Validity of surveys reporting myopia, astigmatism, and presbyopia

Validation of surveys is an important step in establishing the usefulness of questionnaires for gathering information in clinical studies. Often in large studies of refractive error, it is not possible to examine and refract all subjects and their relatives. Therefore, we used 3 methods to evaluate 112 consecutive clinic patients' abilities to report myopia through a survey: (1) The Lay Terms Method asks: "are you nearsighted?"; (2) The Direct Method asks: "are you myopic?"; and (3) The Indirect Method uses a series of questions about the use of eyeglasses and age at first dispensing. Patient responses to the survey before examination were compared to subjective refraction findings at the completion of the eye examination. The Lay Terms Method had the highest sensitivity (0.98) but poor specificity (0.48), and the Direct Method yielded the highest specificity (0.83) but poor sensitivity (0.54). The rate of nonresponse ("don't know") was high (25.9 and 46.4% for the Lay Term and Direct Methods, respectively). The Indirect Method gave both high sensitivity and specificity (0.76 and 0.74, respectively), and only an 8.9% don't know response rate. The survey method that reports the presence or absence of myopia with the best balance of sensitivity and specificity and fewest don't know responses is the Indirect Method, suggesting it is the most suitable of our three approaches for use in clinical studies.