Introduction/Aim: The new simplified obstructive sleep apnoea (OSA) screening tool – the “GPSS” – “General Practice Sleep Scale” is developed for use in GP settings. The GPSS was originally developed against the traditional OSA severity criteria – the apnoea-hypopnoea index (AHI). However, the AHI is increasingly being shown to have poor correlation to outcomes such as incident cardiovascular disease or mortality. Novel oxygen saturation metrics, such as the desaturation severity, or markers such as time under 90% oxygen saturation show much greater associations with these outcomes. Therefore, it is useful to test the predictive capacity of the GPSS against these oxygen saturation measures.
Method: This study utilised first visit data from the Sleep Heart Health Study (SHHS). Questions 5-8 of the GPSS (snoring, witnessed apnoeas, sleepiness and presence of hypertension/diabetes/heart disease/depression) were not directly assessed in the SHHS, thus derived from composites. We compared the GPSS against T90%, desaturation severity, resaturation severity and total severity utilising event specific baselines.
Results: A total of 3.959 patients (47.6% female, median age 63 (IQR 55, 71), median BMI 27.8 (IQR 25, 31), OSA 23.8%). Pearsons pairwise correlation was significant for all outcomes (T90%=0.232, Desaturation severity=0.331, Resaturation severity=0.353, Total severity=0.338, AHI=0.390). 1,496 (37.8%) were assigned a low risk by the GPSS (Pearsons correlation: T90%=0.084, Desaturation severity=0.133, Resaturation severity=0.150, Total severity=0.138, AHI=0.186). 1,839 (46.5%) were assigned a moderate risk (Pearsons correlation: T90%=0.109, Desaturation severity=0.180, Resaturation severity=0.183, Total severity=0.181, AHI=0.168). 624 (15.8%) were assigned a high risk (Pearsons correlation: T90%=0.218, Desaturation severity=0.265, Resaturation severity=0.285, Total severity=0.268, AHI=0.293). In univariate quantile regression, moderate and high risk GPSS significantly increased the absolute T90% (0.48% (95% CI 0.33, 0.63) & 2.09% (95% CI 1.88, 2.29), p < 0.001), and z-scores for each of Desaturation severity (0.17 (95% CI 0.14, 0.19) & 0.45 (95% CI 0.41, 0.49), p < 0.001), Resaturation severity (0.22 (95% CI 0.18, 0.25) & 0.56 (95% CI 0.51, 0.6), p < 0.001) and Total Severity (0.18 (95% CI 0.15, 0.21) & 0.48 (95% CI 0.44, 0.52), p < 0.001).
Conclusion: The GPSS tool shows utility in predicting severity even based upon novel oxygen saturation metrics. A Moderate risk score was associated with significantly greater T90% and oxygen saturation severities compared to a low-risk score, and a High-risk score was associated with significantly heightened values again against a moderate risk score.
Grant Support: Nil to declare