COPD: exacerbations: antibiotics increased recovery and reduced time in hospital

Clinical bottom line (level 1b)

  1. Patients with exacerbations of COPD who took antibiotics compared with placebo were more likely to recover (NNT = 8 at 21 days) , and less likely to require admission to hospital or additional antibiotics (NNT = 11 at 21 days) .
  2. Patients on antibiotics spent less time in hospital (on average 1.5 days).
Anthonisen et al: Annals of Internal Medicine 1987; 106 (2): 196-204
Expires November 2003

The study

Double-blinded ?concealed randomised trial ?with intention-to-treat
Setting: university hospital, Canada

173 patients (aged mean 67, 80% male) with 362 acute exacerbations of chronic obstructive pulmonary disease (mean FEV1 34%, mean PEFR 230 l/min). Exacerbations were defined as increased dyspnoea, sputum volume or purulence.

Excluded if
  • if FEV 1 > 70% or > 70% FVC.
  • other disease serious enough to affect quality of life or clinical course (e.g. cancer, left ventricular failure or stroke)
  • other disease requiring antibiotic therapy (e.g. recurrent sinusitis or urinary tract infection)
  • considered by treating physician to be too sick
  • geographically inaccessible
  • aged < 35
  • clinical diagnosis of asthma, or FEV 1 increased to > 80% predicted following an inhaled bronchodilator


    • Control Group: (n = 180, 180 analysed): placebo
    Experimental Group: (n = 182, 182 analysed): amoxicillin 250 mg four time a day, or trimethoprim-sulfamethoxazole 160/800 mg twice daily, or doxycycline 200 mg initially followed by 100 mg daily for 10 days
    All patients received inhaled salbutamol and oral theophylline. Patients with serious symptoms and thought to be steroid responsive were maintained on 5-10 mg of prednisone, which was increased to 40 mg daily during exacerbations. Patients could take diuretics and home-oxygen if required.
    100% followed for 21 days Only 66% of patients were followed for the total duration of the study (3.5 years)

    The evidence

    Outcome Time to outcome CEREERRRR
    (95% CI)    		ARR
    (95% CI)    		NNT
    (95% CI)
    no resolution or worsening symptoms 21 days 81
    (45.0%)    		 58
    (31.9%)    		 29%
    (8% to 46%)    		 13.1%
    (3.20% to 23.1%)    		 8
    (4 to 31)
    hospitalised or required treatment with another antibiotic 21 days 34
    (18.9%)    		 18
    (9.89%)    		 48%
    (11% to 69%)    		 9.00%
    (1.82% to 16.2%)    		 11
    (6 to 55)
    treatment discontinued due to adverse effects 21 days 5
    (2.78%)    		 3
    (1.65%)    		 41%
    (-150% to 86%)    		 1.13%
    (-1.90% to 4.16%)    		 89
    (NNT = 24 to infinity;
    NNH = 53 to infinity)

    Outcome Control Group
    (SD)    		 Experimental Group
    (SD)    		 Mean Difference
    (95% CI)
    length of stay in hospital 15.5
    (6.1)     		14.1
    (6.3)     		1.4
    (0.12 to 2.7)

    Comments

    1. Since the antibiotics used were not allocated on a randomised basis, no comment about which antibiotic is better can be made.
    2. The emergence of new antibiotics may improve effectiveness, but the spread of antibiotic resistance may decrease the efficacy ... without obvious changes in clinical surveys, it is reasonable to assume these results are valid.

    Citation

    1. Anthonisen NR, Manfreda J, Warren CP, et al: antibiotic therapy in exacerbations of chronic obstructive pulmonary disease. Annals of Internal Medicine 1987; 106 (2): 196-204
    Search Terms: ?
    Contributor: Bob Phillips and Chris Ball, November 1999
    Reviewer: Alec Chessman

    Clinical Question.
    Patient acute exacerbation of COPD
    Intervention or Exposure antibiotics
    Comparison placebo
    Outcome recovery, time in hospital