Can hospital visits help predict the risk of suicide?

TIMELY AID: Healthcare settings play a critical role in suicide prevention
| Photo Credit:
Abdullah Durmaz

Death by suicide is not only a tragic loss of life but also leaves a searing impact on families.

A paper published late last year in the NPJ Mental Health journal showed that probability and machine-learning models built using logistic regression and random forest concepts can predict suicide risk better than existing methods.

A logistic regression estimates the probability of an event occurring, while a random forest model is a machine-learning algorithm that uses an ensemble of ‘decision trees’ to make predictions.

Researcher Roy Adams and others used electronic health record data in the US to test these models, which were able to predict suicide risk 83 per cent of the time, as compared with 64 per cent in the case of the screening methods currently used in the West. A 50 per cent success is as good as leaving things to chance.

The authors emphasise that healthcare settings play a critical role in suicide prevention, noting that about 72 per cent of suicide attempts and 50 per cent of suicide deaths happened within 90 days of contact with the health system. This, they say, highlights the scope for intervention within these settings.

Age factor

The study analysed health record data from over 3,31,000 visits by more than 16,000 adults to American-Indian health service providers between 2017 and 2021. In this period, there were 417 suicide attempts and 37 deaths by suicide.

The study found that suicide attempt or death was associated with younger age, a diagnosis of depression, anxiety, bipolar disorder, post-traumatic stress disorder or suicidal ideation, a diagnosis related to alcohol, cannabis or other substance abuse, and a past positive screen for intimate partner violence, suicide risk or depression.

Though the study was intended to help reduce suicide risk among American-Indian and Alaska Native populations, the approach could well apply to India, which has the highest number of suicides in the world and needs timely and targeted interventions to help save lives.

Deaths by suicide increased from 9.9 per one lakh population in India to 10.4 between 2017 and 2019, and then further to 12.4 by 2022, according to the National Crime Records Bureau data. The latest data show that nearly 34 per cent of suicides was in the 15-30 age group.