Brief report
Screening for bipolar disorders using a French version of the Mood Disorder Questionnaire (MDQ)

https://doi.org/10.1016/j.jad.2005.06.005Get rights and content

Abstract

Background

Bipolar disorders remain much too often unrecognized and subsequently inappropriately treated. This paper presents the translation into French and validation of the MDQ, a screening instrument for bipolar spectrum disorders, in an adult psychiatric sample. Modifications of its criteria for a positive screening as well as its test–retest reliability are also addressed.

Methods

A sample of 96 patients, attending outpatient treatment programs and suffering from mood disorders, completed the MDQ before being interviewed according to the mood module of the structured clinical interview for DSM-IV (SCID). They completed the MDQ a second time 1 month later, in order to examine its stability over time.

Results

According to the SCID interview, 54 patients were suffering from bipolar disorder and 42 from unipolar disorder. Among the bipolar sample, the MDQ identified 74.1% of them, with higher sensitivity in bipolar I (90.3%) than bipolar II (52.4%) and 90.5% specificity. Lowering the level of impairment required for positive screening led to improved sensitivity for bipolar II patients (76.2%). The French MDQ demonstrated adequate internal consistency (Cronbach alpha = 0.89). Its test–retest reliability proved to be satisfactory, with a kappa coefficient of 0.79. Similar stability over a 1-month interval was obtained for bipolar type I and type II (κ = 0.75 and 0.77, respectively).

Limitations

Similarly to the American version, the French MDQ has lower sensitivity for bipolar II disorders.

Conclusions

The performance of the French MDQ is comparable to the one reported in the original American study conducted with a similar patient population. In a psychiatric outpatient sample, the French MDQ proves to be a feasible and reliable screening instrument.

Introduction

With an early age of onset (Suppes et al., 2000) and high rates of recurrence (Keck et al., 2001, Osby et al., 2001), bipolar illness often results in chronic morbidity (Sachs, 2002), as well as functional impairment (Calabrese et al., 2003, MacQueen et al., 2001, Mitchell et al., 2004). It is associated with elevated suicide rates (Harris and Barraclough, 1997) and utilization of mental health systems. Identifying this illness as early as possible should help modifying its natural course. This is of major importance not only because of patients' suffering but also with respect to various public health issues this population raises (Rouillon, 2003). Unfortunately, many patients with bipolar disorders, and especially bipolar II subtype (Benazzi and Akiskal, 2003a, Benazzi and Akiskal, 2003b, Hirschfeld et al., 2003a, Lish et al., 1994), remain unrecognized, often misdiagnosed as unipolar depression, and subsequently inappropriately treated. The wide use of standardized structured diagnostic interviews, prohibitively time consuming (Zimmerman et al., 2004) and requiring well-trained clinicians, is not feasible in clinical practice. Although no screening instrument can replace validated comprehensive diagnostic interview protocols, clinical observation and information from various sources (family members, friends), non-researcher clinicians need to rely on minimally invasive and relatively inexpensive tests to optimize their diagnostic evaluation. Recently, the Mood Disorder Questionnaire (MDQ), a psychometrically validated screening instrument, appeared to be a feasible method for improving identification of bipolar spectrum disorder in psychiatric settings (Hirschfeld et al., 2000). In a general population (Hirschfeld et al., 2003b), its sensitivity was much lower, but specificity remained remarkably high. So far, this instrument has been little investigated by others (Benazzi and Akiskal, 2003b, Benazzi, 2003, Miller et al., 2004) than its developers and only a Finnish version has been validated with satisfactory diagnostic performance (Isometsa et al., 2003).

Section snippets

Method

This study was conducted at the Department of Psychiatry, in a community mental health outpatient clinic, part of the Geneva University Hospitals. This facility is divided into three treatment programs, in charge of unipolar depression, bipolar disorders and other various psychiatric disorders. Patients suffering from mood disorders (either unipolar or bipolar) were selected among those newly referred to treatment or already in contact with the facilities in one of these programs.

The study

Description of the sample

The study included 96 patients (58 females, 38 males). Mean age was 45 (S.D. = 9.8, range = 18–63). Among them, 54 (56.3%) received a SCID diagnosis of bipolar spectrum disorder (bipolar I: N = 31; bipolar II: N = 21 and bipolar not otherwise specified: N = 2).

Concerning the MDQ items, frequency of endorsement ranged from 39.6% (spending money) to 78.1% (thought racing).

Cronbach's alpha coefficient of internal consistency was 0.89 at first completion.

Sensitivity, specificity and predictive values

In order to evaluate the sensitivity, specificity and

Discussion

This first study of a French version of the MDQ is very encouraging as it replicates the good sensitivity (73%) and very good specificity (90%) of the instrument assessed by its developers in a psychiatric sample (Hirschfeld et al., 2000). In other validation studies, either on clinical (Isometsa et al., 2003) or general populations (Hirschfeld et al., 2003b), the MDQ screening always yielded high specificity (90% and 97%, respectively) but lower sensitivity (47% and 28%, respectively) than in

References (19)

There are more references available in the full text version of this article.

Cited by (85)

  • Self-evaluation of negative symptoms in adolescent and young adult first psychiatric episodes

    2020, Progress in Neuro-Psychopharmacology and Biological Psychiatry
  • Validity of the Mood Disorder Questionnaire (MDQ) as a screening tool for bipolar spectrum disorders in anabaptist populations

    2020, Journal of Psychiatric Research
    Citation Excerpt :

    A narrower definition of BSD that required an incapacitating mania led to similar levels of diagnostic agreement between MDQ and BEFD across all scoring methods we investigated, in contrast to the large difference noted between the Hirschfeld and 5-symptom scoring criteria for more widely defined BSD that includes hypomania. This supports past suggestions that the Hirschfeld scoring may be more likely to miss cases that are not characterized by a severe mania (Miller et al., 2004; Twiss et al., 2008; Weber et al., 2005). The 5-symptom threshold may be particularly beneficial in capturing these frequently missed cases.

View all citing articles on Scopus
View full text