so that perhaps even the public may need to be cautious about interpreting reports. In any case, we agree that journalists and scientists share the responsibility for accurate communication to the public . In the case of the two large screening CaP studies, the media did not necessarily report inaccurately because these are complex trials that physicians are divided upon. However, it is apparent that geographical biases contributed to the overall message. Recommendations can be made to reduce miscommunications between the press and the medical world . For example, the availability of researchers who present and publish data should clarify and explain their findings to lay people. In addition, there should be open discussion of research findings, thereby preventing sensationalism that may be provoked by closed discussion.
One limitation of the present study is that it may have missed some news articles by focusing on the two largest search engines rather than using smaller ones. However, the search engines employed represent those that almost all of the adult general population will have used . In addition, it is possible that stories were posted and then removed in a short time, although, in the study period, this was not found to be normal practice, with online articles lasting months on average. Furthermore, it is not possible to directly correlate the broadcasting of media with the retention and incorporation into opinion. Although this is a logical concept, it is difficult to study without surveying huge populations. Moreover, biases such as personal factors (e.g. relative to CaP) probably affect the perception of an article, no matter what the intended message. Finally, the present study did not aim to investigate the impact of other media such as radio or television or printed newspapers in the domains covered to assess for alternate impact and reporting.
Overall, the world’s newsprint media portrayed screening in a negative light after the ERSPC and PLCO studies were published. The general consensus of articles originating in the USA and Canada was that screening practices are currently excessive and should be viewed negatively based on the two trials. The UK media emphasized the need for more screening, with Australia and New Zealand taking an intermediate stance. This may have been related to a bias in reporting only the ERSPC findings in the vast majority of UK articles. What was also obvious is the number of identifiable members of the urology community who were asked to give expert opinion on the trials being disturbingly low. The media has an enormous influence on public opinion and government policy, so that the intervention of specialist organizations such as the AUA in the media is critical. Objective and constructive contribution to the media regarding studies and controversial topics remains our professional duty.