A Short Progressive Supranuclear Palsy Quality of Life Scale: Data from the PSP-NET

Abstract

We read with interest the article by Jensen and colleagues who proposed a condensed version of the Progressive Supranuclear Palsy Quality of Life scale (PSP-ShoQoL) as a reliable and practical tool to evaluate quality of life in PSP patients.¹

The proposed PSP-ShoQoL included 12 items divided into two subscales representing physical (seven items) and mental symptoms (five items) and was administered to 245 patients from the German PSP network. The internal consistency of both total and subscores was high within 0.83 and 0.90. The PSP-ShoQoL significantly correlated with the Progressive Supranuclear Scale-Rating Scale (PSP-RS) and the Geriatric Depression Scale (GDS) but not with the Montreal Cognitive Assessment scale (MoCA). With 12-month follow-up data on a subgroup of 94 patients, the authors showed that the PSP-ShoQoL presented fair sensitivity to change and test–retest reliability.

Herein, we present data on the PSP-ShoQoL on an independent PSP cohort, the Italian PSP-NET supported by Fondazione LIMPE.^{2, 3} 413 PSP patients performed the same evaluations used by Jensen et al. except for the GDS that was substituted by the Hospital Anxiety and Depression Scale (HADS). Compared with the German cohort, the PSP-NET included older (age: mean ± standard deviation [SD] 71.2 ± 8.1 vs. 69.2 ± 7.4) and more severe patients (PSP-RS: 40.56 ± 16.85 vs. 33.8 ± 13.8) while disease duration was similar (years: 4.44 ± 2.70 vs. 4.1 ± 2.6). Accordingly, PSP-ShoQoL total and subscores were higher within the PSP-NET (PSP-ShoQoL total: 25.33 ± 11.3 vs. 19.27 ± 11.10; PSP-ShoQoL Physical: 18.6 ± 8.2 vs. 13.74 ± 8.25; PSP-ShoQoL Mental: 6.7 ± 5.1 vs. 5.53 ± 4.67). We confirm a fair internal consistency for both the total score (Cronbach's alpha: 0.87) and subscores (Physical: 0.89; Mental: 0.80). The PSP-ShoQoL correlated significantly with the original PSP-QoL (r = 0.945, P < 0.001), the PSP-RS (r = 0.646, P < 0.001), the MoCA (−0.340, P < 0.001), and the HADS (r = 0.602, P < 0.001). With 6-month follow-up data available for 80 patients, we revealed a significant increase in both PSP-ShoQoL total score (t = 5.24, P < 0.001) and Physical (t = 5.45, P < 0.001) and Mental (−2.78, P < 0.05) subscores. Test–retest reliability was good both for PSP-ShoQoL total score (intraclass correlation coefficient [ICC] = 0.78, P < 0.001), as well as for its subscales (Physical ICC = 0.80, P < 0.00; Mental ICC = 0.68, P < 0.001). Finally, by analyzing the area under the curve (AUC) we identified a value of 34.5 as a discriminating cutoff for a significant impairment of quality of patients' life measured by the PSP-ShoQoL within the PSP-NET (sensibility: 0.97; specificity: 0.15; AUC: 0.93) (Fig. 1).

Jensen and coworkers proposed a brief instrument with fair psychometric properties for assessing quality of life in PSP patients. Herein, we have demonstrated the application of the PSP-ShoQol in an independent, large PSP cohort. Our results largely replicate those of Jensen et al. except for the relationship between the PSP-ShoQoL and the MoCA. Furthermore, we propose a cutoff of 34.5 as a discriminating value for a significant impairment of quality of patients' life measured by the PSP-ShoQoL.