Modified screening improves prediction of suicidal behavior among adolescents
Click Here to Manage Email Alerts
Key takeaways:
- The PHQ-9M for teens includes four additional questions about depression and suicidal ideation.
- The extra questions on the modified questionnaire had the strongest association with subsequent suicide attempts.
A modified screening with additional questions about suicidal ideation was better at predicting suicide attempts among adolescents than the standard questionnaire, according to findings published in JAMA Network Open.
Researchers performed a retrospective cohort study of 130,028 adolescents (50.4% boys) aged 12 to 17 years who were screened for depression with the Patient Health Questionnaire-9 modified for teens (PHQ-9M) in pediatric outpatient settings between Jan. 1, 2016, and Dec. 31, 2022. Before November 2017, adolescents aged younger than 16 years were not routinely screened for depression during well visits, the researchers noted. They followed up with each participant up to 1 year after their the initial screening to document suicidal behavior.
“The PHQ-9 modified for teens (PHQ-9M), a self-report screen for depression during pediatric care visits, was developed based on the PHQ-9, PHQ for adolescents and the Columbia Depression Scale,” Fuchiang Tsui, PhD, FAMIA, director of the Tsui Laboratory in the department of biomedical informatics and entrepreneurial science at The Children’s Hospital of Philadelphia. “The PHQ-9M retains the nine core items from PHQ-9 and four supplemental items assessing depression in the past year (item 10), functional impairment (item 11), serious suicidal ideation in the past month (item 12) and lifetime history of suicide attempts (item 13).”
Tsui and colleagues compared the PHQ-9M’s ability to predict suicidal behavior vs. the standard PHQ-9. They also identified the top three items for predicting subsequent suicide attempt.
Overall, 549 adolescents (0.4%) attempted suicide a median of 172 days (interquartile range [IQR], 82-267) after being screened. The researchers found that a high score for item 9 on both questionnaires, which indicates daily suicidal thoughts or thoughts of self-harm, performed the best for predicting suicide attempt, with a positive predictive value of 4.8% (95% CI, 3.2%-6.6%).
The additional questions in the modified questionnaire had the strongest association with future suicide attempts, Tsui and colleagues wrote. Previous suicide attempt was the strongest predictor, tripling the odds (adjusted HR = 3.06; 95% CI, 2.47-3.8), followed by depressive symptoms in the past year (aHR = 2.99; 95% CI, 2.32-3.86) and suicidal ideation in the past month (aHR = 1.63; 95% CI, 1.25-2.12). Although it had the highest positive predictive value, the aHR for item 9 was not statistically significant, the researchers noted.
According to Tsui and colleagues, the PHQ-9M was more accurate at predicting suicide attempts within 1 year, compared with the PHQ-9 (area under the receiver operating characteristic curve = 0.8; 95% CI, 0.78-0.82 vs. 0.77; 95% CI, 0.75-0.79).
“All four supplemental items (10-13) had statistically significant aHRs,” Tsui and colleagues wrote. “Thus, the clinical utility of the PHQ-9M for suicide risk prediction was above and beyond information gleaned from PHQ-9, which highlights the importance of incorporating the supplemental items into clinical risk determinations and decision-making.”
Perspective
Back to Top
Of course, there are limitations to a study like this; study populations and data that were collected as part of routine clinical care are less standardized than those collected as part of a research protocol. However, there are some very practical takeaways from this study. Above all, the main takeaway is that screening for suicide can, indeed, identify patients who may be at risk of suicidal self-harm. Many children who die by suicide have encounters with the health care system within several months of their deaths, which means that pediatricians' screening practices can potentially identify these children early enough to intervene.
There are two other considerations for pediatricians in regard to this study.
The first is that although screening may be fruitful in identifying patients at risk of suicide, screening only for depression is not sufficient for identifying suicide risk; in fact, many children with suicidal ideation or intent will not be picked up by depression screening alone. The Tsui study appropriately distinguishes between the traditional PHQ-9, a self-reported depression screening, and the PHQ-9M, which includes questions about suicide risk specifically. In accordance with prior literature, the screening tool with questions specific to suicide identified this risk more effectively than the depression-only tool. In the end, the pediatric care provider is advised to select a tool with good test properties for suicide risk identification.
In a similar vein, the second consideration from this study also has to do with the choice of screening tool. Most suicide risk screening tools are written at about a 5th grade reading level, which may be prohibitive to screening younger children or individuals with disabilities that affect reading cognition or communication. However, a mounting body of evidence suggests that children as young as 5 years old can, indeed, attempt and die by suicide, and individuals with disabilities appear to have not only an increased risk of suicidal thoughts and behaviors but may also have earlier onset of suicidal ideation than the typically developing population. These are certainly patients whose suicide risk should be assessed using language that is appropriate for developmental communication level.
Collapse
Read more about
Click Here to Manage Email Alerts