Optimised thresholds for APSQ in predicting a positively screen on a subsequent questionnaire by sex, games and season
| All athletes | Sex | Games | Season | |||||||||||
| Male | Female | Olympic | Paralympic | Summer | Winter | |||||||||
| Threshold | AUC | Threshold | AUC | Threshold | AUC | Threshold | AUC | Threshold | AUC | Threshold | AUC | Threshold | AUC | |
| GAD-7 | ≥18 | 0.86 | ≥16 | 0.88 | ≥18 | 0.85 | ≥19 | 0.86 | ≥20 | 0.86 | ≥19 | 0.84 | ≥18 | 0.93 |
| PHQ-9 | ≥20 | 0.89 | ≥20 | 0.95 | ≥20 | 0.87 | ≥19 | 0.89 | ≥28 | 0.90 | ≥20 | 0.91 | ≥15 | 0.86 |
| PHQ-9 Q9 | ≥17 | 0.82 | ≥17 | 0.81 | ≥18 | 0.86 | ≥17 | 0.82 | ≥20 | 0.89 | ≥18 | 0.80 | – | – |
| ASSQ | ≥17 | 0.69 | ≥14 | 0.68 | ≥16 | 0.70 | ≥17 | 0.69 | ≥17 | 0.65 | ≥17 | 0.67 | ≥16 | 0.71 |
| AUDIT-C | ≥13 | 0.59 | ≥13 | 0.60 | – | – | ≥13 | 0.59 | – | – | ≥14 | 0.60 | ≥13 | 0.59 |
| CAGE-AID | ≥18 | 0.66 | – | – | ≥19 | 0.76 | ≥18 | 0.67 | – | – | ≥12 | 0.58 | ≥18 | 0.76 |
| BEDA-Q | ≥13 | 0.64 | ≥11 | 0.64 | ≥14 | 0.65 | ≥11 or ≥13 | 0.64 | ≥17 | 0.75 | ≥13 | 0.64 | ≥12 | 0.67 |
Note: AUC of ROC curve for BEDA-Q when modelling Olympic athletes only was equivalent for a threshold of ≥11 and ≥13. Models and therefore optimised thresholds could not be computed for some predictions due to the absence of positive cases or where models’ best threshold performed worse than an AUC of ROC curve=0.5.
APSQ, Athlete Psychological Strain Questionnaire; ASSQ, Athlete Sleep Screening Questionnaire; AUC, area under the curve; AUDIT-C, Alcohol Use Disorders Identification Test Consumption; BEDA-Q, Brief Eating Disorders in Athletes Questionnaire; CAGE-AID, Cutting Down, Annoyance by Criticism, Guilty Feeling and Eye Openings Adapted to Include Drugs; GAD-7, General Anxiety Disorder-7; PHQ-9, Patient Health Questionnaire-9; PHQ-9 Q9, Question 9 of the Patient Health Questionnaire-9; ROC, receiver operating characteristic.