Diagnosis of bipolar disorder in primary and secondary care: what have we learned over a 10-year period?

Abstract Background Studies suggest that misdiagnosis of bipolar disorders (BD) is frequent in primary care. This study aimed to evaluate agreement between referral for BD by general practitioners (GP) and BD diagnosis by secondary care psychiatrists, and to evaluate the impact of age, gender, and BD type on agreement. Methods The study was conducted at Hopital du Sacre-Coeur de Montreal׳s “Module Evaluation/Liaison” (MEL), which establishes/clarifies psychiatric diagnoses requested mainly from GPs and directs patients to appropriate treatment and care. Socio-demographic variables, reason for referral, and psychiatric diagnosis were compiled for patients assessed from 1998 to 2010. GP-psychiatrist agreement was established for BD type, gender, and age group (18–25, 26–35, 36–45, >45) using Cohen׳s Kappa coefficient ( Κ ). Results From 1998 to 2010, MEL psychiatrists received 18,111 requests and carried out 10,492 (58%) assessments. There were 583 referrals for BD suspicion, while 640 assessments (6.1%) received a BD diagnosis (40.3% type I, 40.5% type II). The overall K was 0.35 (95% CI [0.31, 0.38]), and was significantly higher for type I than type II (I=0.35, 95% CI [0.30, 0.39]; II=0.25, 95% CI [0.21, 0.30]), though age group and gender had no impact. Limitations Reasons for referral were converted into keywords and categories to facilitate agreement analyses. Only the main psychiatric diagnosis was available. Conclusions Our study suggests diagnosing BD remains strenuous, regardless of age and gender, though BD type I seems better understood by primary care GPs. The true measure of BD diagnosis remains a critical issue in clinical practice.