Assessing guideline-concordant prescribing for community-acquired pneumonia

Background Current reports of outpatient antimicrobial prescribing practices may overestimate guideline concordance since they address only drug selection. Appropriate stewardship should consider all prescribing criteria (i.e., dose, frequency, duration, and route of administration) to fully assess guideline concordance. Objective Using a community-acquired pneumonia (CAP) example, the aims of this pilot study were to estimate guideline concordance in adult patients 18 years or older when all prescribing criteria are considered, and provide recommendations to optimize treatment. Specific objectives were to determine which medications were most commonly prescribed for high-and low-risk patients, respectively, and determine if prescription parameters typically meet guideline recommendations. Methods This historical (retrospective) chart review at a large, non-emergent, outpatient academic practice included adult cases of CAP identified by ICD-9 codes, 481.x-486.x, 480.x and 487.x, diagnosed between July 1, 2014 and June 30, 2015. Patients were stratified into low- or high-risk categories based on presence of comorbidities and recent antibiotic use. Descriptive statistics were used to profile the sample and estimate aggregate guideline appropriateness, based on Infectious Disease Society of America/American Thoracic Society guidelines. Cases that were not prescribed an antibiotic at the index visit were excluded from assessment of concordance. Results Of the 101 total episodes identified, 49% were treated with an antibiotic. Of the 45 cases that met low-risk criteria, seven of the 24 treated cases (29%) received an appropriate antibiotic. When considering all prescription elements, all seven cases were congruent, for a composite concordance rate of 29%. Of the 56 cases that met high-risk criteria, 13 of the 25 treated cases (52%) received an appropriate antibiotic, although two cases were prescribed a suboptimal dose, and one case was prescribed a suboptimal duration, dropping composite concordance to 40%. Overall, prescribing was concordant in 17 of the 49 treated cases (35%). Conclusion Concordance with current guidelines in this local sample is suboptimal. In the low-risk group, when the correct medication was chosen, dose, duration, and frequency were appropriate. Consideration of dose and duration of treatment decreased the rate of concordant prescribing in the high-risk group.