Patient Health Questionnaire-9 and 17-item Hamilton Depression Rating Scale for detecting post-stroke depression in cerebral infarction

Abstract

BACKGROUND

The prevalence of post-stroke depression (PSD) is approximately 30%-50% among patients with cerebral infarction, which hinders recovery and increases mortality: However the diagnosis rate is low as its symptoms overlap with other conditions. Although 17-item Hamilton Depression Rating Scale (HAMD-17) is the clinical reference standard, its complexity limits its use for screening in routine practice. Patient Health Questionnaire-9 (PHQ-9) is easily accessible but has variable validity in stroke cohorts. This study assessed the diagnostic accuracy of both scales against International Classification of Diseases, 10^th Revision criteria in order to guide PSD screening practices.

AIM

To compare the diagnostic efficacy of the HAMD-17 and the PHQ-9 in screening for PSD in cerebral infarction patients, and to explore the optimal diagnostic cutoff values for both scales.

METHODS

Retrospective analysis was performed on clinical data of 345 cerebral infarction patients who were admitted to our institution between February 2020 and March 2025. Psychiatrists performed structured clinical interviews based on the International Classification of Diseases, 10^th Revision criteria to make the gold-standard diagnosis. HAMD-17 and PHQ-9 were completed by all patients. Diagnostic efficacy of both scales was assessed using receiver operating characteristic curve analysis. Difference in area under the curve (AUC) was compared using the DeLong method. Best diagnostic cutoff values were based on the principle of maximum value of Youden index, and Kappa consistency test was calculated to assess the agreement between scale diagnoses and gold standard.

RESULTS

PSD was found in 112 of 345 patients (incidence: 32.46%). The AUC of HAMD-17 for detecting PSD was 0.891 (95%CI: 0.854-0.928), and the AUC of PHQ-9 was 0.867 (95%CI: 0.826-0.908), and there was no significant difference between them (Z = 1.426, P = 0.154). The optimal diagnostic cutoff value for HAMD-17 was 7 scores; its sensitivity was 88.39%, specificity was 79.83%, Youden index was 0.682, and Kappa value was 0.632. The best cutoff value of PHQ-9 for diagnosis is 7 points, at which sensitivity equals is 83.04%, specificity equals is 81.12%, Youden index equals is 0.642, and Kappa value is 0.606.

CONCLUSION

Both HAMD-17 and PHQ-9 have good diagnostic efficacy for PSD screening without significant difference in AUC. As a simple and easy-to-use self-rating tool, PHQ-9 can be the desired scale for early screening of PSD in neurology wards. We suggest a score of ≥ 7 as the most useful positive screening cutoff and recommend that patients screening positive be assessed by a specialist for confirmation of the diagnosis.

Keywords: Post-stroke depression; Hamilton Depression Rating Scale; Patient Health Questionnaire-9; Diagnostic efficacy; Receiver operating characteristic curve

Core Tip: Post-stroke depression is prevalent and often underrecognized in neurological environments. This study directly compares the diagnostic utility of the 17-item Hamilton Depression Rating Scale clinician-rated and self-administered Patient Health Questionnaire-9 (PHQ-9) in patients with cerebral infarction, using International Classification of Diseases, 10^th Revision structured interviews as reference standard. The diagnostic accuracy of both scales was good, and there was no significant difference in the area under the curve values. The best sensitivity and specificity was at a PHQ-9 cutoff score of ≥ 7. Because it is simple and feasible, PHQ-9 might be preferred to assess post-stroke depression in neurology wards for screening alone and then require confirmation by a specialist assessment.