The impact of endoscopic cultures on care in rhinosinusitis.

OBJECTIVES

The objectives of this retrospective, observational study were primarily to analyze endoscopically guided culture results (EGCRs) from an office setting in patients with subclassifications of chronic rhinosinusitis (CRS) and to determine the frequency EGCR-altered patient management; and secondarily to determine whether purulent secretions are more likely to be positive for a bacterial pathogen.

METHODS

Consecutive symptomatic patients with nasal secretions not currently on antibiotics were cultured endoscopically. Patients were classified as chronic rhinosinusitis (CRS; n = 41) or acute exacerbation of chronic rhinosinusitis (AECRS; n = 27). Initial therapy was individualized and 32 of 68 (47%) had an antibiotic initiated before availability of the EGCR. Patients were reassessed 2 to 4 days later, and therapy was continued or altered based on clinical course and EGCR.

RESULTS

EGCRs were positive in 41 of 68 (60.2%) of all patients. Incidence of positive cultures was not statistically different between CRS (21 of 41 [51%]) and AECRS (20 of 27 [74%]) patients. Initial therapy was changed in 45 of 68 (66%) patients. In 35 of 68 (51.4%), the change was directed by EGCR; in 11 of 77 (14.2%), the change was unsubstantiated by EGCR, and in no cases was an antibiotic initiated inappropriate to EGCR. The frequency of pathogens in the subgroups of CRS and AECRS, respectively, were Staphylococcus aureus (9 of 41 [22%]; 9 of 27 [33%]), acute pathogens (6 of 41 [14.6%], 7 of 27 [25.9%]), and Pseudomonas aeruginosa (5 of 41 [12%], 5 of 27 [18.5%]) and were not statistically different between subgroups. Purulent cultures were not significantly more likely to be Gram stain-positive than nonpurulent cultures (12 of 34 [35%] compared with 6 of 34 [17.6%], respectively; P = .1). Purulent cultures were more likely to be culture-positive (25 of 34 [73.5%]) compared with nonpurulent secretions (17 of 34 [50%]; P < .05). AECRS was significantly more likely to be purulent (21 of 27 cases) compared with CRS (13 of 41 cases, P < .001).

CONCLUSION

EGCR directed a change in therapy in 35 of 68 (51.4%) patients with CRS. The pathogens isolated from CRS and AECRS were similar and predominately S. aureus, acute pathogens, and P. aeruginosa. Purulent secretions are more frequently bacteriologically positive than nonpurulent secretions and more common in AECRS than CRS.