Tools for Practice


#203 Got Depression? I can answer that in two questions!


CLINICAL QUESTION
What is the diagnostic accuracy of the 2-question screen for identifying depression in primary care?

BOTTOM LINE
The 2-question screen is good at ruling out (but not ruling in) depression in primary care. Up to 50% of patients will test positive and should have more thorough evaluation to confirm depression diagnosis. Whether screening alters outcomes is debatable, but the 2-question screen may be reasonable for case-finding or screening higher risk patients. 


CFPCLearn Logo

Reading Tools for Practice Article can earn you MainPro+ Credits

Join Now

Already a CFPCLearn Member? Log in



EVIDENCE
2-Question Screen (positive screen=yes to ≥1 question, negative=no to both): 
  • “During the past month have you often been bothered by: 
    1. Feeling down, depressed or hopeless?
    2. Having little interest or pleasure in doing things?”
Primary care: Three cohort studies1-3 (total 1,893 patients), 23-37% of patients tested positive to screening, 5-18% actually had depression: 
  • Sensitivity: 96-97% (if no to both questions, depression ruled out ~96% of time).  
  • Specificity: 57-78% (if yes to ≥1 question, patients are depressed ~70% of time). 
Systematic reviews of different short depression screens4,5 or pertaining to only geriatric patients6 found similar results, but up to half of geriatric patients may test positive.7   No studies evaluating the effects of screening on patient outcomes in North American context were found.  Context: 
  • Since the 2-question screen is better at ruling-out than ruling-in depression, when a patient answers “yes” to ≥1 question, more formal evaluation (example Patient Health Questionnaire (PHQ)-9 or Geriatric Depression Scale) is needed to diagnose depression. 
    • Example: In adults, PHQ-9 score ³10 has a sensitivity and specificity of 88%:8,9  
      • PHQ-9: Range of scores 0-27, higher worse. 
  • Adding “Is this something with which you would like help?”2 or frequency of symptoms to the 2-question screen (example PHQ-2) may improve depression diagnosis.5,10,11 
    • Asking only 1-question not as accurate in identifying depression.4 
  • Guidelines differ on depression screening recommendations from screening all adults12 to not screening at all,13 to only those with a history of depression, chronic health problems, or post-partum women.14 
    • Effectiveness of screening depends on disease prevalence and impacts opportunity cost.  
  • People with chronic illnesses,15 substance abuse history, First Nations descent, and post-partum women have higher depression rates.14 
January 15, 2018

This content is certified for MainPro+ Credits, log in to access

Author(s):

  • Caitlin R Finley BHSc MSc
  • Michael Kolber BSc MD CCFP MSc

1. Arroll B, Khin N, Kerse N. BMJ. 2003; 327:1144-6.

2. Arroll B, Goodyear-Smith F, Kerse N, et al. BMJ. 2005; 331:884.

3. Whooley M, Alvins AL, Miranda J, et al. J Gen Internal Med. 1997; 12:439-45.

4. Mitchell AJ, Coyne JC. Br J Gen Pract. 2007; 57:144-51.

5. Mitchell AJ, Yadegarfar M, Gill J, et al. B J Psych Open. 2016; 2:127-38.

6. Tsoi KKF, Chan JYC, Hirai HW, et al. Br J Psychiatry. 2017; 210:255-60.

7. Bosanquet K, Mitchell N, Gabe R, et al. J Affect Disord. 2015; 182:39-43.

8. Kroenke K, Spitzer RL, Williams JBW. J Gen Intern Med. 2001; 16:606-13.

9. The Patient Health Questionnaire (PHQ-9) 1999 Pfizer. Available at: http://www.cqaimh.org/pdf/tool_phq9.pdf. Last Accessed: August 28, 2017.

10. Arroll B, Goodyear-Smith F, Crengle S, et al. Ann Fam Med. 2010; 8:348-53.

11. Kroenke K, Spitzer RL, Williams JBW. Med Care. 2003; 41:1284-92.

12. Siu AL, US Preventive Services Task Force. JAMA. 2016; 315:380-7.

13. Canadian Task Force on Preventive Health Care. CMAJ. 2013; 185:775-82.

14. National Collaborating Centre for Mental Health. NICE Clinical Guidelines, no. 90. British Psychological Society; 2009. Available at: https://www.nice.org.uk/guidance/cg90/chapter/Key-priorities-for-implementation. Last Accessed: August 28, 2017.

15. Katon WJ. Dialogues Clin Neurosci. 2011; 13:7-23.

Authors do not have any conflicts of interest to declare.