Community-acquired pneumonia: some antibiotic regimens saved more lives in elderly patients.

Clinical bottom line (level 2b)

  1. Elderly patients with community-acquired pneumonia who received second or third generation cephalosporins with a macrolide or a fluoroquinilone compared with other antibiotics were less likely to die.
  2. Elderly patients who received beta-lactam agents with a macrolide or an aminoglycoside with other microbial agents were more likely to die.
Gleason et al: Archives of Internal Medicine 1999; 159: 2562-2572
Expires March 2003

The study

Retrospective cohort study with objective outcomes, adjusted for confounding factors, not validated in an independent set of patients.

Setting: acute hospitals participating in Medicare progams, USA

12945 patients (aged mean 79, 51% female) admitted to hospital with pneumonia (confirmed on chest X-ray within first 48 hours). 500 patients were randomly selected from each state. 68% were classified as high-risk - 10% had aspiration pneumonia.

Excluded if
  • mortality at 30 days could not be confirmed
  • no antibiotics given within first 48 hours or missing data
  • unknown residence
  • HIV
  • history of organ transplantation or had been exposed to immunotherapy or immunosuppressive therapy within previous 2 months
  • transferred from another acute care facility
  • died or been discharged on date of admission
  • aged < 65
  • experienced acute care hospitalization within previous 10 days



    • Factors studied:
  • anitmicrobial therapy before hospitalisation, pneumonia severity classification, arrival from a long-term care facility, time when antibiotics were started, ICU admission, change in antimicrobial therapy
  • beta-lactam/beta-lactamase inhibitors plus macrolide
  • aminoglycoside plus any other antimicrobial agent(s)
  • fluoroquinolones only
  • second-generation cephalosporin plus macrolide
  • non-pseudomonal third-generation cephalosporin plus macrolide




    • Cox proportional hazards models and logistic regression analysis were performed to adjust for confounding factors.

    100% followed for 30 days
    Outcomes studied:
  • death

    • The evidence

    outcome time to outcome number of patients/total number %
    (95% CI)    		 NNF
    (95% CI)
    death 30 days 1980/12945 15.3%
    (14.7% to 15.9%)    		 7
    (6 to 7)

    prognostic factor for
    death     		time to outcome control rate (%) adjusted OR
    (95% CI)    		 NNF+
    (95% CI)
    beta-lactam/beta-lactamase inhibitors plus macrolide 30 days 1980/12945
    (15.3%)     		1.77
    (1.28 to 2.46)     		11
    (6 to 29)
    aminoglycoside plus any other antimicrobial agent(s) 30 days 1980/12945
    (15.3%)     		1.21
    (1.02 to 1.43)     		38
    (19 to 390)
    fluoroquinolones only 30 days 1980/12945
    (15.3%)     		0.64
    (0.43 to 0.94)     		-20
    (-130 to -12)
    second-generation cephalosporin plus macrolide 30 days 1980/12945
    (15.2%)     		0.71
    (0.52 to 0.96)     		-25
    (-190 to -15)
    non-pseudomonal third-generation cephalosporin plus macrolide 30 days 1980/12945
    (15.3%)     		0.74
    (0.60 to 0.92)     		-28
    (-95 to -18)

    Comments

    1. This study suggests that the initial antibiotic regimen matters, and this warrants a prospective randomised study.
    2. The reported variation in prescribed antibiotic regimens may reflect patient-related factors (prescription bias) not controlled in the factorial analysis.

    Citation

    1. Gleason PP, Meehan TP, Fine JM, et al: associations between initial antimicrobial therapy and medical outcomes for hospitalized elderly patients with pneumonia. Archives of Internal Medicine 1999; 159: 2562-2572
    Search Terms: hand-search
    Contributor: Chris Ball and Clare Wotton, November 1999
    Reviewer: Chris van Weel

    Clinical Question.
    Patient elderly with community-acquired pneumonia
    Intervention or Exposure antibiotics
    Outcome mortality