Brief Questionnaire Derived from PANSS Using a General Probability Model to Assess and Monitor the Clinical Features...

Patients with schizophrenia require continuous treatment much longer than the duration of their hospitalization, which makes their family members essential in their medical care. However, the evaluation of the disease state could only be done by professionals. This prompted us to seek potent indicators of disease states that are understandable and easy to use for the patients' family.

Specific items were firstly extracted from the total PANSS scale. Then 3 096 PANSS scores were analyzed using a nonlinear mixed-effects model (NONMEM). A questionnaire was subsequently developed for family members to assess and monitor the overall severity of schizophrenia. Finally this questionnaire was validated in 33 patients.

2 items (P1 and N4) were extracted from the 8 effective remission items according to the correlation coefficients between the total PANSS score and different combinations of items. P1N4 was defined as the sum of these 2 items. A model was then developed to describe the probability of PANSS≥60, with P1N4 as indicators. The results indicated that P1N4 could make a good predictor of the overall probability of PANSS≥60, which was independent of treatment. A brief questionnaire with 7 questions was developed based on the results. External validation results indicated the questionnaire's suitability for a good assessment.

Questionnaire developed based on P1 and N4 may facilitate the patients' family members to better understand the disease state and help to prevent relapse.