COPD: a clinical prediction rule may help diagnose it

Clinical bottom line (level 2a)

  1. A clinical prediction rule may be able to rank patients for risk of COPD.
Badgett et al: Chest 1994; 106 (5): 1427-1431
Expires November 2002

The study

Setting:

342 patients (aged >40 yr, ?% male) who were current or former smokers, or had a self-reported diagnosis of chronic bronchitis, emphysema or asthma.

Excluded if
  • aged < 40



Independent blinded reference standard, applied in all patients from a non-consecutive inappropriate spectrum.
Reference standard:
  • spirometry: FEV1<80% predicted without restrictive lung disease, or FEV/FVC <70%.
Diagnostic test: Clinical examination and questionnaire history-taking. Used to create a clinical prediction rule with "modified recursive partitioning analysis" (form of logistic regression).
  • Clinical prediction rule: score one point for any of the following
    • smoked more than 30 pack-years
    • diminished breath sounds
    • peak flow < 350 l/min

The evidence

pre-test probability of COPD: 36%, (95% CI: 31% to 41%)

diagnostic test COPD no COPD LR
(95% CI)		 post-test probability
score 3 24 1 43
(5.9 to 320) 		96%
score 2 56 31 3.3
(2.2 to 4.8) 		64%
score 1 39 86 0.82
(0.60 to 1.1) 		31%
score 0 3 102 0.053
(0.017 to 0.16) 		3%
total 122 220

Comments

  1. Patients came from a primary care setting by notices posted in outpatient clinics, and perhaps some caution should be applied in generalizing results to hospital or emergency settings.
  2. This clinical prediction rule needs to be prospectively validated.

Citation

  1. Badgett RG, Tanaka DJ, Hunt DK, et al: the clinical evaluation for diagnosing obstructive airways disease in high-risk patients. Chest 1994; 106 (5): 1427-1431
Search Terms: ?
Contributor: Bob Phillips and Chris Ball, November 1999
Reviewer: Santiago Alvarez Montero

Clinical Question.
Patient suspected COPD
Intervention or Exposure clinical prediction rule
Outcome COPD