Article

Retinoscopy Measurement Differences as a Variable of Technique John W. Streff, OD and Verne E. Claussen, OD Gesell Institute of , New Haven, Connecticut

This article originally appeared in the American Many optometrists express a lack of confidence Journal of Optometry and Archives of American Academy in their ability to interpret the results of dynamic of Optometry 1971;48(8):671-6. With permission from skiametry, particularly the techniques commonly the publisher this article has been reproduced to honor referred to as “book retinoscopy” and “dangle bell the late Dr. John Streff. COVD would like to thank retinoscopy.” They question the reliability of these Stephanie McLin for her assistance in the reproduction techniques. One of the more disturbing elements of of this article. these private “confessions” is that the doubters are often the most competent and thorough clinicians. Abstract The present study was carried out to determine A study was made to check the reliability of the reliability of modified bell retinoscopy findings. dynamic retinoscopy findings in a modification of the We tested the same subjects, using presumably the bell retinoscopy technique. The study was to compare same procedures and identical instruments, under the findings between two examiners, the authors. Selected same conditions. The tests were modifications of the subjects in the test sample were retested to determine bell retinoscopy procedure. if the retinoscopy techniques were a variable. One The data revealed that the findings of the two examiner was measuring the light shadow edge of the optometrists were not the same, despite the standard beam with rather large sweeps of the instrument. The conditions. It was then hypothesized that the different other was measuring the very center of the beam with results were due to variation of retinoscopy procedures very small, rapid sweeps of the instrument. The re- of the examiners. test data confirm that the technique differences could Additional testing did show that there were two account for at least 3.50 D difference at the 40 cm significant test factors not initially recognized. These test distance. were: 1) what area of the reflex examiner was looking Retinoscopy is one of the basic optometric test at during the administration; 2) how he moved his procedures for objectively evaluating the optical instrument during the procedure. status of a patient. All of the optometry schools in the United States emphasize both the theoretical Test Description and applied aspects of this test procedure in their The test target was a chrome-plated sphere, curricula. Yet some men in optometric practices view one-half inch in diameter. It was attached to an retinoscopy findings with doubts and reservation. anodized aluminum rod, 3/16” in diameter. This target works well for two reasons. First it enables the subject to report clarity, while maintaining its desired three-dimensionality. Second, a sphere maintains a Correspondence regarding this article should be emailed to [email protected] or sent to Dominick M. Maino, OD, MEd, Illinois constancy of shape when seen from any vantage point. College of Optometry, 3241 S. Michigan Ave., Chicago, IL 60616. All Therefore, the spherical target allows the subject to be statements are the author’s personal opinions and may not reflect the opinions more aware of the spatial changes directly related to of the College of Optometrists in Vision Development, Optometry & Vision the retinoscopy procedure. The subject, then, is not Development or any institution or organization to which the author may be affiliated OVD is indexed in the Directory of Open Access Journals. Online distracted by any change in target shape. access is available at http://www.covd.org. Fig. 1 presents the equipment diagram. The test target (T) was suspended in a clamp holder so Streff J. and Claussen V. Retinoscopy measurement differences asa variable of technique. Optom Vis Dev 2011;42(2):99-102. that it moved parallel to the carrier track and floor, and could be adjusted to the subject’s level. The

Volume 42/Number 2/2011 99 stopped when the first persistent against motion was observed by the examiner. This distance was measured from the subject’s head rest, in the nearest centimeter units. The distance closely corresponded to the corneal plane of the subject. Reflex change measurements were made in the two principle meridians: 180˚ and 90˚. The same procedure was repeated for 50 cm and 40 cm test distances. Figure 1: VisionPrint System: The subject wears a headset with sensors The subjects were nine third-grade students in (as shown) and sits approximately 40 cm from the machine. the Doolittle School, Cheshire, Connecticut. The children were selected from a group of subjects who carrier track extended beyond both the subject’s (S) had participated in visual testing projects for two and the examiner’s (E) positions. A forehead rest years. They were all known to measure standard was mounted on the underside of the carrier track acuity and have a static retinoscopy within the range so that the subject’s position could be controlled. A of -0.25 to +1.25 D at 6 M. Subjects were limited to loop which was adjustable for different test distances those measuring 0.50 D or less of either with-the-rule extended below the track on the examiner’s end. This or against-the-rule . loop was designed to accept the forehead rest of the Table 1 presents the results and differences of American Optical spot retinoscope. In this way the the right-eye examinations of the nine subjects by desired retinoscopy distance could be maintained. Examiners 1 and 2. They are recorded in spherical The target clamp was free to move on the overhead equivalent dioptric values equal to the distance track from the retinoscope towards the subject. A measures. There were no astigmatic differences greater continuous cord belt (B) allowed the examiner to than 0.50 D when distance measures were converted move the target by reaching up and pulling the cord. to dioptric values. Since the left-eye measures were In this way, there was no need for him to touch the comparable to those of the right eye, there was no target or the target holder. reason to report them in this study. The subject was seated on a flat-top school desk It was apparent that there was a significant with an appropriate foot rest. The carrier track was discrepancy between the findings of the two examiners. adjustable in height so that the subject’s forehead rest The differences ranged from no disagreement for could be positioned for comfortably straight sitting subject #7 to as much as 3.69 D at 40 cms for subject posture. #1. The mean dioptric differences between examiners for subjects 1 through 9, at all three retinoscopy First Experiment Test Procedure distances were close to 1.00 D. The tests were administered in a large utility An examination of the raw data brought only one room of a public school. Each optometrist tested each discernible pattern of examiner differences. It was child separately. When one examiner completed his noted that in all cases Examiner 2 measured a later work with a particular subject, the other optometrist against motion than Examiner 1. Nothing further examined the child. The examining optometrist was could be made of this information. neither aided nor observed by the unoccupied one. It was hypothesized that the differences in results One-half of the subjects were first tested by Examiner were due to a varied use of the retinoscope. An inquiry 1 and the other half first by Examiner 2. lead to the discovery that Examiner 2 was using his Testing was done at three retinoscopy distances: instrument essentially for clinical purposes. Examiner 67cms, 50cms, and 40 cms. These distances were used 1, on the other hand, was using his retinoscope to in the same order for all subjects. The procedure was determine subtle retinoscopy changes while engaged as follows: The retinoscope was placed at the first test in problem solving talks. distance (67 cms). The test target was positioned at The two optometrists demonstrated their the same distance from the subject, at a height just retinoscopic techniques to each other. It was apparent above the light source of the retinoscope. While the that Examiner 2 was making greater sweeps with examiner scanned the reflex of the subject’s , the the retinoscope than Examiner 1. In addition, the test target was moved slowly towards the subject. It was demonstration revealed that the two optometrists

100 Optometry & Vision Development Table 1 • Individual Findings, Mean Scores and Differences Between Two Examiners’ Retinoscopy Findings for Nine Subjects Measured at Three Distances. (Scores reported are centimeter values converted into dioptric equivalents.) Item: Test Distance

Individual 67 cm 50 cm 40 cm Findings Ex. 1 Ex. 2 Diff. Ex. 1 Ex. 2 Diff. Ex. 1 Ex. 2 Diff. Subj. 1 1.61 3.50 1.89 2.19 4.17 1.98 2.98 6.67 3.69 Subj. 2 1.64 2.33 0.69 2.19 3.27 1.06 2.78 4.08 1.30 Subj. 3 1.69 2.23 0.54 2.10 2.63 0.53 2.82 3.03 0.21 Subj. 4 2.08 2.56 0.48 2.35 3.03 0.68 2.94 3.27 0.34 Subj. 5 1.68 2.33 0.65 2.23 2.78 0.55 2.66 3.57 0.91 Subj. 6 1.89 4.44 2.55 2.50 4.65 2.15 3.03 4.88 1.85 Subj. 7 1.64 1.71 0.07 2.23 2.17 0.04 2.78 2.78 0.00 Subj. 8 1.62 2.94 1.32 2.23 2.94 0.71 2.82 3.45 0.63 Subj. 9 1.71 2.47 0.76 2.33 2.90 0.57 2.94 3.28 0.34

Mean Score 1.73 2.72 0.99 2.25 3.17 0.92 2.86 3.89 1.03 in Diopters SD 0.22 0.21 0.35 watched different parts of the reflex during the test. Examiner 2 neutralized the changes in motion of the light-shadow border lines as they passed through the central pupillary area. Examiner 1 neutralized the center of the reflex and did not sweep the instrument enough to see the light-shadow edge of the beam.

Second Experiment Test Procedure The validity of the hypothesis that technique Figure 2: a. (Left) Subject No. 1. b. (Middle) Subject No. 2. c. differences accounted for the discrepancies in the (Right) Subject No. 3. Findings for Examiner 1 (circles) and Examiner results of Examiners 2 and 1 was tested. Three of the 2 (triangles) on Day 1 (solid lines) and Day 2 (broken lines) using original subjects were re-examined: one whose results peripheral (open figures) and central (closed figures) retinoscopy techniques at three distances. Graphed on full logarithmic scale. revealed a great difference between the two examiners; one whose results showed an average difference; one whose results showed little difference. The test was The measurements of the previous day were not administered at the same three retinoscopic distances. consulted by Examiner 1. The only exception to the schedule was that subject The results of the second testing procedure 3 was not tested at 50 cms. Retesting was done by are plotted (Fig. 2) and are compared to the initial Examiner 1 alone (because Examiner 2 was not findings of the two examiners. The data supported available for the second day of testing). the hypothesis that differences in the examiner’s Two different retinoscopy techniques were used techniques were responsible to a large extent for the to retest subjects 1, 2, and 3. These two techniques discrepancies in the initial test results. employed the differing procedures of Examiner 1 and Table 2 presents possible ways of comparing the Examiner 2. Two measurements were taken: 1) of mean differences in results of the findings. From the the distance at the time that the center of the reflex results it can be seen that: 1. The two procedures changed from a with to an against motion, as the in retinoscopy technique give findings which differ instrument was moved in very small, rapid sweeps and closely from one another, each resembling one of the 2) of the distance at the time that the light-shadow two examiner’s findings, 2. Examiner 1’s findings edge of the beam changed from a with to an against differ only slightly from the central reflex findings motion with wider, slower sweeps of the instrument. (which could be expected since this is testing test-

Volume 42/Number 2/2011 101 Table 2 • Mean Differences Between Different Acknowledgments Examiners and Different Retinoscopy Procedures for The authors thank the Gesell Institute Visual Three Subjects Research Fund, which supported the work for this Test Distance Techniques Compared study. 67 cm 50 cm 40 cm Ex. 1 vs. Ex.2, First Day 1.04 D 1.52 D 1.40 D * Read before the Section on Binocular Vision and Ex. 1, Central Reflex vs. Ex. 1, 1.16 D 1.58 D 1.84 D Peripheral Reflex, Second Day Perception at the Annual Meeting of the American Academy of Optometry, Miami Beach, Florida, Ex. 1, First Day vs. Ex. 1, Central 0.01 D 0.03 D 0.09 D Reflex, Second Day December 13, 1970. For publication in the August, Ex. 2, First Day vs. Ex. 1, 1971, issue of the American Journal of Optometry 0.12 D 0.36 D 0.20 D Peripheral Reflex, Second Day and Archives of American Academy Of Optometry. Optometrist. retest reliability on different days) and Examiner 1’s peripheral reflex (shadow-light) neutralization on the Selected Bibliography second day findings closely resemble Examiner 2’s • Apell, R. J., Book retinoscopy, Eastern States Visual Training Conference, findings on the first day. 1958. • Apell, R. J., and R. W. Lowry, Jr., Preschool Vision; Tests, Diagnosis, Guidance. St. Louis, American Discussion • Optometric Assn., 1959. It would have been desirable for both examiners • Apell, R. J., and John W. Streff. The performance Skills, Optometric Child to have retested the children. It would also be Care and Guidance. O.E.P., Vol. desirable to have pupillary size measurements for the • 34, Ser. 6, No. 4, 1962. different subjects at each test distance. It was possible, • Gesell, Arnold, Frances L. Ilg and Glenna E. Bullis, Vision; its Development however, for Examiner 1 to simulate the technique in Infant and Child. New York, used by Examiner 2 initially. The data supported • Paul B. Hoeber, Inc., 1949. the hypothesis that differences in the examiners’ • Henry, W. R., Dangled bell with retinoscope, St. Louis Visual Training Conference, 1959. retinoscopy procedure and technique were responsible • Nott, Ivan S., Convergence, and fusion, Am. J. Optom., to a large extent for the discrepancies in the initial test 6 (1): 19-29, 1929. results. • Pascal, Joseph I., Modern Retinoscopy. London, Hatton Press, 1930. The study shows that the technique of retinoscopy • Streff, J. W., Some early observations with a bright bell and retinoscope, in this test situation did account for relatively large Washington Visual Training differences in test results (as much as 3.82 D for subject • Conference, 1962. 1 at 40 cm). It is understandable that optometrists lack • Tait, E. F., A quantitative system of dynamic skiametry, Am. J. Optom., 6 confidence in the retinoscopy procedure and results. (12): 669-693, 1929. The implication of this study raised several questions. How can retinoscopy findings be reported in a more standard manner? How can retinoscopy findings yield two optical neutrals with a spread of 3.82 D, both findings being relatively reliable on a test-retest basis? What is the difference between subjects who show large difference between central reflex neutralization and shadow-light neutralization and subjects who show little difference? It is clear that optometrists must understand and recognize different techniques and their results in retinoscopy. Further, reports of retinoscopy findings should be accompanied by clear descriptions of the procedural techniques employed. In this way, retinoscopy can be more meaningful and communicable, and confidence in the findings can be restored.

102 Optometry & Vision Development