The Influence Of Ciprofloxacin-Dexamethasone Ear Drops On Perforation Closure Rates After Endoscopic Tympanoplasty

By June 7, 2019

Sharma, Sunil1,2; Cushing, Sharon1,2,3,4; Papsin, Blake1,2,3,4; James, Adrian1,2,4
1 Department of Otolaryngology, Head & Neck Surgery, Hospital for Sick Children, Toronto, ON Canada
2 Department of Otolaryngology, Head & Neck Surgery, University of Toronto, Toronto, ON Canada
3 Archie’s Cochlear Implant Laboratory, Hospital for Sick Children, Toronto, ON Canada
4 Institute of Medical Science, University of Toronto, Toronto, ON Canada

Topical ciprofloxacin and dexamethasone have both been shown to disrupt healing of tympanic membrane perforations in animal models. There have been no clinical studies evaluating the effect of ciprofloxacin-dexamethasone ear drops (CD) on success of tympanoplasty. We compare perforation closure rates in pediatric endoscopic tympanoplasty with and without use of post-operative CD.

Study Design: Retrospective comparative cohort study
Setting: Tertiary referral centre
Patients: 127 children, mean age 12.3 years (range 2.3 – 17.5 years) undergoing totally endoscopic pediatric tympanoplasty with porcine-derived collagen graft.
Intervention: Prescription of CD versus no ear drops in the immediate post-operative period.

Main outcome measure: Perforation closure rate 2 months after totally endoscopic tympanoplasty.

Post-operative CD were given to 54 (43%) ears and no drops given to the remainder. Overall, successful closure of tympanic membrane perforation was achieved in 109 (86%) of ears.  The closure rate was significantly lower in those ears given CD post-operatively than those not given CD (41/54 (76%) vs 68/73 (93%), p=0.009, X2 test). Number needed to treat analysis indicates prescription of CD to around 3 ears results in one more persistent perforation (NNTH = 2.9; 95%CI 1.8 – 15).

Our results suggest that prescription of drops containing ciprofloxacin and dexamethasone can have an adverse effect on success of perforation closure after tympanoplasty. Allocation to treatment in this retrospective study was non-randomized but was predominantly based on a change in practice: no other variables are known to have influenced this finding.