Assessing the quality of suicide reporting in Portugal: 6 case studies of celebrities and non-celebrities over 25 years

Abstract

Multiple studies from various countries have found evidence of suicide increases after media reports of suicide, which are known as the Werther effect, but responsible suicide reporting can contribute to suicide prevention, a phenomenon known as the Papageno effect. This study aims to assess adherence by the Portuguese media to the suicide reporting guidelines released by the World Health Organization (WHO) and to determine whether there are differences in reporting style between them and over time. News items about six suicides of public and non-public figures between 1996 and 2020 were collected from several Portuguese media—two newspapers, three radio stations, three television channels, and a few websites—in the periods following those deaths, and they were subsequently examined for adherence to the WHO guidelines. A total of 374 news items were analyzed. The majority of them complied with most of the WHO recommendations for suicide reporting, especially the ‘don’t do’ guidelines. However, most of the news items that were analyzed did not follow the recommendations regarding putatively preventive information, notably the inclusion of support contacts and citation of mental health experts. The Portuguese media mostly adhered to the WHO suicide reporting guidelines and there was a positive evolution over time in the level of compliance with them. Nevertheless, there is room for improvement.

Keywords:

• Suicide
• media reporting
• content analysis
• guidelines
• Portugal

Introduction

Suicide is a global health problem, causing more than 700,000 deaths worldwide per year (World Health Organization, 2023). In Portugal, there are more deaths by suicide than by transport accidents (Instituto Nacional de Estatística [National Institute of Statistics], 2022a) and some regions are particularly affected by historically high suicide numbers (Instituto Nacional de Estatística [National Institute of Statistics], 2022b). However, there has never been an analysis of how different types of Portuguese media report suicide cases both of celebrities and non-celebrities, nor a study about this reality in different years. The present study aims to address exactly this gap.

Multiple studies from different decades, and several countries, have documented that media reports of suicides can stimulate imitation acts (Cheng et al., 2011; Gould, 2001; Niederkrotenthaler et al., 2012; Pirkis et al., 2018; Sinyor et al., 2018; Stack, 2003; Ueda et al., 2014), a phenomenon known as the Werther effect (Phillips, 1974). The risk of copycat suicide behavior is exacerbated in vulnerable individuals by an extensive, prominent, and sensationalist coverage that explicitly describes the method used and repeats myths about suicide (WHO, 2008, 2017). In addition, several studies found a relation between media coverage of celebrity suicides and following rises in suicides among the population (Fink et al., 2018; Ladwig et al., 2012; Niederkrotenthaler et al., 2020; Pirkis et al., 2020; Sinyor et al., 2020; Whitley et al., 2019).

However, there is also research showing that the media can contribute to suicide prevention, a phenomenon known as the Papageno effect (Niederkrotenthaler et al., 2010), if they follow certain guidelines, notably by publishing alternative actions to suicidal behavior, emphasizing positive coping strategies, and presenting support contacts and other information about where people at risk can find help (Etzersdorfer & Sonneck, 1998; Jobes et al., 1996; Niederkrotenthaler et al., 2014; WHO, 2017).

Within the scope of its worldwide initiative for the prevention of suicide, the WHO first released guidelines for media professionals in 2000, with updates in 2008 and 2017.

This study aims to analyze the news coverage of suicide deaths, both of celebrities and non-celebrities, in some of the main Portuguese media—newspapers, radio stations, television channels, and websites—against the WHO guidelines for responsible suicide reporting, which are the basis for the Portuguese recommendations. The selected case studies cover a period of 25 years in search of possible changes in the Portuguese suicide coverage over time.

Materials and methods

Case selection

We selected six cases of Portuguese suicides covered by national media.

1. On the night of February 3, 1996, three young people threw themselves from the Duarte Pacheco Viaduct onto Ceuta Avenue, in Lisbon, the capital city of Portugal: Ricardo, 21 years, Paulo, 19 years, and Alice, 16 years.

2. Journalist Miguel Ganhão Pereira, 29 years, died by suicide on December 4, 2000, at night, by jumping off Lisbon’s (famed) 25th of April Bridge into the Tagus River below. He was the anchor for the main nightly newscast at the TVI television channel.

3. Singer Cândida Branca Flor was found dead, next to pills and alcohol, on July 11, 2001, at her home in Massamá (in the outskirts of Lisbon), where she lived alone. The only family she had was her mother, who lived in Alentejo. She was called the ‘Portuguese Marylin’ (Expresso, 2001), given some similarities with Marylin Monroe, who died in 1962.

4. Gil Sousa, a 48-year-old jeweler, died by suicide in the early hours of October 30, 2018, with a shot to the head, in the house where he lived in Porto. He was facing economic problems, having been declared insolvent.

5. Fashion designer João Branco, one half of Storytailors, created in 2001, died on Monday, December 17, 2018, aged 40. The stylist threw himself from his bedroom window at his mother’s house, on the 10th floor.

6. Famous actor Pedro Lima was found dead at around 10 a.m. on June 20, 2020, at Praia do Abano, a beach in Cascais. Pedro Lima had a contractual relationship with the TVI television channel for more than two decades.

We looked for news items about these deaths, whether or not they mentioned that it was a suicide, to assess how each death was framed in each piece. Whenever possible, we sought to analyze all the news pieces released since these deaths occurred until the time of the current study.

Media selection

With the purpose of assessing whether there were differences in the reporting style of different media and in the same media over time, from 1996 to 2020, we selected the following media outlets for analysis:

1. The newspapers Público and Correio da Manhã (CM), which are the dailies with the largest circulation in Portugal (Associação Portuguesa para o Controlo de Tiragem e Circulação [Portuguese Association for Print Run and Circulation Control], 2023). They are, respectively, an example of a newspaper associated with ‘quality’ journalism and an example of a newspaper associated with ‘popular’ journalism (Traquina, 2004);

2. The generalist national television channels RTP, SIC, and TVI;

3. The radio stations TSF, Antena 1, and Rádio Renascença (RR), which are the Portuguese radio stations that dedicate more time of their programming to the news;

4. The websites of Público and CM, to assess whether there were differences in suicide coverage between the print and digital editions; and other Portuguese news websites concerning cases D and E.

We considered that a sample of eight media outlets (without counting the news websites) would give scope and variety to the study, providing an overview of the media landscape and ensuring a focused analysis and data control. Besides, it is a similar number to that of other studies on suicide media coverage and mental health (Carmichael & Whitley, 2019; Chen & Lawrie, 2017; Kamboh & Ittefaq, 2019). In Table 1, we present details about the media sources consulted for each case under analysis.

Table 1. Media sources consulted by case study.

Media	Case studies
	Triple youth suicide	Journalist Miguel Ganhão Pereira	Singer Cândida Branca Flor	Jeweler Gil Sousa	Fashion designer João Branco	Actor Pedro Lima
Newspapers (CM and Público) (1)	From February 5 to February 20, 1996.	From December 5 to December 12, 2000.	From July 12 to July 20, 2001.	From October 29 to November 5, 2018.	From December 17 to December 24, 2018.	From June 20 to June 30, 2020.
Radio stations (TSF, RR, and Antena 1)	TSF had ‘no files available.’			The TSF news bulletins available online were analyzed:		All the news bulletins of the three radio stations were made available for this research.
				From October 29 to November 5, 2018.	From December 17 to December 24, 2018 (2).
	RR had ‘no files available.’
	At Antena 1, ‘nothing was found’ in the archive.		At Antena 1, there were 3 newscasts available.	At Antena 1, ‘nothing was found’ in the archive.
Television channels (RTP, SIC, and TVI) (3)	RTP’s and TVI’s archives were consulted in analog cassette systems, under the guidance of one of the staff members responsible for the archive.			News bulletins broadcast at 1 p.m. and 8 p.m. were available on the RTP and TVI websites: RTP Play and TVI Media Player, respectively:
				From October 29 to November 5, 2018.	From December 17 to December 24, 2018.	From June 20 to June 28, 2020.
	At SIC, the archive was consulted in a digital system called Invenio, under the guidance of one of the staff members responsible for the archive. All news programs of SIC and SIC Notícias (SIC’s news channel) up until the date of the consultation were available.
Websites	Search for ‘triple suicide’ on the Público and CM websites. No news items found.	Search for ‘Miguel Ganhão Pereira’ on the Público and CM websites.	Search for ‘Cândida Branca Flor’ on the Público and CM websites.	Search for ‘Gil Sousa’ on Google. All news items up until the date of the search were selected.	Search for ‘João Branco’ on Google. All news items up until the date of the search were selected.	Search for ‘Pedro Lima’ on the Público and CM websites. All news items up until the date of the search were selected.

Note.

(1) Newspapers were consulted at a physical newspaper archive, called Hemeroteca, and on the newspapers’ websites.(2) ‘TSF Mornings’ from December 21 to 24, 2018, were not available online. The deputy director of TSF explained to us that any newscasts that were unavailable online and more than one year old would not be available elsewhere, since that is the legal period of retention of the records.

(3) The consultation of the television channels’ archives was coordinated by their respective file section managers, upon prior request by email.

For the deaths that occurred during the week—those of designer João Branco (Monday, December 17, 2018) and of jeweler Gil Sousa (Monday, October 29, 2018)—, we listened to the radio stations’ morning news editions, specifically the 7 a.m., 8 a.m., and 9 a.m. news broadcasts, which correspond to the time window with the largest audience (Marktest, 2022). For Pedro Lima’s case study, we listened to all the radio newscasts from 6 a.m. on Saturday, June 20, to 11 p.m. on Sunday, June 21. We also listened to the 7 a.m., 8 a.m., and 9 a.m. news broadcasts of the following five working days.

For the most recent case studies, we also analyzed, whenever possible, the television newscasts aired on the thematic channels included in the same television network as the selected generalist channels, for instance, SIC Notícias [SIC News channel].

Analysis

Each news piece was assessed for adherence to the WHO guidelines for responsible reporting on suicide (see Table 2), which determined 27 criteria for the analysis. We considered the WHO’s resources for media professionals released in 2000, 2008, and 2017.

Table 2. Analysis criteria drawn from the WHO guidelines.

Analyzed criteria	WHO recommendations for suicide reporting
A) Analysis of location and images
1. No prominent location (1)	Don’t place stories about suicide prominently. (2017)
2. No images of the dead 3. No images of the suicide method 4. No images of the suicide scene/place	Photographs of the deceased, of the method used and of the scene of the suicide are to be avoided. (2000)
B) Content analysis
5. The word suicide is not used in the title (2)	The word ‘suicide’ should not be used in the headline. (2017)
6. No use of sensationalist language/ sensational presentation of suicide (3)	Don’t use language which sensationalizes or normalizes suicide or presents it as a constructive solution to problems. (2017)
7. No glorification of suicide victims as martyrs or objects of adulation	Glorifying a celebrity’s death may suggest that society honors suicidal behavior. For this reason, particular care should be taken in reporting celebrity suicides. (2008)
8. Mention of the existence of a mental health problem (4)	Any mental health problem the celebrity may have had should also be acknowledged. (2000)
9. Citation of a mental health professional	Media professionals should seek advice from local suicide prevention experts when preparing stories about suicide. (2017)
10. Suicide is not presented as a solution to problems	Don’t use language which sensationalizes or normalizes suicide or presents it as a constructive solution to problems. (2017)
11. No simplistic reasons given for suicide	Language that misinforms the public about suicide, normalizes it or provides simplistic explanations for a suicide should also be avoided. (2017)
12. No use of language that normalizes suicide
13. The method of suicide is not described explicitly	Don’t explicitly describe the method used. (2017)
14. No detailed information about the location of the suicide	Don’t provide details about the site/location. (2017)
15. No citation of a suicide note	Suicide notes, final text messages, social media posts and emails from the deceased individual should not be published. (2017)
16. Suicide is either confirmed or not referred to (not presented as hypothesis)	Speculation about suicide on the part of the media can be harmful, and it is preferable to wait for the cause of death. (2008)
17. Due consideration for people bereaved by suicide	Show due consideration for people bereaved by suicide. (2008)
18. Support contacts or information on where to seek help are provided	Information about support resources should be provided at the end of all stories about suicide. (2017)
19. Alternatives to suicide are presented	Highlight alternatives to suicide. (2000)
20. The news piece educates the audience about suicide	Do educate the public about the facts of suicide and suicide prevention, without spreading myths. (2017)
21. Indication of risk factors and warning signs of suicidal behavior	Publicize risk indicators and warning signs of suicidal behavior. (2000)
22. There is an association between depression and suicidal behavior	Conveying the message that depression is often associated with suicidal behavior and that depression is a treatable condition. (2000)
23. Description of physical consequences of non-fatal suicide attempts	Describing the physical consequences of non-fatal suicide attempts (brain damage, paralysis, etc.) can act as a deterrent. (2000)
24. Stories about coping with stressors and suicidal thoughts are presented	Do report stories of how to cope with life stressors or suicidal thoughts, and how to get help. (2017)
C) Analysis of the repetition of news pieces about a reported suicide
25. Number of news items published 26. Number of days with news items published 27. Period of news coverage	Don’t unduly repeat stories about suicide. (2017)

Note.

(1) Regarding TV and radio news bulletins, we considered the beginning and the end, as well as the last news pieces to be shown before the breaks, as a prominent location.(2) We considered the title and pre-title as a unit of analysis.

(3) We deemed sensationalist the use of words such as ‘tragedy,’ ‘shock,’ and ‘devasted’; the overly detailed description of the suicide method; and the use of images that showed the suffering of bereaved family and friends. We also considered the rules of Luce (2016) to detect stigmatizing and sensationalistic reporting.

(4) We expanded the ‘Is the existence of a mental health problem mentioned?’ criterion to include not only the references to the mental health problems of the people who died by suicide but also the discussions around mental health elicited by those suicides.

All articles were read and coded by the first author, who also viewed and listened to the television and radio news bulletins. The first author is an experienced journalist, who has led studies related to media coverage of suicide and was closely supervised by the second author, who helped resolve ambiguities during all the stages of data analysis.

We recovered Van Dijk (1988, p. 4) definition of news as ‘a news item or news report, i.e. a text or discourse on radio, on TV or in the newspaper, in which new information is given about recent events.’ In this sense, we considered each of the following as a unit of analysis: a news item/piece (written or broadcast); a highlight at the beginning of a newscast; a news read by the news anchor; an interview; a text box in a newspaper forming an independent unit; the first page of the selected newspapers.

We decided to include in our analysis not only pieces about the selected suicide deaths, but also follow-up pieces, and those using the analyzed suicides as a starting point to address mental health issues.

Ethical approval has not been necessary since our study involved information freely available in the public domain and the analysis of data obtained from public and private entities whose written informed consent was obtained for their data to be used in the research before its original collection.

Results

We analyzed a total of 374 news items from the selected media. News items not directly related to the analyzed suicide deaths, or their follow-up, were excluded.

The amount of news items concerning the death of actor Pedro Lima—a total of 211, mainly published on CM’s website (102)—stands out. Through the 10 days that followed the actor’s death, the theme was on the front page of CM for eight days, with great prominence on three of them, and we gathered a total of 36 news items. In Público, there were only two news items during this period, none of which in a prominent place.

Afterwards, we analyzed the news items published on CM’s website and on Público’s website, whether only available online or taken from the print edition, the latter being exclusively for subscribers. While 102 news items about the death of Pedro Lima were published on CM’s website, there were only 12 news items on Público’s website. We assumed the duplication of content on the online and print editions of Público and CM, since Público’s readers are mainly digital, while CM’s readers are mainly readers of the print edition (APCT, 2023). Besides, the WHO recommends that media refrain from unduly repeating stories about suicide (WHO, 2017).

After the websites, with a total of 174 news items about the six case studies and a strong dominance of CM online, the three television channels were the second type of media with the largest number of news items—a total of 114—, mainly about the triple youth suicide in 1996 (51). The number was lower for the most recent case regarding the death of Pedro Lima, in 2020; nevertheless, this case was second in terms of television news items (32), even though their number was not much different from the number of news items about Pedro Lima’s case collected from the three radio stations (27).

In Table 3, we can identify changes in the level of compliance with the WHO guidelines between 1996 and 2020 or different trends depending on the reported cases. The values are expressed in percentages relative to the total number of news items about each case study. They appear in green when the results mostly meet the WHO recommendations (values above 50%) and in red when they do not.

Table 3. Compliance with the WHO guidelines by case study.

	Triple youth suicide (1996)	Miguel Ganhão Pereira (2000)	Cândida Branca Flor (2001)	Gil Sousa (2018)	João Branco (2018)	Pedro Lima (2020)	Total	Total news items analyzed
1. No prominent location (1)	37%	27%	60%	100%	100%	60%	55%	303
2. No images of the dead	97%	100%	100%	100%	100%	100%	99%	347
3. No images of the suicide method (If applicable)	79%	100%	100%	100%	100%	100%	96%	347
4. No images of the suicide scene/place	68%	100%	84%	100%	100%	86%	86%	347
5. The word suicide is not used in the title (If applicable)	25%	67%	100%	82%	88%	98%	91%	233
6. No use of sensationalist language/sensational presentation of suicide	27%	75%	96%	41%	72%	73%	65%	374
7. No glorification of suicide victims as martyrs or objects of adulation	100%	50%	80%	82%	70%	65%	73%	374
8. Mention of the existence of a mental health problem	5%	0%	28%	6%	4%	27%	19%	374
9. Citation of a mental health professional	13%	0%	0%	0%	0%	3%	4%	374
10. Suicide is not presented as a solution to problems	81%	100%	72%	76%	100%	99%	93%	374
11. No simplistic reasons given for suicide	57%	100%	52%	35%	100%	92%	82%	374
12. No use of language that normalizes suicide	100%	100%	100%	100%	100%	100%	100%	374
13. The method of suicide is not described explicitly	35%	75%	68%	88%	87%	88%	77%	374
14. No detailed information about the location of the suicide	40%	75%	20%	29%	96%	44%	48%	374
15. No citation of a suicide note	40%	100%	100%	94%	100%	87%	82%	374
16. Suicide is either confirmed or not referred (not presented as hypothesis)	100%	100%	100%	94%	100%	94%	97%	374
17. Due consideration for people bereaved by suicide	56%	92%	96%	82%	100%	86%	83%	374
18. Support contacts or information on where to seek help are provided	2%	0%	0%	0%	2%	4%	3%	374
19. Alternatives to suicide are presented	14%	0%	0%	0%	0%	6%	6%	374
20. The news piece educates the audience about suicide	38%	8%	0%	0%	0%	11%	13%	374
21. Indication of risk factors and warning signs of suicidal behavior	60%	0%	0%	0%	0%	10%	16%	374
22. There is an association between depression and suicidal behavior	11%	0%	32%	6%	4%	25%	19%	374
23. Description of physical consequences of non-fatal suicide attempts	0%	0%	0%	0%	0%	0%	0%	374
24. Stories about coping with stressors and suicidal thoughts are presented	2%	0%	0%	0%	0%	5%	3%	374

Note. (1) The links to internet news, apart from those of CM, do not provide any indication as to the positioning of those news in the corresponding websites. Therefore, it is not possible to assess whether they were published in a prominent place. The links to CM’s online news do contain the indication ‘highlights’ or ‘main highlights.’

The case of the triple youth suicide presented a total of 14 mostly negative criteria; the cases of journalist Miguel Ganhão Pereira and singer Cândida Branca Flor had 10; the case of jeweler Gil Sousa had 12; the case of designer João Branco had nine; and the case of actor Pedro Lima had 10.

Regarding the first criterion, most of the news about the triple youth suicide and the death of Ganhão Pereira were found in a prominent location, while the news about the other four cases were not. As for the fifth criterion, when reporting the death of singer Cândida Branca Flor, none of the newspaper articles or newspaper websites used the word suicide in the title. By contrast, in the online news items concerning the death of Gil Sousa, if we were to consider the use of synonyms or expressions that convey the idea of suicide as cases of a reference to suicide in the title, we would have a tie of 8:8.

Concerning the triple youth suicide, most of the news items were in a prominent place, used sensationalist language, described the method of suicide, and quoted suicide notes. The majority also used the word suicide in the title, and there was not a big difference between those that fulfilled and those that did not fulfil the criterion of showing consideration for the bereaved (56% ‘Yes’ vs. 44% ‘No’). Regarding the coverage of Gil Sousa’s death, in 2018, most of the news items also used sensationalist language and presented simplistic reasons for the suicide.

Apart from these situations, there seems to be consistency among the various case studies in terms of compliance with most of the WHO recommendations: there is mostly no use of images of the deaths, nor of the methods or location of the suicide; there is no glorification of the victim; suicide is not presented as a solution to problems, nor it is normalized; the suicide method is not mentioned, nor are suicide notes cited; suicide is not presented as an hypothesis or supposition; and consideration is shown for the bereaved.

On the negative side, however, we found detailed information about the suicide location in 52% of the news items, while potentially helpful characteristics were largely absent from the reporting. Most news items did not mention the existence of a mental health problem, nor did they cite a mental health professional; no support contacts nor other information on where to seek help were included; no alternatives to suicide, nor risk factors and warning signs, were presented; no association was made between depression and suicidal behavior (Table 4).

Table 4. Compliance with the WHO guidelines by media outlet.

	TVI	RTP	SIC	Público	CM	Público website	CM website	Other websites	RR	Antena 1	TSF	Total	Total news items
1. No prominent location	50%	70%	43%	45%	38%	–	79%	–	7%	43%	50%	55%	303
2. No images of the dead	100%	82%	100%	100%	100%	100%	100%	100%	–	–	–	99%	347
3. No images of the suicide method (If applicable)	95%	82%	84%	100%	98%	100%	100%	100%	–	–	–	96%	347
4. No images of the suicide scene/place	81%	64%	64%	100%	85%	100%	89%	100%	–	–	–	86%	347
5. The word suicide is not used in the title (If applicable)	–	–	–	55%	92%	79%	100%	84%	–	–	–	91%	233
6. No use of sensationalist language/ sensational presentation of suicide	59%	73%	68%	45%	42%	95%	69%	58%	93%	100%	100%	65%	374
7. No glorification of suicide victims as martyrs or objects of adulation	68%	64%	93%	73%	67%	74%	78%	72%	57%	29%	33%	73%	374
8. Mention of the existence of a mental health problem	0%	9%	14%	9%	29%	32%	38%	6%	0%	0%	0%	19%	374
9. Citation of a mental health professional	5%	9%	5%	18%	2%	21%	1%	0%	0%	0%	0%	4%	374
10. Suicide is not presented as a solution to problems	90%	100%	82%	73%	100%	89%	97%	94%	100%	100%	100%	93%	374
11. No simplistic reasons given for suicide	81%	100%	57%	64%	73%	84%	92%	82%	100%	100%	100%	82%	374
12. No use of language that normalizes suicide	100%	100%	100%	100%	100%	100%	100%	100%	100%	100%	100%	100%	374
13. The method of suicide is not described explicitly	69%	82%	70%	36%	71%	63%	85%	84%	100%	100%	100%	77%	374
14. No detailed information about the location of the suicide	54%	45%	43%	27%	54%	42%	46%	76%	14%	0%	0%	48%	374
15. No citation of a suicide note	69%	82%	66%	82%	83%	79%	84%	98%	100%	100%	100%	82%	374
16. Suicide is either confirmed or not referred (not presented as hypothesis)	100%	100%	93%	100%	98%	95%	93%	98%	100%	100%	100%	97%	374
17. Due consideration for people bereaved by suicide	88%	82%	70%	55%	75%	95%	81%	94%	100%	100%	100%	83%	374
18. Support contacts or information on where to seek help are provided	0%	0%	0%	9%	2%	53%	0%	0%	0%	0%	0%	3%	374
19. Alternatives to suicide are presented	7%	0%	5%	18%	6%	32%	4%	0%	0%	0%	0%	6%	374
20. The news piece educates the audience about suicide	27%	0%	9%	27%	10%	37%	12%	0%	0%	0%	0%	13%	374
21. Indication of risk factors and warning signs of suicidal behavior	42%	0%	14%	45%	15%	32%	10%	0%	0%	0%	0%	16%	374
22. There is an association between depression and suicidal behavior	5%	9%	16%	9%	29%	37%	32%	6%	0%	0%	0%	19%	374
23. Description of physical consequences of non-fatal suicide attempts	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	374
24. Stories about coping with stressors and suicidal thoughts are presented	0%	0%	2%	9%	4%	26%	3%	0%	0%	0%	0%	3%	374

SIC, Público, CM, RR, and Antena 1 mostly published news items about the case studies in a prominent place. More than half of Público’s and CM’s news items used sensationalist language. None of the media mentioned the existence of a mental health problem in the majority of the news items, nor did they cite a mental health professional. Furthermore, most of them presented detailed information on the location of the suicides. None of the media, apart from Público’s website, included support contacts or information on where to seek help on most of the news items. Nor did most of the news items from all the analyzed media present alternatives to suicide. They neither educated the audience about it, nor indicated risk factors or warning signs of suicidal behavior. More than 90% of the analyzed news items from all the media, except Público’s website, did not present stories about coping with stressors and suicidal thoughts.

Despite its negative performance, Público’s website stands out for presenting alternatives to suicide and indicating risk factors in 32% of the analyzed news items; educating the audience about suicide and associating it with depression in 37% of them; and presenting stories about coping with stressors in 26% of the cases. In fact, of the 12 pieces about Pedro Lima’s death published on Público’s website, more than half (seven) used his death as a starting point to address suicide and mental health issues, such as depression. The Público newspaper also presented suicide risk factors in 45% of the analyzed news items.

On the positive side, only Público’s website stands out with more than half of its news items showing support contacts. In all the other media, between 91% and 100% of published news items did not present support contacts or information on where to seek help.

On the negative side, the CM exploited Pedro Lima’s death to a fairly large extent, both on its newspaper and website. This was the media outlet that published news items about the actor’s death on its website for a longer period, a total of 51 days. All the other media presented a short news coverage (no more than 8 days) of all the analyzed suicides.

Discussion

Our findings suggest that the Portuguese media displayed suicide news with some moderation and in compliance with many of the WHO guidelines for responsible suicide coverage. The fact that 77% of the news items did not describe the suicide method explicitly, with only 23% violating this WHO recommendation, is in line with the numbers registered in Ireland (20%) (McTernan et al., 2018) and Canada (24%) (Creed & Whitley, 2017) and indeed better than those found in the UK (31%) (Utterson et al., 2017), Pakistan (34%) (Kamboh & Ittefaq, 2019), South Korea (37%) (Lee et al., 2014), the US (46%) (Carmichael & Whitley, 2019), and India (93%) (Armstrong et al., 2018).

Although they did not include information that contributes to suicide prevention, the Portuguese media did comply with most of the ‘don’t do’ guidelines. The fact that suicide is still a taboo topic in most Portuguese newsrooms (Ribeiro & Granado, 2022) is probably contributing to this type of coverage. Nevertheless, this attitude towards suicide may also be inhibiting improvements in suicide reporting, such as the inclusion of preventive information that could help people at risk.

Few of the analyzed news items had the potential to contribute to suicide prevention, and only 3% presented support contacts or information on where to seek help, which is a worrying finding. This result is lower than the 11% of US articles (Carmichael & Whitley, 2019) and the 10.5% of South Korean news items (Lee et al., 2014) and far lower than the 27% of Canadian articles (Creed & Whitley, 2017) but close to the 2.5% of Indian articles that provided contact details for a suicide support service (Armstrong et al., 2018). Furthermore, only 16% of all the news items provided information on how to recognize warning signs and risk factors, while mental health professionals were cited in only 4% of the total sample, which is a very negative performance, although better than the 1.1% of Indian articles that featured opinions from mental health professionals (Armstrong et al., 2018).

These results are also in line with those of another study with Canadian newspapers and news websites that concluded that there was high adherence to several of the recommendations, such as not presenting a monocausal explanation or not including sensational language, but at the same time there was low adherence to recommendations related to protective content (Antebi et al., 2020).

Focusing on the Portuguese case studies, the triple youth suicide stood out negatively from the global tendency regarding the prominent location criterion, the use of the word suicide in the title, the traces of sensationalist language, the description of the suicide method and location, and the citation of suicide notes. While it bears noting that this suicide occurred before the WHO released the guide with recommendations for media professionals, in 2000, it should also be pointed out that as early as 1993, in a set of guidelines dedicated to suicide (WHO, 1993), the WHO had already devoted some paragraphs to press reports on suicides.

Furthermore, as regards the news coverage of journalist Miguel Ganhão Pereira’s death in December 2000, at the age of 29, none of the analyzed news items presented comments from mental health experts nor information on suicide prevention. By contrast, a study about the suicide coverage of a popular television reporter in Quebec, Canada, who died in January 1999, at 33 years old, found that about one third of the analyzed articles (98) presented information on suicide and comments from experts (Tousignant et al., 2005).

During the ten days that followed the death of actor Pedro Lima, we found significant differences between the coverage of the CM newspaper and website and that of all the other media. CM’s front-page headlines were particularly eye-catching, sometimes mentioning a negative life event purported to be a causal factor in the suicide: ‘Look After My Children’ (CM, June 21, 2020), ‘Leaves family in financial distress’ (CM, June 22, 2020), and ‘Unbearable pain’ for fear of failure’ (CM, June 23, 2020).

Only CM published details about the suicide method used by actor Pedro Lima and presented simplified causes for the actor’s death, specifically, his depression and worries about his future. This is in line with a recent study that showed that almost 90% of Portuguese journalists consider that suicide stories should use selective disclosure of data, even though there is a minority for whom the media should publish all the information and details about a suicide if they have access to it (Ribeiro & Granado, 2022). In CM, we also found traces of sensationalism beyond the exploitation of detail, notably the use of images of the suicidal person’s grieving family or friends and images of the funeral showing their suffering.

In addition, although the CM and Público newspapers present similar percentages for sensationalist language, their total numbers in that respect are quite different. In Público, only six news items (concerning the two older case studies) presented this feature, against five (regarding the more recent cases) that did not. In CM, there were 28 news items (24 of which about Pedro Lima) that used sensationalist language, against 20 that did not. In this sense, we believe there was a ‘dilution effect’ in the results, that is, the media with smaller numbers of news items about the analyzed case studies showed higher percentages in some negative criteria, while the CM newspaper and the CM website seem to have benefited from a high total number of news items in the results.

Another conclusion proceeding from the content analysis is that CM never used the word suicide in the titles on Pedro Lima’s death, either in the newspaper or on the website, while in Público’s website we found the word suicide in three titles, but in pieces that used the actor’s death as a starting point to address mental health issues.

Despite this specific media performance, the Portuguese media performed well in a total of 14 criteria, against 10 criteria in which it performed poorly. This is a much better performance than the one registered in the media reports from seven television news channels and all newspapers in Taiwan, following the suicide of a famous television actor there, characterized by sensational words and the repeated and detailed indication of the method used (Cheng et al., 2007).

Conclusion

To the best of our knowledge, this is the first study to analyze suicide media coverage in several types of mass media and making the necessary methodological adjustments to take their idiosyncrasies into account. In addition, since we included case studies from different moments over 25 years, this study also allowed us to look for possible changes over time. We can conclude that there was a positive evolution in suicide reporting in Portugal between our first case study, from 1996, which presented mainly negative results for 14 of the 24 criteria drawn from the WHO guidelines for suicide reporting, and the more recent case of Pedro Lima, from 2020, which presented mostly negative results for 10 of those criteria.

However, most of the guidelines concerning potentially preventive information were not followed by the Portuguese media. We found few examples of news items that used the analyzed suicides to start a discussion on suicide prevention and mental health, which could have raised the public’s awareness and contribute to behavior change (Carmichael & Whitley, 2019; Wakefield et al., 2010). In this sense, there is much room for improvement.

This study has several limitations. First, despite our best efforts, it seems almost impossible to remove all subjectivity from the news analysis, which introduces the possibility of observer bias in this study. Nevertheless, we took several steps to decrease such bias. The first author has years of study of media coverage of suicide and was thoroughly supervised by the second author in all the stages of data analysis. Second, our goal of analyzing the news items released since the date of each studied death until the present proved to be much too ambitious, since we depended on the archives of the selected radio stations and television channels, which, in many cases, were not available. In addition, even the online search was conditioned by the apparent change of website servers over time. Third, we considered news items instead of news articles, since our media selection involved more than newspapers, but we have put our results in relation to those of other studies that analyzed only newspapers to have a broader basis for comparison.

Future research in Portugal could focus on social media, since some studies stress the potential for copycat behavior in connection with social media posts on celebrity suicides (Ueda et al., 2017). It would also be interesting to have more international studies involving various types of media besides newspapers, specifically, television, radio, and online news.

Authors’ contributions

Eudora Ribeiro: conceptualization (lead); data curation (lead); formal analysis (lead); investigation (lead); methodology (lead); project administration (lead); resources (lead); writing – original draft (lead). António Granado: conceptualization (supporting); investigation (supporting); methodology (supporting); project administration (supporting); supervision (lead); validation (lead); writing – review & editing (lead).

Disclosure statement

The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data that support the findings of this study are available on request from the corresponding author.

Additional information

Funding

This study was supported by the FCT – Fundação para a Ciência e a Tecnologia under Grant SFRH/BD/143780/2019 and Project UIDB/05021/2020.

Notes

1 To visualize the data, it is necessary to change the selection conditions of the table. To do so, click on “Alterar condições de seleção”; select “Local de residência (NUTS – 2013)”; click on “+Portugal”; click on “+ Continente”; select the five regions below; and then select “Visualizar quadro.”.