Exercise therapies play an indispensable part in managing certain types of health issues, but there is no definite method that can determine the adherence of these exercises except self-report diaries. However, these diaries involve self-presentation bias and lack of standardization. The study forms an Adherence Rating Scale (ARS) of good external and internal validity.

Knee osteoarthritis (KOA) counts among the most frequently occurred chronic diseases among the elders and the principal reason for the decline in motor function and disability. One of the fundamental, core and mature ways to treat KOA is exercise therapy. In the diagnosis and management guidelines of osteoarthritis, exercise therapy is "recommended" or "strongly recommended" in many countries. Most scholars favoured exercise therapy as the best rehabilitation program for KOA patients. But, the feeble long-term adherence is a significant issue that affects the treatment effects. The study aimed to construct a compact and practical adherence rating scale (ARS) based on the exercise therapy adherence prediction model.

By considering 218 KOA patients adherence as the dependent variable, a binary logistic regression model was established. The independent variables taken for the analysis were exercise habits, patients' general information, attitude, knowledge and exercise therapy. The variables of the model allotted with regression coefficients and the ARS was created subsequently. Receiver operating characteristic curves and curve fitting was used to investigate the ARS impact on finding the adherence and to discover the goodness of adherence fit. The randomized controlled trials were used to estimate the external validity of the ARS.

The development of the ARS and the  adherence model involved the following variables: knowledge of exercise therapy (4 points), treatment confidence (3 points), exercise habits (3 points), treatment (2 points), degree of social support (2 points), degree of care required to manage the disease (1 point), education level (1 point) and age (1 point). The critical value of the total score of the ARS was reported to be 6.50, with the specificity of 76.34% and sensitivity of 87.20%.

A simple and practical ARS and an adherence model were created. The ARS has good external and internal validity and can be applied to estimate the exercise therapy adherence in KOA patients.