DARWIN EU® - Suicidality following exposure to doxycycline

Study type

• Study topic:
  Human medicinal product
• Study type:
  Non-interventional study
• Scope of the study:
  Safety study (incl. comparative)
• Data collection methods:
  Secondary use of data

Non-interventional study

• Non-interventional study design:
  Case-control

  Cohort

Study drug and medical condition

• Study drug International non-proprietary name (INN) or common name:
  DOXYCYCLINE
• Anatomical Therapeutic Chemical (ATC) code:
  (J01AA02) doxycycline

  doxycycline
• Medical condition to be studied:
  Suicidal behaviour

  Suicidal ideation

  Suicide attempt

  Suicide threat

  Completed suicide

  Depression suicidal

  Depression

  Anxiety

Population studied

• Short description of the study population:
  The study population is new users of doxycycline (SCCS and cohort study) or the comparators (cohort). The new-user cohorts will be per indication: acne vulgaris: doxycycline, erythromycin or isotretinoin; rosacea: doxycycline, erythromycin or isotretinoin; chlamydia: doxycycline, azithromycin, erythromycin or amoxicillin; and lower respiratory tract infection (CAP and bronchitis): doxycycline, azithromycin, or amoxicillin.

Study design details

• Study design:
  New-user cohort study with active comparator (objective 1) and self-controlled case series (objective 2)
• Main study objective:
  1. To use a new-user cohort study to assess the association between doxycycline and completed suicide, composite suicide and suicide-related events (completed suicide, suicide ideation and suicide attempt, self-harm), composite suicide-related events (suicide ideation, suicide attempt, self-harm), depression and anxiety, compared to active comparators, stratified by indication of acne vulgaris, rosacea, chlamydia and lower respiratory tract infection (CAP or bronchitis) ; 2. To use a self-controlled case series study to assess the association between use of doxycycline and composite suicide-related events (including suicide ideation, suicide attempt, self-harm), depression and anxiety.;
• Setting:
  study was conducted using routinely collected data from 3 primary care data sources in 3 European countries
• Outcomes:
  The primary outcomes for the SCCS and the new-user cohort of the study were different. In the SCCS study, completed suicide was not included because this type of outcome affects subsequent follow-up time such as death breaches the assumptions of the SCCS study design.[28] ; The new-user cohort design outcomes were as follows: Completed suicide (condition record of suicide plus; death date in ±30 days); composite suicide-related events (includes completed suicide, suicide attempt, suicide ideation and self-harm); composite non-fatal suicide-related events (suicide attempt, suicide ideation and self-harm); depression (for general study population, not the subgroup with history of depression); and anxiety. ; The outcomes for the SCCS study were as follows: composite non-fatal suicide-related events (suicide attempt, suicide ideation and self-harm); depression; and anxiety. Due to the risk of duplicate recording events, we only assessed the first outcome occurring after entry into the study.
• Summary results:
  CONCLUSION ; In this drug safety study using two study designs across three European electronic healthcare databases, we identified a two-fold increased association of suicide-related events with doxycycline use compared to erythromycin use in acne patients in the cohort study. However, these findings were not consistent and we; also found an increased risk of non-fatal suicide-related events in the pre-exposure period in the SCCS study, showing there is uncertainty around the causality of the association with non-fatal suicide-related events we find in doxycycline use. We found a smaller but increased association of anxiety with doxycycline use compared to erythromycin and isotretinoin, as well as an increased risk of depression with doxycycline use compared to erythromycin in patients with acne. We did not observe these increased associations for other indications in the cohort study, and in addition, many of the analyses remained blinded due to non passing of diagnostic tests. In contrast to the increased risk of psychiatric outcomes we found in acne patients, we observed a reduced risk of suicide-related events in doxycycline versus amoxicillin in LRTI patients in the cohort study. Although we observed an association between use of doxycycline and suicide related events in individuals with acne, this could not be confirmed in the SCCS, where an increased risk in the pre-exposure period was observed. Studying the association between use of doxycycline and suicide related events in a population of individuals with acne might have been hindered by confounding. It is possible that conditions such as (severe) acne that impact self-esteem, especially in adolescents, might ultimately result in suicidal ideation and attempts.