Nomogram model for predicting medication adherence in patients with various mental disorders based on the Dryad...

Area of Science:

• Psychiatry
• Pharmacology
• Medical Informatics

Background:

• Treatment compliance significantly impacts disease outcomes in psychiatric patients.
• Assessing and predicting patient compliance remains a critical challenge in managing psychotic disorders.

Purpose of the Study:

• To establish a predictive model for medication compliance in patients with psychotic disorders.
• To provide a tool for early intervention against treatment non-compliance behavior.

Main Methods:

• Utilized clinical data from 451 patients with psychotic disorders from the Dryad database.
• Employed Least Absolute Shrinkage and Selection Operator (LASSO) and logistic regression to build the predictive model.
• Validated the model using bootstrap resampling and evaluated its performance with concordance statistics, Brier score, ROC curve, and decision curve analysis.

Main Results:

• The final model included 432 patients, with an overall compliance rate of 61.3%.
• Independent risk factors for non-compliance identified were low Drug Attitude Inventory-10 (DAI-10) scores, high Brief Psychiatric Rating Scale (BPRS) scores, multiple hospitalizations within a year, and prior use of long-acting injectables.
• The nomogram achieved a concordance statistic of 0.709 and an area under the ROC curve of 0.716, indicating good predictive ability.

Conclusions:

• Low DAI-10 scores, high BPRS scores, frequent hospitalizations, and previous long-acting injectable use are significant predictors of medication non-compliance in psychotic disorders.
• The developed nomogram demonstrates good sensitivity and specificity for predicting treatment adherence, offering a valuable tool for clinical practice.