Abstract
Aim: To ascertain the association between antidepressant and suicidal ideation, self-harm, suicide attempt or death by suicide (ideation and death by suicide) and self-harm) within a secondary psychiatric healthcare setting.Methods: A mental healthcare database with appended text-mined data was used for analyses (including validation exercises and an exploratory study). A retrospective cohort study was carried out, in patients who received a diagnosis of depression during their first referral to psychiatric care services, in a series of studies aimed at understanding factors influencing antidepressant use, factors associated with new instances of suicidal ideation, self-harm, suicide attempt or death by suicide, identifying common antidepressant treatments and assessing the role of antidepressants in future instances of recorded suicidal ideation, self-harm, suicide attempt or death by suicide.
Main analysis: Sensitivity and specificity tests were conducted to evaluate text-mining applications. Logistic regression analyses were conducted to determine factors associated with antidepressant treatment and suicidal ideation, self-harm, suicide attempt or death by suicide. Cox regression analyses were conducted to longitudinally assess associations between antidepressants with future recorded suicidal ideation, self-harm, suicide attempt or death by suicide.
Results: For the text-mining applications, I found a precision of 91.7% and 82.8% for identifying suicide ideation and suicide attempt, respectively. The study ascertaining factors associated with antidepressant treatment found indications of comorbidity were associated with newer antidepressant use (e.g. SNRIs or mirtazapine). Different depressive symptom clusters were associated differently with antidepressant treatment use, suggesting clinicianpatient communication. In addition, in patients older than 25 years, past suicide attempts were associated with increased use of fluoxetine relative to sertraline. The prevalence of new suicidal ideation, self-harm and deaths by suicide were observed in 8.76%, 1.67% and 0.24% of the cohort, respectively. All three clusters of depressive symptoms – affective, somatic and cognitive – were associated with increased risk of suicidal ideation, self-harm, suicide attempt or death by suicide, compared to their patients who did not experience these symptoms.
In terms of the longitudinal association of specific antidepressant treatment with suicidal ideation, self-harm, suicide attempt or death by suicide, results were found to be in line with current literature but also novel: i) There was no association of most of the specified antidepressants with suicidal ideation, self-harm, suicide attempt or death by suicide in young or older adults; ii) however, results suggest that compared to citalopram, mirtazapine and sertraline may be associated with increased risk of suicidal ideation in young people (HR 5.99; 95% CI 1.30 – 27.70; p-value 0.021 and 4.66; 1.52 – 14.24; 0.007, respectively), who experience cognitive symptoms (i.e. helplessness, worthlessness and hopelessness) during antidepressant treatment, iii) whereas mirtazapine (0.46; 0.22 – 0.95; 0.04) or dual combinations of antidepressant treatment (0.40; 0.18 – 0.90; 0.025) is protectively associated with ideation in adults, who experience cognitive symptoms, iv) contrastingly compared to citalopram, mirtazapine may increase the risk of self-harm in adults (23.7; 1.42 – 394.8; 0.02) who experience cognitive symptoms while and v) collectively, the results suggest a general role for mirtazapine increasing the risk of ideation compared to citalopram in adults, in the presence of certain symptoms. Hence, no direct evidence was found between antidepressants and suicidal ideation, self-harm, suicide attempt or death by suicide however there may be an indirect role via depressive symptoms.
Conclusions: The current study is one of the first studies to use a naturalistic clinical psychiatric database in research studying the potential role of antidepressants in treatmentemergent suicidal ideation, self-harm, suicide attempt or death by suicide (suicidal ideation, self-harm and death by suicide) (See Chapter 2 and 7). The study calls for attention on patient symptom profiles over the course of mirtazapine treatment in secondary mental healthcare (Chapter 8). The limitations faced in this study (Chapter 5 and 6) requires all conclusions to be interpreted with caution and primarily calls for further studies to validate the current findings in particular dictated by substantial qualitative work given the inconsistencies in clinical text (Chapter 9).
Date of Award | 1 Jun 2021 |
---|---|
Original language | English |
Awarding Institution |
|
Supervisor | Robert Stewart (Supervisor) & Rina Dutta (Supervisor) |