From the perspective of an idealistic researcher, research in the field of suicide prevention has obvious merit. Understanding this subject and what works to prevent it will eventually lead to better help and care for individuals who are at-risk. Potentially, we can save lives.
There are important ethical questions to consider when conducting research that involves individuals who are at-risk of suicide. To what extent is participating unpleasant or distressing? Is this burden acceptable and is it outweighed by the benefits of the study? There’s another important question, but I’ll take a look at the benefits and the burden first.
Benefits can be discussed in terms of direct benefits for the participants in the study, e.g. participating in a study might give quick access to care. More broadly, research should also eventually lead to something useful for individuals at-risk of suicide in general. For example, better access to care, better preventive activities, more effective treatment, better understanding among the general public. Fundamental research can also be beneficial, for example if it can inform development of new treatment. The burden, on the other hand, is that a researcher asks someone to do something they would not normally do and/or causes someone to get a treatment that is different from the treatment they would normally get. The extent of this burden differs a lot per study, of course. It can just take time, can be bothersome, can be distressing. Sometimes participating in research comes down to answering a few questions, and sometimes it involves trying an experimental drug where the effect on mental wellbeing or suicide risk is unknown.
Whether the benefits of a study outweigh the burden is a complicated question and that’s why study protocols are carefully examined by ethics boards. There is, however, an ethical question I recently thought of that ethics board have never asked (in my experience), which is: can the research question be answered by analysing existing data? If so, burdening people is unnecessary and conducting the study would be unethical. Adding to that, I would say that if by analysing existing data I could answer a research question that is closely related to my research question, I should do that first, and then decide whether gathering new data is necessary. The same applies to replicating a study’s results. Replication might necessitate gathering new data, but first the researcher should see whether a study can be replicated by using an existing dataset.
Sources of existing data include health records and data gathered for previous studies. Imagine how many patients have been in mental health care in, let’s say, the past 5 years. Consider that each year there are more studies on suicide prevention published than in each previous year. Imagine how many people have been included in these studies. Getting the data you want may not be easy, but neither is gathering new data. I still have to meet a researcher in the field of suicide prevention (or mental health in general) who says that finding participants and gathering new data is easy.
Analysing existing data comes with additional ethical questions concerning privacy and consent, because it is usually not possible to ask people whether you can use personal data that concerns their mental health. If you cannot ask permission, you can still use the data for research, provided that the data are completely anonymous. That’s because international laws on privacy (such as the GDPR) apply only to data that relate to an identifiable individual. Under which conditions data can be considered to be anonymous is a different subject, and I won’t get into that now. Compared with health records, perhaps data gathered for previous studies is the most ethical source of data. Participants in those studies knowingly and willingly participated in research. It can be argued that the consent they gave may not have included using their data for other research, but it can also be argued that the study that re-uses their data can be considered an extension of the study for which they consented. The purpose of the new study is (likely to be) very similar to the purpose of the original study. Some good news is that I’ve noticed that institutions that provide research grants are recently warming up for the idea to use existing sources of data for answering research questions, and require researchers to make sure the gathered data are FAIR (findable, accessible, interoperable and reusable).
To conclude, I would make the case that one shouldn’t burden vulnerable people unless it is absolutely necessary. Researchers should analyse, re-analyse, meta-analyse and again re-analyse all existing data of individuals at-risk of suicide, until every bit of information that can advance the field is known. Therefore, sharing data is essential. Not only from the open science perspective and being transparent, but also to enable analysis for answering other research questions, which may prevent unnecessary data gathering.