Home | Pre GDS-15 | PredMe?

Pre GDS-15

Questionnaire-free method to predict 15-item Geriatric Depression (GDS-15)

This application only uses demographics and physical health to predict GDS-15, that avoids frequently asking an elderly patient the same questions. During regular health visit (e.g. monthly), a healthcare giver may input the existing routine information to this application. Only if the GDS-15 is predicted positive, then the questionnaire is given to confirm further diagnostic procedures. Patients aged 60 years or older in community-health center (or equivalent) are eligible.A patient is unlikely well-predicted if the patient input data are not provided by the input options in this application. The exception is religion due to ethical reason, but, religious believers are likely well-predicted, particularly two of the religion options.




Disclaimer

This web application is intended as a prototype to showcase a tool for diagnostic prediction of depressive symptoms by GDS-15. A predicted positive is not a definitive diagnosis of depression. Do not use this web application without supervision from a competent healthcare giver. Implementation of this predictive system should be incorporated to an electronic health record (EHR) system. Our system should also be conducted in automatic manner; thus, a GDS-15 evaluation is recommended by our system based on pre-existing, required information in EHR. Manual input by clinicians considerably cancels out the objective of this predictive system.

References

Sri Susanty, Herdiantri Sufriyana, Emily Chia-Yu Su, Yeu-Hui Chuang. Questionnaire-free machine-learning method to predict depressive symptoms among community-dwelling older adults. PLoS One 2023 Jan 25;18(1):e0280330. doi: 10.1371/journal.pone.0280330 PMID: 36696383

PredMe Apps are currently self-funded and maintained alongside active research work. Your donation helps keep PredMe up and running.
$1 $4 $16 $64


Home | Pre GDS-15 | PredMe?

Predict

Input your patient data.

Please kindly wait the processing for approximately 5 minutes after starting to predict. For re-prediction, it will take approximately 30 seconds.





Please hide all details...

Please wait...