P032 Comparison of the GPSS between indigenous and non-indigenous Australian patients

Abstract Aim OSA screening tools may require adjustments for use between differing populations. The General Practice Sleep Scale (GPSS), developed for GP settings, was tested for its accuracy between Indigenous and non-Indigenous Australian OSA patients to determine factors which may warrant population specific thresholds. Methods This study included all patients from our centre in the NT of Australia with data to derive the GPSS score. Indigenous and non-Indigenous patients were matched on sex, age, BMI, and presence of COPD, hypertension, heart disease, diabetes or depression. Results 234 patients (117 (50%) Indigenous, 56.4% female, median age 48.5 years, BMI 33.9 kg/m2) were included. More non-Indigenous patients than Indigenous had BMI ≥40kg/m2 (35 vs. 13%) – although 70% of both Indigenous and non-Indigenous had a BMI ≥ 30kg/m2. Indigenous patients reported less morning tiredness (38.5% vs. 50.4%, p=.066) and more daytime sleepiness (39.3% vs. 25.6%, p=.026). 62.4% of Indigenous and 55.6% of non-Indigenous recorded moderate-to-severe OSA (AHI ≥ 15) respectively. Median GPSS scores were 10 (IQR 7, 13) for Indigenous (22.2% low risk (≤7), 49.6% moderate risk (8-13), 28.2% high risk (≥14)) and 11 (IQR 8, 14) for non-Indigenous (26.5% low, 48.7% moderate, 24.8% high risk). The area under the curve (AUC) for moderate-to-high risk score was 0.616 (95% CI 0.530, 0.701) for Indigenous and 0.664 (95% CI 0.589, 0.738) for non-Indigenous patients (sensitivity and specificity of 82.2 & 40.9% and 92.3 & 40.4% respectively). Conclusion Differences in symptoms, anthropometrics and OSA prevalence between Indigenous/non-Indigenous patients may lead to reduced accuracy of OSA screening tools.