I am deeply troubled by your publication of “Occurrence of Physiologic Gaze-Evoked Nystagmus at Small Angles of Gaze,” in the May issue.
1 In multiple areas, it falls well below any acceptable standard of peer review and especially is not in keeping with the high standards that are typical of the papers published in this journal. The research and analysis are flawed and the conclusions are not supported by the references or evidence presented.
The authors make no reference to any accepted, peer-reviewed literature that supports the horizontal gaze nystagmus (HGN) test
2 –5 or its underlying principles.
6,7 Numerous additional research papers and reports, extending back over 50 years, also could be cited.
The authors cite incorrect references in several instances or misstate what they are citing. For example, they list Leigh and Zee
8 as the source of the “facts” stated in the second sentence of the paper: “Studies suggest that [gaze-evoked nystagmus] is present in over 50% of the normal population and is more common in fatigued subjects.”
1 However, this information comes essentially from Abel et al.
9 Also,
fatigue nystagmus occurs with maintenance of lateral gaze for an extended time. It has nothing to do with sleep deprivation, as the term “fatigued subjects” suggests.
Other errors include the authors' count of the subjects used in the research—they claim that data from 56 subjects were analyzed, but their breakdown of ages totals 63—and the lack of a detailed description of Figure 1.
As to the research itself, the authors used an infrared eye-tracking system with an accuracy of no better than 0.5°, and that varies with gaze angle. Yet, they reported a single mean amplitude of nystagmus of 0.22° across all gaze angles. In any case, a measuring system should be at least twice as sensitive as the effect that one intends to measure. Consequently, one can have no confidence in any reported nystagmus with amplitude of 1° or less. In addition, the authors provide no information about the parameters of the targets that the subjects fixated.
With regard to data analysis, reporting of means and standard deviations implies that the data are normally distributed. The reported mean amplitude of nystagmus is an absolute value, and a negative amplitude has no meaning. Consequently, the SD 0.33° gives a 2-SD range (that should include approximately 95% of the sample) of +0.88° to −0.44°. Thus, the data are not normally distributed.
Finally, and most important, is the issue of the authors' conclusion that their findings somehow relate to the HGN test conducted by police officers. The authors state that they did not visually observe the nystagmus that was being recorded. Yet the HGN test is one of visual observation.
2 –5 Even the most experienced clinician under ideal conditions can recognize an eye movement of no less than approximately 2 prism diopters, or approximately 1.1°.
10 The only conclusion that the authors can make is that they measured small-amplitude physiologic nystagmus using a non–visually observed method. This method has nothing to do with the HGN test.