Towards standardization of Prakriti Evaluation: A scoping review of modern assessment tools and their psychometric properties in Ayurvedic medicine

Abstract

Background

Ayurveda is an ancient Indian medical system that emphasizes individualized care based on the concept of Prakriti, which represents an individual's relative proportion of three doshas (humour) at the time of conception. Prakriti is believed to remain unchanged throughout one's life span, and based on the relative preponderance of doshas, it has been classified into ten types. Researchers and practitioners have developed various assessment tools, such as questionnaires, algorithmic and machine learning-based methods, and devices to bring objectivity and replicability to the Prakriti evaluation procedure. However, a systematic evaluation of their effectiveness and psychometric properties is currently lacking in the literature.

Objective

To identify Prakriti, classical Ayurveda texts suggest various techniques, but subjective variations and bias in interpretation have been reported among practitioners.

Methods

This scoping review aims to identify modern Prakriti assessment tools available in scientific literature and describe their psychometric properties to contribute to the development of more effective and standardized methods for Prakriti evaluation. Employing Arksey and O'Malley's five-stage methodological framework, the review critically assesses Prakriti evaluation tools, such as questionnaires, machine learning algorithms, and devices, to provide insights into the robustness and accuracy of these assessment methods.

Results

Thirty-two studies meeting the inclusion criteria were included in the review. Sixteen studies utilized questionnaires for Prakriti assessment, of which three were validated, six had established reliability, and three had both reliability and validity confirmed. Five studies employed algorithm-based methods, with three of these using validated machine learning models. Eleven studies utilized devices to assess Prakriti types, with only one device validated and providing access to data for reproducibility. Questionnaires were the most commonly used tools for Prakriti evaluation, followed by algorithm-based methods and devices.

Conclusion

Questionnaires are the most validated tools presently for Prakriti assessment despite their limited psychometric robustness. Machine learning based models show potential but suffers challenges in accuracy, replicability and finding the ground truth in a robust manner, and devices are the most objective assessment tools though none of them have been appropriately validated. Despite these limitations, Prakriti diagnosis is a promising field that requires robust procedures to establish standardized, replicable and reliable measures for its evaluation.