Using network analysis to identify factors differentiating antipsychotic treatment-resistant and treatment-responsive patients with psychosis

There is an increasing evidence that suggest mental disorders are caused by the complex interaction of various symptoms rather than isolated effects of a limited number of latent factors. The complex interactions among symptoms of mental disorders can be studied using network science. In this study, we applied analytical tools of network science to understand the mechanism of symptoms interaction in treatment-resistant schizophrenia patients. In this regard, we constructed a network of schizophrenia risk factors and symptoms using the CATIE dataset and analyzed different network properties to infer the difference of symptomatic interaction between treatment-resistant and treatment-responsive patients. According to the results of our study, "Preoccupation" is the most central node in the symptom interaction network of treatment-resistant patients, whereas "Unusual Thought Content" and "Stereotyped Thinking" are the most central symptoms in treatment-responsive patients. These central symptoms are worth further analysis for effective planning of treatment strategy for schizophrenia patients.