Pretreatment workup
A pertinent historical account of medical history,
review of ophthalmologic history, medication
allergies, and current medical history were
obtained. A slit-lamp examination, including an
evaluation of the eyelids, sclera, conjunctiva,
cornea, iris, and lens, was performed. Additional
pertinent examination findings were noted.
Visual acuity for each eye was determined and
fundoscopy was performed. Slit-lamp examination
was normal for all except three patients. Ttyo
patients were disqualified from the study due to
a diagnosis of keratoconjunctivitis sicca, and one
patient due to the presence of dystrichiasis. One
additional patient was disqualified based on med-
ication history. Females of childbearing age
underwent a urine pregnancy test to confirm non-
pregnancy (Savon Osco, One Step Pregnancy Test).
Statistical analysis
The primary endpoints of the trial were the
alleviation of visual discomfort following
administration of the study products and the
incidence of adverse events associated with
study product use. Visual discomfort was
assessed through a subjective ordinal scoring
system of:
0 = very comfortable,
1 = slightly uncomfortable eye symptoms,
2 = uncomfortable eye symptoms, and
3 = very uncomfortable eye symptoms,
with outcome variables dichotomized by combining the latter three categories. A generalized
estimating equation (GEE) modeling method was
used to analyze data. In the GEE models, logitlink function and binomial distribution of the
outcome variable were assumed. Statistical
analysis was conducted with the use of SAS statistical software. Statistical significance was presumed at a P value I 0.05. For the 50 patients (25 per group), we assumed a compound symmetry correlation of 0.24 to detect a smallest meaningful difference of 5% in 20 measurements and a two-tailed alpha level of 0.05. This achieved a
power of 90%.