Article Text
PDF
Statistics from Altmetric.com
Question
Question:
Is the Patient Health Questionnaire 9 (PHQ-9) an accurate tool to screen for, diagnose and measure the severity of depressive disorder in groups that are at high risk for depressive disorder?
Patients:
A total of 664 high-risk patients aged between 18 and 70 years selected from patients enrolled at 6 health centres with 23 GPs and falling into one of 3 high-risk groups for depressive disorder: patients with unexplained somatic complaints; frequent attenders, defined as patients with consultation rates in the highest 10% in the year prior to study allocation; patients with a new mental health problem up to 3 months prior to the selection date. All participants completed the PHQ screening questionnaire.
Setting:
Amsterdam; primary care patients from two academic medical centres, The Academic Medical Centre in Amsterdam and the Radbound University Nijmegen Medical centre; recruited from March 2006 through July 2007.
Test:
The Patient Health Questionnaire-9, which evaluates the presence of one of the nine DSM-IV criteria of a depressive episode in the past 2 weeks: depressed mood, anhedonia, having trouble sleeping, feeling tired, change in appetite or weight, feelings of guilt and worthlessness, having trouble concentrating, feeling slowed down or restless, and having suicidal thoughts. Four response categories: 0 (not at all), 1 (few days), 2 (more than half the days), 3 (almost every day).
Diagnostic standard:
For assessment of screening and diagnostic properties: clinical diagnosis by trained researchers using the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) interview and for assessment of severity: the 17-item Hamilton Depression Rating Scale (HDRS-17).
Outcomes:
Sensitivity and specificity for screening and diagnosing depressive disorders.
Methods
Design:
Diagnostic cohort study.
Main results
An initial 2,659 patients met at least one criteria for high risk, but after exclusions by physicians, failure to return the questionnaire, lack of informed consent and failure to contact, 479 patients were available. PHQ-9 and SCID-I were analyzed in 440 patients, 17% of the original eligible group; another 82 of 664 (12.3%) of patients were diagnosed with depression using SCID-I. PHQ-9 performed well as a screening instrument: sensitivity: 0.93, 95 % CI 0.84 to 0.97; specificity: 0.85, 95% CI: 0.82 to 0.97; and as a diagnostic tool, although there was a higher false-negative rate (lower sensitivity) and other measures, for example, sensitivity and specificity (sensitivity: 0.68, 95 % CI 0.57 to 0.78; specificity: 0.95, 95% CI 0.93 to 0.97). As a tool for assessing severity, correlation between scores on the PHQ-9 and HDRS-17 was low, r=0.52 (p<0.01).
Conclusions
The PHQ-9 performs well as a screening instrument; however, in diagnosing depressive disorder, a formal diagnostic process following PHQ-9 remains imperative. The PHQ-9 does not seem adequate for measuring severity, compared to the HDSR-17.
Abstracted from
Commentary
The identification of depression in primary care patients by family/general practice physicians has been an active focus of research and debate for the past decade. Spitzer and colleagues1 developed the Patient Health Questionnaire-9 as a brief self-report inventory derived from the PRIME-MD clinical interview to measure the nine core DSM-IV diagnostic symptoms associated with a depressive episode. The utility of measures such as the PHQ-9 rests in their ability to accurately and quickly screen for cases that merit further diagnostic evaluation by physicians. In addition to its use as a self-report severity measure, the PHQ-9 also includes a scoring algorithm that applies DSM-IV disorder criteria to the 9 items along with a 10th item which examines symptoms' impact on functioning.
Wittkampf and colleagues provide strong and substantial evidence for the clinical utility of the PHQ-9 as a screening measure of depression for use in primary care. On the basis of initial administration of the PHQ-9 to 980 patients of family physicians in the Netherlands, 3 initial groups were created with the PHQ-9, which was also re-administered to 440 completed cases who were interviewed with the SCID-I, in order to obtain reference diagnoses. Useful diagnostic efficacy data were obtained that suggested that the PHQ-9 severity cutoff score of 10 or greater is superior to the PHQ-9 diagnostic scoring algorithm. The latter finding is particularly noteworthy for clinical use and the application of the PHQ-9 in primary care patient screening.
A relatively moderate relationship between the PHQ-9 and HDRS-17 was found, which the authors suggest limits its usefulness for measuring depression severity. However, this issue aside, this study provides strong evidence in support of the PHQ-9 as a brief screening measure for cases of depression in patients who are frequent health care seekers, have mental health problems, or present unexplained somatic complaints. The authors appropriately note that patients with PHQ-9 scores at or above the cut-off require subsequent diagnostic evaluation to determine presence or absence of a formal depressive disorder.
References
Footnotes
-
Sources of funding The Netherlands Organization for Health Research and Development (ZonMw) Mental Health Program and the academic Medical Center/University of Amsterdam
Footnotes
-
Competing interests WR is the co-author of the Hamilton Depression Inventory, published by Psychological Assessment Resources.