Journal of Depression and Anxiety
Open Access

Abstract

Predictors of Non-response and Persistent Functional Impairments in Treatment Adhering to Evidence-based Practice Guidelines for Anxiety Disorders

Maarten K van Dijk, Marc JPM Verbraak, Desiree B Oosterbaan, Adriaan W Hoogendoorn and Anton JLM van Balkom

Background: Several countries have developed guidelines for anxiety disorders containing algorithms that summarize the recommended treatment steps for these disorders. It is important to know which patients have a poor prognosis for treatment according to such algorithms.

Aims: To investigate the predictive power of variables known to be able to influence treatment prognosis in situations where practice guidelines for anxiety disorders are adhered to.

Method: To study the predictive power of variables that are known to be able to influence treatment prognosis, 81 patients who participated in a guideline implementation study and whose treatment was found to adhere to available guidelines were selected. Using logistic regression analysis two models were constructed: one to predict treatment nonresponse; another to predict persistent functional impairments at the 1-year follow-up.

Results: The final prediction model for treatment non-response contains only gender and secondary gain variables. It appears that: males have a higher likelihood (p=.074), and patients that report hopes of obtaining external benefits by seeking treatment have a lower likelihood (p=.054) of showing treatment non-response at the 1-year follow-up. The discriminatory power of this model was found to be poor, however. The model for persistent functional impairments includes gender, satisfaction with the accessibility of healthcare services and the presence of a comorbid anxiety disorder. It appears that: males (p=.87) and patients who express dissatisfaction with the accessibility of care (p=.008) have a higher likelihood, and that; patients who suffer from an additional comorbid anxiety disorder have a lower likelihood (p=.079) of persistent functional impairments. The discriminatory power of this model is excellent.

Conclusion: It remains difficult to predict which anxiety disorder patients will not benefit from treatment that is tailored according to available practice guideline recommendations, therefore no one should be prevented from being offered such treatment, if one removes barriers in attending treatment.