To evaluate the effect of tobacco smoking on healing of corneal abrasions and keratitis in humans.
A retrospective chart review of patients diagnosed with corneal abrasions and corneal ulcers from 1990 to 2010 at the Dean McGee Eye Institute was performed. The primary outcome measured was time from initial examination to epithelial closure; tobacco smoking was the primary exposure variable. Comorbidity variables were also examined as potential confounding factors. The data were analyzed using Kaplan–Meier curves and Cox proportional hazards models.
Eighty-seven patients met the inclusion criteria for corneal abrasion and 52 for keratitis. The mean healing time was 4.8 days in nonsmokers and 5.9 days in smokers in the abrasion arm, and 15.5 days in nonsmokers and 39.4 days in smokers in the keratitis arm. After controlling for comorbidities, treatment, and demographic variables, the healing time in both arms was significantly delayed in smokers when compared with nonsmokers (P = 0.01 and P = 0.03, respectively). The comorbidities were also studied after adjusting for smoking. A statistically significant delay was seen with previous keratoplasty and steroid treatment for the abrasion arm. In the keratitis arm, neurotrophic corneas and fungal infections were associated with a delay in healing after correcting for smoking.
Epithelial healing is delayed by 1.1 days on average in smokers when compared with nonsmokers with corneal abrasions. Patients who had undergone a previous keratoplasty had delayed healing from corneal abrasions regardless of the smoking status. Epithelial healing is delayed by 23.9 days on average in patients with keratitis who report smoking. Neurotrophic corneas and fungal infections also had prolonged time to healing.
Departments of *Ophthalmology, Dean A. McGee Eye Institute; and
†Biostatistics and Epidemiology, University of Oklahoma, Oklahoma City, OK.
Reprints: Donald U. Stone, Department of Ophthalmology, Dean A. McGee Eye Institute, University of Oklahoma, 608 Stanton L. Young Boulevard, Oklahoma City, OK 73106 (e-mail: email@example.com).
The authors have no conflicts of interest to disclose.
Supported in part by an unrestricted grant from Research to Prevent Blindness to the Department of Ophthalmology, University of Oklahoma, and the Dean A McGee Eye Institute.
Received September 02, 2013
Accepted January 28, 2014