Effective screening for Parkinson disease (PD) is important for symptom management and patient recruitment into disease-modifying trials. Here, we examined the performance of the PD-RISK Index toolkit in identifying PD patients at home without supervision.
Between 2022-2024, we recruited 305 participants (ages, 40-85 years): 93 with typical PD; 66 had other neurological conditions; and 146 were healthy controls. Two versions of the toolkit were completed: An original, 69-item-long questionnaire with the 40-scent University-of-Pennsylvania-Smell-Identification-Test (UPSIT); thereafter, a simplified, 11-item-short questionnaire with a newly-developed 8-scent smell test and responses submitted online. The corresponding summary scores ( PD-RISK Indices) were calculated for each subject to evaluate diagnostic performances.
In both versions, PD patients had higher questionnaire scores and worse olfaction than healthy controls (P<0.001), as expected; scores for subjects with other neurological conditions ranked at intermediate levels. The simplified 11-item questionnaire outperformed the original version in identifying PD. The abbreviated 8-scent smell test and the UPSIT booklet performed equally well in detecting hyposmia. The simplified PD-RISK Index showed the highest diagnostic performance in identifying PD patients amongst all participants; at a value of 23 (range, 0-100, where higher scores indicate higher PD risk), we measured a sensitivity of 0.98, specificity of 0.83, and positive and negative likelihood ratios of 5.88 and 0.02, respectively. The simplified PD-RISK Index showed stable performance across sexes and disease duration.
Our study reveals that unsupervised screening for manifest PD can be effectively carried out at home using an 11-item questionnaire, 8-scent smell test and online data entry.
The plain language titleThe PD-RISK Index study: Facilitating an earlier diagnosis of adults with Parkinson disease by screening at home with a short questionnaire and a simplified smell test
Plain language summariesEffective screening for Parkinson disease (PD) is important for symptom management by healthcare workers and the recruitment of patients into disease modifying trials. We previously published a two-step model for this purpose, which has been validated using several established study cohorts. The study below represents a field-test of this model and the toolkit to implement it (i.e. a self-report questionnaire and a smell test) using a real-world setting.
Between 2022-2024, 305 participants (ages, 40-85 years) were recruited into the study: 93 with typical PD; 66 had other neurological conditions; 146 were healthy. Two versions of the toolkit were completed at-home without supervision by a healthcare worker: (1) an original, 69-item-long questionnaire and a 40-scent smell test; (2) a simplified, 11-item-short questionnaire with a newly created, 8-scent smell test. Accordingly, two versions of summary scores ( PD-RISK Index) were calculated for each subject to examine group classification. Several performance metrics were used to evaluate diagnostic performances and to suggest clinically relevant thresholds.
In both versions, PD patients had significantly higher questionnaire scores (meaning higher risk of having PD) and lower smell test scores (meaning more impaired sense of smell) than healthy controls; scores for subjects with other neurological conditions ranked at intermediate levels. The simplified questionnaire outperformed the original version in diagnostic accuracy. The abbreviated smell test performed as well as the 40-item version in identifying hyposmia. The simplified PD-RISK Index showed stable performance across sexes and disease duration.
This study shows that non-invasive and unsupervised screening for manifest PD can be effectively carried out remotely using a simple 11-item questionnaire, an abbreviated 8-scent smell test and an online data entry system, thereby generating a PD-RISK Index for each individual.