Methods

A commercial service supplied newspaper clippings reporting unintentional injury events during a three month period (July to September 1995) from 17 daily and 55 weekly Connecticut newspapers. There were three daily newspapers with a circulation >50 000 that were classified as large. Since they represented the largest readership, we compared their news accounts of unintentional injuries with the smaller newspapers to determine if there were differences in reporting injury prevention information. We reviewed each clipping for the following: (1) an accompanying photograph, (2) time and date, (3) location of injury event (street address, town name), (4) personal and demographic information of injured person(s) (name, age, sex, town of residence), (5) cause of injury, (6) alcohol use, (7) vehicle, and vehicle damage, (8) road/weather conditions for motor vehicle collisions, (9) type of collision and collision sequence, (10) use of preventive measures (seat belts, airbags, helmet use, window guards, smoke detectors, safety gates, pool fencing and supervision, poison center contact, use of gun safety devices), and (11) outcome (killed or injured).

We classified motor vehicle related articles for persons under 21 years old in the following manner: motor vehicle collisions with bicycles (n=40) and bicycle only injuries (n=4) were classified as bicycle related (n=44). Motor vehicle collisions with motorcycles (n=10) and motorcycle only injuries (n=7) were classified as motorcycle related (n=17). Motor vehicle v motor vehicles and motor vehicle only collisions were classified as occupants (n=204).

Results

In total, there were 962 newspaper articles. Thirty five editorials were excluded because they did not report on specific injury events. Thus, we based our results on 927 articles describing 774 events. Relatively few articles (11%) included a photograph. In contrast, most included personal and demographic data about the injured persons (76%), time and date (89%), and location (84%). All specified the cause and most (96%) reported on whether the victim was killed. There was no significant difference between large and small newspapers.

Of the 342 clippings involving persons under 21 there were 322 injuries and 58 deaths reported. Only one article involving a death, however, stated that the victim (a motorcyclist) was not wearing a helmet. Alcohol was noted to be a factor in just two fatalities.

Table 1 provides a summary on injury cause, contributing factors, and mention of preventive measures. Two hundred and four (60%) dealt with motor vehicle collisions, 17 were motorcycle injuries (5%), 44 bicycle injuries (13%), and 16 house fires (5%). Stories describing falls, drowning, poisoning, and unintentional firearm injury each accounted for 2% or less of the total.

Among the 204 motor vehicle collisions, 157 (77%) included information about the type of collision and the collision sequence. Only six (3%) mentioned driver alcohol use, 19 (9%) seat belt use, and one (0.5%) mentioned airbags. Of the 17 motorcycle and 44 bicycle articles, only 29% and 21%, respectively stated whether a helmet was used. In the 16 articles about house fires, 13% mentioned smoke detectors.

Discussion

Newspapers provide health information to the public and could influence the way in which both the public and professionals respond to the injury problem.² The media communicate individual injury events daily and such articles generate considerable public interest. Although public health professionals produce mass media safety campaigns to increase knowledge, to change attitudes, and to produce behavior change, these campaigns are often under funded, take place during off hours, and compete with many other commercial messages for the public's attention.^3–5

This study suggests that injury reporting in newspapers contains many details about unintentional injury events, but few details about prevention measures or specific protective devices. Most articles presented the view that injuries were primarily due to human error, bad luck, or unpredictable equipment malfunctioning.

A Canadian study involving a three month content analysis of traffic injury reporting also showed that the public was well informed about the demographic characteristics and vehicle descriptions but were not adequately informed about the conditions surrounding the injury, including alcohol involvement (2–3%) and seat belt use (3–6%).⁶

Another study, from North Carolina, followed newspaper coverage of childhood fatalities from fires and drowning compared with medical examiner reports, to assess the extent, completeness, and accuracy of the reports. Their results demonstrated accurate coverage of demographic information and cause and circumstances, but did not accurately report alcohol involvement.⁷

Frost et al recently conducted an analysis to compare newspaper representations of causes of death with actual causes. They concluded that print media disproportionately represent the leading causes of death and their risk factors, and that this misrepresentation may contribute to distorted perceptions of health threats.⁸ In analyzing news coverage of health related topics, Chapman and Lupton also concluded that the media may deliver contradictory reports of health risks or may distort information by emphasizing extreme views over moderate ones. In addition, reporters often do not include constructive background or prevention information or enough details to help the public to evaluate their own risks.⁹

Journalists must deal with the difficult task of balancing the commercial interests of their newspapers with public health interests. Reporters face time and space constraints, and writing articles about injury prevention are far less compelling than more sensational stories often involving rare diseases. Thus, dramatic or “catchy” issues, such as extreme sports, receive much media coverage compared with everyday issues such as the use of child restraints.⁹ Research on the influence of mass media on societal and personal risk judgments support the “optimistic bias” hypothesis, that is, that people tend to underestimate their own vulnerability to harm but overestimate the level of risk to others. The public generally regards unintentional injuries like car crashes as random, or involving others. For example, after a media campaign cautioning about the dangers of drinking and driving on holiday weekends, one study revealed many people became less concerned about driving because they believed this risk did not apply to them.³

Newspapers have often been referred to as “gate keepers” of information. Editors determine what is newsworthy and their choices in turn influence what people think or talk about. Thus, the more safety and health information newspapers print, the more likely they will positively influence the general public's knowledge, attitudes, and observable behavior.³

Implications for prevention

News reports about injuries are widely read and popular items. Our data show these reports tend to focus on the circumstances or aftermath of these events but rarely provide sufficient information about prevention. Rather than viewing injuries as understandable, predictable, and often preventable, they continue to regard them as acts of fate.¹⁰ Public health professionals should become more involved in advising newspapers to set proactive health and safety agendas. This could be accomplished by providing their research findings to newspapers, thereby increasing reporters' understanding of the problem and opportunities for prevention. Reporters could then ask informed questions of police or fire departments: “Was alcohol involved?” “Where was it obtained?” “Were safety devices used?” “Are law enforcement, public health agencies and community organizations taking any preventive actions?”¹¹

One Canadian newspaper collaborated with public health professionals to produce a prototype of a more educational reporting style. For eight weeks, reporters included more detail on the causal chain of unintentional injury events, described these events in the context of antecedents and consequences, and provided weekly accounts of local statistics, including date, location, and severity. In addition, they reported information from research as well as data from specific traffic injury events, including the role of alcohol, driver age and sex, and seat belt use. Subsequent telephone interviews showed a significant shift in the public's opinion about these reports and a positive change in the public's perception of the injury problem.³

In Seattle, Bergman et al conducted an educational campaign advocating bicycle helmet use and enlisted the help of the media to carry out this message. “Victim” news reports had a significant impact on the public and were instrumental in increasing helmet use among school age children from 5% to 16% over a three year period.¹² Newspapers worked with the community, bike manufacturers, and local and state legislators to improve safety.¹³ Newspapers often report year-to-date statistics on injuries sustained from violent crimes, fires, suicides, and domestic violence and should provide similar summaries of unintentional injuries to increase public awareness of the magnitude of the injury problem. Reporters should write more articles with “human context” information—for example, details about the reason for a car trip, seat restraint use, and describing the emotional shock of the event, the insurance implications for the driver, or the medical costs.

In summary, newspapers appear to neglect the powerful part they could play in educating the public about injury risks and the effectiveness of protective measures.