Of paramount importance in the inability of the eye to maintain a consis­tent lateral gaze, or HGN, is a docu­mented correlation between a de­crease in the degree to which the eye can gaze to the side, and an increase of blood alcohol content (BAC). First not­ed by Lehti in 1976,^[iv] Tharp, et. al., de­vel­oped the fol­lowing equa­tion to de­scribe the degree to which lat­eral gaze was af­fect­ed:

O ' 51 G 100 (BAC as % w/v)

Solving for the angle of onset [O] or the point where pulsation is first observed, the following simplification has been de­vel­oped for use in the field:

BAC ' 50 G Angle of onset.

In theory, this formula is fairly easy to employ. The administering officer con­structs a 90 de­gree angle by first drawing an imaginary line through the mid­line of the skull. This line repre­sen­ts zero de­grees. A second imaginary line perpen­dicular to the first and drawn through the ears becomes 90 de­grees. Thereafter, the officer, by smoothly moving a target such as a pen or penlight watches for the onset of pul­sation. The point at which this action is first noted is the "angle of onset." This means that if the onset angle was 40 de­grees, then the BAC would be expected to be approx­i­mate­ly .10% (50 G 40). Likewise, with an on­set an­gle of 35 de­grees one would expect to find a BAC of .15% (50 G 35).

In a study conducted by Goding and Dobie,^[v] 37 patients in an emergency room were tested using the above for­mula. Of 20 patients with a BAC in ex­cess of .20% w/v, 17, or 85%, had an an­gle of onset of less than 30 degrees and all 20 showed onset at 40 degrees or less. The measured BAC of six patients was between .10% and .20% and five of the six had an angle of onset of 40 degrees or less. In one patient with a BAC of .155%, onset occurred at 45% suggesting a BAC of .05%.

In the eleven patients that re­mained, the results were not as good. Four patients with a BAC of less than .10% had an angle of onset of less than 40 degrees, thus falsely indicating intox­ication. These false positive were found as a result of drugs such as di­phenhydramine, phenytoin and me­peri­dine.

When the authors conducted a simi­lar examination using experienced members of the Seattle police depart­ment, BAC esti­mating errors were had in approximately 10% of the cases, of which 86.86% resulted in overestima­tion, in one instance as great as .09%.

While Goding and Dobie achieved results greater than the 78% predicted by Burns & Moskowitz in their 1981 study,^[vi] it must be recognized that this was a comparatively small sample and that the an­gle of on­set was mea­sured us­ing a nys­tag­mus pro­tractor, whereas in actual prac­tice nys­tagmus is crudely esti­mated us­ing crutches such as the disap­pear­ance of the white of the eye. In­deed, even NHTSA recog­nizes the need for accura­cy in adminis­tration, not­ing that "nys­tagmus an­gle of onset is an ex­cellent tool for pre­dicting the BAC when it is measured with suffi­cient pre­cision."^[vii]

Moreover, not all researchers agree on the existence of a relationship be­tween the angle of onset and BAC. Norris,^[viii] for instance, rejected the exis­tence of virtually any relationship be­tween onset and BAC. Others, most notably Aschan, the grandfa­ther of Nys­tagmus research, described a 40 degree angle of onset at a BAC of .06% and found nystagmus to persist for more than five hours after the disappearance of all alcohol from the body.^[ix] Al­so, there exists a lively debate as to occur­rence of "normal nystagmus." While NHTSA places this figure at 4.4%, Toglia states that nystag­mus at 40 de­grees and great­er exists in 50-60% of the sober popula­tion.^[x] Even the hour of administra­tion may have a role to play. Tharp, et. al., in a 1981 NHTSA study^[xi] reported that "the ef­fects of fatigue or cir­ca­dian rhythms on gaze nystagmus could be significant," observing that "the alcohol dose de­creased the angle of on­set by an addi­tional five degrees after mid­night." Finally, de­tec­tion of HGN is com­pli­cat­ed by the pres­ence of end point nys­tag­mus or pseudonystag­mus. This phenom­enon, which occurs nor­mal­ly in many individ­uals, disap­pears when the gaze is shifted centrally, or toward the nose a few de­grees. Pseudonystag­mus, there­fore, strongly resembles HGN and has been described in the literature as dif­ficult to deter­mine from true nystag­mus.^[xii]

While several New York police agen­cies, most notably the New York State Police, routinely employ HGN for detecting and measur­ing intox­ication, a close examination of the tech­nique as em­ployed raises grave concerns. For instance, NHTSA instructs the of­ficer to observe the subject and take note of "the first sign of jerking." HGN, how­ever, is far more than "jerking." An of­ficer looking merely for "jerking" may record rotary or PAN, although both are undesired side affects. Whereas NHTSA adminis­tered placebos to those with natural nystagmus or who were taking prescription drugs known to pro­duce such results, in practice officers do not take an adequate history prior to administration of the test. Troublesome is that given the "jerking" definition that is taught, a history should include all alcohol consumption within the last 24 hours. Given the circum­stance in which the test is administered, it does not seem probable that a forth­right history can be obtained. Contact lenses also create problems. Whereas NHTSA now recommends that they be re­moved,^[xiii] officers frequently fail to ask if they are being worn and of course re­moval under the circumstances creates difficulties that are perhaps insur­mountable.

Perhaps the greatest impediment to the use of HGN is its unquestioned lack of specificity. As recently recog­nized by the Supreme Court of Kansas in State v Witte, (1992) 836 P.2d 1110: "Nystag­mus can be caused by problems in an individual's inner ear * * *. Physio­logi­cal problems such as certain kinds of diseas­es may also result in gaze nystag­mus. Influen­za, streptococcus infec­tions, ver­tigo, measles, syphilis, arterio­sclerosis, muscu­lar dystrophy, multiple sclerosis, Korsakoff's Syndrome, brain hemor­rhage, epilepsy, and other psycho­genic disorders all have been shown to cause nystagmus. Furthermore, condi­tions such as hypertension, motion sick­ness, sunstroke, eyestrain, eye muscle fatigue, glaucoma, and changes in atmo­spheric pressure may result in gaze nys­tagmus. The consumption of common substances such as caffeine, nicotine, or aspirin also lead to nystagmus almost identical to that caused by alcohol con­sumption."

While cases regulating HGN run the gamut, in New York, the test has been rejected on foundational grounds. In People v Erickson, (1989, 3rd Dept) 156 AD2d 760, 549 NYS2d 182, the Court found ad­mis­sion to be error, albeit harm­less. A similar out­come was reached in People v Torrey, (1988, 3rd Dept) 144 AD2d 865; 534 NYS2d 807. In Torrey, the Third De­part­ment, cit­ing the Frye stan­dard, found the proof in­suf­fi­cient to warrant admis­sion, although such ad­mis­sion was ulti­mately held to con­stitut­e harm­less er­ror.

When confronted with HGN, ques­tion the officer as to his or her training. If the officer will admit that the NHTSA publica­tions are authoritative, explore with particularity the degree to which he or she followed its procedures. If the officer ac­knowledges the danger presented by contact lenses or organic conditions such as seizure disorders, see if an adequate history was taken. If it appears that HGN will be an issue at trial, explore the possibility of retaining an ophthalmologist or an optometrist to testify as to the nonspecificity of the tech­nique. While we will leave it to others to characterize the officer's tech­nique as "eye-balling," it is help­ful to distinguish the officer's examination from the sophisticated tech­niques medi­cally employed for nys­tagmus mea­sure­ment.