Article Text

‘It was a freak accident’: an analysis of the labelling of injury events in the US press
  1. Katherine C Smith1,2,
  2. Deborah C Girasek3,
  3. Susan P Baker1,4,
  4. Jennifer A Manganello5,
  5. Stephen M Bowman1,4,
  6. Alicia Samuels1,
  7. Andrea C Gielen1,2
  1. 1Johns Hopkins Center for Injury Research and Policy, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA
  2. 2Department of Health, Behavior and Society, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA
  3. 3Department of Preventive Medicine and Biometrics, Uniformed Services University of the Health Sciences, Bethesda, MD, USA
  4. 4Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA
  5. 5School of Public Health, Department of Health Policy, Management and Behavior, University at Albany, SUNY Rensselaer, NY, USA
  1. Correspondence to Katherine Clegg Smith, Johns Hopkins University, 624 N. Broadway, Room 726 Hampton House, Baltimore, MD 21205, USA; kasmith{at}


Objectives Given that the news media shape our understanding of health issues, a study was undertaken to examine the use by the US media of the expression ‘freak accident’ in relation to injury events. This analysis is intended to contribute to the ongoing consideration of lay conceptualisation of injuries as ‘accidents’.

Methods LexisNexis Academic was used to search three purposively selected US news sources (Associated Press, New York Times and Philadelphia Inquirer) for the expression ‘freak accident’ over 5 years (2005–9). Textual analysis included both structured and open coding. Coding included measures for who used the expression within the story, the nature of the injury event and the injured person(s) being reported upon, incorporation of prevention information within the story and finally a phenomenological consideration of the uses and meanings of the expression within the story context.

Results The search yielded a dataset of 250 human injury stories incorporating the term ‘freak accident’. Injuries sustained by professional athletes dominated coverage (61%). Fewer than 10% of stories provided a clear and explicit injury prevention message. Stories in which journalists employed the expression ‘freak accident’ were less likely to include prevention information than stories in which the expression was used by people quoted in the story.

Conclusions Journalists who frame injury events as freak accidents may be an appropriate focus for advocacy efforts. Effective prevention messages should be developed and disseminated to accompany injury reporting in order to educate and protect the public.

  • Journalism
  • mass media
  • newspaper
  • qualitative research
  • media
  • qualitative

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


The use of ‘accident’ to describe injury events has been problematised by many injury experts because it is thought to convey notions of randomness and unpredictability, thus discouraging injury prevention initiatives.1–3 Debate over this issue continues, despite mounting evidence that such concerns may be misplaced. Public surveys conducted in Brazil, Canada and the USA have all demonstrated that the expression ‘accident’ does not appear to be associated with an assessment that an injury-producing event was unpreventable.4–6 We examine how the news media use ‘freak accident’ in covering injury events. We will consider what meaning this expression conveys in the context of print journalism and what objectives are served by its use.

News organisations are influential in shaping societal discourse on health issues. News media are a significant source of health information and have been associated with important behavioural and policy changes.7–11 In relation to injury, news media contribute to defining what events and issues should be considered ‘risky’12 and where and with whom risk lies. Scholarly work documenting news media coverage of injuries is, however, still in its infancy, and the relationship between media framing and public attitudes towards injury topics has not been sufficiently explored.13–15

To our knowledge, this is only the second examination of media characterisations of ‘accidents’12 and is the first to consider any modifier of the expression ‘accident’ in relation to injury events. This study is therefore intended to be entirely exploratory, aiming to generate hypotheses and avenues for further research rather than definitive findings. Our exploration is shaped by the following research questions:

  • What types of injury events are described as ‘freak accidents’ in the US news media?

  • In the news media environment, who uses the expression ‘freak accident’ to describe injury events?

  • Do ‘freak accident’ news stories convey prevention information?

  • What purpose does using the expression ‘freak accident’ seem to serve for those who use it?


Qualitative sample construction

LexisNexis Academic ( was used to collect news data. A small and purposive sample of three US news sources (Associated Press (AP), New York Times and Philadelphia Inquirer) that serve as a proxy for the wider news media was compiled for the purpose of this study. We selected AP because of its considerable influence as a story source for print and other news media in the US context and beyond. We added two daily newspapers because they were easily accessible and because their mandate includes coverage of local events, which would include injury events in their locality. We used a single search term ‘freak accident’, identifying stories that included this expression anywhere in the text. The initial dataset included 400 stories from which 150 were excluded because they dealt with fictional injuries (such as reviews of novels or plays), non-human injuries (such as an injury to a race horse) or an ‘accidental’ event that was unrelated to injury. Distinct stories of a single injury event (such as a developing story over time) were retained, but exact replica stories on the same event in one outlet were excluded.

Text analysis

A text analysis was conducted that combined structured and open coding. First, story subsets were collectively reviewed by the authors and a consensus as to areas and issues of interest was developed. During this initial review it was discovered that many stories focused on injuries sustained by professional athletes. This finding guided further development of appropriate and detailed coding categories for sporting injuries in addition to overall coding.

To facilitate formal coding, author A developed a codebook that included eight quantitative and two qualitative measures. The codebook was then provided to two independent coders, one of whom is a working journalist and the other a health communication expert (author D). The coders applied the framework to a set of 30 news stories and then discussed coding discrepancies with author A, after which the codebook was refined and finalised (table 1). Subsequent inter-rater reliability was calculated for the quantitative variables by double-coding a sample of 10% of stories (half at the outset of coding and half at the completion); kappa statistics demonstrated excellent concordance (>0.75) for all but one quantitative variable (level of sports engaged in), which was subsequently dropped from the analysis.

Table 1

Descriptive coding

The two qualitative measures employed in the analysis (‘prevention content’ and ‘phenomenological analysis’) were refined following the initial structured coding using open-ended notes made by the coders when reviewing each story. We created the ‘prevention content’ variable by recoding notes made by the coders pertaining to prevention content into three categories: ‘clear prevention message’, ‘no prevention message’ and ‘ambiguous’ (the ‘ambiguous’ category captured stories such as those where the description of the event included some reference to potential contributing factors, but without any explicit messages about prevention). The creation of a categorical variable facilitated consideration of whether prevention messages were more prevalent in stories about children's injuries, from a particular news outlet or in which the expression was employed by a journalist versus a stakeholder in the injury event (table 2).

Table 2

Prevention content

The ‘phenomenological analysis’ of ‘freak accident’ was an explicit consideration of the use of this expression to construct meaning around the injury event. We used this variable to explore the boundaries of the concept as employed in the news media rather than to quantify frequency of occurrence. Our ultimate aim in this was to construct a typology of the meanings and uses of ‘freak accident’ in the news media (table 3). The qualitative coding of this variable took place in three stages: (1) author A elucidated a range of ‘meanings and uses’ based on a review of all articles; (2) the coders used these to guide (but not determine) their open coding of the articles; (3) author A used coders' notes to develop a typology of meanings and uses of the expression. The larger authorship team also provided input at each stage.

Table 3

Meanings and uses of ‘freak accident’


The final dataset contained 250 relevant news stories. The AP contributed 175 stories (70%), the New York Times 37 (15%) and the Philadelphia Inquirer 38 (15%) (table 1). Just over half of the stories (n=140, 56%) were focused primarily on the injury event described as a ‘freak accident’, rather than simply mentioning an injury event as part of a larger story.

From our dataset of 250 stories, 30 were excluded from the analysis of specific injury events because, while they made mention of ‘freak accidents’ in relation to human injury, they did so without reference to one single injury event. For example, our dataset included a story discussing the content of morning television shows and mentioned that victims of a ‘freak accident’ were typical guests. This story was relevant and thus included in the overall sample, but it could not be coded in terms of who was the injured party as injury was talked about in concept only. Similarly, while we included stories such as one describing the practice of decorating graves of people who had died from ‘freak accidents’, we could not fully code such articles as there was no specific injury event described.

Approximately 60% (n=134) of the stories dealing with a specific event focused on professional athletes' injuries. These athletic injuries occurred most frequently during competition (n=54, 40%), but athletes were also injured during practice (n=32, 34%) and during activities unrelated to paid work, such as in ‘horseplay’, fights or while engaged in other recreational sports (n=29, 21%).

Of the 220 stories that focused on a specific injury event, 122 (55%) described some provision of medical treatment (eg, getting stitches or an x-ray). In 21 stories (10%) there was discussion of a child's injury and, in 24 stories (11%), injuries were reported as occurring at home. Of the 24 ‘home’ injuries, seven involved children. Thus, 33% of child injuries reported on as ‘freak accidents’ occurred at home compared with only 8% of adult injuries. The three news outlets differed (approaching significance) in relation to the proportion of the ‘freak accident’ stories published pertaining to children, with a greater proportion of such stories appearing in the Philadelphia Inquirer (χ2=5.32, df(2), p=0.07).

In 102 stories (41%) the expression ‘freak accident’ appeared in the story as part of the journalist's prose rather than as part of a quote from a source. The following is from one such story.Rose scored 15 points and had seven assists, along with a few ankle-breaking moves, playing in the aftermath of a freak accident. Rose said he rolled over on a knife that he had left on his bed after cutting an apple on Monday evening. The resulting wound slashed into his left forearm, requiring 10 stitches. (New York Times, 10 December 2008).

Injured parties used the expression in 47 stories (19%) and coaches, managers and team mates in 30 (12%).Myers was scheduled to pitch Saturday, but he fell stepping out of his car Friday night and injured his eye. Initially Myers said he was hurt playing catch with his 4-year-old son. ‘It was a freak accident; that is all that I can say about it’, Myers said. ‘It's something you don't expect to happen and now you just try and be more cautious’. (Associated Press, 21 August 2009).Holtz (coach) confirmed that running back Dominique Lindsay, who was expected to replace NFL first-round draft pick Chris Johnson at tailback, will miss the season after injuring a knee during a ‘freak accident’ collision with linebacker Quentin Cotton during preseason camp. (Associated Press, 25 August 2008).

In 22 stories (9%), ‘freak accident’ was used to describe an injury event by someone in an ownership or spokesperson role.‘I think that it is just one of those terrible, sad freak accidents’, said Lynne Loresch, executive director of the Mental Health Association of Washington County, the group home operator. (Associated Press, 18 April 2008, story about fatal minibus crash).

Among the subset of 220 stories that discussed a specific injury event, 185 (84%) included no prevention information, 19 (9%) included ambiguous information (such as warnings that people might have been given or safety measures that had been in place and presumably had failed) and 17 (8%) included clear prevention messages (such as discussions about the protection afforded by wearing ski helmets following Natasha Richardson's death). Stories in which the journalist employed the expression ‘freak accident’ were significantly less likely to include any prevention content than those in which the expression was used by a quoted stakeholder (χ2=5.4, df(1), p<0.05).

Stories in which the injured person was a professional athlete were less likely than stories about any other injured party to include any prevention content (χ2=3.24, df(1), p=0.07). There were no differences in prevention information by media outlet, nor was the likelihood of prevention content associated with child injuries (table 2).

Qualitative coding revealed five distinct meanings and three different purposes incorporated in the use of ‘freak accident’ to describe an injury event either by the journalist or someone quoted in the story (table 3). Our analysis showed that this expression could be used to present injury events as unpredictable, unexplainable, bizarre, grotesque, rare or severe. In many stories, journalists and various stakeholders were able to employ ‘freak accident’ in order to deflect responsibility, deny intent or minimise the severity of the event.


We examined how and why injury events are framed as ‘freak accidents’ in the news media. This analysis was not designed to determine whether the appearance of ‘freak accident’ in news media has a negative public health impact. The aim was to contribute to the limited empirical literature on lay discourse on injury in order to inform prevention efforts.

Our analysis of ‘freak accident’ news stories revealed a general absence of explicit injury prevention content in such stories. Moreover, stories in which journalists labelled an event a ‘freak accident’ (rather than the expression appearing in a quote from someone involved in the event) were significantly less likely to include prevention information. We also saw little or no evidence of injury prevention specialists contributing to these stories. It seems clear that the stories in this sample represent missed opportunities for promoting safety, which is consistent with the findings of previous studies of injury coverage.14 15

Notions of rarity and unpredictability evident in this analysis suggest that there may be potential benefit in educational outreach with journalists (especially sports journalists) with regard to injury epidemiology and prevention, and a potential need for injury prevention specialists to create relationships with these journalists. Such partnerships would facilitate the use of potentially newsworthy injury events as opportunities for prevention communication.

The idea of issue ‘newsworthiness’ came to shape our analysis in important ways. The label ‘freak accident’ can be understood as a tool employed by journalists to create interest and convey newsworthiness for an injury event that would otherwise not warrant coverage. A routine injury is, by its very nature, not ‘news’ whereas a ‘freak accident’, given the right injured person and context, may be a story worthy of attention. The expression ‘freak accident’ also conveyed severity of outcome, sometimes calling attention to the gruesome nature of an injury. Threats that are out of the ordinary (unusual or exotic) are more likely to provoke concern than risks that are faced on a regular basis.16 The extent to which framing an injury event as a ‘freak accident’ in order to convey newsworthiness is palatable to the injury prevention community is worth further consideration.

We were struck by the dominance of injuries to professional athletes in our news sample—both during competition and practice as well as injuries occurring ‘off the field’. This finding suggests that partnering with sports journalists may prove particularly fruitful in terms of shaping injury-pertinent discourse in the media. While we are aware that these data are the result of a search for a ‘freak accident’ rather than all news reports on injuries, we conclude that athletes' injuries may be more newsworthy than those of people whose work does not depend on peak physical fitness. The dominance of athletes' injuries draws attention to the practice of framing injuries that occur in the course of routine play as ‘freak accidents’.

Guttman17 argued that societal fondness for sports is paradoxical, as many sports entail a level of interpersonal violence that would be unsanctioned in most other areas of social life. Moreover, the physical exertion necessary for professional success frequently pushes athletes to the point of physical harm. Framing sporting injuries as ‘freak accidents’ might be appropriately understood as an attempt (albeit possibly subconscious) to mitigate the normalised extreme risk-taking that occurs in popular sports, thus making it comfortable for audiences to enjoy watching and following highly risky sporting endeavours. In contrast, coverage of injuries sustained off the field as ‘freak accidents’ may be understood in relation to the contracts signed by many professional athletes that preclude engaging in risky activities not associated with their professional sport. Such restrictions may create motivation to misrepresent the circumstances surrounding injuries resulting from prohibited activities as ‘freak accidents’.

The typology that we developed offers a new way to study news coverage of injury events and insights into the potential risks and benefits of labelling an injury event a ‘freak accident’. The creation of a two-dimensional typology of the framing of injury events as ‘freak accidents’ revealed that the expression takes on different meanings depending on context. Our qualitative analysis revealed considerable scope for incorporating a public health frame into these types of stories. Use of the expression ‘freak accident’ was seemingly called upon to deflect responsibility from one or more parties involved in the event. The possibility of ‘accident’ being associated with an alleviation of responsibility has not yet been a dominant theme in the ongoing debate of ‘accident’.

Limitations of the study

These data are taken from only three news sources, one of which is explicitly national in scope (AP) and two that are associated with large eastern cities in the USA. The nature of the injury events covered elsewhere (eg, small town newspapers, newspapers from other regions or in other countries) may differ, which would impact the conclusions drawn. It is also possible that these findings are not generalisable to other media sources (such as television and internet news). Our sampling decisions were guided by both feasibility considerations and the fact that this is an exploratory study in an area where little or no previous research has been conducted.

Our decision to exclude exact replica stories but to include multiple distinct stories on a single injury event may affect our findings. Each of the stories in the sample does, however, represent a discreet opportunity to include a prevention message. Finally, our analysis is explicitly bound by the absence of any attempt to measure what impact journalists' framing has on readers' perceptions.


This analysis suggests a potentially missed opportunity for public health professionals to capitalise on the influence of the media by working to incorporate an injury prevention frame into stories of human injury. The subsequent impact of such efforts on public attitudes towards injury prevention would also warrant empirical evaluation.

What is already known on this subject

  • Many injury experts and institutions have rejected the use of the word ‘accident’ because it is thought to convey randomness and unpredictability, undermining notions of preventability.

  • There is a disconnect between professional assumptions made about the connotations of ‘accidents’ and empirical studies of how audiences interpret this word in relation to injury events.

What is this study adds

  • The majority of news stories on injury events that were framed as ‘freak accidents’ pertained to injuries sustained by professional athletes.

  • Fewer than 1 in 10 news stories on injury events framed as ‘freak accidents’ included a clear prevention message.

  • News stories in which the journalist employed the expression ‘freak accident’ to describe an injury event were significantly less likely to include prevention information than those in which the expression came from a quote from a stakeholder involved in the event.


Supplementary materials


  • Funding This paper was supported by the cooperative agreement number SR49CE001507 from the Centers for Disease Control and Prevention. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centres for Disease Control and Prevention.

  • Provenance and peer review Not commissioned; externally peer reviewed.