In 1853, a German ear specialist named Moritz Heinrich Romberg developed a test to diagnose diseases through balance assessment. Today this “Romberg Test” is used by clinicians as a non-specific test of neurological or inner ear dysfunction.
These clinical tests have been modified and are now used by police officers in field sobriety testing. The Romberg test is not one of the Standardized Field Sobriety Tests, but it is widely used by the police.
The basic Romberg Test has subjects stand with their feet together, hands at their side, their head tilted back, and their eyes closed.
Out of this basic format several variations have arisen. These variants are sometimes referred to as the “Sharpened” Romberg or the “Modified Position of Attention.” These variants may include standing heel to toe rather than feet side by side, holding the arms straight out in front of the body rather than touching the sides of the body, tilting the head back, estimating when 30 seconds have elapsed, and standing heel to toe with the head back, or standing in the basic Romberg position while performing the finger to nose test.
Clinically, the degree of sway is used as an assessment of certain illnesses, aging, exposure to drugs or toxins, injury, fatigue or stress. In a clinical setting, a positive Romberg test is considered to be an actual loss of balance.
In a clinical setting the test is always performed with the eyes open at first, in order to establish a performance baseline. In Law Enforcement, the amount of sway is assumed to correlate to alcohol impairment, without the establishment of a baseline.
However, the key to reliability in the Romberg Test hinges on whether a nexus between normal performance and alcohol-induced performance has been established. The measurement of a normal baseline performance for the test subject is critical in the determining whether there is a change from normal performance and whether any change can be attributable to alcohol ingestion.
Since determining a baseline performance is virtually impossible in a law enforcement setting, data on the normal range of performance for the general population is essential. However, it is well known that “balance tests of various sorts show large individual differences in the performance of sober individuals.”
The use of the Romberg Test as applied in law enforcement has been studied with 104 subjects, ages 18-52. The subjects stood with their feet together and their arms down at the sides, and maintained that position while an officer gave instructions. The officer’s instructions were as follows:
Do not start the test until I say ‘begin’; tilt your head back slightly and close your eyes; when I say ‘start’ keep your head tilted back with your eyes closed until thirty seconds have elapsed; and when thirty seconds have been estimated bring your head forward, open your eyes, and say ‘stop.’
The officer was instructed to be sure that the subjects understood the procedure, and to keep track of the time while the subjects performed the test. The officer also was instructed to ask the subject “How much time was that?” when the subject opened his/her eyes. Data was collected on the amount of sway, eyelid flutter and accuracy of time estimation.
All subjects estimated the elapse of 30 seconds time within +/- 23 seconds; only 4.8% estimated 30 seconds at exactly 30 seconds, 78.9% (cumulative) estimated within +/- 10 seconds, 87.5% (cumulative) within +/- 15 seconds.
Only 28.9% had no eyelid flutters; over 71% exhibited eyelid flutters. 56.7% had a sway of 1-2 inches, with almost an equal distribution of the remaining percentage between subjects who swayed over 2-4 inches and subjects who swayed under one inch.
This degree of variation among the general population is expected. Also, increased body sway can be present for a variety of reasons, including weight, age and physical condition.
The Federal studies also seem to indicate that the Romberg test is unreliable in identifying individuals who are over a 0.08% BAC. In a Florida study there were 53 evaluations where the battery of three Standardized Field Sobriety Tests were not given or were augmented by the Romberg and the Finger to nose. In the 7 cases where drivers were actually less than a 0.08%, 4 were incorrectly arrested as being over a 0.08%, for a 43% accuracy rate.