Patient Health Questionnaire-9 (PHQ-9)

# of studies reporting psychometric properties: 8

Interpretability

MCID: not established for the SCI population, but for a sample of older primary care patients (n = 434, mean age = 71 (7.4) years, all participants enrolled in the Improving Mood-Promoting Access to Collaborative Treatment (IMPACT)):

MCID = 5 points

Reference: Lowe, B., Unutzer, J., et al. (2004). “Monitoring depression treatment outcomes with the patient health questionnaire-9.” Med Care 42(12): 1194-1201.

SEM: not established for the SCI population, but for a sample of older primary care patients (see Lowe et al. 2004 reference above): SEM for change due to treatment and no control of prior depression = 2.44 SEM for the same number of DSM-IV depressive symptoms at both assessments = 1.32

MDC: not established

• Higher scores indicate increased severity of depression
• Mean (SD) scores: 5.57 (5.74)
  (Krause et al. 2009; PHQ-9; n=727, 70.2% males; 53.3% cervical injuries; mean time since injury = 18.2 years)
• A cut-off score of 10 has been reported to indicate major depression for the general population (Kroenke et al. 2001: n=6000).

Reliability

• High Internal Consistency:
  Cronbach’s a = 0.71-0.87

(Richardson & Richards 2008: PHQ-9; n=2570, 2013 males; time post-injury range = 1-25 years)
(Summaka et al. 2019: PHQ9-Arabic version; n=51; 51 males; 37 paraplegia and 14 tetraplegia)

• High Test-Retest Reliability:
  ICC = 0.88 (0.711-0.955); P < 0.001

(Summaka et al. 2019: PHQ9-Arabic version; n=51; 51 males; 37 paraplegia and 14 tetraplegia)

Validity

• Correlation of the PHQ-9 is:
  • High with the Older Adult Health and Mood Questionnaire (Spearman’s r=0.781)
  • High with major depressive disorder (MDD) (Spearman’s r=0.530)
  • High with the Hamilton Depression Rating Scale (Spearman’s r=0.713)
  • Moderate with the Satisfaction with Life Scale (Spearman’s r=-0.477)
  • Low with the Connor-Davidson Resilience Scale (CD-RISC-8) (Correlation=-0.39)
• PHQ-9 scores were inversely and Moderately correlated with subjective health on the SF-1 (Spearman’s r=0.37).
• For a 3-item screening test with a score cutoff of 3, a sensitivity of 0.87 and specificity of 0.93 were reported; with a score cutoff of 4, a sensitivity of 0.82 and a specificity of 0.95 were reported (Graves & Bombardier 2008).
• For the total PHQ-9, a cutoff of 11 was determined to have optimal diagnostic accuracy of MDD. At this cutoff, the PHQ-9 detected 100% (sensitivity) of those with a diagnosis of MDD and had a specificity of 84% (Bombardier et al. 2012).

(Bombardier et al. 2004: PHQ-9; n=849, 645 males, 204 females; 47.6% ASIA A; 45.5% paraplegia; all 1 year post-SCI)
(Bombardier et al. 2012: PHQ-9; n=142, 111 males, 31 females; injury level: 95 cervical, 32 thoracic, 11 lumbar, 4 sacral)
(Krause et al. 2009: PHQ-9; n=727, 70.2% males; mean time since injury = 18.2 years; 53.3% cervical injury)
(Richardson & Richards 2008: PHQ-9; n=2570, 2013 males; time post-injury range = 1-25 years)
(Summaka et al. 2019: PHQ9-Arabic version; n=51; 51 males; 37 paraplegia and 14 tetraplegia)
(Chiu et al. 2024: PHQ-9; n=93, 58 males; mean (SD) time since injury = 2.43 (8.29) years)

Responsiveness

No values were reported for the responsiveness of the PHQ-9 for the SCI population.

Floor/ceiling effect

Floor: 22% reported no depressive symptoms
(Williams et al. 2009: PHQ-9; n=202, 77% males; median time since injury = 7 years)

Reviewer

Jane Hsieh, Dr. Carlos L. Cano-Herrera, Elsa Sun, Tyra Chu

Date Last Updated

31 December 2024