Concurrent validity and screening accuracy of the Malay Ask Suicide-Screening Questions (ASQ) among a Malaysian psychiatric sample

The Ask Suicide-Screening Questions (ASQ) is a practical and brief suicide risk screening tool. However, the ASQ has never been validated in the Malay language. Therefore, this study aimed to translate the ASQ into Malay and assess its concurrent validity and screening accuracy in detecting suicide risk among a sample of Malaysian patients with psychiatric illness in a low- and middle-income setting. We performed forward-backward translation, and assessment of face and content validity of the ASQ. A total of 133 psychiatric inpatients and outpatients were included. Participants first completed the Malay version of the ASQ, administered by trained medical students. Participants were then assessed with a clinical interview guided by the Brief Suicide Safety Assessment (BSSA), conducted by psychiatry medical officers blinded to ASQ results. The ASQ was scored dichotomously (positive/negative; acute/non-acute risk), while the BSSA served as the gold standard. Participants were aged 19–70 years (mean = 36.1 ± 12.9), with most females (73.7%) and Malay (77.4%). Positive screens were significantly younger than negatives (33.14 ± 11.98 years versus 41.06 ± 12.93 years, t (131) = 3.58, p < 0.001. Compared with the clinical interview guided by the BSSA, the ASQ showed a sensitivity of 98.3%, specificity 65.3%, positive predictive value 68.7%, negative predictive value 98.0%, overall accuracy 79.7% and moderate agreement (κ = 0.606). Receiver Operating Characteristic (ROC) analysis demonstrated good validity [Area under the curve (AUC) = 0.818 (95% CI [0.745–0.891], p < 0.001)]. The Malay ASQ was found to be a valid and reliable screening tool for detecting suicide risk among psychiatric patients. This supports its use as a culturally relevant and practical screening instrument to facilitate early identification and timely management of at-risk patients with psychiatric illness in a Malaysian sample. Not applicable.