日本語版Primary Care Evaluation of Mental Disorders Patient Health Questionnaire-9(PRIME-MD PHQ-9)の精神尺度としての有用性に関する検討

Psychometric Evaluation for the Japanese Version of the Primary Care Evaluation of Mental Disorders Patient Health Questionnaire-9 (PRIME-MD PHQ-9) as a Depression Assessment Tool

抄録

　日本語版PHQ-9の信頼性・妥当性を検討した。精神科医院外来通院中の患者471人にPHQ-9とHAMD-17構造化面接を施行し，一定間隔後追跡調査を行った。PHQ-9は2回の検査とも単因子構成で，負荷係数・分散ともに高値を示した。HAMD-17の因子構成は安定せず，負荷係数の小さな項目が多数みられた。PHQ-9とHAMD-17の得点間には強い相関が得られ，得点分布からPHQ-9のほうがうつの重症度をより詳細に反映していた。PHQ-9により対象を大うつ病群・その他のうつ病群・非うつ群に分けてHAMD-17の得点を一元配置分析したところ，この分類法にも信用性があることが明示された。Cronbachのα係数・再検査の級内相関係数ともに高値であった。PHQ-9はHAMD-17より有用性の高い評価尺度として臨床での活用が期待される。

　To date, the validity and reliability for the Japanese version of the Primary Care Evaluation of Mental Disorders Patient Health Questionnaire-9 (PHQ-9) has not been reported. We conducted a psychometric evaluation study to examine its utility as a tool to assess depression levels. The participants comprised 471 outpatients of 2 psychiatric clinics, and the follow-up data were collected at a minimum interval of 4 weeks. Principal component analysis revealed that the PHQ-9 had a unifactorial nature and it had higher item-factor loadings than the more commonly used 17-item Hamilton Depression Rating Scale Interview Guide (HAMD-17). For the HAMD-17 the factor structure was inconsistent and a substantial number of items contributed only moderately, even to a lesser extent, to the total score. The high correlation between the PHQ-9 scores and the HAMD-17 scores supported convergent validity for the PHQ-9. Analysis of score distribution indicated that the PHQ-9 could measure very minute differences in the depression levels unlike the HAMD-17. According to the diagnostic classification by the PHQ-9 scores, the participants were categorized into 3 groups：major depression, other depression, and non-depression. Contrasting group analysis using one-way analysis of variance showed that the major depression group had higher scores on the HAMD-17 than the other depression group, and the non-depression group had even lower scores. Moreover, Cronbach's alpha and test-retest coefficients indicated sufficient reliability for the PHQ-9. Thus the PHQ-9 is recommended for use in research and in clinical practice as a more precise and convenient substitute to the HAMD-17.