Abstract
Abstract: :
Purpose: To create a questionnaire to identify and evaluate the main factors determining the impact of tearing on one's daily life. Methods: The process of developing a questionnaire began with interviewing 10–15 patients presenting to an Oculoplastics Surgeon with tearing. They were asked to identify and prioritize their symptoms. The item list was refined through further consultation with the Oculoplastic Surgeon. An initial questionnaire was then drafted and administered to 10 patients. The questionnaire was further revised and then administered by one technician, trained to provide identical, specific instructions to 28 consecutive patients presenting with tearing. Each patient was asked to fill out the same questionnaire approximately one week later. In addition, a short questionnaire representing the objective signs in assessing tearing was constructed and filled out by the surgeon. The statistical analysis included using Cronbach's Alpha to measure the internal consistency both for the initial administration of the revised questionnaire (T1) and for the second administration one week later (T2). In order to assess test–retest reliability, intraclass correlation coefficient (ICC) was used. Finally, to assess the correlational validity between the patient's and surgeon's assessment, a Pearson Correlation Coefficient with a 2–tailed construct was used. Results:The internal consistency measured by Cronbach's Alpha for T1 was 0.8130 and for T2 was 0.8191. These values reflect adequate internal consistency. The test–retest reliability was measured for each item on the questionnaire. The average intraclass correlation coefficient (ICC) was 0.9830. Also, each individual item had an ICC >0.9. These results are consistent with excellent reproducibility. The Pearson Correlation comparing the patients and surgeon's assessment of tearing was 0.378 at T1 and 0.379 at T2. The level of significance was 0.052 at T1 and 0.051 at T2. These p–values were just short of statistical signicance but demonstrated a strong trend towards correlation between the patient's and surgeon's assessment of tearing. A slight increase in the sample size would achieve statistical significance. Conclusions: In assessing tearing, it is important to evaluate the extent to which the tearing interferes with a patient's quality of life. This questionnaire appears to be a useful tool in assessing tearing and has demonstrated adequate internal consistency, excellent reproducibility and a strong trend towards correlating to the surgeon's objective assessment. The next step is to determine if the questionnaire can detect clinical change.