Equating the PHQ-9 and GAD-7 to the HADS depression and anxiety subscales in patients with major depressive disorder

Highlights

• Used Item Response Theory-based method to link self-rating scales PHQ-9/GAD-7 to HADS.

• Generated crosswalk tables to directly convert raw scores from one scale to the other.

• Compared actual and converted scores to confirm that the linkages were successful.

Abstract

Objective

The present study aimed to equate the 9-item Patient Health Questionnaire (PHQ-9) and 7-item Generalized Anxiety Disorder Scale (GAD-7) to the Hospital Anxiety and Depression Scale (HADS) depression and anxiety subscales (HADS-D,HADS-A) respectively in patients with major depressive disorder (MDD) and generate crosswalks of raw scores.

Methods

As it is a single group design that adopts common-person equating method, a total of 460 patients with MDD completed the PHQ-9, GAD-7 and HADS at the same time. Rasch analysis was used to filter out invalid participants, investigate the psychometric properties of test items and participants, link the PHQ-9 and HADS-D as well as GAD-7 and HADS-A, and produce conversion tables respectively. The differences between original scores and converted scores were analyzed to validate the crosswalks.

Result

401 samples of depression part and 396 samples of anxiety part were left for final samples. Both the PHQ-9 / HADS-D and GAD-7 / HADS-A combined analysis adequately fit the unidimensional Rasch model, demonstrated acceptable reliability and item-person targeting and showed no disordering category. Slight differential item functioning across gender was found in item PHQ9 and item GAD6. The crosswalks were generated and verified to be validity.

Limitations

The results might be restricted to patients with MDD recruited in a single mental health center.

Conclusion

The PHQ-9, GAD-7 and HADS depression and anxiety subscales were successfully linked, producing conversion tables that could be used for directly converting raw score from one instrument to the other.

Introduction

Major depressive disorder (MDD) is globally prevalent, which may severely impair psychosocial functioning and reduce quality of life. In clinical practice, it takes challenges in detection and diagnosis because the symptoms of depression can present different forms of combinations and none of the symptoms is pathognomonic (Malhi and Mann, 2018).

The key symptoms of MDD can be broadly classified into emotional domain (depressed mood, anhedonia, feelings of worthless or guilty), neurovegetative domain (fatigue, insomnia, weight or appetite), and cognitive domains (disability to think or concentrate). Some symptoms are more specific to MDD such as depressed mood and anhedonia. Some symptoms such as fatigue and insomnia are also common in other mental and physical diseases, which make MDD patients more difficult to identify. Moreover, individuals with MDD often have features of anxiety disorders. Previous studies have reported that a majority of MDD patients experience significant anxiety and the comorbidity of depression and anxiety can increase impairment (Goldberg and Fawcett, 2012; Tiller, 2013). According to the Diagnostic and Statistical Manual of Mental Disorders (DSM)-5, a ‘with anxious distress’ specifier is included in the MDD section that demonstrates the necessity to measure anxiety symptoms for MDD patients (American Psychiatric Association, 2013).

Self-rating instruments are commonly used in measuring depression and concurrent anxiety symptoms for the purpose of guiding decision making in treatment. There are multiple well accepted and psychometrically tested instruments currently in use. For instance, the 9-item Patient Health Questionnaire (PHQ-9) and 7-item Generalized Anxiety Disorder Scale (GAD-7) are globally popular and strongly associated with depression and anxiety symptoms for being developed based on DSM criteria (Boulton et al., 2019). In addition, the Beck Depression Inventory (BDI), Beck Anxiety Inventory (BAI), Center for Epidemiological Studies Depression Scale (CES-D), Hospital Anxiety and Depression Scale (HADS) and the like are widely used as well. Though each of these distress measures is brief, freely available for clinical and research use, and shows robust measurement properties, it is unlikely that any of them could be regarded as a ‘gold standard’ to provide accurate assessments (Uher et al., 2012) and hence offers a variety of options for clinical use. Given that different instruments produce scores on different metrics, it is hard to compare results of different instruments directly. In practical, clinicians and investigators need a crosswalk to transform raw scores from one instrument to the other across situations that use either one.

The methodology for linking and equating scores from instruments that measure similar construct has been made greater use in clinical fields. Boulton et al. (2019) developed linkages between GAD-7, PHQ-9 and the anxiety and depression items of the Traumatic Brain Injury Quality of Life (TBI-QOL) measurement system respectively. Some other studies connected instruments for assessing functional status in activities of daily living (Velozo et al., 2007; Hong et al., 2018; Li et al., 2018). Some previous studies had done comparisons between the HADS depression subscale (HADS-D) and PHQ-9 (Hansson et al., 2009; Kendel et al., 2010; Cameron et al., 2013; Yuan et al., 2019), as well as between the HADS anxiety subscale (HADS-A) and GAD-7 (Esser et al., 2018; Hartung et al., 2017; Snijkers et al., 2021). However, to the best of our knowledge, neither the HADS-D and PHQ-9 nor the HADS-A and GAD-7 has been equated and employed to generate a crosswalk in either Chinese populations or those from other countries.

This study aimed to link the PHQ-9 and GAD-7 respectively to the HADS depression and anxiety subscales in patients with MDD and, thus, establish crosswalks. The HADS was firstly developed to screen for anxiety and depression symptoms in patients with somatic problems, which is composed of a depression subscale (HADS-D) and an anxiety subscale (HADS-A) (Beekman and Verhagen, 2018). Different from the historically used PHQ-9 and the GAD-7, the HADS excludes somatic symptoms related to emotional and physical disorders such as insomnia and weight loss. Therefore, the HADS may avoid confounding the diagnosis in patients who have both emotional and physical illnesses and then improve the sensitivity in screening for depression and anxiety (Brennan et al., 2010). Since the PHQ-9, GAD-7 and HADS are widely used, the establishment of an effective crosswalk would be able to compare or convert their relative raw scores directly and improve efficiency in clinical assessment facilities.