Evaluating Cutpoints for Mental Health Scales
Author Information
Author(s): Kelly Mark J, Dunstan Frank D, Lloyd Keith, Fone David L
Primary Institution: Dept. of Primary Care and Public Health, Centre for Health Sciences Research, School of Medicine, Cardiff University
Hypothesis
This paper aims to produce cutpoints for the MHI-5 and MCS by comparison with the General Health Questionnaire (GHQ-12).
Conclusion
The Youden Index and (0,1) methods are most suitable for determining a cutpoint for the MHI-5, since they are least dependent on population prevalence.
Supporting Evidence
- The MHI-5 performs remarkably well against the longer MCS.
- The correlation between the MHI-5 and the MCS was 0.88.
- The Youden Index and (0,1) methods both gave a cutpoint of 76 for the MHI-5.
- The choice of method is dependent on the intended application.
- Using the sample prevalence of 25.3%, the corresponding cutpoint for minimising the error rate was 60.
Takeaway
Researchers wanted to find the best scores to identify mental health issues using two different tests, and they found that a score of 76 on one test is a good indicator.
Methodology
Data were analysed from wave 9 of the British Household Panel Survey (2000) using Receiver Operating Characteristic (ROC) curves to compare the scales and define cutpoints.
Potential Biases
There is a potential for overestimating sensitivity and specificity since the same dataset was used to define cutpoints and assess their performance.
Limitations
The study used the GHQ-12 as a comparative gold standard, which may have lower sensitivity and specificity than a standardised interview schedule.
Participant Demographics
The dataset includes individuals from England, Scotland, and Wales, with a representative sample of the UK population.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website