Introduction: Neuropsychiatric (NP) involvement in Systemic Lupus Erythematosus (SLE) presents with a variety of neurologic and psychiatric syndromes. Their relationship with the underlying disease could be difficult to assess especially for mild, diffuse and nonspecific NP syndromes such as, headache, mild depression and anxiety. Aim: To test the performance of an attribution model applied to mild NP events in a large mono centric cohort of SLE patients. Patients and Methods: NP events were defined according to the 1999 ACR case definitions and re-assessed using a previously developed and validated attribution model, including a weighted score of 4 different items: 'onset time' of the NP event; presence of non-SLE 'confounding factors' (i.e. associations as suggested by the 1999 ACR criteria); presence of specific NP events according to Aniala; presence of 'favouring factors' (general and disease specific). Expert opinion was set as reference standard of the attribution of NP event to SLE. The relationship between the attribution model and the reference diagnosis was evaluated by logistic models, and results presented as odds ratio (OR) and 95% confidence intervals (CI), and standardised coefficients. Performance was evaluated by the Area Under the ROC Curve (AUC). Results: A total of 126 patients with a diagnosis of SLE according to the revised ACR classification criteria with one NP minor event (86 headache, 31 mild depression, 9 anxiety) were included: M/F 5/121 females, with a median age at first NP event of 38 years. The overall performance of the model was good (AUC 0.754 (0.675-0.832). The most important items for correct classification was the 'confounding factor' sub-score OR (95%CI)/stand. coeff. 15.86 (3.59-70.03)/1.27, with a lower contribution for 'onset time' 2.18 (1.22-3.92)/0.61, and 'additional factors' 2.71 (1.51-4.85)/0.75. Conclusion: Despite expert's opinion remains the gold standard, using an assignment algorithm may improve the correct attribution of minor NP events to SLE.
Relationships between minor neuropsychiatric events and systemic lupus erythematosus: Performance of an attribution model
Bortoluzzi A.Primo
;Sciré Carlo Alberto;Padovan M.;Govoni M.Ultimo
2014
Abstract
Introduction: Neuropsychiatric (NP) involvement in Systemic Lupus Erythematosus (SLE) presents with a variety of neurologic and psychiatric syndromes. Their relationship with the underlying disease could be difficult to assess especially for mild, diffuse and nonspecific NP syndromes such as, headache, mild depression and anxiety. Aim: To test the performance of an attribution model applied to mild NP events in a large mono centric cohort of SLE patients. Patients and Methods: NP events were defined according to the 1999 ACR case definitions and re-assessed using a previously developed and validated attribution model, including a weighted score of 4 different items: 'onset time' of the NP event; presence of non-SLE 'confounding factors' (i.e. associations as suggested by the 1999 ACR criteria); presence of specific NP events according to Aniala; presence of 'favouring factors' (general and disease specific). Expert opinion was set as reference standard of the attribution of NP event to SLE. The relationship between the attribution model and the reference diagnosis was evaluated by logistic models, and results presented as odds ratio (OR) and 95% confidence intervals (CI), and standardised coefficients. Performance was evaluated by the Area Under the ROC Curve (AUC). Results: A total of 126 patients with a diagnosis of SLE according to the revised ACR classification criteria with one NP minor event (86 headache, 31 mild depression, 9 anxiety) were included: M/F 5/121 females, with a median age at first NP event of 38 years. The overall performance of the model was good (AUC 0.754 (0.675-0.832). The most important items for correct classification was the 'confounding factor' sub-score OR (95%CI)/stand. coeff. 15.86 (3.59-70.03)/1.27, with a lower contribution for 'onset time' 2.18 (1.22-3.92)/0.61, and 'additional factors' 2.71 (1.51-4.85)/0.75. Conclusion: Despite expert's opinion remains the gold standard, using an assignment algorithm may improve the correct attribution of minor NP events to SLE.I documenti in SFERA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
