Abstract
Background:
Several countries have prescribed standard guidelines for media professionals on suicide reporting. However, the implementation of these guidelines has been varied. Suicide rates in South Asia are one of the highest in the world, and it is known that media guidelines for suicide reporting are not followed adequately. However, there are no published reports available from this region.
Aim:
This study aimed at assessing newspaper reports of suicide for quality of reporting based on standard reporting guidelines and to study differences between English and vernacular (Kannada) newspapers in Bangalore, South India.
Methods:
A total of 341 newspaper reports of suicide from 550 newspapers (3 English and 3 Kannada) over 3 months were systematically assessed for compliance with reporting guidelines. Each report was evaluated on 2 domains and 36 parameters. Data were analyzed for frequency of inappropriate reporting and patterns compared between vernacular and English newspapers.
Results:
In all, 87% of the reports were those of completed suicide. Non-compliant reporting – method of suicide was reported in 89% and 32% of reports were in prominent pages of the newspaper, 95% mentioned gender, 90% reported the name, 80% reported age and suicide location, 75% reported life events related to suicide, 70% reported occupation, 69% had headline explicity on suicide and 61% reported monocausality. Only 16% reported mental disorder related to suicide, and less than 3% included information on suicide prevention and helplines. Vernacular papers showed significantly better compliance in 16 of the 20 areas. However, protective characteristics were better reported in English newspapers.
Conclusion:
Majority of reports on suicides in newspapers from Bangalore did not comply with standard guidelines of reporting. There is a strong need to evolve local guidelines and mechanisms for ensuring responsible reporting which have important implications in prevention of suicide.
Keywords
Introduction
Every year, almost one million people die from suicide, and in the last 45 years, suicide rates have increased by 60% worldwide. Suicide is among the three leading causes of death among those aged 15–44 years in some countries, and the second leading cause of death in the 10- to 24-years age group (World Health Organization (WHO), n.d.). The suicide rate in India has been increasing steadily and has reached 11.2 (per 100,000 of population) in 2011 (National Crime Records Bureau (NCRB), 2011). The rates have been reported to be highest among young women (Patel et al., 2012). These figures highlight the need for effective prevention strategies.
One of the important preventive measures recommended is responsible media reporting of suicides (Bohanna & Wang, 2012; Gunnell & Lewis, 2005; Patel et al., 2012; Ram, Darshan, Rao, & Honagodu, 2012; Vijaykumar, 2007, 2010). Research has demonstrated the important relationship between suicide rates and media reporting (Pirkis & Blood, 2001; Stack, 2005; Yip et al., 2006). Portraying suicide by print and visual media in a manner that overemphasizes unnecessary details has been found to lead to imitative behavior in vulnerable individuals (Pirkis, 2009).
Several studies done in the United States, Japan and Australia have indicated that suicide rates significantly increase after publication or publicity of suicide stories in the media (Hagihara, Tarumi, & Abe, 2007; Hassan, 1995; Romer, Jamieson, & Jamieson, 2006). A systematic review that analyzed findings from different studies on this topic concluded that there was an increase in number of suicides resulting from media accounts of suicide that romanticize or dramatize the description of suicidal deaths (Sudak & Sudak, 2005).
Recognizing the implications of media reporting on suicide rates, several countries and the WHO have prescribed standard guidelines for media professionals on reporting of suicide. These guidelines for responsible reporting are expected to educate the public about suicide, and may also encourage those at risk of suicide to seek help (WHO, 2000). However, the implementation of these guidelines by different countries has been varied.
Pirkis, Blood, Beautrais, Burgess, and Skehan (2006) reviewed nine guidelines developed by different countries to promote responsible suicide reporting on the following parameters – imitation effect, content of guidelines, development and implementation process. The study found that the guidelines had almost similar content (e.g. avoiding sensationalization and stressing the importance of providing information about help services were emphasized). However, they differed in the way they were developed and implemented. The authors felt that ‘journalist-led’ guidelines and proper dissemination were more likely to have desirable impact.
In countries where guidelines are available, the compliance has been varied. Tatum, Canetto, and Slater (2010) evaluated adherence level to media guidelines in the United States, examining 968 newspaper articles on suicide published in 2002 and 2003, and found inconsistent adherence. The most striking feature was the absence of helpful suicide prevention information.
Literature from Asia on this topic is limited to a report from China that studied newspaper articles of suicides from Hong Kong, Taiwan and Guangzhou (1,086, 967 and 226 newspaper reports, respectively). The results found adequate compliance in only one characteristic (only 2% of the articles were prominently placed). The study also found that newspapers emphasized more on suicides of young people (Fu, Chan, & Yip, 2011).
A detailed literature search revealed that there has been no systematic study from any South Asian country on this topic, except for a study from India that assessed the impact of a workshop for media professionals on quality of reporting. This study, however, did not provide any details of compliance to reporting guidelines (Ramadas & Kuttichira, 2011).
This study was conducted in Bangalore in South India, in an attempt to generate some data on this topic. Bangalore is a rapidly growing city and has been labeled the suicide capital of India (Deccan Herald, 2012). Based on the Population Census (2011), the population in Bangalore is around 9 million, and literacy rate is 87.51%. According to the latest NCRB report (NCRB, 2012), Bangalore has a share of 10.4% in total suicides in India, ranking second highest in India. The city has been recording an unprecedented increase in population, industrialization, urbanization and migration of families from rural areas to city along with the ongoing sociodemographic and epidemiological transition, resulting in increase of social, economic and health problems in the city. In this scenario, suicides have emerged as a leading public health problem in Bangalore (Gururaj & Isaac, 2001).
Newspapers are a popular source of information in India. The National Readership Survey (NRS, 2005) revealed that newspaper readership had grown from 131 million in 1999 to 155 million in 2002. According to NRS 2005, the number of readers of newspapers and magazines in rural India (98 million) became almost equal to that of urban India (101 million).
The aims of this study were twofold:
To assess newspaper reports of suicide for quality of reporting based on standard suicide reporting guidelines among six most widely circulated newspapers in Bangalore city;
To compare the quality of reporting between English and Vernacular (Kannada) newspapers.
The newspaper industry in India is divided into the vernacular (based on region) and English newspapers. The daily circulation figures of the six daily newspapers used in this study are among the highest in this region. English daily 1 has a circulation of 562,832, English daily 2 has a circulation of 246,346 and English daily 3 has a circulation of 111,005. The vernacular daily 1 has a daily circulation of 612,658, vernacular daily 2 has a circulation of 528,893 and vernacular daily 3 has a circulation of 182,385. In India, the dailies in regional languages command more than half of the Average Issue Readership (AIR) as per Indian Readership Survey (IRS) 2010 (Audit Bureau of Circulations, 2012).
Readership figures are usually higher than circulation figures because of the assumption that a typical copy of the newspaper is read by more than one person. The number of readers per copy varies considerably by publication (Audit Bureau of Circulations, 2012). According to IRS, the AIR figures of the daily newspapers used in this study are as follows: English daily 1, AIR of 612,000; English daily 2, AIR of 423,000; and English daily 3, AIR of 78,000. AIR figures of vernacular dailies are much higher and range from 1,221,000 to 3,327,000 (Audit Bureau of Circulations, 2012).
Methods
The study assessed compliance to suicide reporting guidelines in six highly circulated newspapers of Bangalore which included three English and three vernacular (Kannada) daily newspapers. All reports and articles on suicide which were published during a 3-month period, that is, 1 October 2011 to 31 December 2011, were included. A total of 552 newspapers were hand searched, and the search yielded 341 reports and articles on suicide. Of these, 157 were English newspaper reports, and 184 were vernacular newspaper reports. Suicide articles were defined as ‘all reports of suicide attempted or completed’, whereas articles that had metaphorical words like ‘suicide bombing’ or the word suicide was located in a movie, TV listing or advertisement were excluded from the study.
Each report was evaluated by two independent assessors on two domains and 36 parameters – harmful characteristics (27 items) and protective characteristics (9 items). The variables under these two headings were based on a similar study done by Niederkrotenthaler et al. (2010). The authors in the study had reviewed 15 international guidelines in addition to the one prescribed by WHO and categorized these variables into harmful and protective items.
Harmful item characteristics included mention of the name, age, gender and occupation of the person involved in suicidal act; description of character of the person; mention of the method of suicide in the headline; step-by-step details of the act; suicide location/site; suicide pact/mass suicide reported; life events related to suicide reported; societal problems reported as having increased; monocausality, that is, item attributes a single cause for the act; citation from suicide note; effects on bereaved, that is, item includes impact of the act on close relatives/friends; effects of suicidal act on public life, that is, item includes impact of the act on public; immediate death without suffering reported; photograph/images included; several independent suicidal acts reported; references to suicide epidemic; headline explicity on suicide; public myths reported; number of sentences; use of emotionally explicit words; repetitive reporting, that is, follow-up report of an event on subsequent days in the same newspaper; density of reporting, that is, presence of more than one suicide-related articles on the day of publication; reported methods, that is, hanging, jumping, shooting, drowning, poisoning, burning, unusual method; and prominence of the report – front page/third page/top of an inside page.
Protective item characteristics included focus on suicidal ideation, focus on prevention program, focus on suicide research, expert opinion reported, mental disorder related to suicide, myths debunked, contact information for support service given, suicide-related statistical data reported from reliable sources and suffering of suicidal person resulting from suicidal act reported.
Each variable was scored as 0 (Absent) or 1 (Present). It was also noted whether it was a completed suicide, attempted suicide or a celebrity suicide, and whether it included a homicide.
In order to assess the inter-rater reliability, the assessors coded 100 suicide articles independently. The Kappa value for harmful items was .918 and was significant at .001 level. The Kappa value for protective items was .457 and was significant at .02 level. This indicates that the inter-coder reliability of the assessors was significant and reliable.
Statistical analysis
The data were analyzed for frequency of inappropriate reporting and protective reporting. Reporting was compared between Kannada (vernacular) and English newspapers. Chi-square test was applied to compare the quality of reporting between English and vernacular reports. Some of the cell frequencies were <5. However, it was noted that none of those cells had expected frequency <5, and hence, Pearson’s chi-square test was used.
Results
Of the 341 reports, completed suicides were 297 (87%), attempted suicides were 37 (11%), a combination of homicides and suicides were 32 (9%), and celebrity suicide reports were 10 (3%). Suicide pacts were reported in 48 (14%). Suicidal ideation was mentioned in 12 (3.5%), articles on prevention were 8 (2%) and articles on suicide research were 15 (4%). Method of suicide was reported in 304 (89%) reports and included hanging in 158 (46%), poisoning in 57 (17%), burning in 33 (10%), jumping in 22 (7%), drowning in 17 (5%), shooting in 2 (1%) and other unusual methods in 15 (4%). Reports of suicide in children below 14 years were 11 (4%), and reports of suicide in children and adolescents between the ages of 0 and 18 years were 24 (9%).
Table 1 indicates the harmful characteristics on 26 items. Among the harmful characteristics, the important findings include the following: 326 (95%) of the reports mentioned gender; 308 (90%) reported the name; 272 (80%) reported age and suicide location; 256 (75%) reported life events related to suicide; 239 (70%) reported occupation; 234 (69%) had headline explicity on suicide; 210 (61%) had density of reporting, that is, presence of more than one suicide-related articles on the day of publication; 207 (61%) reported monocausality for the suicide; 151 (44%) had description of character; 136 (40%) reported several independent suicidal acts, and 110 (32%) reports were prominently placed.
Frequency and percentage distribution of harmful characteristics.
Some harmful characteristics that were found in less than one-quarter of reports included the following: 78 (23%) had method in headline, 71 (21%) reported immediate death without suffering, 66 (19%) reported step-by-step guide, 65 (19%) included photograph or image, 50 (15%) reported suicide pact or mass suicide, 49 (14%) included citation from suicide note, 36 (11%) used emotionally charged words, 10 (3%) repetitive reporting and 6 (2%) reported public myths.
Among the protective characteristics (Table 2), 54 (16%) reported mental disorder related to suicide, 14 (4%) of the reports included suicide-related statistical data, 12 (3.5%) reports focused on suicidal research, 9 (2%) reports focused on suicide prevention programs, 11 (3%) reports focused on suicidal ideation, 11 (3%) reports included expert opinion, 11 (3%) reported suffering of suicidal person resulting from suicidal act, 4 (1.2%) debunked myths and information related to helplines and details of support services were found only in 2 (0.6%).
Frequency and percentage distribution of protective characteristics.
Among the harmful characteristics, vernacular newspapers had better reporting compared to English newspapers for 16 of 26 items. Among the nine protective characteristics, however, English newspaper reporting was significantly better for six of the items – focus on suicidal ideation, focus on prevention program, focus on suicide research, expert opinion reported, mental disorder related to suicide and suicide-related statistical data reported from reliable sources.
Discussion
This study systematically assessed 341 newspaper reports of suicide for quality of reporting and compared the quality of reporting between English and Vernacular (Kannada) newspapers published in Bangalore over a 3-month period using specific criteria adopted from the study by Niederkrotenthaler et al. (2010). The results of this study suggest that there was fairly poor adherence to the guidelines for media reporting on suicide during the study period on most of the harmful characteristics (Table 1). Some aspects of the guidelines showed better adherence than others. Specifically, less than one-quarter reported method in headline, immediate death without suffering, step-by-step guide, photograph or image, suicide pact or mass suicide, citation from suicide note and use of emotionally charged words. Less than 10% reported effects on bereaved, effect of suicidal act on public life and had repetitive reporting. Furthermore, while only 16% reported mental disorder related to suicide, all other protective characteristics of reporting were found in less than 4% of reports (Table 2). These findings indicate that Indian newspapers during the study period showed fairly poor adherence to guidelines of reporting and also did not report protective aspects.
A social learning explanation of imitation suggests that the risk of suicide imitation increases with the degree to which vulnerable audiences can identify with a suicide model (Pirkis, 2009). Therefore, from a social learning perspective, many reports in our study that included information about gender, occupation and age of the suicidal person may provide vulnerable readers with models for imitative suicidal behavior. This is a matter of great concern.
How do our findings compare with similar studies done around the world?
Some measures that were comparable across studies done in the United States, Austria, Australia, China and our study included the use of photographs, placement of the suicide report in the newspaper, mentioning method of suicide, mentioning suicide in the headline and mentioning where the suicide was committed (location).
In all, 19% of the reports in our study had photographs, which is similar to those found by Niederkrotenthaler et al. (2010) (13%) in Austria and by Tatum et al. (2010) (21%) in United States. However, a study from China (Fu et al., 2011) reported much higher rates of photographs and images (57.5%). In all, 89% of reports from Bangalore mentioned method of suicide compared to 21.5% from the Austrian study (Niederkrotenthaler et al., 2010); 23% of our reports mentioned method of suicide in the headlines, which is slightly higher than findings (14%) of the Austrian study (Niederkrotenthaler et al., 2010). However, studies from China (Fu et al., 2011) and the United States (Tatum et al., 2010) showed much higher headline reporting (80.5% and 50%, respectively).
Our study showed that 32% of the reports were prominently placed (front page, third page); however, studies done in Austria, China and the United States (Fu et al., 2011; Niederkrotenthaler et al., 2010; Tatum et al., 2010) showed better compliance in newspaper placing of the suicide reports.
Nearly 80% of the reports mentioned suicide location which was much higher than the US and Austrian findings (Niederkrotenthaler et al., 2010; Tatum et al., 2010). In all, 19% of our reports mentioned the method in detail, giving step-by-step guide which is similar to the findings of the Austrian study (21.5%) (Niederkrotenthaler et al., 2010). However, in the US study, 56% of reports included details of the method (Tatum et al., 2010).
This study showed poor compliance with the guidelines in reporting monocausality (61%) and life events related to suicide (75%) compared to the Austrian study (30%). Overall, the newspaper reports in Bangalore during this period showed poor adherence to guidelines in several areas compared to reports from other parts of the world.
With regard to the protective variables, it appears that around the world, these are inadequately reported and our study too had the same findings. Only 1% of the reports provided information related to helplines and support services, which can be considered almost negligible. The rates of protective reporting were not much better from other parts of the world. In all, 8.6% of reports provided details on support services in the study from China (Fu et al., 2011), while in the US and Austrian studies, only 3% and 5.6% of reports provided details on support services, respectively (Niederkrotenthaler et al., 2010; Tatum et al., 2010). In all, 16% of the Bangalore reports mentioned a mental disorder related to the suicide compared to 12% in the United States and 19% in Austria (Niederkrotenthaler et al., 2010; Tatum et al., 2010).
Table 1 shows that vernacular newspapers were better than the English newspapers in some aspects of reporting of the harmful characteristics. However, English reports were better than the vernacular reports in reporting protective characteristics.
The reasons for slightly better reporting in several domains in vernacular newspapers are not very clear. We speculate that vernacular reports might be giving more importance to local political news and may hence not have enough space. Results of this study cannot be generalized for all the vernacular newspapers since it has evaluated only Kannada language newspapers.
Comparing our data with the rest of Asia (Fu et al., 2011) indicates that adherence to guidelines appears to be particularly low in Asia, where suicides are highly newsworthy events. Considering the high suicide rate in some areas of Asia, and the prominent influence of media on imitative suicidal behavior, it is crucial to develop strategies that effectively engage the media industry in the prevention of suicide (Chen et al., 2010; Cheng, Hawton, Lee, & Chen, 2007).
Our study has several strengths. It systematically analyzed 552 newspapers and 341 reports from six highly circulated newspapers. The reports were rated on standard guidelines by two independent raters, thereby reducing the researcher bias. This is also the first study reported from South Asia on this topic, a region which has high rates of suicide. The limitations include a short period of study (3 months) and the fact that it was done only in one city and used only newspapers of one vernacular language in addition to English. We also did not analyze reports from tabloids and evening newspapers which also have a wide readership.
What are the implications of our findings? It has been found that prescribing media guidelines in Asian countries has led to some improvement in reporting, but the results are far from satisfactory. Fu and Yip (2008) examined report quality in five Hong Kong newspapers between 2004 and 2006, that is, before and after launch of media guidelines. The results showed significant reductions in articles with photos and articles mentioning suicide in headline; however, rates were still high. There was no reduction, however, in front page placement, graphical presentation, mentioning suicide in headline and also no increase in positive messages, and there was an increase in number of suicide articles. A study done in Switzerland by Michel, Wyss, Frey, and Valach (2000) has, however, reported significant improvement in reporting, following media sensitization and training. These included the number of headlines on the front page (20% to 4%), use of sensational headlines (62% to 25%) and text (49% to 14%), glorifying headlines (17% to 3%) and text (33% to 6%) and inclusion of pictures (43% to 8%). No change was observed in the mentioning of name, time or place in the headline, repetition of articles about the same case, or the mentioning of suicide method in the headline or text.
Pirkis et al. (2009) analyzed newspaper, television and radio report quality over two 12-month periods before (2001–2002) and after (2006–2007) guideline implementation. Decrease in use of inappropriate language, use of images/footage, details of method, reference to celebrity status and use of suicide in headline were important outcomes. There was also an increase in inclusion of help service information.
Bohanna and Wang (2012) found that guidelines can change reporting style and prevent imitative suicide provided the approaches center on consultation, collaboration, media ownership and training.
The inconsistent compliance to the guidelines in reporting suicide in newspapers from India may be related to lack of awareness about existing guidelines among media professionals and viewing suicide as a criminal offense rather than a mental health issue. In India attempt to suicide is still a crime under Section 309 of the Indian Penal Code (Law Commission of India, 2008). As a result, the reporters who cover media are crime reporters rather than those who report health. This might also be a contributing factor to the high rates of nonadherence to guidelines and sensationalization.
There is hence a strong need for sensitization of the media about its responsibility and for appropriate dissemination of media guidelines. This is especially true when one looks at the high readership figures for the six newspapers included in the study. A collaborative effort with the mental health professionals, media professionals, policy makers and other gate keepers can facilitate this process.
For the success of such an endeavor, it is also important to understand what journalists and newspaper editors think about the guidelines and their acceptance of the same (Collings & Kemp, 2010). More research is also needed to determine the optimal frequency of guideline dissemination campaigns necessary to sustain changes in reporting behavior over the long term.
Conclusion
This study found poor adherence to prescribed media guidelines for reporting of suicide in six newspapers from Bangalore, India, a city with high rates of suicide. In addition, the reporting of protective factors such as information about helplines and mental health support was very low. We recommend a collaborative approach with the media in enhancing implementation of guidelines. We also advocate the role of organizations like the World Psychiatric Association in providing support to countries on methodologies that might enhance successful mental health and media partnerships for suicide prevention.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.