Measuring Mental Health by Questionnaires in Primary Care – Unidimensionality, Responsiveness and Compliance

Kaj Sparle Christensen, Per Bech, Per Fink
European Psychiatric Review, 2010;3(1):8-12
pdf icon download pdf ejournal view ejournal

Abstract

Background: The aim of this study is to compare the homogeneity and responsiveness of established self-rating scales measuring mental health by family doctors, namely selected Symptom Checklist-90 (SCL-90) subscales, the General Health Questionnaire (GHQ-12), WHO-5 and Short Form 36 (SF-36). Methods: One thousand, seven hundred and eighty-five primary care patients 18–65 years of age were included in the study and followed up three months later. The homogeneity of the different scales was assessed using the Mokken analysis. In the GHQ-12, factor analysis was used to evaluate the compliance of the respondents. Responsiveness indices were calculated in terms of standardised response means (SRM). Results: Factor analysis identified two factors within the GHQ-12, one of which included the positively phrased items and the other the negatively phrased items, i.e. acceptable compliance. The Mokken analysis showed that all of the included scales had an acceptable unidimensionality, i.e. their total scores are sufficient statistics. The SRM showed the highest response in the diagnostic groups of depression and anxiety. The WHO-5 showed that the caseness of minor affective states followed the ICD-10 hierarchical structure so that depression had the highest rate of caseness, followed by anxiety disorder and somatisation disorder. Conclusions: The homogeneity, responsiveness and compliance of the common questionnaires for measuring mental health in the primary care setting were found to be acceptable.

Acknowledgements: The authors wish to thank Ove Aaskoven for the requested statistical analysis.
Authors’ contributions: Kaj Sparle Christensen acquired the data for the study, assisted in data analysis and drafted the manuscript. Per Bech conceived the idea for the manuscript. Per Bech and Per Fink helped to draft the manuscript. All authors read and approved the final manuscript.
Keywords
Depression, screening questionnaires, responsiveness, compliance
Disclosure The authors have no conflicts of interest to declare.
Received: July 14, 2009 Accepted November 16, 2009
Correspondence: Kaj Sparle Christensen, Research Unit for General Practice, Aarhus University, Vennelyst Boulevard 6, 8000 Aarhus C, Denmark. E: kasc@alm.au.dk

Mental disorders are very common among primary care patients, but often remain unrecognised by family doctors.1,2 Easy-to-use instruments are needed for assistance in case identification, for measuring of symptom severity (unidimensionality) and for monitoring changes in mental health (responsiveness over time).3 The reluctance to use patient-administrated questionnaires measuring mental disorders is often due to the perceived problem of the respondents’ compliance; that is, their test-taking behaviour.

The Common Mental Disorder Questionnaire (CMDQ) has recently been introduced as an instrument for measuring mental health status in the primary care setting.4 The CMDQ is a short composite instrument including subscales from the Hopkins Symptom Checklist (SCL-90),5,6 the Whiteley-7 scale7 and the alcohol abuse disorder questionnaire ‘Cut down, Anger, Guilt and Eye-opener’ (CAGE).8

The CMDQ was primarily designed to provide the general practitioner with specific clinical measures of somatisation including illness worry, emotional distress, depression, general anxiety and alcohol problems. All of the scales used in the CMDQ are in the public domain and no permission to use is necessary unless used in a profit-making endeavour.5,7,8

The General Health Questionnare (GHQ-12)9 has also been designed to screen for mental disorders in the medically ill, but has mainly been employed as a screening instrument in epidemiological studies or primary care research; it has rarely been used in clinical practice. However, in contrast to the CMDQ, the GHQ-12 includes both positively formulated items (measuring positive wellbeing) and negatively formulated items (measuring symptoms of anxiety or depression). A recent study found it of psychometric importance to separate these two dimensions by factor analysis in order to ensure the test-taking behaviour of the respondents as this problem of compliance is actually behind the strategy of having both positively and negatively formulated items.10 Epidemiological studies with the GHQ-12 have shown that a score of 12 or higher indicates a minor psychiatric case.9 The WHO-511 is a scale designed to measure positive psychological wellbeing.12 Henkel et al. showed that the WHO-5 was superior to GHQ-12 in identifying depression in the family doctor setting.13 A score below 50 indicates a minor psychiatric case.

References:
  1. Toft T, Fink P, Oernboel E, et al., Psychol Med, 2005;35:1175–84.
  2. Christensen KS, Toft T, Frostholm L, et al., Fam Pract, 2005;22;428–34.
  3. Guyatt G, Walter S, Norman G, J Chronic Dis, 1987;40:171–8.
  4. Christensen KS, Fink P, Toft T, et al., Fam Pract, 2005;22:448–57.
  5. Derogatis LR, Lipman RS, Covi L, Psychopharmacol Bull, 1973;9:13–28.
  6. Fink P, Jensen J, Borgquist L, et al., Acta Psychiatr Scand, 1995;92:409–18.
  7. Fink P, Ewald H, Jensen J, et al., J Psychosom Res, 1999;46:261–73.
  8. Ewing J, JAMA, 1984;252:1905–7.
  9. Goldberg DP, Gater R, Sartorius N, et al., Psychol Med, 1997;27:191–7.
  10. Bech P, Gormsen L, Loldrup D, Lunde M, J Affect Disord, 2009 Apr 4 (Epub ahead of print).
  11. Bech P, Olsen LR, Kjoller M, Rasmussen NK, Int J Methods Psychiatr Res, 2003;12:85–91.
  12. McCabe CJ, Thomas KJ, Brazier JE, Coleman P, Br J Psychiatry, 1996;169:516–21.
  13. Henkel V, Mergl R, Kohnen R, et al., BMJ, 2003;326:200–1.
  14. Ware JE Jr., Spine, 2000;25:3130–39.
  15. Hays RD, Hadorn D, Qual Life Res, 1992;1:73–5.
  16. Derogatis LR, Lipman RS, Rickels K, et al., Behav Sci, 1974;19:1–15.
  17. Mollica RF, Caspi-Yavin Y, Bollini P, et al., J Nerv Ment Dis, 1992;180:111–16.
  18. Fink P, Orbol E, Hansen MS, et al., J Psychosom Res, 2004;56:371–5.
  19. Goldberg D, Lancet, 1974;2(7891):1245–8.
  20. Keller SD, Ware JE Jr, Bentler PM, et al., J Clin Epidemiol, 1998;51:1179–88.
  21. WHO. SCAN, version 2.1, Geneva: World Health Organization, Introduction to Non-parametric Item Response Theory, London: Sage Publications, 2002.
  23. Licht RW, Qvitzau S, Allerup P, Bech P, Acta Psychiatr Scand, 2005;111:144–9.
  24. Mokken RJ, Theory and Procedure of Scale Analysis, Paris: Mouton, 1971.
  25. DeJong A, Molenaar IW, J Psychiatr Res, 1987;21:137–49.
  26. Lowe B, Kroenke K, Herzog W, Grafe K, J Affect Disord, 2004;81:61–6.
  27. Revicki DA, Cella D, Hays RD, et al., Health Qual Life Outcomes, 2006;4:70.
  28. Child D, The Essentials of Factor Analysis, second edn. London: Cassed Educational, 1990.
  29. Boyer P, Montgomery S, Lepola U, et al., Int Clin Psychopharmacol, 2008;23:243–53.
close

This content is copyright protected

However, if you would like to share the information on this webpage, you may use the headline and link below:

Measuring Mental Health by Questionnaires in Primary Care – Unidimensionality, Responsiveness and Compliance

Read more: Measuring Mental Health by Questionnaires in Primary Care – Unidimensionality, Responsiveness and Compliance