NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Cataract Management Guideline Panel. Cataract in Adults: Management of Functional Impairment. Rockville (MD): Agency for Health Care Policy and Research (AHCPR); 1993 Feb. (AHCPR Clinical Practice Guidelines, No. 4.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Cataract in Adults: Management of Functional Impairment

Cataract in Adults: Management of Functional Impairment.

Show details

4Guideline: Preoperative Ophthalmic Testing in Patients with Cataract

The decision to perform cataract surgery is generally made after judging the effect of the cataract on the patient's visual function and/or the cataract's contribution to functional impairment, after assessing the patient's visual needs, and after a thorough consideration of the potential risks associated with surgery.

Background

The decision to perform cataract surgery is generally made after a judgment of the effect of the cataract on the patient's visual function and assessment of the patient's visual needs. Visual complaints are compared with visual acuity measurements and the results of the eye examination to confirm the presence of a cataract and rule out or otherwise gauge the potential impact of other causes of visual loss. Snellen visual acuity is the most universally accepted method of assessing visual function, but in a number of patients, it may not accurately or adequately reflect the level of impairment caused by cataracts.

Newer techniques for evaluating visual dysfunction caused by cataract include contrast sensitivity and glare testing. Like Snellen acuity, contrast sensitivity testing is an index, or measure, of visual function. However, contrast sensitivity testing attempts to determine the ability to perceive objects of varying contrast. Glare testing is used clinically in an effort to assess the functional impairment of a patient who is disabled by glare. Certain types of cataracts, such as posterior subcapsular (PSC) cataract, may cause disabling glare in daylight conditions or from oncoming headlights during night driving despite relatively good Snellen visual acuity measured in a darkened room.

In addition, potential vision tests and specular photographic microscopy have been developed to determine the likelihood of a successful outcome after cataract surgery. Potential vision tests attempt to predict the visual acuity that might be expected as a result of cataract surgery. Specular photographic microscopy provides information on endothelial cell counts and morphology, which are thought to be useful in identifying patients in whom the cornea is unlikely to withstand surgery, i.e., patients likely to develop corneal edema.

Assessment of Preoperative Tests

The approach taken in assessing the value of all of these tests for the purpose of development of this guideline is that typically applied to diagnostic tests in any area of medicine. A number of criteria must be satisfied for the test to be considered appropriate for use.

First, the test must be valid, that is, both acceptably sensitive and specific. Sensitivity is defined as the proportion of patients a test correctly identifies as having the disorder or condition of interest. Specificity is the proportion of patients a test correctly calls "negative."

The assessment of a test's sensitivity and specificity depends on a comparison of the test's results and "the truth" with regard to whether a disorder or condition (e.g., disabling glare) is actually present. Unfortunately, a true "gold standard" is often unavailable or not used properly in research studies of this type. The gold standard must be judged for its appropriateness and quality based on prior evidence and clinical experience. If the gold standard is inappropriate or inadequate, the estimated sensitivity and specificity of the test being evaluated will be inaccurate.

Second, the test must be repeatable (reliable or precise). Variation in the observer (interobserver or intraobserver) and variation related to the test itself (lighting in room, size of the pupil) can affect the test's reliability.

The yield of a test is defined as the amount of previously unrecognized disease diagnosed and brought to treatment as a result of performing the test. Yield is dependent on both sensitivity and the prevalence of the disorder of interest. Prevalence is defined as the proportion of the population with the disorder within a specified time period. The higher the prevalence, the higher is the yield from the test.

The concept of the predictive value of a test unites the concepts of sensitivity, specificity, and prevalence. A positive predictive value is defined as the proportion of patients testing positive who actually have the condition or disorder of interest. If the prevalence of a disorder is low, even a very specific test will produce a large number of false positives because most people being tested, including many of those found to be positive, have no disease.

Before being accepted as clinically useful tools, diagnostic procedures should be evaluated in well-designed studies. Results from examinations on patients diagnosed through the use of the "new" test should be compared with results on the same patients diagnosed by an already established technique. When this type of comparison is being made, the order in which the new and old test are given should be randomly assigned to avoid learning curve effects on the data collected. In addition, data from the new test or from the gold standard test should be masked as to the values obtained for the comparison test performed on the same patient.

Literature Review Methods

In developing these guidelines, the methods used by the literature review teams for evaluating contrast sensitivity, glare testing, and potential acuity were nearly identical. The panel developed topic-specific questions. The National Library of Medicine designed and ran literature searches. The printouts generated by the three searches were grouped together and reviewed as a whole by the same literature review team. Part of the reason for this approach was that many articles address more than one topic.

Additional articles of potential relevance, not identified by the National Library of Medicine search, were gleaned from reference lists of relevant articles and by notification from other investigators and clinicians.

Many more articles on these tests were published than are included here. The articles not included did not meet the general selection criteria outlined in Appendix A in the Guideline Report or the specific criteria described for evaluation of each test (Appendices D-G in the Guideline Report).

The literature review proceeded as described in Appendices A and D-F in the Guideline Report. Once the final selection of articles for all three topics was made (65 articles in all), reviews were conducted separately for each topic. Some reports were included in reviews for more than one topic. The methods used for evaluating specular microscopy were different and are described in full in Appendices A and G in the Guideline Report.

This section of the guideline addresses the use of preoperative testing in clinical practice for managing the adult patient with functional impairment potentially due to cataract. The tests and their intended uses are described, and the methods and findings of a formal literature review and consensus reached by the panel are detailed, together with recommendations for future research.

The guideline makes the distinction between visual impairment and functional impairment. Visual impairment refers to patient status measured in terms of the results of visual testing (i.e., Snellen visual acuity). Functional impairment refers to patient status in terms of the ability of the patient to perform everyday activities (see Organization of the Clinical Practice Guideline in Chapter 1), as reported by the patient or health care provider.

The purpose of this literature review was to identify and review those articles that contained information on contrast sensitivity, glare, and potential vision testing as tests of visual function for patients with functional impairment due to cataract. (For a full description of the literature reviews and their results, see Appendices D-F in the Guideline Report.) The questions for specular photographic microscopy are different and are detailed later in the chapter.

Contrast Sensitivity Testing

Test Description

Contrast sensitivity is a measure of the degree of contrast required to detect a test object. Snellen visual acuity measures the eye's ability to perceive high-contrast letters but does not adequately assess its ability to see low-contrast patterns. Contrast sensitivity testing attempts to determine the eye's ability to detect subtle variations of shading by presenting letters, figures, or sine wave gratings that are varied in contrast, luminance, and spatial frequency. Contrast sensitivity is determined by the lowest detectable contrast in which a subject is able to identify targets.

Numerous devices and methods have been developed to test contrast sensitivity. The simplest and least expensive methods employ reading cards or eye charts similar to those used in Snellen testing. More elaborate (and more expensive) methods involve the use of electronic devices (such as oscilloscopes) to generate sine wave gratings.

Literature Review Results

The results of the literature review are given here and summarized in Attachment 2 at the end of this document.

One hundred and nineteen papers on contrast sensitivity testing were reviewed by the contrast sensitivity content literature review teams. Ninety-eight of these were excluded because they did not meet the inclusion/ exclusion criteria set forth by the guideline panel. A total of 21 papers were reviewed for content and methodologic quality. Of these, only 10 provided data relating to the questions specified by the panel.

For the 21 articles reviewed for content, there was, in general, excellent agreement among content reviewers for items on the content evaluation form. In a few cases where there was disagreement, one of the reviewers changed his or her mind on reevaluating the study in question, or a third content reviewer was asked to review the article and cast the deciding vote. There was very poor agreement among content reviewers about the type of study (e.g., clinical trials), and these questions were therefore referred to the methodology reviewers with no attempt made to achieve consensus among content reviewers.

The results of the methodologic review of the 10 articles relevant to the questions posed by the panel are summarized in Attachment 2.

Very few of the 10 relevant articles met any of the methodologic criteria pertinent to studies that assess a preoperative test. Only two studies set out to assess contrast sensitivity testing as a test to document patients' complaints of visual impairment that may be useful in a preoperative evaluation (Elliott, Hurst, and Weatherill, 1990; Koch, 1989). Eight studies described contrast sensitivity testing in patients with different ocular pathologies or different types and severity of cataracts (Elliott, Gilchrist, and Whitaker, 1989; Hess and Woo, 1978; Howes, Caelli, and Mitchell, 1982; Lempert, Hopcroft, and Lempert, 1987; Maudgal, Stout, and vanBalen, 1988; Morris, Klett, Gieser et al., 1991; Singh, Cooper, Alder et al., 1981; Skalka, 1981a). These studies either did not use a gold standard for comparison of contrast sensitivity results or used one that was inappropriate for answering the questions posed in this literature review.

Question 1: Is there a correlation between the results of the test and the patient's level of functional impairment?

One study provided information regarding a correlation between contrast sensitivity test results and functional impairment due to cataract. This article (Elliott, Hurst, and Weatherill, 1990) described a case series of 33 patients with cataract in at least one eye and correlated contrast sensitivity with patient-reported functional disability. Contrast sensitivity was measured for both eyes (binocular) using a Pelli Robson letter chart (both sides). Functional disability was measured using a patient questionnaire of 20 questions. The questions were grouped into mutually exclusive categories pertaining to mobility, near vision, and discrimination. The correlations between binocular contrast sensitivity and mobility, near vision, and discrimination were -0.65, -0.60, and -0.49, respectively. (The negative correlation coefficient implies that functional disability decreases as contrast sensitivity score increases.) For the worse eye, the correlations between contrast sensitivity and mobility and near vision were -0.50 and -0.35, respectively; data were not provided for discrimination because the correlation was "not significant." For the better eye, the correlations between contrast sensitivity and mobility, near vision, and discrimination were -0.50, -0.67, and -0.43, respectively. Thus, the correlation between contrast sensitivity and functional disability was highest when binocular contrast sensitivity was used. The magnitude of several correlations was higher for contrast sensitivity than for visual acuity.

One study (Koch, 1989) provided information regarding the correlation between contrast sensitivity test results and patients' complaints of glare. The proportion of patients with false positive and false negative results for contrast sensitivity was neither better nor worse than for Snellen acuity testing, when those results were compared with the patient's complaint of glare as the gold standard. In addition, the correlation between contrast sensitivity test results and glare complaints was no higher than correlations between Snellen acuity and glare complaints.

Three studies measured correlations between contrast sensitivity and Snellen visual acuity (Elliott, Gilchrist, and Whitaker, 1989; Howes, Caelli, and Mitchell, 1982; Lempert, Hopcroft, and Lempert, 1987). These studies showed a correlation between contrast sensitivity and Snellen visual acuity, but there was no standardization of methods of contrast sensitivity or Snellen acuity measurement across studies. One study found a moderate to high correlation (r=0.44 to 1.0) between contrast sensitivity and Snellen acuity (Elliott, Gilchrist, and Whitaker, 1989) for cortical, nuclear, and PSC cataracts. Contrast sensitivity decreased as the severity of cataract increased. This latter result was also obtained by a second study (Howes, Caelli, and Mitchell, 1982). In a third study, as contrast sensitivity decreased, mean Snellen acuity decreased in eyes with PSC cataract. For normal eyes, the trend was the same, but the decrease in Snellen values was less pronounced (Lempert, Hopcroft, and Lempert, 1987).

A fourth study (Hess and Woo, 1978) involving only 10 subjects used a ratio of cataract to normal eye acuity as a comparison. This ratio is not clinically meaningful, and its correlation with the contrast sensitivity measure was poor (r=0.1). Furthermore, its use would be limited to patients with uniocular cataract. Although all of these studies are based on the premise that functional visual impairment involves more than poor acuity, none of them assesses patients' overall function or ability to perform everyday activities.

One study (Maudgal, Stout, and vanBalen, 1988) provided information on the sensitivity and specificity of contrast sensitivity testing for detection of cataracts. The gold standard for comparison was the clinical examination. This information is not applicable, therefore, to the setting where it is already known that the patient has a cataract and where the purpose of the test is to provide information preoperatively regarding the potential value of cataract surgery.

In two studies of contrast sensitivity (Arden gratings), no visual acuity data were provided (Singh, Cooper, Alder et al., 1981; Skalka, 1981a). One study reported that Arden scores correlated better than Snellen visual acuity with subjective patient complaints (Skalka, 1981a). In another study, eyes with cataract had a mean score that was significantly different (worse score) from normal eyes (14.5+/-3 vs. 12.25+/-1.44, respectively, for "plate 6" [ Singh, Cooper, Alder et al., 1981]). After pupil dilation, there was an improvement in Arden scores for eyes with cataract but not for normal eyes.

Question 2: Is there information on the type or degree of functional impairment detected by the test that is not detected by history and routine physical examination alone?

Question 3: Is there evidence to support the value of the test in detecting patients who are not suitable for surgery?

Question 4: Is there evidence to identify a relationship between the use of the test and the timing of surgery?

Question 5: Is there evidence to identify a relationship between the use of the test and the volume of surgery performed annually by the ophthalmologist ordering the test?

Question 6: Is there information on costs and/or cost-benefit issues?

No articles were found addressing these questions.

Question 7: What is the relationship between preoperative test results and postoperative test results in patients in whom no intraoperative or postoperative complication occurred?

No article contained information on the relationship between preoperative and postoperative contrast sensitivity results in at least 10 patients in whom no complications occurred.

One article (Morris, Klett, Gieser et al., 1991) contained information on this relationship for a group of patients with and without complications. In this study, postoperative contrast sensitivity was accurately predicted in 10 of 15 patients. Four of the five inaccurate predictions were in patients with visual acuity worse than 20/200. Data were available for only 15 of 20 (75 percent) patients originally enrolled in the study.

Conclusions of the Panel

Based on the findings from the literature review, the panel reached the following conclusions.

  • In patients with cataract, the relationship between contrast sensitivity test results and functional impairment has not been adequately examined, although one study (Elliott, Hurst, and Weatherill, 1990) has demonstrated a correlation between binocular and better eye contrast sensitivity and patient-reported functional impairment due to vision. There is also published evidence in the low-vision literature, as summarized by the American Academy of Ophthalmology (1990), that suggests that contrast sensitivity test results are correlated with the degree of functional impairment in patients without cataract. These data and their relevance to the patient with cataract will be reviewed in a subsequent update of the guideline.
  • Contrast sensitivity testing does not differentiate between visual loss due to cataract and visual loss from other causes.
  • There is no commonly used contrast sensitivity test for which preoperative and postoperative results have been compared except the interferometer. In this case, only one published article exists (Morris, Klett, Gieser et al., 1991). It indicated that postoperative contrast sensitivity was accurately predicted in 10 of 15 cases. Inaccurate predictions occurred in patients with visual acuity worse than 20/200.
  • In the absence of other studies comparing preoperative with postoperative contrast sensitivity results and in the absence of preoperative vs. postoperative assessments of functional impairment, the panel is unable to assess the clinical usefulness of information from contrast sensitivity testing, above and beyond that obtained from routine history and ocular examination, in determining whether a patient would benefit from cataract surgery.

Recommendations for Research

Further research should be undertaken to determine the clinical utility of contrast sensitivity testing as a source of valid information with respect to functional impairment due to cataract above and beyond that obtained through history and routine ocular examination.

Glare Testing

Test Description

Glare is an impairment in visual function caused by the presence of a source of light located in the visual field. Patients with a cataract often complain of glare because cataracts and other opacities in the ocular media cause light entering the eye to be scattered, thereby producing glare. Glare testing attempts to reproduce the symptom of glare and to quantify the degree of visual impairment it causes. A typical patient with disabling glare complains of poor vision in well-lighted situations but may have nearly normal acuity when measured in a darkened room. Visual acuity may fall precipitously when retested in ambient light. Glare disability is generally determined by comparing a subject's visual acuity with and without the presence of a bright-light source directed at the patient's eye.

A variety of methods and devices for testing glare are described in the published literature. Some of the devices used for this purpose are simple and inexpensive, such as a penlight or room lights. More elaborate instruments costing hundreds of dollars, such as the brightness acuity tester (BAT), or thousands of dollars, such as the Miller-Nadler glare tester, are also available.

In the United States, glare disability testing is widely used to evaluate patients with cataract, in part because in many States, peer review organizations request the results of this test at the time of precertification for cataract surgery, particularly in patients with good visual acuity.

Literature Review Results

The results of the literature review are given here and summarized in Attachment 3.

Question 1: Is there a correlation between the results of the test and the patient's level of functional impairment?

One study examined whether there was a correlation between the results of glare testing and the patient's level of functional disability. This article (Elliott, Hurst, and Weatherill, 1990) described a case series of 33 patients with cataract in at least one eye and correlated glare scores with patient-reported functional disability. Glare disability was measured using a Mentor BAT, which measured reduction in visual acuity due to glare (GDVA) and reduction in letter contrast sensitivity due to glare (GDCS). The BAT can take only monocular measurements, so all testing was done on the eye with the worse visual acuity. Functional disability was measured with a patient questionnaire of 20 questions. The questions were grouped into mutually exclusive categories pertaining to mobility, near vision, or discrimination. The correlation between the worse eye's GDVA and mobility was "not significant," although no correlation coefficient was provided. The correlations between the worse eye's GDVA and near vision and discrimination were 0.61 and 0.51, respectively. Only the correlation between the worse eye's GDCS and discrimination was provided (r=0.45) because the corresponding values for mobility and near vision were "not significant." The correlation between patient-reported functional disability related to near vision and discrimination was higher for worse eye glare disability than for visual acuity.

One study (Koch, 1989) examined the correlation between the results of glare testing and patients' glare complaints. The correlations between the results of the Baylor visual function tester and glare complaints were 0.36 and 0.61 at low and high levels of glare, respectively. The correlation between the Stereo Optical Glare Tester and glare complaints was 0.68. The article did not specify the impact of glare on the patients' everyday activities and did not indicate how the glare complaints were measured.

Six studies examined the association between the results of glare testing and outdoor visual acuity (Hirsch, Nadler, and Miller, 1984a; Holladay, Prager, Trujillo et al., 1987; Neumann, McCarty, Locke et al., 1988; Neumann, McCarty, Steedle et al., 1988a and 1988b; Prager, Urso, Holladay et al., 1989). The BAT was used in three of these studies (Holladay, Prager, Trujillo et al., 1987; Neumann, McCarty, Locke et al., 1988; Prager, Urso, Holladay et al., 1989). In one study (Holladay, Prager, Trujillo et al., 1987), normals showed no decrease in visual acuity caused by either the BAT or outdoor testing. For patients with cataract, the correlation between outdoor visual acuity and BAT was r=0.84. The BAT predicted outdoor acuity to within one line in 73 percent of the subjects in a second study (Neumann, McCarty, Locke et al., 1988) and overpredicted glare disability in 81 percent of subjects at the high setting and 42 percent at the medium setting in a third study (Prager, Urso, Holladay et al., 1989).

In one article, the correlation between the results of glare testing (using author's device) and outdoor acuity was better for those facing the sun than for those not facing the sun (Hirsch, Nadler, and Miller, 1984a), both for subjects with and without cataracts. In another, Snellen visual acuity was decreased by two lines in 70 percent of subjects when they faced the sun (Neumann, McCarty, Steedle et al., 1988a). A third study found that the Miller-Nadler glare tester predicted outdoor acuity within one line for 47 percent of all eyes in the study.

Two other studies relate to this question. One study (Maltzman, Horan, and Rengel, 1988) compared visual acuity with and without penlight glare in 114 cataract patients. All patients with nonglare visual acuities of 20/40 or worse also saw 20/40 or worse with penlight glare. Forty-five percent of patients with nonglare visual acuities of 20/40 or better (sic) had an acuity of 20/40 or worse with penlight glare. Another study (Hard, Abrahamsson, and Sjostrand, 1990) found decreased visual acuity and contrast sensitivity in the presence of glare. The decreased contrast sensitivity was independent of visual acuity.

Several studies examined the correlation between the results of glare testing and visual acuity by the type of glare test used. One study found that the correlation varied depending on the glare test used (Neumann, McCarty, Locke et al., 1988). Outdoor visual acuity was predicted to within one line in 73 percent of the eyes for BAT, 69 percent for True Vision Analyzer (TVA), 56 percent for VisTech, 47 percent for Miller-Nadler, and 15 percent for Eye Con tests, in which all tested eyes were diagnosed with cataract. The same results were reported in a second article (Neumann, McCarty, Steedle et al., 1988b), perhaps on the same patients, for the Miller-Nadler test alone. (It predicted outdoor visual acuity to within one line in 47 percent of the eyes.) The authors also provided information regarding the cataract type and the ability to predict outdoor acuity (see below); however, a third study (Prager, Urso, Holladay et al., 1989) found that in eyes with cataract, accuracy to within one line occurred in 36 percent of eyes for Miller-Nadler, 17 percent for high luminance BAT, and 45 percent for medium luminance BAT. The Miller-Nadler test tended to underestimate glare disability (as measured by outdoor visual acuity) in 62 percent of eyes, and the BAT overpredicted in 81 percent of eyes at high settings and 42 percent at medium settings. This study also found significant interobserver variability in glare test results.

In terms of the correlation between glare test and visual acuity for specific types of cataract, one study showed high correlations for cortical (r=1.0) and nuclear (r=0.77) cataract but not for PSC (r=0.12) (Elliott, Gilchrist, and Whitaker, 1989). A second study demonstrated that the Miller-Nadler test predicted outdoor acuity to within one line in 50 percent of eyes with nuclear cataract, 40 percent with nuclear plus PSC, and 52 percent with other cataracts (Neumann, McCarty, Steedle et al., 1988b). A third study found a weak correlation between glare score and visual acuity in subjects with PSC cataracts or normal eyes (Abrahamsson and Sjostrand, 1986). Thus, glare scores have the lowest correlation with visual acuity in patients with PSC cataract. Therefore, the test may be of greatest value in these patients.

Question 2: Is there information on the type or degree of functional impairment detected by the test that is not detected by history and routine physical examination alone?

Question 3: Is there evidence to support the value of the test in detecting cases that are not suitable for surgery?

Question 4: Is there evidence to identify a relationship between the use of the test and the timing of surgery?

Question 5: Is there evidence to identify a relationship between the use of the test and the volume of surgery performed annually by the ophthalmologist ordering the test?

Question 6: Is there information on costs and/or cost-benefit issues?

No articles were found providing useful information on these questions.

Question 7: What is the relationship between preoperative test results and postoperative test results in patients in whom no intraoperative or postoperative complication occurred?

Three studies addressed the relationship between preoperative and postoperative glare test results (Levin, 1989; ; Masket, 1989; Weiss, 1990), and all three showed an improvement between preoperative and postoperative scores related to cataract surgery. In one study, mean decimal acuity in conditions of glare improved from 0.06 preoperatively to 0.22 postoperatively (Weiss, 1990). Mean Miller-Nadler glare scores improved from 55.2 percent preoperatively to 7.6 percent postoperatively in a second study (Masket, 1989) and from 15 percent preoperatively to 10 percent postoperatively, using TVA, in a third study (Levin, 1989).

No studies examined a correlation between glare scores and functional impairment. Although these studies demonstrated that subjects' scores on the glare test improved after surgery, the studies do not provide evidence that the patients' complaints of glare or their functional impairment also improved. Thus, no study demonstrated that patients' complaints of impairment due to glare were accurately detected by the glare tester and that both the complaints and glare scores improved after surgery.

Two studies examined the test-retest reliability of glare testing. One showed an excellent correlation between first and subsequent glare tests when the test was performed on patients with cataract (r=0.95) but a less good correlation when patients were aphakic (r=0.54) or had normal vision (r=0.06) (Hirsch, Nadler, and Miller, 1984b). A second study (van Lith and Hekkert-Wiebenga, 1983) examined test-retest reliability for letter contrast sensitivity with and without glare in patients with cataract and found the reliability to be very good (r=0.83 without glare; r=0.92 with glare).

One major deficit noted throughout the published literature on glare testing is the lack of established standard equipment, conditions of administration, and scoring of the test. For example, there is no consensus regarding the most appropriate type of glare source, background illumination, target configuration, or criterion for abnormality justifying surgery.

Conclusions of the Panel

Based on the findings from the literature review, the panel reached the following conclusions.

  • A number of devices are available for testing glare. It is difficult to compare them, however, because standards have not been established for glare source, background illumination, target configuration, and other factors.
  • There is some evidence in the published literature that glare test results are reproducible in patients with cataract. More data are needed to evaluate the reproducibility of glare test results in patients with different types of cataract and different types of symptoms potentially attributable to glare.
  • Glare disability symptoms are not specific to cataract. As with most tests of vision, glare disability testing is not able to differentiate between visual loss due to cataract and visual loss due to other causes.
  • Several studies have adequately shown that abnormal glare test results improve after cataract surgery. Whether symptoms potentially attributable to glare and glare-related functional disability decrease postoperatively in such patients, however, has not been reported in the published literature.
  • The correlation between glare test results and functional impairment in patients with cataract was examined in one study (Elliott, Hurst, and Weatherill, 1990). In that study, there was a correlation between glare disability, as measured by a reduction in visual acuity in the worse eye, and functional disability related to near vision and discrimination. The magnitude of these correlations was greater than the magnitude of the correlations between visual acuity and these functional disabilities. One cannot determine from the data that are reported, however, whether the differences in the magnitude of these correlations are either statistically or clinically significant.
  • The correlation between glare test results and visual acuity is high in patients with cortical and nuclear cataracts and low in patients with PSC cataracts. This suggests that, if glare testing provides useful information regarding functional disability due to glare, it is most likely to be of value in patients with PSC cataracts.
  • There are three ways that glare testing might be used in the context of a patient with functional impairment potentially due to cataract.

Application 1: As a stand-alone objective measure of functional impairment due to cataract.

The panel agrees that there is no evidence in the published literature that the test can be used as a stand-alone objective measure of functional impairment due to cataract justifying cataract surgery.

Application 2: As a source of valid information concerning functional impairment vs. visual impairment due to cataract, over and above information provided by the history and physical examination.

The panel agrees that there is no evidence from the published literature to support the use of the test as a source of valid information concerning functional impairment due to cataract, over and above that provided by the history and ocular examination.

Application 3: As a source of valid information on visual impairment that will assist the ophthalmologist in advising the patient regarding the suitability or deferral of surgery when used in conjunction with the history, ocular examination, and measurement of visual acuity.

The panel believes that glare testing may be useful in the clinical evaluation of patients who complain of glare, or who have symptoms potentially attributable to glare, and who have a cataract (particularly a PSC cataract), yet measure good Snellen acuity in the testing circumstances of an office. However, the published literature has not yet provided data that demonstrate that postoperative improvements in glare test results are associated with improvement in glare symptoms or with symptoms potentially attributable to glare.

Given the current state of knowledge, it is reasonable to perform a glare test as part of the evaluation of patients who complain of glare, or who have symptoms potentially attributable to glare, and who have a cataract (particularly a PSC cataract) and good Snellen acuity. However, glare testing should not be required by utilization review or quality assessment organizations, such as peer review organizations, as objective documentation of visual disability sufficient to justify the potential benefit of surgery.

The panel believes that the glare test is not useful in patients with cataract who do not complain of symptoms potentially attributable to glare, regardless of their visual acuity.

Recommendations for Research

Glare testing may be of value in corroborating glare symptoms in some patients. The panel recommends that more research be performed on glare testing to clarify its role in the evaluation of the patient with functional impairment potentially due to cataract.

Potential Vision Testing

Test Description

Tests of potential vision have been developed in an effort to determine whether individuals with obviously impaired vision have a potential to see well after cataract surgery (i.e., the significant cause of their visual impairment is cataract rather than other pathology). Two basic types of tests, subjective and objective, are available. Subjective tests require the individual being tested to respond to questions asked about visual stimuli that are presented, and objective tests do not.

Subjective tests include the suprathreshold pinhole device, Maddox rod test, laser interferometer, Guyton-Minkowski potential acuity meter, and a number of devices that use blue field entoptic images. Some of these are simple and inexpensive, such as entoptic imagery and the Maddox rod test, whereas others are more expensive because of the cost of the instruments needed to perform them, such as the laser interferometer and potential acuity meter (PAM). Objective tests are electrophysiologic tests in which the response to visual stimuli is measured electronically. Those evaluated in the literature include electroretinography and visually evoked potential.

Objective tests are generally more expensive than subjective ones because of the cost of electrophysiologic equipment and the expertise necessary to administer and interpret them. Such tests are generally not available in the private clinician's office. Instead, patients are usually referred to teaching institutions to obtain them.

Literature Review Results

The results of the literature review are given here and summarized in Attachment 4.

Question 1. Is there a correlation between the results of the test and the patient's level of functiional impairment?

One study (Graney, Applegate, Miller et at., 1990) provided information regarding the association between the results of potential vision testing and the patient's level of functional impairment. This study provided measures of the impact of visual impairment on the ability of patients to perform their everyday activities. It measured everyday activities and mental status (Short Portable Mental Status Questionnaire) in 36 patients with cataracts and retinal disease. The impact of cataract surgery on 10 of the 14 activities could not be assessed because 85-100 percent of the subjects were already independent in these activities. In the remaining four activities, significant improvement occurred in traveling independently beyond walking distance and in independent shopping. Statistically significant changes did not occur in mental status after cataracvt surgery.

Of the remaining articles in this review, 29 provided some information on the relationship between the result of potential vision testing and postoperative visual acuity. These will be discussed under Questions 3 and 7.

Question 2: Is there information on the type or degree of functional impairment detected by the test that is not detected by history and routine physical examination alone?

Three articles (Graney, Applegate, Miller et al., 1988 and 1990; Miller, Graney, Elam et al., 1988) provided information potentially pertinent to the type and degree of functional impairment detected by otential vision testing. Two of the studies (Graney, Applegate, Miller et al., 1988 and 1990) reported predictive models based on clinical and functional data. The third study (Miller, Graney, Elam et al., 1988) compared results from laser interferometry, PAM testing, and an ophthalmologist's clinical judgment using the results of these two tests. The ophthalmologist predicted postoperative acuity exactly in 15 percent of cases, the interferometer did so in 15 percent, and the PAM in 11 percent. Predictions to within two lines of actual postoperative acuity were 57 percent, 27 percent, and 26 percent, respectively. The study did not examine results from an ophthalmologist's clinical judgment alone.

Question 3: Is there evidence to support the value of the test in detecting cases that are not suitable for surgery?

The literature identified seven studies that suggest that subjective tests of potential vision, particularly the PAM and the laser interferometer, can provide an accurate indication of potential acuity in eyes with "relatively clear media" and a Snellen acuity better than 20/200 and no other eye disease (Christenbury and McPherson, 1985; Faulkner, 1983; Graney, Applegate, Miller et al., 1988; Ing, 1986; Minkowski, Palese, and Guyton, 1983; Spurny, Zaldivar, Belcher et al., 1986; Tabbut and Lindstrom, 1986). "Relatively clear media" was not well defined in the articles reviewed. The panel took the position that, for the purposes of drawing conclusions and recommendations, clear media would be defined as media through which the person performing the ocular examination could visualize the posterior pole. The studies applying this definition confirm that, in such eyes, history and clinical evaluation alone are as good as tests of potential vision in predicting visual outcome of cataract surgery.

In eyes with relatively clear media and clinical evidence of macular disease, it was not possible to determine from the available published literature whether any of the tests of potential vision were of value, although a number of studies concluded specifically that they were not (Bernth-Petersen and Naeser, 1982; Faulkner, 1983; Goldstein, Jamara, Hecht et al., 1988; Tabbut and Lindstrom, 1986). Most of these comparisons involved only small numbers of patients and did not have enough power to detect true differences. These studies failed to clearly define macular disease, however. Attachment 5 summarizes the results of testing in the presence of macular disease. (See Appendix F in the Guideline Report for a detailed discussion of the literature reviewed.)

In eyes with grossly opaque media, subjective tests of potential acuity are not accurate. In such eyes, objective tests (electroretinography or flash-evoked cortical potentials) can determine whether light signals are being received by the retina or brain, but these modalities do not measure fine visual function and their role is therefore limited.

Testing with Retinometers, Visometers, and Interferometers (see Attachment 5): The specificity of the potential vision test (i.e., the ability of the test to identify correctly patients who did have a poor surgical outcome) was presented, or could be calculated, in 11 of the 15 articles that assessed retinometers, visometers, and interferometers (Bernth-Petersen and Naeser, 1982; Cohen, 1976; Datiles, Edwards, Kaiser-Kupfer et al., 1987; Dubey, Masani, and Shroff, 1983; Enoch, Bedell, and Kaufman, 1979; Goldstein, Jamara, Hecht et al., 1988; Graney, Applegate, Miller et al., 1988; Grignolo, Moscone, Sobrero et al., 1988; Halliday and Ross, 1983; Spurny, Zaldivar, Belcher et al., 1986; Tabbut and Lindstrom, 1986). The specificities from these studies ranged from 0.20 to 1.0, with little consistency across studies. Thus 20-100 percent of the patients with bad outcomes were correctly identified by these tests. The wide variability in these results was most likely due to variation in the study designs, the specific instrument used, procedures used for the preoperative testing and the postoperative evaluation of visual acuity, severity and type of cataract, and presence of comorbid ocular pathology.

The sensitivity of the tests (the ability of the test to correctly identify patients having "good" surgical outcomes) reported in these 11 articles also had a wide range. From 22 to 90 percent of the time, good outcomes were correctly identified before surgery by these tests. Nine of these studies reported a sensitivity of 0.70-0.90.

The remaining 4 of these 15 articles also reported a wide variety of results. In one study (Faulkner, 1983), the interferometer predicted visual acuity to within one line 88 percent of the time in patients with immature cataracts and no other ocular pathology, but it predicted visual acuity to be better than was actually achieved postoperatively (by greater than or equal to two lines) 96 percent of the time in patients with eyes having other pathology. The interferometer predicted postoperative acuity to within one line in 27 percent of cases (Miller, Graney, Elam et al., 1988) and to within two lines in 88 percent of cases with normal maculas and 66 percent of cases with abnormal maculas (Bryant, 1985). A fourth study (Angra and Pal, 1990) of patients undergoing cataract surgery indicated that the laser interferometer was "accurate" 35 percent of the time, falsely predicted a "poor outcome" 50 percent of the time, and falsely predicted a "good outcome" 15 percent of the time.

Potential Acuity Meters (see Attachment 5): The specificity of potential vision testing was presented, or could be calculated, in three of the nine articles assessing the PAM (Graney, Applegate, Miller et al., 1988; Minkowski, Palese, and Guyton, 1983; Spurny, Zaldivar, Belcher et al., 1986); specificity ranged from 0.71 to 0.92. Thus, the PAM correctly identified preoperatively 71-92 percent of cases that had poor acuity postoperatively. The PAM correctly identified preoperatively 33-94 percent of patients who had good postoperative visual acuity (Datiles, Edwards, Kaiser-Kupfer et al., 1987; Graney, Applegate, Miller et al., 1988; Minkowski, Palese, and Guyton, 1983; Spurny, Zaldivar, Belcher et al., 1986).

In addition, one study (Miller, Graney, Elam et al., 1988) reported that the PAM predicted visual acuity to within one line in 26 percent of cases; a second study (Severin and Severin, 1988a) reported that the PAM predicted visual acuity precisely in 26 percent of cases and to within two lines in 78 percent; and a third (Carpel and Henderson, 1986) reported predictions to within three lines in 86 percent of cases. One article (Ing, 1986) reported an accuracy of prediction to within two lines in 31 of 33 eyes (96 percent) tested.

Testing with Blue Field Entoptoscopy (BFE) (see Attachment 5): The specificity of potential vision testing was presented, or could be calculated, for three of the five articles assessing BFE (Grignolo, Moscone, Sobrero et al., 1988; Miris and Missotten, 1982; Sinclair, Loebl, and Riva, 1979). The BFE identified preoperatively 47-100 percent (specificity range 0.47-1.0) of those patients who ultimately had poor outcomes. Although the study by Lischwe and Ide (1988) provided information regarding both sensitivity and specificity of the BFE, the study included 8 patients with capsular opacities, excluded 15 patients with preoperative visual acuities worse than 20/200, and miscalculated the false positive and negative rates. Given these factors, the study's results could not be interpreted with any confidence. Murphy (1983)also provided some information, but sufficient reliable data were not provided to allow calculation of specificity.

The sensitivity of the BFE ranged from 0.5 to 0.95 for the three studies with data that could be evaluated. Thus, the BFE correctly identified 50-95 percent of patients who eventually had good surgical outcomes.

Clinical Indices and Judgment: Two studies examined clinical indices as predictors of good and bad surgical outcomes. One study of 36 patients with retinal disease (Graney, Applegate, Miller et al., 1990) found that the combination of three clinical variables (age, preoperative visual acuity, and retinal disease only in the periphery) tended to be a good predictor of surgical outcome (sensitivity = 0.83 and specificity = 0.85). Of the 13 patients who had successful surgeries (postoperative acuity 20/40 or better), 83 percent were predicted to have success by the index. Of the 23 patients who had unsuccessful surgeries, 85 percent were predicted not to have success by the index.

A second study (Graney, Applegate, Miller et al., 1988) reported the development of a predictive model on one sample of patients (a training set) and its validation on a second sample (a testing set). The clinical index, which included the patients' age, Snellen acuity, number of prescription medications, and frequency of reading the newspaper, had a sensitivity of 0.91 and specificity of 0.26. Thus, it was a good predictor of good but not poor surgical outcomes. On the other hand, both the interferometer and PAM used in the study were poor predictors of good outcomes (Se =0.43 and 0.50, respectively) but good predictors of poor outcomes (Sp = 0.90 and 0.81, respectively).

A third study (Miller, Graney, Elam et al., 1988) compared the results of laser interferometry, the PAM, and an ophthalmologist's clinical judgment using data from both tests. Prediction of postoperative visual acuity to within one line occurred in 27 percent of cases with the interferometer alone, 26 percent with the PAM alone, and 57 percent for the ophthalmologist using all the data. This study was not designed to examine the ophthalmologist's clinical judgment alone. No data were provided on the distribution of the preoperative visual acuities of the patients.

Visually Evoked Potentials, Visually Evoked Responses, and Electroretinogram:Six of the 39 articles evaluated electro-ophthalmologic testing (Cruz and Adachi-Usami, 1989; Fukuhara, Oozato, Nojima et al., 1983; Skalka, 1981b; van Lith and Hekkert-Wiebenga, 1983; Vrijland and van Lith, 1983; Weinstein, 1977). One study did not provide postoperative data (van Lith and Hekkert-Wiebenga, 1983). A second (Skalka, 1981b) presented only negative predictive values, which are greatly dependent on the prevalence of the outcome of interest. Of 12 patients with preoperative acuities of 20/400 or better and a negative test result, 5 (42 percent) actually had postoperative acuities worse than 20/40. The negative predictive value dropped to 28 percent for the 25 patients who had preoperative acuities of count fingers or worse. The other four studies reported sensitivities that ranged between 0.1 and 1.0 and specificities that ranged between 0.5 and 1.0 (see Attachment 5).

Other Potential Vision Tests: Nine studies used a variety of tests to assess potential vision. Studies included: Baraldi, Enoch, and Raphael (1986) (gap test); Dubey, Masani, and Shroff (1983) (Maddox rod and two-point discrimination); Elliott and Hurst (1989), which was rejected because it did not meet all of the inclusion criteria (opacity lensometer 701); Enoch, Williams, Essock et al. (1985a) (vernier acuity, gap acuity, perimetry); Enoch, Williams, Essock et al. (1985b) (gap and perimetry test); Essock, Williams, Enoch et al. (1984) (2 dot vernier acuity); Miris and Missotten (1982) (color perception); Sinclair, Loebl, and Riva (1979) (two-light discrimination, color perception, Purkinje vascular entoptic phenomenon); and Whitaker and Deady (1989) (displacement threshold acuity). Most reported on 10-15 patients or on a subgroup of a larger sample.

One study (Elliott and Hurst, 1989), which was rejected because it did not meet all of the inclusion criteria, did not use postoperative outcomes and found no correlation (r=}0.09) between the opacity lensometer and Logmar acuity for 45 cataract patients. The Maddox rod in one study (Dubey, Masani, and Shroff, 1983) of 100 cataract eyes had a sensitivity of 1.0 and specificity of 0.95 (see Attachment 5). Although the potential vision tests in this group were intriguing, there were not enough data from well-conducted studies or clinical trials to come to firm conclusions concerning their use in clinical practice.

Question 4: Is there evidence to identify a relationship between the use of the test and the timing of surgery?

Question 5: Is there evidence to identify a relationship between the use of the test and the volume of surgery performed annually by the ophthalmologist ordering the test?

Question 6: Is there information on costs and/or cost-benefit issues?

No articles were found addressing these questions.

Question 7: What is the relationship between preoperative test results and postoperative test results in patients in whom no intraoperative or postoperative complication occurred?

Eight articles (Adams and Shock, 1986; Bryant, 1985; Dubey, Masani, and Shroff, 1983; Faulkner, 1983; Goldstein, Jamara, Hecht et al., 1988; Lischwe and Ide, 1988; Murphy, 1983; Sinclair, Loebl, and Rivera, 1979) summarized under question 3 provide information on the relationship between preoperative potential vision measures and the postoperative visual acuity obtained in patients in whom no intraoperative or postoperative complication occurred, although no articles looked at this issue directly.

Conclusions of the Panel

Based on the findings from the literature review and panel discussion, the panel reached the following conclusions.

  • Tests of potential vision can predict postoperative outcome reasonably accurately in eyes with a preoperative vision of better than 20/200 and media that are clear enough to confirm by examination that the posterior pole is normal. Evidence is lacking in the published literature that potential vision measurement increases the accuracy of the predicted outcome beyond that based on history and ocular examination alone.
  • In eyes with opaque media and vision of 20/200 or worse, none of the tests of potential vision provides an accurate estimate of visual outcome after uncomplicated cataract surgery, although electrophysiologic tests may be useful in determining whether light signals are being received by the retina or brain (i.e., whether there is gross retinal disorganization).
  • For eyes in which the posterior pole can be visualized, no conclusions can be drawn from the literature review about the influence of specific types of macular disease on the accuracy and value of tests of potential vision beyond what can be determined by ophthalmoscopic examination.
  • There were no studies identified that address the issue of the predictive value of a collection or battery of potential vision tests in relation to postoperative visual acuity.
  • There are three possible ways in which tests of potential vision could be used in evaluating a patient with functional impairment potentially due to cataract.

Application 1: As a stand-alone objective measure of expected visual outcome after surgery for cataract.

The panel concluded that tests of potential vision do not function as a stand-alone measure of predicted visual outcome after surgery for cataract, since they appear to be accurate only when:

  • Visual acuity is 20/200 or better.
  • The media are clear.
  • The posterior pole is normal on clinical examination.

Application 2: As a source of valid information concerning the visual outcome vs. function after cataract surgery, over and above that provided by the history and ocular examination.

The panel concluded that tests of potential vision do not provide information over and above that provided by the history and ocular examination to justify their use as a basis for decisionmaking about the visual outcome after cataract surgery.

Application 3: As a source of valid information on visual outcome after cataract surgery that will assist the ophthalmologist in advising the patient about the potential value of cataract surgery when used in conjunction with the history and ocular examination.

The panel concluded that the available evidence in the published literature suggests that tests of potential vision do not assist the ophthalmologist in predicting the outcome of surgery when used in conjunction with history and ocular examination.

Tests of potential vision add nothing to the history and ocular examination of a patient with cataract when the posterior pole can be visualized and is normal. Tests of potential vision are not accurate in predicting visual outcome after surgery in a patient with cataract when the posterior pole cannot be visualized. It is impossible to determine from the literature whether any of the tests of potential vision is of value when the posterior pole can be visualized and macular disease is present.

Based on this literature review and clinical judgment, the panel concluded that a test of potential vision should not be included as a preoperative test in the routine management of patients with functional impairment due to cataract. The panel recognizes that there may be a subset of patients (patients with specific types of macular degeneration and clear media) in whom such a test might be of value, but this has yet to be shown.

Many ophthalmologists find potential acuity measurements reassuring in situations in which they are uncertain whether other diseases, particularly mild to moderate atrophic macular degeneration, might limit postoperative vision rehabilitation. However, there are no data that have substantiated the value of potential vision testing in this circumstance, even though such testing might be helpful.

Recommendations for Research

The panel recommends that properly designed studies be performed to assess the value of preoperative tests for potential vision in eyes with cataract and macular disease. These should include further investigation of the methods of potential vision testing that are already in existence. Reports of the results of evaluations of potential vision tests in patients with cataract should be more specific with regard to definition of media clarity, the presence of and type of macular disease, and the time at which patients are classified as to the presence of macular disease (i.e., preoperatively or postoperatively).

The panel also recommends that future studies address the usefulness of individual tests as well as a battery or collection of potential vision tests and that additional methods be developed to predict visual outcome after cataract surgery.

Specular Photographic Microscopy

Test Description

Specular photographic microscopy is a research tool used to measure and record endothelial cell counts and to evaluate endothelial cell morphology. When endothelial cells are destroyed by disease or surgery, the remaining cells normally do not divide. Instead, they enlarge and spread out to cover the posterior corneal surface, thus decreasing the cell density (cell count) in a given area. Corneas with extremely low endothelial cell densities may decompensate, swell, and become cloudy with time. Although a precise lower limit cannot be defined, there is a variable lower limit of cell density in different patients below which the corneal endothelial cell layer cannot function adequately to maintain corneal clarity.

A cataract extraction procedure will usually result in some decrease in the endothelial cell density. Therefore, corneas with very low cell densities before cataract extraction are thought to be at greater risk for postoperative corneal decompensation and permanent corneal edema than are corneas with "normal" cell densities. Thus, there is a conceptual rationale for advocating the use of specular microscopy preoperatively to ascertain the cell density of the corneal endothelium and thereby assess a patient's risk of postoperative corneal decompensation.

The specular microscope provides a magnified view of a small area of corneal endothelial cells that is obtained by specular (mirrorlike) reflection. The specular image can be seen on routine slit lamp examination, but when a slit lamp is used, the magnification is less than when a specular microscope is used. Consequently, it may be difficult to discern individual cells with a slit lamp.

A specular microscope may be either a "contact" type (an applanating lens touches the cornea) or "noncontact" type (the microscope does not contact the eye but merely magnifies the image normally seen through the slit lamp). The contact specular microscope provides more magnification and is easier to focus, but it requires the use of a topical anesthetic on the eye. Both types of microscopes can be used by a trained technician.

Images of the endothelium seen with specular microscopy can be recorded on video tape or photographic film. The endothelial cell density and configuration can then be estimated from the video or photographic image. The cell density can also be estimated during examination with the slit lamp, but this technique is more difficult to perform and less accurate.

Quantitative analyses of the morphology of endothelial cells also can be performed by digitizing the image obtained with the specular microscope. A variety of morphologic measures can be defined, such as the mean area per cell or the variation in cell area.

Literature Review

Methods.

The purpose of the literature review was to identify and review those articles that contained information on specular photographic microscopy of the corneal endothelium in relation to functional impairment due to cataract. (For a full description of the literature review, see Appendix G in the Guideline Report.)

Results.

The results of the literature review are given here and summarized in Attachment 6

Question 1: Is there evidence to support or refute an association between the results of preoperative specular photographic examination of the endothelium and the outcome of cataract surgery?

None of the eight studies (Azen, Hurt, Steel et al., 1983; Bates and Cheng, 1988; Bates, Cheng, and Hiorns, 1986; Bourne and Kaufman, 1976; Irvine, Kratz, and O'Donnell, 1978; Kraff, Sanders, and Lieberman, 1980; Rao, Aquavella, Goldberg et al., 1984; Stur, 1988) reported data enabling the reader to calculate the sensitivity or specificity of preoperative specular microscopy results in predicting the occurrence of postoperative corneal decompensation. The proportion of patients with postoperative corneal decompensation in the eight studies varied from 0 to 34 percent.

  • Two studies found that preoperative cell count was not associated with postoperative corneal decompensation (Bates, Cheng, and Hiorns, 1986; Rao, Aquavella, Goldberg et al., 1984).
  • One study reported that no patients in the study had postoperative corneal decompensation and that all had "normal" preoperative cell counts (Irvine, Kratz, and O'Donnell, 1978).
  • One study reported that no patients had postoperative corneal decompensation, although three patients had "abnormal" preoperative cell counts. These three patients had corneal problems, including penetrating keratoplasty for Fuchs' dystrophy and keratoconus, and history of blunt trauma (Bourne and Kaufman, 1976).
  • One study of 161 patients found that 2 of 3 patients with postoperative corneal decompensation had "abnormal" preoperative cell counts (Stur, 1988).
  • One study had five patients with postoperative corneal decompensation. Of these, preoperative specular data were available on three, and all three had abnormal specular microscopy; one had uncountable cells and two had low cell counts (Kraff, Sanders, and Lieberman, 1980).
  • One study had a single patient with corneal decompensation but provided no preoperative data for that patient (Azen, Hurt, Steel et al., 1983).
  • One study suggested that the preoperative cell area may vary more in patients who develop postoperative corneal decompensation (Rao, Aquavella, Goldberg et al., 1984), and another that the mean area of the smallest cell is greater in patients who develop postoperative corneal decompensation (Bates and Cheng, 1988).

The results of the methodology review indicated, overall, relatively poor quality scores for the eight studies, regardless of which evaluation criteria were applied (see Appendix G in the Guideline Report). This is, in part, because the primary objectives of the studies reviewed were different from the objective of the questions posed by the panel for development of the Guideline Report.

Question 2: Is there evidence to support or refute an association between the results of preoperative specular microscopy and indications or contraindications for surgery?

No article contained data suggesting that the results of specular microscopy are useful in defining indications or contraindications for surgery.

Question 3: Is there information on cost and/or cost-benefit issues?

No article contained information on cost and/or cost-benefit issues.

Conclusions of the Panel

The results of the literature review failed to establish an association between the results of preoperative specular photography of the corneal endothelium and the outcome of surgery for cataract. The panel concluded that there is currently no evidence in the published literature or compelling rationale to support routine use of specular microscopy in patients for whom cataract surgery is being considered. Most patients at risk of corneal decompensation from surgery can be identified through history and clinical examination. There is no empirical evidence from the published literature that suggests that use of specular photographic microscopy in routine preoperative assessment adds information of value, beyond that gained through careful history and clinical examination, in deciding whether to proceed with cataract surgery or which procedure to perform.

The panel further concurred that, although the published literature does not provide evidence supporting use of specular photographic microscopy in the routine workup, there may be specific instances in which the endothelial cell count can contribute information pertinent to clinical decisionmaking. For example, there are patients with moderate degrees of endothelial disease and/or a history of previous intraocular surgery or intraocular inflammation for whom the test might be warranted. It is in these patients that a test to determine the status of the endothelium would be most helpful. However, the irregularity of the posterior surface layer in such patients may render the endothelium difficult to visualize. Further research on this subgroup of patients is warranted.

Other Preoperative Tests

The following tests are not indicated as part of the preoperative workup for cataract surgery unless special circumstances, as documented in the patient's chart, justify them.

  • Formal visual fields.
  • Fluorescein angiography.
  • External photography.
  • Corneal pachymetry.
  • B-scan ultrasonography.
  • Specialized color vision tests.
  • Tonography.
  • Electrophysiologic tests.

Views

  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...