How to compare scores from different depression scales: equating the Patient Health Questionnaire (PHQ) and the ICD‐10‐Symptom Rating (ISR) using Item Response Theory

Abstract

A wide range of questionnaires for measuring depression are available. Item Response Theory models can help to evaluate the questionnaires exceeding the boundaries of Classical Test Theory and provide an opportunity to equate the questionnaires. In this study after checking for unidimensionality, a General Partial Credit Model was applied to data from two different depression scales [Patient Health Questionnaire (PHQ-9) and ICD-10-Symptom Rating (ISR)] obtained in clinical settings from a consecutive sample, including 4517 observations from a total of 2999 inpatients and outpatients of a psychosomatic clinic. The precision of each questionnaire was compared and the model was used to transform scores based on the assumed underlying latent trait. Both instruments were constructed to measure the same construct and their estimates of depression severity are highly correlated. Our analysis showed that the predicted scores provided by the conversion tables are similar to the observed scores in a validation sample. The PHQ-9 and ISR depression scales measure depression severity across a broad range with similar precision. While the PHQ-9 shows advantages in measuring low or high depression severity, the ISR is more parsimonious and also suitable for clinical purposes. Furthermore, the equation tables derived in this study enhance the comparability of studies using either one of the instruments, but due to substantial statistical spread the comparison of individual scores is imprecise.