Beyond Snellen Acuity: Title and subTitle BreakThe Assessment of Visual Function After Refractive Surgery

THE GOAL of refractive surgery is to improve unaided vision in ametropic patients without the aid of spectacles and contact lenses. Refraction for the prescription of these appliances has traditionally been based on high-contrast distance visual acuity using Snellen charts. It is under such conditions that residual spherocylindrical refractive error and visual acuity following refractive surgery are usually assessed, and from these measures, the patient's visual function is inferred. It is now generally accepted that high-contrast distance visual acuity and residual refractive error are indeed correlated with overall patient visual function and satisfaction following surgery¹ ;however, there are many refractive surgery patients with minimal residual spherocylindrical error and excellent uncorrected high-contrast distance visual acuity who are dissatisfied with their postoperative quality of vision. Refractive surgeons will be familiar with a variety of problems expressed by such patients, ranging from general, unspecific complaints, to specific phenomena, such as halo formation that arises consistently under mesopic conditions.

Given that the goal of refractive surgery is to improve unaided vision, understanding the nature of this discordance between simple high-contrast distance visual acuity and visual function is of seminal importance. A major problem, however, is that there are no established objective measures by which the necessarily subjective visual experiences of patients can be characterized and quantified. It is likely that a wide range of distinct symptoms are similarly described using undefined terms such as "glare," "blur," or "haze," but most studies of patient function and satisfaction following refractive surgery make little effort to elicit more specific descriptions that could be used to categorize outcome groups and thereby identify risk factors.

In an attempt to identify more sensitive measures of visual function following refractive surgery, psychophysical tests such as low-contrast acuity, low-luminance acuity, disability glare, contrast sensitivity, postoperative corneal opacification or "haze," corneal topographic irregularities, and intraocular light scatter have been studied. Until recently, perhaps the greatest attention has been paid to measures of contrast sensitivity and low-contrast visual acuity. Numerous reports have documented a decrease in some measures of contrast sensitivity in the first few months following excimer laser surface photorefractive keratectomy (PRK)^{2 - 5} and laser-assisted in situ keratomileusis (LASIK).^{6 - 7} However, the persistence of reduction in the measure of contrast sensitivity employed by a given study group has been variable, with some reporting an increase in function over time to near-preoperative levels,^{2 ,4 ,6} and others reporting a sustained reduction for up to a year following surgery.^{3 ,5 ,7} Although some studies have found low-contrast visual acuity to be highly correlated with high-contrast visual acuity, and therefore to provide little additional information,³ others^{8 - 12} have suggested that low-contrast visual acuity might be a particularly sensitive measure of visual function following refractive surgery. In such cases, a decrease in low-contrast visual acuity immediately following surgery has been demonstrated that persists to some degree for as long as 12 months following surgery in some groups,^{8 - 11} but returns to near-normal levels in others.¹²

Unfortunately, poor performance on these tests of contrast sensitivity has not been shown to correlate with patients' subjective visual function. In fact, few studies have even attempted to establish a relationship between performance on psychometric testing, the subjective experience, and functional satisfaction of patients with postoperative changes in contrast sensitivity measures or low-contrast visual acuity.^{2 - 3 ,8 - 12} In a meticulous study of active-duty US military personnel who underwent excimer laser PRK, Schallhorn et al⁴ performed a battery of functional and diagnostic tests, including glare disability, near-contrast acuity with glare, and intraocular light scatter. They also administered a questionnaire designed to elicit patient complaints of overall dissatisfaction with quality of vision, glare or halos around lights, or difficulty with night driving. These investigators found no correlation between the objective measures studied and the subjective responses to quality of vision. Since decentration of the treatment zone has been associated with disturbing visual symptoms¹³ and decreased low-contrast visual acuity,⁹ by extension, it has been suggested that untoward visual symptoms might be marked by reduced low-contrast visual acuity,⁹ but this has not been proven convincingly.

The causes for a decrease in contrast sensitivity function following excimer laser treatment are not clearly understood and probably represent a complex interplay of factors. Forward light scatter is to some extent related to the backscatter seen as haze in corneas following PRK and is expected to degrade the retinal image and affect contrast sensitivity function.¹² Although the degree of corneal haze has been correlated with a reduction in contrast sensitivity,¹⁴ the course of clinically apparent haze does not match that of changes in contrast sensitivity. Contrast sensitivity function is noted to decrease immediately following surgery, but haze commonly appears many weeks to months after PRK. Furthermore, decreased contrast sensitivity is commonly seen in eyes that have had LASIK and that presumably do not demonstrate haze.^{6 - 7} Alternative suggestions are that changes in the normal physiologic cell structure and extracellular matrix, including intracellular vacuole formation, proteoglycan content, and irregular spacing of collagen fibers, could all contribute to atypical corneal optical quality and decreased visual function.³ Furthermore, optical aberrations due to surface irregularity (caused by epithelial irregularity following PRK, or flap folds following LASIK), inclusion of the excimer ablation zone edge within the entrance pupil, or alteration of the normal prolate corneal shape, might all degrade the retinal image, leading to decreased contrast sensitivity function and visual perturbations.⁷

Of the aberrations produced by ablation profile in excimer laser treatments, it is probably the effect of the edge of the optical zone that has been best characterized. O'Brart et al^{15 - 16} have demonstrated that smaller treatment zone diameters are associated with an increased incidence of visually disturbing halo formations. As the pupil dilates, light is admitted to the retina not only through the treated cornea but through the untreated cornea that retains its original power, and that therefore creates a defocused image on the retina. This defocused image is superimposed on the sharper image created by the treated cornea and produces a halo of large diameter. Multizone treatments can also result in significant image degradation if a relatively small optical zone is combined with a broad transition zone created within the boundaries of the entrance pupil producing a band of variable defocus.¹⁷ Recognition of this phenomenon and an understanding of its origins have allowed surgeons to reduce the incidence of disturbing halo formation by increasing the ablation zone diameter of the optical zone.¹⁶

Unfortunately, many other geometric aberrations are not so well characterized, do not produce such distinctive symptoms, and are not so strongly correlated with subjective patient experience. Based on an analysis of corneal videokeratographs following excimer PRK, Martinez et al¹⁸ have demonstrated significant increases in coma-like and spherical-like aberrations for both 3- and 7-mm pupils. However, it cannot be assumed that these measures are directly related to visual experience, since qualitative topographic patterns have been shown to be poorly predictive of subjective glare or halo.¹⁹ Likewise, Oshika et al²⁰ found that both PRK and LASIK significantly increased the wavefront aberrations as calculated from corneal topography measurements for 3- and 7-mm pupils. However, although they found that the aberrations seemed to be significantly larger in the LASIK group (which was treated with a smaller ablation-zone diameter), a subsequent study comparing the 2 groups failed to demonstrate a difference in patient satisfaction with regard to quality of uncorrected vision.²¹ While Holladay et al⁷ have demonstrated an increase in corneal asphericity values following LASIK, and shown a corresponding small decrease in contrast threshold,⁷ the relevance of these measures to patients' visual experience is not clear.

Most recently, great interest has been generated by the prospect of comprehensively evaluating optical aberrations of the eye as a system and using this analysis to evaluate and refine laser refractive techniques. However, this approach does not escape the dilemma of relating objective measures of geometric optical quality to the subjective visual experience of patients. Using a Tscherning-type aberroscope, Seiler et al¹¹ have demonstrated that after PRK, the wavefront error of the eyes studied increased on average by a factor of 17.65, and that there was a significant correlation with best-corrected visual acuity, low-contrast visual acuity, and glare. Consistent with prior studies of corneal aberration following excimer ablation, the increase in overall ocular aberration was related to virtual pupil size. This study demonstrated the usefulness of wavefront techniques in correlating corneal anatomy with objective functional findings, but given current methodological limitations, it is difficult to move to the next stage of assessing the effect of wavefront error on subjective vision.

Ideally, such comprehensive optical analyses should be correlated with specific subjective visual experiences or with an overall measure of subjective quality of vision. The first step in this process is to gain a better understanding of the patient's visual experience. One approach is to establish a reliable and valid questionnaire that is sensitive to clinically relevant elements of the subjective function. Standard performance-based clinical measures of vision, such as high-contrast visual acuity, glare disability, and contrast sensitivity, have long been recognized as inadequate in describing visual disability associated with cataract. This led to the development of instruments such as the VF-14,²² its abbreviated form the VF-7,²³ the Visual Disability Assessment,²⁴ and the Activities of Daily Vision Scale.²⁵ A similar instrument, the Glaucoma Symptom Scale, has also been created to examine the functional effects of visual disability associated with glaucoma.²⁶ Although instruments designed specifically for the evaluation of cataract-related impairment have been shown to provide a valid measure of impairment in other conditions such as corneal disease²⁷ and glaucoma,^{28 - 29} it is expected that the specificity of these instruments might result in their failure to capture the effects of the broad range of eye disorders that might affect quality of vision.

The National Eye Institute Visual Function Questionnaire (NEI-VFQ) was therefore constructed as a broadly applicable instrument, explicitly designed to capture the effects of a wider range of eye diseases.³⁰ However, in its development, this instrument did not include refractive error or perceived disability following refractive surgery. Unfortunately, the effect of visual symptoms unique to refractive error and its modes of correction(eg, halo formation under mesopic conditions or fluctuations in vision, and issues of convenience and utility that have been shown to be significant factors for refractive surgery patients³¹ ) are unlikely to be captured by nonspecific instruments. It is these considerations that have motivated the National Eye Institute to sponsor the development of the Refractive Error Correction Questionnaire, which was specifically designed to assess the subjective visual and functional effects of refractive error and its correction, both by appliances and surgery. The design process, patterned after the NEI-VFQ, is based on a review of current instruments. It also relies heavily on focus groups to determine the importance of elements identified in previous visual function instruments, and to identify elements of unique importance to patients with refractive error or patients who have undergone surgical refractive error correction. Using similar methods and with similar goals, Vitale et al³² have constructed a 42-item questionnaire (the Refractive Status and Vision Profile) designed to measure the visual, functional, and psychological consequences of refractive error. These authors reported that certain subscales, as well as the overall test score, were more strongly correlated with patient satisfaction and overall assessment of vision than were visual acuity or refractive error measures.

Such instruments should be useful not only in providing measures of visual function that can be used to characterize a patient's quality of vision, but they may also establish a basis of comparison for subjective visual quality associated with various appliances and refractive surgery techniques. Furthermore, by engaging in a detailed and specific exploration of patients' visual experiences, clues to the relationship between performance on various psychophysical tests and day-to-day visual function (as described by elements of these instruments) might hopefully be unearthed. Ultimately, what we must work toward is a convergence of sophisticated diagnostic tests that enable a comprehensive objective analysis of optical aberration; performance-based psychophysical tests; and valid, reliable, and sensitive instruments for the description and assessment of subjective patient visual experience. Armed with these tools, we can expect to better evaluate current refractive surgery outcomes, compare existing technologies and techniques, and in the future, design procedures that optimize visual function and patient satisfaction.