Shorter, faster, but still accurate: using ant colony optimisation to develop and validate the Athlete Mental Health Screener-18 (AMHS-18) as a brief alternative to Sport Mental Health Assessment Tool 1 (SMHAT-1).

The SMHAT-1 triage has variable predictive validity with false negative rates ranging from 5% to 60% depending on the mental health domain.

• False negative rates for SMHAT-1 triage vary between 5% and 60% depending on the mental health domain assessed.
• The full SMHAT-1 algorithm imposes a substantial time burden for routine screening.
• These limitations motivated development of a briefer, more accurate alternative.

An ant colony optimisation algorithm was used to identify the optimal subset of items from SMHAT-1 for predicting post-interview mental health recommendations.

• The algorithm evaluates item combinations holistically rather than sequentially.
• The algorithm was trained on two datasets: N1=1121 and N2=803 Polish Olympic athletes.
• Data were collected over three waves of routine medical check-ups.
• SMHAT-1 and subsequent brief clinical interviews were administered to all participants.

The optimal solution identified by ant colony optimisation comprised 18 items, forming the Athlete Mental Health Screener-18 (AMHS-18).

• The 18-item solution achieved 81% accuracy in predicting post-clinical interview mental health evaluation in the training datasets (N1=1121, N2=803).
• The AMHS-18 achieved 80% accuracy in the independent validation dataset (N3=1134).
• The AMHS-18 overperformed the SMHAT-1 triage on all performance indicators.
• The optimal cut-off score was ≥19.

The AMHS-18 demonstrated superior sensitivity and specificity compared to SMHAT-1 triage at the optimal cut-off.

• Sensitivity was 94% at the optimal cut-off of ≥19.
• Specificity was 72% at the optimal cut-off of ≥19.
• The AMHS-18 overperformed the SMHAT-1 triage on all performance indicators in both training and independent validation datasets.

The AMHS-18 demonstrated strong internal consistency.

• Internal consistency was measured using omega (ω) = 0.82.
• This was assessed in the independent validation dataset (N3=1134).

The authors recommend using AMHS-18 as a practical alternative to SMHAT-1, particularly in resource-limited or high-frequency screening settings.

• The AMHS-18 is described as 'time-efficient, valid and reliable' for mental health screening in athletes.
• The authors recommend using AMHS-18 instead of SMHAT-1's triage.
• They also recommend alternating SMHAT-1 with AMHS-18 as a starting point for routine brief clinical intake interviews.
• The tool is considered particularly useful 'in settings where time and resources are limited or when athletes are screened regularly.'