The issue of suicidality (suicidal thoughts and/or behaviors) in relation to medications is challenging. Medications might increase or decrease suicidality. When suicidality increases with medications, the size of the increase is usually small, which makes it hard to identify. Decreases in suicidality would be a highly desirable attribute of medications, but very few medications have been found that have this desirable effect to date. Further complicating this issue is the finding that suicidality may be increased not only by antidepressants, but also by antiepileptics, as well as some drugs for smoking cessation, acne, asthma and allergy. Medications that have a direct or indirect effect on the brain through interactions with bodily chemicals could, theoretically, cause increased suicidality. Defining the magnitude of any such risks is part of the role of the Food and Drug Administration, just as FDA tries to identify any adverse effects of all medications to increase the benefit-risk ratio for patients receiving them. In summary, the small size of increased suicidality risk if any is present at all and the possibility that increases may occur in many classes of medications makes the identification of these risks and their magnitude difficult but important.

A person experiencing suicidal thoughts or behaviors is the best source of information about his or her thoughts and behaviors – occasionally additional information is available from others. Suicidality is stigmatized and people experiencing suicidality are often reluctant to reveal their thoughts and behaviors to clinicians. Some clinicians are also reluctant to ask patients about suicidality. Research over the past 35 years has consistently shown that people disclose more
suicidality in computer interviews than when asked the same questions by clinicians. Even when clinicians do their very best assessing suicidality, being human, they inevitably do it somewhat differently than other clinicians and even differently themselves from time to time. The inevitability of limited human assessment reliability puts an upper limit on the accuracy possible in any kind of assessment, including suicidality. Beyond problems of reliability, human record keeping is variable in accuracy, calculations, completeness, and legibility, problems computers largely overcome.

For all these reasons, use of computer interviews to assess suicidality offer substantial advantages compared with human interviewers. The interviews are the result of careful composition of wording and branching by experts on suicidality, faithfully conduct each interview as an ideal human interviewer would, are accessible any time and any where a computer or telephone is available and store data in electronic data bases, where they can be used to alert clinicians of results and preserved for analyses as data accumulate. The goal of accurately identifying suicidality signals, positive or negative, is well-served by the use of computer interviews of patients at possible risk of
suicidality.