Identifying patients with deteriorating generalized pustular psoriasis: Development of a prediction model

Generalized pustular psoriasis (GPP) is a life-threatening condition; however, little is known about the factors that can predict GPP patients manifesting a deteriorating course. To investigate the demographics and clinical features of adult inpatient GPP and propose a prediction model for detecting fatal GPP (fGPP) and GPP requiring intensive care unit admission (iGPP) patients, a nationwide population-based retrospective cohort study was conducted. The adult inpatients with GPP from January 2007 to December 2020 were assessed. The 800 cases were aged 51.0 years (median [interquartile range, 37.0-64.0]). Overall, 21 iGPP (64.0 years [54.0-77.0]) and 17 fGPP (75.0 years [68.0-77.0]) cases were identified as deteriorating GPP. Renal disease (odds ratio [OR], 7.31), myocardial infarction history (OR, 4.29), liver disease (OR, 2.82), and diabetes mellitus (OR, 2.34) were identified as predictors for iGPP. For fGPP, myocardial infarction history (OR, 5.10) and psoriasis history (OR, 3.13) were established as predictors. A prediction model with scores ranging 0-11 points showed a reliable diagnostic value in detecting deteriorating GPP (area under the curve = 0.75 for iGPP and 0.83 for fGPP). In conclusion, this study provides the clinical features of deteriorating GPP. A prediction model may help physicians to identify patients with deteriorating GPP.