- Email: [email protected]
Validation of the Mood Disorder Questionnaire for screening for bipolar disorder in a UK sample
Journal of Affective Disorders 110 (2008) 180 – 184 www.elsevier.com/locate/jad
Brief report
Validation of the Mood Disorder Questionnaire for screening for bipolar disorder in a UK sample James Twiss a , Steven Jones a , Ian Anderson b,⁎ a
School of Psychological Sciences, Clinical and Health Psychology, The University of Manchester, Manchester, UK b Neuroscience and Psychiatry Unit, The University of Manchester, Manchester, UK Received 10 October 2007; received in revised form 19 December 2007; accepted 19 December 2007 Available online 7 February 2008
Abstract Background: The Mood Disorder Questionnaire (MDQ) was designed as a screening questionnaire for bipolar disorder. Previous research has raised questions about the suitability of the MDQ structure for screening for bipolar II disorder. This study investigated the optimal sensitivity and specificity cut-off thresholds for the MDQ in bipolar I and bipolar II patients in a UK sample. Methods: The MDQ was administered to patients before attending a tertiary mood disorders clinic. Diagnostic interviews were used to determine DSM-IV diagnoses and these were used as the gold standard against which to investigate the performance of the MDQ. Results: 54 patients with bipolar spectrum disorder and 73 patients with unipolar depressive disorder completed the MDQ. With the original scoring criteria (symptoms and supplementary questions) the sensitivity for bipolar disorder was 0.76 (bipolar I disorder 0.83, bipolar II disorder 0.67) with specificity 0.86. The optimal cut-off score in the current sample was a score of 9 or more endorsed symptoms without applying the supplementary questions (sensitivity of 0.90 and 0.88 for bipolar I and bipolar II groups respectively with a specificity of 0.90). Limitations: The sample was drawn from a tertiary mood disorders clinic. Conclusions: The MDQ appears to be a useful screening tool for bipolar spectrum disorder in UK psychiatric practice with sensitivity for bipolar II disorder improved by dropping the supplementary sections. Further investigation of the optimal cut-off scores of the MDQ is needed to determine its utility in non-specialist and community based samples. © 2008 Elsevier B.V. All rights reserved. Keywords: Bipolar disorder; Mood Disorder Questionnaire; Sensitivity; Specificity
1. Introduction Bipolar disorder often goes unrecognised with delays of 8 years or more common before accurate diagnosis ⁎ Corresponding author. University of Manchester, Neuroscience and Psychiatry Unit, Room G809 Stopford Building, Oxford Road, Manchester M13 9PT, UK. Tel.: +44 161 275 7428; fax: +44 161 275 7429. E-mail address: [email protected] (I. Anderson). 0165-0327/$ - see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2007.12.235
(Lish et al., 1994) which can result in a greater burden of illness, risk of suicide (Dunner, 2003) and inappropriate treatment (Altshuler et al., 1995). Bipolar spectrum disorders have been estimated to cost the UK around £2 billion per annum (Das Gupta and Guest, 2002) and unrecognised cases may have a greater cost for health services than recognised and treated cases (Li et al., 2002; McCombs et al., 2007). A particular challenge is presented by ‘milder’ forms of bipolar disorder, such as bipolar II disorder and bipolar
J. Twiss et al. / Journal of Affective Disorders 110 (2008) 180–184
NOS. These are typically characterised by hypomania which may not be recognised as pathological by the individual or others (Akiskal et al., 2000). Treatment is likely to be sought for depressive episodes alone (Hirschfeld, 2001) often leading to a unipolar rather than bipolar diagnosis (Ghaemi et al., 2000). The Mood Disorder Questionnaire (MDQ; Hirschfeld et al., 2000) was developed to aid diagnosis of bipolar spectrum disorders. It covers symptom endorsement (Section 1), symptom clustering (Section 2) and severity of problem caused (Section 3) (see Methods). The original validation study reported that the MDQ performed well against DSM-IV diagnosis (sensitivity 0.73, specificity 0.90) in a psychiatric outpatient population. Although questionnaires alone are not an alternative to clinical interview (Miller et al., 2004; Phelps and Ghaemi, 2006) the performance of the MDQ suggested promise for initial screening for bipolar disorder in primary care and busy psychiatric clinics. MDQ validation studies in different US settings (Hirschfeld et al., 2003; Graves et al., 2007; Hirschfeld et al., 2005; Kemp et al., 2008; Miller et al., 2004) and in other countries (Hardoy et al., 2005; Isometsa et al., 2003; Konuk et al., 2007; Vieta et al., 2007; Weber Rouget et al., 2005) indicate that its psychometric properties vary somewhat according to the population studied. Its sensitivity, especially for bipolar II disorder, has been reported to be improved by exclusion or modification of Section 3 without markedly impairing its specificity (Benazzi, 2003; Isometsa et al., 2003; Kemp et al., 2008; Miller et al., 2004; Weber Rouget et al., 2005). No validation studies of the MDQ have been reported for a UK sample. Given the importance of early detection of bipolar spectrum disorder, the aims of the present study were to validate the MDQ in a UK sample, to assess sensitivity and specificity scores of the MDQ for bipolar I and bipolar II groups using the original cut-off guidelines and to assess its performance without Sections 2 and 3.
181
older; the study was approved by the Local NHS Research Ethics Committee. The MDQ comprises 3 Sections. In Section 1 participants are asked “Has there ever been a time when you were not your usual self and…” followed by 13 manic/ hypomanic symptoms (e.g. “…you had much more energy than usual”) answered ‘yes/no’. Section 2 asks whether more than one of these symptoms were experienced at the same time (answered ‘yes/no’). Section 3 asks participants what degree of problem the symptoms caused on a fourpoint scale ranging from ‘no problem’ to ‘serious problem’. In the original validation report the cut-off for a positive screen for bipolar disorder was ≥7 symptoms, ‘yes’ to Section 2 and a problem severity of ‘moderate’ or ‘serious’ in Section 3 (Hirschfeld et al., 2000). Sixteen of our subjects (13%), mostly unipolar patients with few symptoms endorsed, left Section 2 and/or 3 blank. The missing data were taken as ‘no’ (Section 2) or ‘no problem’ (Section 3) to avoid inflating sensitivity. Analysis without these subjects made little difference to the results. Data were analysed using SPSS 14 for windows. χ2 tests were used to test categorical data. Continuous data were analysed with independent t-tests (two tailed) or ANOVA with p b 0.05 taken as significant. Sensitivity and specificity are reported with 95% confidence interval (95%CI) for the main findings and relevant positive and negative predictive values. 3. Results The bipolar group included 30 bipolar I, 21 bipolar II and 3 bipolar NOS participants; the later two were combined for analysis (called bipolar II). The unipolar group included 68 patients with major depression (18 single episode, 50 recurrent), 3 with depressive disorder NOS and 2 with dysthymic disorder. Bipolar and unipolar groups did not differ with respect to age (means (sd): bipolar group 44.0 (11.8) years;
2. Methods Participants were sequential outpatient attendees of a tertiary NHS Specialist Service for Affective Disorders who completed the MDQ and then received a semistructured clinical interview covering current and past mood disorder (DSM-IV-TR diagnosis; American Psychiatric Association, 2000) conducted by experienced psychiatrists who had access to the MDQ results. The sample consisted of 127 patients; 54 had a bipolar spectrum disorder and 73 a unipolar diagnosis. Participants were English speaking, aged 18 years or
Fig. 1. Sensitivity and Specificity of the MDQ for bipolar I and II subjects for different cut-off scores.
182
J. Twiss et al. / Journal of Affective Disorders 110 (2008) 180–184
unipolar group 42.1 (11.8) years) or gender (male 40.7% and 43.8% for bipolar and unipolar groups respectively). Internal reliability of MDQ Section 1 symptoms was good with a Cronbach's alpha coefficient of 0.91. Item– item total correlations ranged from 0.41–0.81. Using the original cut-off guidelines (Hirschfeld et al., 2000) the overall MDQ sensitivity was 0.76 95% CI 0.63–0.85), higher for bipolar I disorder (0.83 95% CI: 0.66–0.93) than for bipolar II disorder (0.67 95% CI: 0.47–0.82). Specificity was 0.86 (95% CI: 0.77–0.92). Overall positive and negative predictive values were 0.80 and 0.83 respectively. The optimum cut-off for symptom number in our population was 9 (specificity 0.90, 95% CI: 0.82–0.95) without altering sensitivity. Analyses were then re-run using only Section 1. Fig. 1 shows the sensitivity and specificity of the MDQ for various cut-off scores for the bipolar I group and bipolar II groups with the optimal cut-off score 9 symptom items. Compared with the original scoring guidelines sensitivity was slightly improved for bipolar I (0.90, 95% CI: 0.74– 0.97) and greatly improved for bipolar II disorder (0.88, 95% CI: 0.69–0.96). Specificity also improved slightly (0.90, CI: 0.80–0.95). Overall positive and negative predictive values were 0.79 and 0.92 respectively. Table 1 shows frequency scores for the MDQ and how often each of the Sections were answered by each of the groups. There was a significant difference in total symptoms scores between unipolar and both bipolar groups (p b 0.001 for both), but not between bipolar I and bipolar II groups. Section 2 screened out 10% of the bipolar I group but none of the bipolar II group. Section 3 screened out 3.3% of the bipolar I group and Table 1 MDQ Section scores by group
Section 1 Symptom score Mean (SD)
Bipolar I
Bipolar II
11.07 (2.68)
10.12 (2.89)
Unipolar
4 (2.82)⁎⁎
Section 2 Clustering of symptoms? (%) Yes 28 (90%) No 3 (10%) Missing 0 (0%)
23 (95.8%) 49 (67.1%) 0 (0%) 12 (16.4%) 1 (4.2%) 12 (16.4%)
Section 3 Problem severity (%) No problem Minor problem Moderate problem Serious problem Missing
3 (12.5%) 4 (16.7%) 6 (25%) 10 (41.7%) 1 (4.2%)
⁎⁎p b 0.001.
0 (0%) 1 (3.3%) 10 (33.3%) 18 (60.0%) 1 (3.3%)
13 (17.8%) 16 (21.9%) 11 (15.1%) 22 (30.1%) 11 (15.1%)
29.2% of the bipolar II group using the standard scoring guidelines. Including Section 2 gave the same results as using Section 1 alone. Modifying Section 3 to include ‘minor problem’ improved the sensitivity for bipolar II disorder (0.79) with minor changes in sensitivity for bipolar I disorder (0.87) and specificity (0.85) compared with the original scoring guidelines. 4. Discussion The MDQ showed good reliability and validity in identifying bipolar disorder in UK patients attending a tertiary clinic. Internal consistency was good (0.91) and comparable with that reported in the original validation study (0.90). Our finding of a higher sensitivity for bipolar I disorder (0.83) than bipolar II disorder (0.67) using the original scoring guidelines is similar to that reported in previous studies (Isometsa et al., 2003; Kemp et al., 2008; Miller et al., 2004; Weber Rouget et al., 2005; Benazzi and Akiskal, 2003). The DSM-IV definition of hypomanic episodes includes an association with a ‘change in functioning’ rather than with ‘marked impairment’ which suggests that the problem severity question (Section 3) may incorrectly screen out individuals with bipolar II disorder. Miller et al. (2004) found that the reduction in sensitivity caused by Section 3 was equal for both bipolar I and bipolar II participants whereas we found it reduced sensitivity to a greater degree in the bipolar II group; this is consistent with other studies (Weber Rouget et al., 2005; Isometsa et al., 2003; Benazzi and Akiskal, 2003). However some of these studies changed the severity criterion to ‘minor problem’ rather than omitting it altogether; Isometsa et al. (2003) found that omission reduced specificity whereas we found only a slight decrease in specificity. It is therefore unclear whether omission or modification of Section 3 is the preferable general strategy to optimise the MDQ for detecting bipolar II disorder and may depend on population studied; in our patients omission was better. Including Section 2 did not improve the performance of the MDQ in our study. It is notable that a sizeable minority of the unipolar group did not complete Sections 2 and 3 suggesting some found these questions difficult to answer or not applicable. The bipolar I and bipolar II groups had a similar number of endorsed symptoms. DSM-IV does not differentiate between the symptomatology of hypomanic and manic episodes apart from ‘psychotic features’ specific for mania. The main distinction between bipolar I and II disorder is severity and length of episode. Differences in symptom patterns of hypomania vs. mania may however exist
J. Twiss et al. / Journal of Affective Disorders 110 (2008) 180–184
(Serretti and Olgiati, 2005) and our finding of a similar number of symptoms does exclude this possibility. A limitation of the study is the setting of a tertiary mood disorder service so that patients’ knowledge of diagnosis is likely to be higher than in other settings. The median sensitivity and specificity of the MDQ in psychiatric settings across different countries have been 0.68 (range 0.58 to 0.74) and 0.91 (range 0.67 to 0.96) respectively ( Miller et al. 2004; Hirschfeld et al. 2000; Weber Rouget et al., 2005; Konuk et al., 2007; Corona et al., 2007; Vieta et al., 2007). In our population the sensitivity was a little higher but its performance was largely comparable. In depressed patients treated with antidepressants in a primary care setting the sensitivity and specificity were 0.58 and 0.93 respectively (Hirschfeld et al., 2005) but a much lower sensitivity of 0.28 (specificity 0.97) was found in a community sample (Hirschfeld et al., 2003). However the latter study only identified 8 out of the 78 bipolar subjects as having bipolar II disorder which is not consistent with epidemiological evidence (Akiskal et al., 2000; Benazzi 2007) suggesting a problem with the method of identification of bipolar II disorder in that study. Our study suggests that the MDQ might be a useful tool in identifying bipolar II disorder in patients presenting with depression but further studies in nonspecialist psychiatric services and general practice using modified/omitted Section 3 criteria are needed to test this. Recent studies suggest that the Hypomania Checklist (HCL-32) is more sensitive at detecting hypomania than the original MDQ (Vieta et al., 2007); our results suggest that it should also be tested against the MDQ with Section 3 modified or omitted. The performance of the MDQ as a population screening instrument without Sections 2 and 3 also warrants further study. Role of funding source Nothing declared. Conflict of interest No conflict declared.
Acknowledgements We thank Kirsten Bond for assisting with the collection of the data and Manchester Mental Health and Social Care trust for their facilitation in the study. References Akiskal, H.S., Bourgeois, M.L., Angst, J., Post, R., Moller, H., Hirschfeld, R., 2000. Re-evaluating the prevalence of and
183
diagnostic composition within the broad clinical spectrum of bipolar disorders. J. Affect. Disord. 59, S5–S30. Altshuler, L.L., Post, R.M., Leverich, G.S., Mikalauskas, K., Rosoff, A., Ackerman, L., 1995. Antidepressant-induced mania and cycle acceleration: a controversy revisited. Am. J. Psychiatry 152, 1130–1138. American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental Disorders Fourth Edition. American Psychiatric Association, Washington DC. Text Revision. Benazzi, F., 2003. Improving the Mood Disorder Questionnaire to detect bipolar II disorder. Can. J. Psychiatry 48, 770–771. Benazzi, F., 2007. Bipolar II disorder: epidemiology, diagnosis and management. CNS Drugs 21 (9), 727–740. Benazzi, F., Akiskal, H.S., 2003. The dual factor structure of self-rated MDQ hypomania: energized-activity versus irritable-thought racing. J. Affect. Disord. 73, 59–64. Corona, R., Berlanga, C., Gutierrez-Mora, D., Fresan, A., 2007. Detection of bipolar disorder with a clinical screening questionnaire: the Spanish version of the Mood Disorder Questionnaire. Salud Ment. 30, 50–57. Das Gupta, R., Guest, J.F., 2002. Annual cost of bipolar disorder to UK society. Br. J. Psychiatry 180, 227–233. Dunner, L.D., 2003. Clinical consequences of under-recognized bipolar spectrum disorder. Bipolar Disord. 5, 456–463. Ghaemi, S.N., Boiman, E.E., Goodwin, F.K., 2000. Diagnosing bipolar disorder and the effect of antidepressants: a naturalistic study. J. Clin. Psychiatry 61, 804–808. Graves, R.E., Alim, T.N., Aigbogun, N., Chrishon, K., Mellman, T.A., Charney, D.S., Lawson, W.B., 2007. Diagnosing bipolar disorder in trauma exposed primary care patients. Bipolar. Disord. 9, 318–323. Hardoy, M.C., Cadeddu, M., Murru, A., Dell'Osso, B., Carpiniello, B., Morosini, P.L., Calabrese, J.R., Carta, M.G., 2005. Validation of the Italian version of the “Mood Disorder Questionnaire” for the screening of bipolar disorders. Clin. Pract. Epidemol. Ment. Health. 1, 8. Hirschfeld, R.M., 2001. Bipolar spectrum disorder: improving its recognition and diagnosis. J. Clin. Psychiatry 62 S, 5–9. Hirschfeld, R.M.A., Williams, J.B.W., Spitzer, R.L., Calabrese, J.R., Flynn, L., Keck Jr., P.E., Lewis, L., McElroy, S.L., Post, R.M., Rapport, D.J., Russell, J.M., Sachs, G.S., Zajecka, J., 2000. Development and validation of a screening instrument for bipolar: the Mood Disorder Questionnaire. Am. J. Psychiatry 157, 1873–1875. Hirshfeld, R.M.A., Holzer, C., Calabresse, J.R., Weissman, M., Reed, M., Fyre, M.A., Keck, P., McElroy, S., Lewis, L., Tierce, J., Wagner, K.D., Hazard, E., 2003. Validity of the mood disorder questionnaire: a general population study. Am. J. Psychatry 160, 178–180. Hirschfeld, R.M., Cass, A.R., Holt, D.C., Carlson, C.A., 2005. Screening for bipolar disorder in patients treated for depression in a family medicine clinic. J. Am. Board Fam. Pract. 18, 233–239. Isometsa, E., Suominen, K., Mantere, O., Valtonen, H., Leppamaki, S., Pippingskold, M., Arvilommi, P., 2003. The mood disorder questionnaire improves recognition of bipolar disorder in psychiatric care BMC Psychiatry. 3, 8. Kemp, D.E., Hirschfeld, R.M., Ganocy, S.J., Elhaj, O., Slembarski, R., Bilali, S., Conroy, C., Pontau, J., Findling, R.L., Calabrese, J.R., 2008. Screening for bipolar disorder in a county jail at the time of criminal arrest. J. Psychiatr. Res. 42, 778–786. Konuk, N., Kiran, S., Tamam, L., Karaahmet, E., Aydin, H., Atik, L., 2007. Validation of the Turkish version of the mood disorder
184
J. Twiss et al. / Journal of Affective Disorders 110 (2008) 180–184
questionnaire for screening bipolar disorders. Turk. Psikiyatri. Derg. 18, 147–154. Lish, J.D., Dime-Meenan, S., Whybrow, P.C., Price, R.A., Hirschfeld, R.M., 1994. The National Depressive and Manic depressive Association (DMDA) survey of bipolar members. J. Affect. Disord. 31 (4), 281–294. Li, J., McCombs, J.S., Stimmel, G.L., 2002. Cost of treating bipolar disorder in the California Medicaid (Medi-Cal) program. J. Affect. Disord. 71 (1–3), 131–139. McCombs, J.S., Ahn, J., Tencer, T., Shi, L., 2007. The impact of unrecognized bipolar disorders among patients treated for depression with antidepressants in the fee-for-services California Medicaid (Medi-Cal) program: a 6-year retrospective analysis. J. Affect. Disord. 97, 171–179. Miller, C.J., Klugman, J., Berv, D.A., Rosenquist, K.J., Ghaemi, N., 2004. Sensitivity and specificity of the Mood Disorder Questionnaire for detecting bipolar disorder. J. Affect. Disord. 81, 167–171.
Phelps, J.R., Ghaemi, S.N., 2006. Improving the diagnosis of bipolar disorder: predictive value of screening tests. J. Affect. Disord. 92, 141–148. Serretti, A., Olgiati, A., 2005. Profiles of ‘‘manic’’ symptoms in bipolar I, bipolar II and major depressive disorders. J. Affect. Disord. 84, 159–166. Vieta, E., Sanchez-Moreno, J., Bulbena, A., Chamorro, L., Ramos, J.L., Artal, J., Perez, F., Oliveras, M.A., Valle, J., Lahuerta, J., Angst, J., 2007. Cross validation with the mood disorder questionnaire (MDQ) of an instrument for the detection of hypomania in Spanish: the 32 item hypomania symptom check list (HCL-32). J. Affect. Disord. 101, 43–55. Weber Rouget, B., Gervasoni, N., Dubuis, V., Gex-Fabry, M., Bondolfi, G., Aubry, J.M., 2005. Screening for bipolar disorders using a French version of the Mood Disorder Questionnaire (MDQ). J. Affect. Disord. 88, 103–108.
Brief report
Validation of the Mood Disorder Questionnaire for screening for bipolar disorder in a UK sample James Twiss a , Steven Jones a , Ian Anderson b,⁎ a
School of Psychological Sciences, Clinical and Health Psychology, The University of Manchester, Manchester, UK b Neuroscience and Psychiatry Unit, The University of Manchester, Manchester, UK Received 10 October 2007; received in revised form 19 December 2007; accepted 19 December 2007 Available online 7 February 2008
Abstract Background: The Mood Disorder Questionnaire (MDQ) was designed as a screening questionnaire for bipolar disorder. Previous research has raised questions about the suitability of the MDQ structure for screening for bipolar II disorder. This study investigated the optimal sensitivity and specificity cut-off thresholds for the MDQ in bipolar I and bipolar II patients in a UK sample. Methods: The MDQ was administered to patients before attending a tertiary mood disorders clinic. Diagnostic interviews were used to determine DSM-IV diagnoses and these were used as the gold standard against which to investigate the performance of the MDQ. Results: 54 patients with bipolar spectrum disorder and 73 patients with unipolar depressive disorder completed the MDQ. With the original scoring criteria (symptoms and supplementary questions) the sensitivity for bipolar disorder was 0.76 (bipolar I disorder 0.83, bipolar II disorder 0.67) with specificity 0.86. The optimal cut-off score in the current sample was a score of 9 or more endorsed symptoms without applying the supplementary questions (sensitivity of 0.90 and 0.88 for bipolar I and bipolar II groups respectively with a specificity of 0.90). Limitations: The sample was drawn from a tertiary mood disorders clinic. Conclusions: The MDQ appears to be a useful screening tool for bipolar spectrum disorder in UK psychiatric practice with sensitivity for bipolar II disorder improved by dropping the supplementary sections. Further investigation of the optimal cut-off scores of the MDQ is needed to determine its utility in non-specialist and community based samples. © 2008 Elsevier B.V. All rights reserved. Keywords: Bipolar disorder; Mood Disorder Questionnaire; Sensitivity; Specificity
1. Introduction Bipolar disorder often goes unrecognised with delays of 8 years or more common before accurate diagnosis ⁎ Corresponding author. University of Manchester, Neuroscience and Psychiatry Unit, Room G809 Stopford Building, Oxford Road, Manchester M13 9PT, UK. Tel.: +44 161 275 7428; fax: +44 161 275 7429. E-mail address: [email protected] (I. Anderson). 0165-0327/$ - see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2007.12.235
(Lish et al., 1994) which can result in a greater burden of illness, risk of suicide (Dunner, 2003) and inappropriate treatment (Altshuler et al., 1995). Bipolar spectrum disorders have been estimated to cost the UK around £2 billion per annum (Das Gupta and Guest, 2002) and unrecognised cases may have a greater cost for health services than recognised and treated cases (Li et al., 2002; McCombs et al., 2007). A particular challenge is presented by ‘milder’ forms of bipolar disorder, such as bipolar II disorder and bipolar
J. Twiss et al. / Journal of Affective Disorders 110 (2008) 180–184
NOS. These are typically characterised by hypomania which may not be recognised as pathological by the individual or others (Akiskal et al., 2000). Treatment is likely to be sought for depressive episodes alone (Hirschfeld, 2001) often leading to a unipolar rather than bipolar diagnosis (Ghaemi et al., 2000). The Mood Disorder Questionnaire (MDQ; Hirschfeld et al., 2000) was developed to aid diagnosis of bipolar spectrum disorders. It covers symptom endorsement (Section 1), symptom clustering (Section 2) and severity of problem caused (Section 3) (see Methods). The original validation study reported that the MDQ performed well against DSM-IV diagnosis (sensitivity 0.73, specificity 0.90) in a psychiatric outpatient population. Although questionnaires alone are not an alternative to clinical interview (Miller et al., 2004; Phelps and Ghaemi, 2006) the performance of the MDQ suggested promise for initial screening for bipolar disorder in primary care and busy psychiatric clinics. MDQ validation studies in different US settings (Hirschfeld et al., 2003; Graves et al., 2007; Hirschfeld et al., 2005; Kemp et al., 2008; Miller et al., 2004) and in other countries (Hardoy et al., 2005; Isometsa et al., 2003; Konuk et al., 2007; Vieta et al., 2007; Weber Rouget et al., 2005) indicate that its psychometric properties vary somewhat according to the population studied. Its sensitivity, especially for bipolar II disorder, has been reported to be improved by exclusion or modification of Section 3 without markedly impairing its specificity (Benazzi, 2003; Isometsa et al., 2003; Kemp et al., 2008; Miller et al., 2004; Weber Rouget et al., 2005). No validation studies of the MDQ have been reported for a UK sample. Given the importance of early detection of bipolar spectrum disorder, the aims of the present study were to validate the MDQ in a UK sample, to assess sensitivity and specificity scores of the MDQ for bipolar I and bipolar II groups using the original cut-off guidelines and to assess its performance without Sections 2 and 3.
181
older; the study was approved by the Local NHS Research Ethics Committee. The MDQ comprises 3 Sections. In Section 1 participants are asked “Has there ever been a time when you were not your usual self and…” followed by 13 manic/ hypomanic symptoms (e.g. “…you had much more energy than usual”) answered ‘yes/no’. Section 2 asks whether more than one of these symptoms were experienced at the same time (answered ‘yes/no’). Section 3 asks participants what degree of problem the symptoms caused on a fourpoint scale ranging from ‘no problem’ to ‘serious problem’. In the original validation report the cut-off for a positive screen for bipolar disorder was ≥7 symptoms, ‘yes’ to Section 2 and a problem severity of ‘moderate’ or ‘serious’ in Section 3 (Hirschfeld et al., 2000). Sixteen of our subjects (13%), mostly unipolar patients with few symptoms endorsed, left Section 2 and/or 3 blank. The missing data were taken as ‘no’ (Section 2) or ‘no problem’ (Section 3) to avoid inflating sensitivity. Analysis without these subjects made little difference to the results. Data were analysed using SPSS 14 for windows. χ2 tests were used to test categorical data. Continuous data were analysed with independent t-tests (two tailed) or ANOVA with p b 0.05 taken as significant. Sensitivity and specificity are reported with 95% confidence interval (95%CI) for the main findings and relevant positive and negative predictive values. 3. Results The bipolar group included 30 bipolar I, 21 bipolar II and 3 bipolar NOS participants; the later two were combined for analysis (called bipolar II). The unipolar group included 68 patients with major depression (18 single episode, 50 recurrent), 3 with depressive disorder NOS and 2 with dysthymic disorder. Bipolar and unipolar groups did not differ with respect to age (means (sd): bipolar group 44.0 (11.8) years;
2. Methods Participants were sequential outpatient attendees of a tertiary NHS Specialist Service for Affective Disorders who completed the MDQ and then received a semistructured clinical interview covering current and past mood disorder (DSM-IV-TR diagnosis; American Psychiatric Association, 2000) conducted by experienced psychiatrists who had access to the MDQ results. The sample consisted of 127 patients; 54 had a bipolar spectrum disorder and 73 a unipolar diagnosis. Participants were English speaking, aged 18 years or
Fig. 1. Sensitivity and Specificity of the MDQ for bipolar I and II subjects for different cut-off scores.
182
J. Twiss et al. / Journal of Affective Disorders 110 (2008) 180–184
unipolar group 42.1 (11.8) years) or gender (male 40.7% and 43.8% for bipolar and unipolar groups respectively). Internal reliability of MDQ Section 1 symptoms was good with a Cronbach's alpha coefficient of 0.91. Item– item total correlations ranged from 0.41–0.81. Using the original cut-off guidelines (Hirschfeld et al., 2000) the overall MDQ sensitivity was 0.76 95% CI 0.63–0.85), higher for bipolar I disorder (0.83 95% CI: 0.66–0.93) than for bipolar II disorder (0.67 95% CI: 0.47–0.82). Specificity was 0.86 (95% CI: 0.77–0.92). Overall positive and negative predictive values were 0.80 and 0.83 respectively. The optimum cut-off for symptom number in our population was 9 (specificity 0.90, 95% CI: 0.82–0.95) without altering sensitivity. Analyses were then re-run using only Section 1. Fig. 1 shows the sensitivity and specificity of the MDQ for various cut-off scores for the bipolar I group and bipolar II groups with the optimal cut-off score 9 symptom items. Compared with the original scoring guidelines sensitivity was slightly improved for bipolar I (0.90, 95% CI: 0.74– 0.97) and greatly improved for bipolar II disorder (0.88, 95% CI: 0.69–0.96). Specificity also improved slightly (0.90, CI: 0.80–0.95). Overall positive and negative predictive values were 0.79 and 0.92 respectively. Table 1 shows frequency scores for the MDQ and how often each of the Sections were answered by each of the groups. There was a significant difference in total symptoms scores between unipolar and both bipolar groups (p b 0.001 for both), but not between bipolar I and bipolar II groups. Section 2 screened out 10% of the bipolar I group but none of the bipolar II group. Section 3 screened out 3.3% of the bipolar I group and Table 1 MDQ Section scores by group
Section 1 Symptom score Mean (SD)
Bipolar I
Bipolar II
11.07 (2.68)
10.12 (2.89)
Unipolar
4 (2.82)⁎⁎
Section 2 Clustering of symptoms? (%) Yes 28 (90%) No 3 (10%) Missing 0 (0%)
23 (95.8%) 49 (67.1%) 0 (0%) 12 (16.4%) 1 (4.2%) 12 (16.4%)
Section 3 Problem severity (%) No problem Minor problem Moderate problem Serious problem Missing
3 (12.5%) 4 (16.7%) 6 (25%) 10 (41.7%) 1 (4.2%)
⁎⁎p b 0.001.
0 (0%) 1 (3.3%) 10 (33.3%) 18 (60.0%) 1 (3.3%)
13 (17.8%) 16 (21.9%) 11 (15.1%) 22 (30.1%) 11 (15.1%)
29.2% of the bipolar II group using the standard scoring guidelines. Including Section 2 gave the same results as using Section 1 alone. Modifying Section 3 to include ‘minor problem’ improved the sensitivity for bipolar II disorder (0.79) with minor changes in sensitivity for bipolar I disorder (0.87) and specificity (0.85) compared with the original scoring guidelines. 4. Discussion The MDQ showed good reliability and validity in identifying bipolar disorder in UK patients attending a tertiary clinic. Internal consistency was good (0.91) and comparable with that reported in the original validation study (0.90). Our finding of a higher sensitivity for bipolar I disorder (0.83) than bipolar II disorder (0.67) using the original scoring guidelines is similar to that reported in previous studies (Isometsa et al., 2003; Kemp et al., 2008; Miller et al., 2004; Weber Rouget et al., 2005; Benazzi and Akiskal, 2003). The DSM-IV definition of hypomanic episodes includes an association with a ‘change in functioning’ rather than with ‘marked impairment’ which suggests that the problem severity question (Section 3) may incorrectly screen out individuals with bipolar II disorder. Miller et al. (2004) found that the reduction in sensitivity caused by Section 3 was equal for both bipolar I and bipolar II participants whereas we found it reduced sensitivity to a greater degree in the bipolar II group; this is consistent with other studies (Weber Rouget et al., 2005; Isometsa et al., 2003; Benazzi and Akiskal, 2003). However some of these studies changed the severity criterion to ‘minor problem’ rather than omitting it altogether; Isometsa et al. (2003) found that omission reduced specificity whereas we found only a slight decrease in specificity. It is therefore unclear whether omission or modification of Section 3 is the preferable general strategy to optimise the MDQ for detecting bipolar II disorder and may depend on population studied; in our patients omission was better. Including Section 2 did not improve the performance of the MDQ in our study. It is notable that a sizeable minority of the unipolar group did not complete Sections 2 and 3 suggesting some found these questions difficult to answer or not applicable. The bipolar I and bipolar II groups had a similar number of endorsed symptoms. DSM-IV does not differentiate between the symptomatology of hypomanic and manic episodes apart from ‘psychotic features’ specific for mania. The main distinction between bipolar I and II disorder is severity and length of episode. Differences in symptom patterns of hypomania vs. mania may however exist
J. Twiss et al. / Journal of Affective Disorders 110 (2008) 180–184
(Serretti and Olgiati, 2005) and our finding of a similar number of symptoms does exclude this possibility. A limitation of the study is the setting of a tertiary mood disorder service so that patients’ knowledge of diagnosis is likely to be higher than in other settings. The median sensitivity and specificity of the MDQ in psychiatric settings across different countries have been 0.68 (range 0.58 to 0.74) and 0.91 (range 0.67 to 0.96) respectively ( Miller et al. 2004; Hirschfeld et al. 2000; Weber Rouget et al., 2005; Konuk et al., 2007; Corona et al., 2007; Vieta et al., 2007). In our population the sensitivity was a little higher but its performance was largely comparable. In depressed patients treated with antidepressants in a primary care setting the sensitivity and specificity were 0.58 and 0.93 respectively (Hirschfeld et al., 2005) but a much lower sensitivity of 0.28 (specificity 0.97) was found in a community sample (Hirschfeld et al., 2003). However the latter study only identified 8 out of the 78 bipolar subjects as having bipolar II disorder which is not consistent with epidemiological evidence (Akiskal et al., 2000; Benazzi 2007) suggesting a problem with the method of identification of bipolar II disorder in that study. Our study suggests that the MDQ might be a useful tool in identifying bipolar II disorder in patients presenting with depression but further studies in nonspecialist psychiatric services and general practice using modified/omitted Section 3 criteria are needed to test this. Recent studies suggest that the Hypomania Checklist (HCL-32) is more sensitive at detecting hypomania than the original MDQ (Vieta et al., 2007); our results suggest that it should also be tested against the MDQ with Section 3 modified or omitted. The performance of the MDQ as a population screening instrument without Sections 2 and 3 also warrants further study. Role of funding source Nothing declared. Conflict of interest No conflict declared.
Acknowledgements We thank Kirsten Bond for assisting with the collection of the data and Manchester Mental Health and Social Care trust for their facilitation in the study. References Akiskal, H.S., Bourgeois, M.L., Angst, J., Post, R., Moller, H., Hirschfeld, R., 2000. Re-evaluating the prevalence of and
183
diagnostic composition within the broad clinical spectrum of bipolar disorders. J. Affect. Disord. 59, S5–S30. Altshuler, L.L., Post, R.M., Leverich, G.S., Mikalauskas, K., Rosoff, A., Ackerman, L., 1995. Antidepressant-induced mania and cycle acceleration: a controversy revisited. Am. J. Psychiatry 152, 1130–1138. American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental Disorders Fourth Edition. American Psychiatric Association, Washington DC. Text Revision. Benazzi, F., 2003. Improving the Mood Disorder Questionnaire to detect bipolar II disorder. Can. J. Psychiatry 48, 770–771. Benazzi, F., 2007. Bipolar II disorder: epidemiology, diagnosis and management. CNS Drugs 21 (9), 727–740. Benazzi, F., Akiskal, H.S., 2003. The dual factor structure of self-rated MDQ hypomania: energized-activity versus irritable-thought racing. J. Affect. Disord. 73, 59–64. Corona, R., Berlanga, C., Gutierrez-Mora, D., Fresan, A., 2007. Detection of bipolar disorder with a clinical screening questionnaire: the Spanish version of the Mood Disorder Questionnaire. Salud Ment. 30, 50–57. Das Gupta, R., Guest, J.F., 2002. Annual cost of bipolar disorder to UK society. Br. J. Psychiatry 180, 227–233. Dunner, L.D., 2003. Clinical consequences of under-recognized bipolar spectrum disorder. Bipolar Disord. 5, 456–463. Ghaemi, S.N., Boiman, E.E., Goodwin, F.K., 2000. Diagnosing bipolar disorder and the effect of antidepressants: a naturalistic study. J. Clin. Psychiatry 61, 804–808. Graves, R.E., Alim, T.N., Aigbogun, N., Chrishon, K., Mellman, T.A., Charney, D.S., Lawson, W.B., 2007. Diagnosing bipolar disorder in trauma exposed primary care patients. Bipolar. Disord. 9, 318–323. Hardoy, M.C., Cadeddu, M., Murru, A., Dell'Osso, B., Carpiniello, B., Morosini, P.L., Calabrese, J.R., Carta, M.G., 2005. Validation of the Italian version of the “Mood Disorder Questionnaire” for the screening of bipolar disorders. Clin. Pract. Epidemol. Ment. Health. 1, 8. Hirschfeld, R.M., 2001. Bipolar spectrum disorder: improving its recognition and diagnosis. J. Clin. Psychiatry 62 S, 5–9. Hirschfeld, R.M.A., Williams, J.B.W., Spitzer, R.L., Calabrese, J.R., Flynn, L., Keck Jr., P.E., Lewis, L., McElroy, S.L., Post, R.M., Rapport, D.J., Russell, J.M., Sachs, G.S., Zajecka, J., 2000. Development and validation of a screening instrument for bipolar: the Mood Disorder Questionnaire. Am. J. Psychiatry 157, 1873–1875. Hirshfeld, R.M.A., Holzer, C., Calabresse, J.R., Weissman, M., Reed, M., Fyre, M.A., Keck, P., McElroy, S., Lewis, L., Tierce, J., Wagner, K.D., Hazard, E., 2003. Validity of the mood disorder questionnaire: a general population study. Am. J. Psychatry 160, 178–180. Hirschfeld, R.M., Cass, A.R., Holt, D.C., Carlson, C.A., 2005. Screening for bipolar disorder in patients treated for depression in a family medicine clinic. J. Am. Board Fam. Pract. 18, 233–239. Isometsa, E., Suominen, K., Mantere, O., Valtonen, H., Leppamaki, S., Pippingskold, M., Arvilommi, P., 2003. The mood disorder questionnaire improves recognition of bipolar disorder in psychiatric care BMC Psychiatry. 3, 8. Kemp, D.E., Hirschfeld, R.M., Ganocy, S.J., Elhaj, O., Slembarski, R., Bilali, S., Conroy, C., Pontau, J., Findling, R.L., Calabrese, J.R., 2008. Screening for bipolar disorder in a county jail at the time of criminal arrest. J. Psychiatr. Res. 42, 778–786. Konuk, N., Kiran, S., Tamam, L., Karaahmet, E., Aydin, H., Atik, L., 2007. Validation of the Turkish version of the mood disorder
184
J. Twiss et al. / Journal of Affective Disorders 110 (2008) 180–184
questionnaire for screening bipolar disorders. Turk. Psikiyatri. Derg. 18, 147–154. Lish, J.D., Dime-Meenan, S., Whybrow, P.C., Price, R.A., Hirschfeld, R.M., 1994. The National Depressive and Manic depressive Association (DMDA) survey of bipolar members. J. Affect. Disord. 31 (4), 281–294. Li, J., McCombs, J.S., Stimmel, G.L., 2002. Cost of treating bipolar disorder in the California Medicaid (Medi-Cal) program. J. Affect. Disord. 71 (1–3), 131–139. McCombs, J.S., Ahn, J., Tencer, T., Shi, L., 2007. The impact of unrecognized bipolar disorders among patients treated for depression with antidepressants in the fee-for-services California Medicaid (Medi-Cal) program: a 6-year retrospective analysis. J. Affect. Disord. 97, 171–179. Miller, C.J., Klugman, J., Berv, D.A., Rosenquist, K.J., Ghaemi, N., 2004. Sensitivity and specificity of the Mood Disorder Questionnaire for detecting bipolar disorder. J. Affect. Disord. 81, 167–171.
Phelps, J.R., Ghaemi, S.N., 2006. Improving the diagnosis of bipolar disorder: predictive value of screening tests. J. Affect. Disord. 92, 141–148. Serretti, A., Olgiati, A., 2005. Profiles of ‘‘manic’’ symptoms in bipolar I, bipolar II and major depressive disorders. J. Affect. Disord. 84, 159–166. Vieta, E., Sanchez-Moreno, J., Bulbena, A., Chamorro, L., Ramos, J.L., Artal, J., Perez, F., Oliveras, M.A., Valle, J., Lahuerta, J., Angst, J., 2007. Cross validation with the mood disorder questionnaire (MDQ) of an instrument for the detection of hypomania in Spanish: the 32 item hypomania symptom check list (HCL-32). J. Affect. Disord. 101, 43–55. Weber Rouget, B., Gervasoni, N., Dubuis, V., Gex-Fabry, M., Bondolfi, G., Aubry, J.M., 2005. Screening for bipolar disorders using a French version of the Mood Disorder Questionnaire (MDQ). J. Affect. Disord. 88, 103–108.