Discontinuation of Antibiotic Therapy Based on Clinical Stability Criteria May Reduce Treatment Duration in Adults Hospitalized for Community-Acquired Pneumonia

EBM Focus - Volume 11, Issue 32

Read the full EBM Focus and earn CME credit.

Reference - JAMA Intern Med 2016 early online (level 2 [mid-level] evidence)

  • Shortening the duration of antibiotic therapy for some conditions by discontinuing it once the patient is clinically stable may help reduce the emergence of drug-resistant pathogens.
  • Antibiotic therapy with duration based on clinical stability criteria significantly reduced the duration of antibiotic use with similar clinical success rates compared to antibiotic duration based on physician’s discretion in a randomized trial of 312 adults hospitalized for community-acquired pneumonia (CAP).
  • Most patients were prescribed quinolones in this trial, potentially limiting the generalizability of the findings.

The emergence of drug-resistant pathogens is a growing problem, even among common community-acquired infections (WHO 2014 report, CDC 2013 report). Shortening the duration of antibiotic therapy by discontinuing it once the patient is clinically stable has been suggested as a strategy to address this problem (Clin Infect Dis 2008 Feb 15;46(4):491, J Infect Dis 2007 Jun 15;195(12):1818, J Antimicrob Chemother 2012 Nov;67(11):2570). However, evidence assessing the safety and efficacy of this strategy is mainly limited to observational studies. To determine the safety and efficacy of shorter duration therapy in adults with CAP, a trial was conducted in which 312 adults in Spain hospitalized for CAP were randomized on day 5 to one of two strategies for antibiotic therapy discontinuation. The strategy for the intervention group used clinical stability criteria based on recommendations from Infectious Diseases Society of America and American Thoracic Society (IDSA/ATS) to discontinue antibiotic therapy after 5 days if the patient had a temperature below 37.8 degrees C (100 degrees F) for 48 hours and has ≤ 1 CAP-related sign of clinical instability. The control group had antibiotics discontinued at the treating physician’s discretion. The primary outcomes were clinical success, defined as improved symptoms and signs without additional antibiotic therapy, and symptom severity assessed by questionnaire at day 10.

The use of clinical stability criteria significantly decreased the antibiotic therapy duration compared to using the physician’s discretion (median 5 days vs. 10 days, p = 0.001). Despite the shorter duration of treatment, there was no statistical difference in most outcomes between the 2 groups. Clinical success was achieved within 10 days of hospital admission in 56.3% with length of therapy based on clinical stability criteria versus 48.6% with therapy discontinued at physician’s discretion (not significant). At 30 days, the clinical success rate was 91.9% vs. 88.6% and the recurrence rate was 2.4% and 4%, respectively. The 2 groups were similar in other outcomes such as symptom severity at day 10, rates of in-hospital and 30-day mortality, and in-hospital complications. The clinical stability group fared better for 30-day readmission rate (1.8% vs. 6.7% [p = 0.03, NNT 21]), and clinical success rates were higher in a subgroup of patients with higher Pneumonia Severity Index scores.

These results support the IDSA/ATS recommendation to discontinue antibiotic therapy after 5 days in adults hospitalized for CAP if they meet defined clinical stability criteria, but some limitations of this trial temper the support. In particular, steps were not taken to blind patients or physicians to treatment allocation. This introduces a high potential for bias, especially considering that discontinuation was based on physician’s discretion in one group. However, despite the lack of blinding, using clinical stability criteria did shorten median duration of antibiotic therapy, and so some conclusions regarding safety and efficacy can be made. Also, generalizability is limited in two ways: 1) quinolones were used in 79% of patients, so the effect of a shorter duration with other types of antibiotics remains unclear; and 2) patients were excluded if they were admitted to the intensive care unit before randomization, or had other complications such as being immunocompromised or had pneumonia caused by an uncommon pathogen. Finally, although this trial was described as a noninferiority trial, only conventional superiority analyses were reported. Statistical analyses explicitly testing whether or not clinical stability criteria were inferior to physician’s discretion by more than a prespecified margin were not reported. This concern is partly mitigated by the high 30-day clinical success rates in both groups. Overall, in adults hospitalized with CAP, discontinuing antibiotic therapy after 5 days in clinically stable patients appears to have similar outcomes as longer course antibiotic therapy, and potentially reduces the risk of developing drug-resistant pathogens as well as treatment-related adverse events.

For more information, see the Antibiotics for adult inpatients with community-acquired pneumonia topic in DynaMed.