Comparing Ocular Muscle Limitation Tests for Clinical Trial Use

Objective  To evaluate portable methods for documenting ocular muscle limitationthat might be used at multiple sites in a clinical trial.

Methods  In a prospective consecutive case series, 2 examiners independentlyevaluated 3 methods of ocular muscle duction testing: a grading scale of 0to −5 that is in clinical use, the Kestenbaum limbus test using a rulerto measure the millimeters of ocular movement, and an adapted cervical rangeof motion (CROM) device that measures ocular movement in degrees. Twenty consentingpatients (mean age, 55 years) with diplopia, 8 with ocular myasthenia gravis,11 with a cranial nerve III or VI palsy, and 1 with dysthyroid ophthalmopathywere studied.

Results  For Kestenbaum measures, between examiners the standard deviation ofthe difference for all ductions was 1.9 mm (r = 0.75, P = .01); 95% of differences were 4 mm or less. For theCROM device, the standard deviation of the difference was 7.1°; 95% ofdifferences were 15° or less (r = 0.73, P = .01). For each examiner, the CROM standard deviationof the difference was less than 2° (r = 0.98, P = .01). For the grading scale, the 2 examiners had thesame score in 85% of ductions (r = 0.92, P = .01).

Conclusions  The Kestenbaum test and the CROM device gave similar interexaminer repeatability.The repeatability for CROM measures for each examiner was high but was considerablyless between examiners. The grading scale gave similar results between examiners.

Standardized objective testing is needed to gauge the effect of therapyon ocular motor dysfunction, but specific methods have not been used in aclinical trial. The technique of Urist,¹ binocularfield testing,² and Hess screen are not suitablebecause of low retest reliability, poor correlation with the patient's visualdisability, and the need for a stable area of binocular single vision, whichis often absent in a variable disorder such as myasthenia gravis. A photographicsystem used in the assessment of sixth nerve paresis, although reliable andreproducible, is not portable.³ We assessed3 methods that are fast, portable, and readily available and might providemeasurements of ocular muscle limitation in a clinical trial.

We tested supraduction, infraduction, abduction, and adduction of eacheye monocularly by 3 methods: a grading scale⁴ with0 for full excursion and −5 for failure to reach the midline (−4reaching midline, −3 to −1 for excursions in 25% increments);millimeters of excursion by means of the limbus test⁵;and the degrees of movement measured with an adapted cervical range of motion(CROM) device.⁶ Wearing the CROM device, aheadgear and spectacle frame with gravity-driven meters, the patient fixatedon a 20/40 letter as the examiner rotated the patient's head. When the patientnoted optotype blur caused by rotation of the head moving the target off themacula, we recorded the degrees registered on the appropriate meter. EachCROM measure was repeated.

Two examiners independently prospectively tested 20 patients with diplopiacaused by unilateral cranial nerve III or VI palsy (11 patients), dysthyroidophthalmopathy (1 patient), and myasthenia gravis (8 patients) (mean age,55 years). All patients gave informed consent. The repeatability of CROM measuresfor each examiner and the standard deviation of the difference for all resultsbetween examiners were evaluated by means of Pearson correlation or intraclasscorrelation coefficient.

For the limbus test results, the standard deviation of the differencefor 160 ductions between examiners was 1.9 mm (r =0.75, P = .01); approximately 95% of differenceswere 4 mm or less.

For CROM measures between examiners, the standard deviation of the differencewas 7.1°; approximately 95% of differences were 15° or less (r = 0.73, P = .01). For each examiner,the standard deviation of the difference for the repeated CROM measurementswas less than 2° for each examiner (r = 0.98, P = .01) (Figure 1).

For the grading scale, the 160 ductions for each examiner were the samegrade in 85% (r = 0.92, P =.01). The grades differed by 1 grade in 25 measures and by 2 grades in 1 measure.

Our results with the CROM device gave a standard deviation of the differencebetween examiners considerably larger than the 1.1° reported by Kushner.⁶ The variability of the results for 2 examiners suggeststhat neither the limbus test nor the CROM device method as performed in thisstudy is suitable for a clinical trial. In addition, ptosis, poor vision,and bifocal and progressive glasses limit the use of the CROM device. Althoughthe categorical grading system results were consistent between examiners,the coarse determinations may not disclose all clinically relevant changesin a multicenter study. We remain unconvinced that any of these or other existingmethods for assessing extraocular muscle limitation or binocular diplopiacan be used to determine the effects of therapy, particularly in general ophthalmologyor neurology clinics.

Corresponding author: Mark J. Kupersmith, MD, INN at Beth IsraelNorth, 170 East End Ave, New York, NY 10128 (e-mail: mkuper@bethisraelny.org).

Submitted for publication June 12, 2003; final revision received October9, 2003; accepted November 3, 2003.

This study was supported by the Michael Fischer Foundation, Miami, Fla.