COMMONWEALTH vs. DANIEL N. NEAL.

ABRAMS, J. Convicted of operating a motor vehicle on a public way while under the influence of intoxicating liquor, G. L. c. 90, Section 24 (1) (a), the defendant challenges the admission in evidence of the result of his breathalyzer test. [Note 1] The defendant argues that the Commonwealth's failure to preserve samples of his breath for subsequent retesting by his own expert, as well as the Commonwealth's nonpreservation of ampules used during the examination, required exclusion of the breathalyzer evidence. The defendant also argues that the Smith & Wesson Model 900A Breathalyzer, the model used to determine his blood alcohol content, is susceptible to radio frequency interference and is therefore so unreliable as to require suppression of the results of the examination. The defendant finally contends

Page 3

that such evidence should not be admissible because the Commonwealth's practice of conducting only one breathalyzer examination is scientifically deficient.

We conclude that the Commonwealth's failure to retain a sample of the defendant's breath does not require that the examination result be excluded. We also decide that the Commonwealth need not preserve test ampules for inspection by the defense. We hold that, in this case, the prosecution made a sufficient showing that the particular instrument used to measure the defendant's blood alcohol content was not susceptible to radio frequency interference. Finally, we decide that a second breath test, though advisable, is not required. We therefore affirm the judgment.

We summarize the facts. On July 22, 1982, the defendant was arrested following his involvement in a two-vehicle accident in a village of Osterville. The operator of the other vehicle was driving in a northbound lane when he observed a southbound vehicle driving toward him in the northbound lane. The operator unsuccessfully attempted to avoid the oncoming vehicle, which collided with his automobile and then struck a tree on the side of the road. The Centerville-Osterville fire chief came to the scene of the accident and found the defendant, who stated that his vehicle was unregistered and uninsured and that "[y]ou might as well arrest me now." According to the fire chief, the defendant was unsteady and had a strong odor of alcohol on his breath, slurred speech, and glassy eyes. As the fire chief and the defendant walked toward a wooded area where the defendant's vehicle had come to a stop, the defendant ran away and attempted to hide. The fire chief found the defendant and persuaded him to stay at the scene of the accident until a police officer arrived.

When the officer appeared, the defendant was asked to perform several field sobriety tests. The defendant lost his balance while attempting to walk a straight line, could not locate the tip of his nose, and missed several letters while trying to recite the alphabet. The defendant was arrested and transported to the Barnstable police station, where he was booked and given Miranda warnings. The defendant elected to submit to a breathalyzer examination, which yielded a blood alcohol content

Page 4

reading of .14 [Note 2] The defendant was advised of his right to have an independent breathalyzer test administered at his expense by a person or physician of his choice, G. L. c. 90, Section 24 (1) (e), but declined such further examination.

After being arraigned and charged with operating while under the influence of intoxicating liquor, the defendant was tried before a District Court judge and found guilty, receiving a fine and a surfine totaling $500, a suspended six-month sentence to a house of correction, and probation for two years. The defendant appealed to the jury of six session of the District Court, and thereafter filed motions to suppress the breathalyzer result as unreliable and to dismiss the case because of the Commonwealth's failure to obtain and preserve samples of the defendant's breath at the time of his arrest, failure to conduct at least two breathalyzer examinations, and failure to provide exculpatory on the motions over three days, denied the motions on March 28, 1983. The defendant thereupon waived a jury trial and stipulated that a police report containing the above-summarized facts could be read in evidence in lieu of trial. The judge found the defendant guilty and imposed the same sentence as the primary court. [Note 3] The defendant filed a timely appeal to the Appeals Court. We granted the defendant's application for direct appellate review.

1. Operation of the Smith & Wesson Model 900A Breathalyzer [Note 4] We preface our discussion of the specific issues

Page 5

raised by the defendant with a description of the operation of the breathalyzer instrument used to obtain the challenged blood alcohol content measurement. At the suppression hearing, Dr. Harvey Cohen and Dr. James Feldman were qualified by the defendant as experts in the fields of chemistry and electrical engineering, respectively. As explained by Dr. Cohen and Dr. Feldman, the examination performed by the Model 900A unit involves passing a sample of the suspect's breath through a potassium dichromate-sulfuric acid solution, bright yellow in color. A sealed reference ampule and a test ampule, the top of which is manually broken off before the examination, each containing a specified volume of the solution, are situated in holders within the breathalyzer unit, with a photoelectric cell behind each ampule. The ampules are made of glass. Located on a movable carriage between the two ampules is a lamp that casts light in the direction of both ampules. Each photoelectric cell generates a voltage in proportion to the amount of light reaching the cell after passing through the ampule in front of the cell. A device known as a null meter measures the difference, if any, between the voltages generated by the photoelectric cells. Before an examination is conducted, the breathalyzer operator moves the lamp along its carriage by means of a knob until the null meter registers zero, indicating that the cells are receiving equal amounts of light. A pointer on a separate blood alcohol content scale calibrated from .00 to .40 per cent is then manually placed at the .00 mark.

When the instrument is "zeroed," a sample of the suspect's deep lung breath, exhaled through a mouthpiece into the instrument's breath chamber, is directed from the chamber through a bubble tube into the unsealed test ampule. The solution interacts with any alcohol in the breath sample in an oxidation process that lightens the yellow color of the solution in proportion to the amount of alcohol present. If the breath sample contains alcohol, the resulting decrease in the solution's optical density permits an increased amount of light to reach the photoelectric cell behind the test ampule, and the null meter consequently

Page 6

registers a voltage differential. The operator then turns the knob, moving the lamp along the carriage until the null meter again reads zero. The distance the light must be moved to obtain a zero reading on the null meter is directly related to the amount of alcohol in the breath. The knob that moves the lamp carriage also moves the pointer on the calibrated scale, generating a blood alcohol content reading. The location of the pointer after the examination has been administered is imprinted on a test card.

The officer who administered the defendant's breathalyzer examination stated that, pursuant to a written protocol issued by the police, he observed the defendant for twenty minutes prior to conducting the test to ensure that the defendant did not eat or drink anything. During this time, the officer gauged the reference and test ampules to ascertain that each contained the proper volume of solution and, after breaking off the top of the test ampule, inserted both ampules. After "flushing" the breath chamber to remove any residue from a prior examination, the officer "zeroed" the instrument as described above and performed the examination. Having recorded the result, the officer again flushed the chamber and, following standard procedure, conducted a simulator test. Designed to ascertain whether the breathalyzer unit is functioning correctly, the simulator examination is performed by passing a prepared solution containing a .15 per cent concentration of alcohol through the test ampule used during the examination of the suspect's breath sample, [Note 5] and obtaining a blood alcohol content reading. The officer stated that the simulator test conducted immediately following the defendant's examination yielded a correct reading of .15.

2. Due process right to obtain breath samples and ampules. We consider initially the defendant's claim that his due process rights were violated by the Commonwealth's failure to preserve a sample of the defendant's breath for the defendant's independent testing and by the Commonwealth's routine destruction, after the defendant's breathalyzer examination, of the ampules used during the examination.

Page 7

At the hearing, the defense presented, through Dr. Cohen, expert testimony indicating that samples of the defendant's breath taken at the time of arrest or at the time the breathalyzer test was performed, if made available to the defense for independent testing, might reveal inaccuracies in the breathalyzer test result obtained by the police. The validity of the result could also be subjected to scrutiny, according to Dr. Cohen. were the Commonwealth to preserve the ampules used in the breathalyzer test. The defendant argues that the Commonwealth's failure to make available to him such potentially helpful evidence deprived him of a fair trial and entitles him to a dismissal of the charge brought against him. The judge denied the defendant's motion to dismiss, noting that the expert testimony before him indicated that preservation by the Commonwealth of the breath samples and ampules might assist the defendant at trial, but concluding that such preservation is not constitutionally required.

A. Preservation of breath samples for retesting. We review first the judge's decision with regard to the Commonwealth's duty to retain breath samples. Decision of some appellate courts indicate that preservation of breath samples for purposes of retesting is now technologically feasible and can be accomplished at a moderate cost. Anchorage v. Serrano, 649 P.2d 256, 259 (Alaska Ct. App. 1982). Baca v. Smith, 124 Ariz. 353, 356-357 (1979) (en banc) (one week limit to preservation). People v. Trombetta, 142 Cal. App. 3d 138, 142 (1983), cert. granted, 464 U.S. 1036 (1984). Garcia v. District Court, 21st Judicial Dist., 197 Colo. 38, 45 (1979) (en banc). State v. Cornelius, 122 N.H. 925, 927 (1982). See State v. Larson, 313 N.W. 2d 750, 752 (N.D. 1981) (feasibility stipulated). [Note 6]

In support of his request that we impose on the police a constitutional duty to obtain breath samples for a defendant's use, the defendant relies on Brady v. Maryland, 373 U.S. 83, 87 (1963), and subsequent cases establishing a prosecutorial

Page 8

duty to disclose to the defense evidence in the prosecution's possession that is both material to the defendant's guilt or innocence and exculpatory. See Commonwealth v. Gilday, 382 Mass. 166, 175 (1980). Neither the United States Supreme Court nor this court has to date read this duty to extend, beyond the disclosure of evidence already in existence and in the prosecution's control, to the gathering of evidence potentially helpful to the defense. The defendant beckons us to Garcia v. District Court, 21st Judicial Dist., supra, in which the Supreme Court of Colorado held that due process requires the State in driving under the influence cases to preserve a sample of a defendant's breath, taken at the time the breathalyzer test is administered, for separate testing by the defendant. The court reasoned that "[t]he failure of the state to collect and preserve evidence, when those acts can be accomplished as a mere incident to the procedure routinely performed by state agents, is tantamount to suppression of that evidence." Id. at 46.

Garcia, and other cases reaching a similar result, see Baca v. Smith, supra; People v. Trombetta, supra; cf. Anchorage v. Serrano, supra, are informed by a concern that, if breath samples are not preserved, "the defense has little or no recourse to alternate scientific means of contesting the test results." Baca v. Smith, supra at 356. We agree that due process requires that a defendant be given an opportunity to challenge the accuracy of a breathalyzer result.

We believe, however, that the availability to a defendant in this Commonwealth of an independent blood alcohol content test as a matter of statutory right, see G. L. c. 90, Section 24 (1) (e), [Note 7] discharges this due process obligation by granting the defendant access to scientific evidence with which to dispute the Commonwealth's evidence. Cf. Commonwealth v. Alano, 388 Mass. 871, 876 (1983). In such circumstances, we decline

Page 9

to read into the United States Constitution a prosecutorial duty to gather evidence potentially favorable to the accused -- a duty the limits of which are not readily discernible -- or to strain, as the cases supporting the defendant's position have done, to classify the failure to obtain breath samples from a suspect as a "suppression" of evidence. We note that those courts that have considered the issue in light of the availability of an independent test have uniformly held that due process does not require the State to obtain or preserve breath samples where the defendant is given the opportunity to obtain an independent examination. See Anchorage v. Serrano, supra at 258 n.5; State v. Cornelius, supra at 927; State v. Larson, supra at 752-753. See also State v. Young, 228 Kan. 355, 362-363 (1980).

The defendant argues that, to the extent the Commonwealth relies on the availability of an independent test as a due process safeguard, it must establish not only that the defendant was advised of his right to a second test, but also that, as a practical matter, the defendant had access to such a test. Whatever the merits of this argument in a situation where the defendant requests, but is unable to obtain, a second test, see Anchorage v. Serrano, supra at 258 n.5, we find it unpersuasive where, as in this case, the defendant declined the additional test, and where there is no indication that the defendant's decision was influenced by practical obstacles in obtaining the test. Nor do we accept the defendant's contention that, lacking information about potential sources of error in breathalyzer tests, he overestimated the instrument's reliability, and therefore cannot be said to have knowingly and intelligently waived his right to an independent test satisfying due process requirements. We think advice to a person submitting to a breathalyzer examination of the statutory right to a second test constitutes sufficient notice that the breathalyzer is less than infallible; in any event, we perceive no deficiency in the information available to the defendant sufficient to impugn the validity of the waiver. See United States v. Anderson, 533 F.2d 1210, 1212n.3 (D.C. Cir. 1976). [Note 8]

Page 10

We conclude that where a driving under the influence suspect is advised of the right to an independent test and obtains or declines to take an independent test, the gathering by police of breath samples at the scene of arrest or at the time the defendant's breathalyzer examination is conducted is not a constitutional prerequisite to the admissibility of the breathalyzer result. [Note 9]

B. Retention of ampules. Because the defendant's right of access to an independent test is separate from his constitutional right to be provided with material, exculpatory evidence possessed by the prosecution, see Commonwealth v. Gilday, 382 Mass. 166, 175 (1980), we consider whether the destruction of the ampules, which, unlike breath samples, are available to the police as a normal by-product of breathalyzer examinations, deprived the defendant of evidence to which he was constitutionally entitled.

At the hearing, the defendant's expert stated that imperfections in the quality or thickness of glass in the ampules used in the test, deficiencies in the volume of solution contained in the ampules, or contamination of the ampules on the interior or exterior by foreign substances could affect the accuracy of Model 900A test results. According to the expert, the ampules, if preserved, can be examined at a later date for the presence or absence of such defects, and, if frozen, can be retested to establish whether the result obtained on the Model 900A was correct.

The issue whether due process requires the preservation of ampules for inspection by a defendant has been litigated in numerous jurisdictions with conflicting results. Courts in several States have concluded that, because the ampules, if available to the defendant, could at best yield evidence impeaching the accuracy of the breathalyzer result, rather than affirmative

Page 11

evidence of the defendant's guilt or innocence, such ampules are neither material nor exculpatory and therefore are not subsumed within the rationale of Brady v. Maryland, 373 U.S. 83 (1963). See Edwards v. Oklahoma, 429 F. Supp. 668, 671 (W.D. Okla. 1976), rev'd, 577 F.2d 1119 (10th Cir. 1978); State v. Preston, 585 S.W. 2d 569, 571 (Mo. Ct. App. 1979); Edwards v. State, 544 P.2d 60, 64 (Okla. Crim. App. 1975); State v. Helmer, 278 N.W. 2d 808, 812 (S.D. 1979); Turpin v. State, 606 S.W.2d 907, 918 (Tex. Crim. App. 1980) (enbanc); State v. Canaday, 90 Wash. 2d 808, 816 (1978) (enbanc). We have said, however, that evidence tending to impeach the credibility of a key prosecution witness is "clearly exculpatory," Commonwealth v. Collins, 386 Mass. 1, 8 (1982); see Giglio v. United States, 405 U.S. 150, 154 (1972), and think it clear that ampules, to the extent they can be used by the defense to impeach the accuracy of a breathalyzer result that is often the cornerstone of a driving under the influence case, fall within this principle. See Scales v. City Court of Mesa, 122 Ariz. 231, 234 (1979) (en banc); People v. Hitch, 12 Cal. 3d 641, 647 (1974) (en banc); State v. Michener, 25 Or. App. 523, 529-530 (1976); State v. Booth, 98 Wis. 2d 20, 27 (1980). See also Lauderdale v. State, 548 P.2d 376, 381 (Alaska 1976) (right to cross-examine).

If the failure of the police to preserve the ampules deprived the defendant of material. [Note 10] exculpatory evidence, the defendant's

Page 12

due process rights were infringed, notwithstanding the fact that the police acted routinely and in good faith. See Brady v. Maryland, 373 U.S. 83 (1963); Commonwealth v. Pisa, 372 Mass. 590, 594, cert. denied, 434 U.S. 869 (1977); United States v. Bryant, 439 F.2d 642, 652 n.21 (D.C. Cir. 1971). Because the ampules have been destroyed, it is no longer possible to determine whether the defendant would have obtained any evidence of an exculpatory nature had the ampules been made available to him for inspection or examination. To require the defendant at this stage to prove that the ampules were in fact exculpatory would, however, convert the disclosure duty established by Brady and its progeny into "an empty promise, easily circumvented by suppression of evidence by means of destruction rather than mere failure to reveal." United States v. Bryant, supra at 648. The defendant is therefore entitled to relief for the Commonwealth's failure to preserve the ampules if he establishes a "reasonable possibility, based on concrete evidence rather than a fertile imagination," that access to the ampules would have produced evidence favorable to his cause. State v. Michener, 25 Or. App. 523, 532 (1976). See Commonwealth v. Andrade, 389 Mass. 874, 879 (1983). We are not persuaded that either retesting or physical examination of the ampules would yield evidence meeting this standard.

(i) Retesting. In this Commonwealth a prerequisite to the admissibility of scientific evidence is a showing that the process used has attained general acceptance by the relevant scientific community. Commonwealth v. Fatalo, 346 Mass. 266, 269 (1963). "[A]t the present time, a scientifically valid procedure is not known to be available for the reexamination of a Breathalyzer ampoule that has been used in [a] breath test . . . in order to confirm the accuracy and reliability of the original breath analysis." National Safety Council, Formal Statement of Committee on Alcohol and Drugs (Oct. 2, 1975, reaff'd Oct. 21, 1981), published in 3 Am. J. of Forensic Med. & Pathology 273 (1982). Decisions of other jurisdictions indicate that ampule retesting has not, to date, achieved general scientific acceptance. See e.g., Scales v. City Court of Mesa, 122 Ariz.

Page 13

231, 234 (1979); State v. Teare, 135 N.J. Super. 19, 21 (1975); State v. Bryan, 133 N.J. Super. 369, 372-373 (1974); People v. Santiago, 116 Misc. 2d 349 (N.Y. Sup. Ct. 1982); State v. Larson, 313 N.W. 2d 750, 755 (N.D. 1981); State v. Watson, 48 Ohio App. 2d 110 (1975). See also Lauderdale v. State, 548 P.2d 376, 379-380 (Alaska 1976); People v. Hitch, 12 Cal. 3d 641, 645 n.1 (1974). But see State v. Booth, 98 Wis. 2d 20, 29-30 (1980).

The judge was not required to accept the testimony of the defendant's expert as to the accuracy of the retest. Because the defendant has not demonstrated that the result of a retest would be admissible scientific evidence, we reject the claim that the preservation by police of ampules for retesting by the defense is constitutionally required.

(ii) Physical examination. Unlike the retesting process, physical inspection of the ampules for potential sources of inaccuracy, such as improper glass thickness, optical quality, or volume of solution, implicates no scientifically questionable procedure. According to the defendant's expert, police can retain ampules and their contents indefinitely by placing them in covered plastic containers such as those used to house 35 millimeter film. However, the record indicates that the simulator test is performed with the identical ampules used in the defendant's breath examination. A correct simulator reading is thus inconsistent with the defendant's theory that the accuracy of his test may have been affected by defects in the ampule glass, the presence of contaminants, or a deficiency in the volume of solution. We therefore perceive no reasonable possibility that the destruction of the ampules deprived the defendant of evidence of any inaccuracy of his test result.

We conclude that, where the Commonwealth introduces evidence that a simulator test was administered after a breath examination and yielded a correct result, [Note 11] the failure to preserve

Page 14

ampules used in the examination infringes no constitutional right of the defendant. [Note 12]

3. Effect of radio frequency interference on reliability of Model 900A. The defendant moved to suppress the result of the breathalyzer examination, alleging that the Smith & Wesson Breathalyzer Model 900A is inherently unreliable because the null meter reading on which it depends is susceptible to radio frequency interference (RFI). [Note 13]

We summarize the pertinent testimony and exhibits received at the suppression hearing. Tests conducted by Smith & Wesson and an independent testing company in the summer of 1982 revealed that the Model 900A Breathalyzer is susceptible under some circumstances to RFI. [Note 14] On September 10, 1982, Smith & Wesson issued a customer advisory referring to these tests and stating that "[s]ince test conditions were representative of some police environments, the possibility exists, although unlikely, for higher or lower than normal [blood alcohol content] test results."

The advisory outlines procedures that, in Smith & Wesson's view, are sufficient to establish whether specific Model 900A units are being affected by RFI so as to pose a risk of inaccurate results. These procedures consist of two testing programs, one designed to measure susceptibility to "background" RFI (radio

Page 15

frequencies transmitted by sources not controlled by the police department), [Note 15] the other to check the effects of "base-station" or "active" RFI (radio frequencies transmitted by sources affiliated with the police department). [Note 16] The advisory states that, provided the testing programs produce acceptable simulator solution readings, RFI "is not a potential problem at your test location under prevailing conditions." The advisory further asserts that, if these testing procedures yield satisfactory results, subsequent testing of the breathalyzer unit for RFI susceptibility is not required unless there is (1) a change in the department's radio operating frequencies or power output, (2) a change in the position, location, or type of base-station antenna, (3) repair or calibration of the breathalyzer, (4) a change in the unit's location or position, or (5) a change in the "general background RFI environment." The advisory recommends that police cruiser transmitters not be operated within 150 yards of the breathalyzer, and that portable transmitters (walkie-talkies) not be operated within twenty-five yards of the breathalyzer, absent test results demonstrating that operation at closer distances produces no RFI.

The Smith & Wesson advisory was received by the Barnstable police department on January 26, 1983. During the midnight to 8 A.M. shift on January 27, a State-certified breathalyzer operator tested the Model 900A for background RFI susceptibility as well as for potential RFI from the central

Page 16

transmitter, cruiser transmitters, or walkie-talkies. No significant deviations in the test readings were recorded. In testing for RFI caused by the station's central transmitter, see note 16, supra, the operator performed simulator tests while the transmitter was broadcasting on the frequency channel regularly used by the department but did not test the breathalyzer for potential RFI resulting from broadcasts over three other infrequently used channels. The operator said that, apart from the testing conducted pursuant to the advisory, he had performed approximately 100 simulator tests in the regular course of duty and had observed an inaccurate simulator reading on one occasion.

Testing for RFI susceptibility was performed at an earlier date by the owner of the Law Enforcement Maintenance Company, who, pursuant to a service contract with the police department, calibrated the Barnstable unit on May 6, 1982, and October 15, 1982, and certified the unit to be in compliance with Federal standards. On both dates, the maintenance company owner checked the unit for potential RFI from the central transmitter by means of a test he had independently developed which essentially followed the Smith & Wesson protocol. The maintenance company owner did not test all of the central transmitter's frequencies, nor did he test the breathalyzer unit for susceptibility to RFI from cruiser transmitters and walkie-talkies. On October 15, 1982, he also performed routine maintenance involving disassembly of the breath chamber.

The defendant's expert estimated that, in a police environment, most breathalyzer results would be reliable, but stated his opinion that the testing protocol in the Smith & Wesson advisory is insufficient to establish that a particular instrument is impervious to RFI. The fact that the breathalyzer is not affected by the particular background frequencies present at the time the RFI test program is conducted does not, in his opinion, eliminate the possibility of inaccurate results at other times resulting from radio waves transmitted at other frequencies or power levels. He opined that, because of the transient nature of radio frequency environments, a simulator test conducted following the performance of a breathalyzer test on a

Page 17

suspect would not necessarily indicate whether the instrument was affected by RFI at the time the suspect's breath sample was analyzed.

Stating that he would believe no single reading, the expert noted the reliability of Model 900A tests could be improved by performing, in addition to the simulator test, two separate tests of the suspect's breath and comparing the results for consistency. The 900A can be immunized from susceptibility to RFI, he said, if it is subjected to a procedure called the "Minnesota modification" or "hardening." That procedure, which has been approved by Smith & Wesson, involves the insertion of RFI filter components and the sealing of openings in the instrument at a cost estimated by the maintenance company owner at under $100.

The defendant does not dispute that the Model 900A Breathalyzer and other scientific instruments that measure blood alcohol content on the basis of breath samples have for some time been deemed to satisfy this Commonwealth's "general acceptance" standard for admissibility of scientific evidence, [Note 17] Commonwealth v. Fatalo, 346 Mass. 266, 269 (1963). See Commonwealth v. Whynaught, 377 Mass. 14, 17 (1979). See also Commonwealth v. Kater, 388 Mass. 519, 527 (1983); Commonwealth v. Vitello, 376 Mass. 426, 441-442 (1978); Commonwealth v. Lykus, 367 Mass. 191, 196 (1975); Commonwealth v. A. Juvenile, 365 Mass. 421, 425 (1974). The defendant contends instead that the recent discovery of the Model 900A's susceptibility to RFI requires reconsideration of the admissibility of examination results obtained from Model 900A units.

The body of knowledge regarding the precise effects of RFI on the accuracy of the Model 900A is not yet well developed. Literature on the topic in scientific and other journals is sparse, as are judicial decisions treating the subject. [Note 18] Compare Commonwealth

Page 18

v. Lykus, supra at 198. See Commonwealth v. Kater, 388 Mass. 519, 527 (1983). The pertinent information in the record consists principally of Smith & Wesson's test results establishing the susceptibility of the 900A to RFI at specific distances, frequencies, and power outputs in what a Smith & Wesson product manager described as "the worst possible [RFI] environment [to which] we could subject the instrument," and the opinion of the defendant's expert that no single blood alcohol content reading from an "unhardened" 900A unit is reliable. Because general acceptance of scientific evidence by the relevant body of scientists establishes the admissibility of such evidence, the opinion of one expert that scientific evidence has become unreliable does not by itself require exclusion of the evidence. We note that the defendant's expert stated that, notwithstanding potential RFI susceptibility, results obtained from the Model 900A would be reliable in most instances. See Romano v. Kimmelman, 96 N.J. 66, 82 (1984). See also Commonwealth v. Lykus, supra at 201; United States v. Williams, 583 F.2d 1194, 1198 (2d Cir. 1978), cert. denied, 439 U.S. 1117 (1979) (potential rate of error a factor in deciding admissibility of scientific evidence). On this record, we cannot conclude that the judge below erred in rejecting the defendant's argument that the Model 900A Breathalyzer is no longer generally accepted by the scientific community. [Note 19]

We believe, nonetheless, that the evidence presented regarding the potential effect of RFI on Model 900A units is sufficient

Page 19

to require that the admission of the result of a breath test performed on a 900A instrument be conditioned on a demonstration to the judge of the accuracy of the particular unit at the time the test was performed. See Commonwealth v. Whynaught, 377 Mass. 14, 18 (1979). We reject the Commonwealth's suggestion that the question whether a specific breathalyzer was rendered unreliable by the potential effects of RFI should be resolved by the fact finder on the basis of evidence introduced in the course of the trial. In deciding "whether judges (at least initially) or finders of fact should sort out any controversy over the acceptability" of scientific evidence, Commonwealth v. A Juvenile, 381 Mass. 727, 731 (1980), we have often emphasized the desirability of a judicial determination in order to avoid "the danger that on the introduction of such evidence a trial could descend into a battle of experts on the probative value of the . . . test rather than a determination of the guilt or innocence of a defendant." Commonwealth v. Fatalo, 346 Mass. 266, 269 (1963). See Commonwealth v. Kater, 388 Mass. 519, 526 (1983). Were we to leave to the jury the highly technical issue of the reliability of the Model 900A, "[t]he end result, in all likelihood, would be confusion instead of enlightenment." Commonwealth v. Fatalo, supra.

We therefore hold that, in any trial involving a Model 900A Breathalyzer test administered after the date of this opinion, and in any pending case in which an objection to the admission of a Model 900A breath test result based on potential RFI susceptibility has been made and properly preserved, the admissibility of such a result is contingent on presentation by the Commonwealth of an adequate foundation establishing that the instrument used was not so susceptible to RFI as to create a significant risk that the result was inaccurate on that basis. Although we do not at this time specify an exclusive method, see Commonwealth v. Whynaught, supra at 19, we believe a showing that the breathalyzer used was, at the time of the examination, "hardened" according to the procedure approved by Smith & Wesson, would satisfy the foundation requirement. A second test of the suspect's breath, conducted after a correct simulator reading, would, if corroborative of the initial test,

Page 20

also suffice as a foundation. Cf. Romano v. Kimmelman, supra at 87. At a minimum, the Commonwealth should be prepared to demonstrate that RFI testing procedures, sufficient under the terms of the Smith & Wesson advisory to establish nonsusceptibility to RFI at the time the result in question was obtained, have been followed. Cf. id. at 87-88 & n.6.

In denying the defendant's motion to suppress, the judge found that the breathalyzer operator had tested the RFI susceptibility of the department's breathalyzer in compliance with the Smith & Wesson directive. The judge also found that these tests determined the unit was not susceptible to RFI. The judge decided that evidence relating to the breathalyzer's susceptibility to RFI, although admissible to impeach the breathalyzer result, did not require exclusion of the result as a matter of law. [Note 20] We believe the judge properly allowed the introduction of the test result in this case.

The record supports the judge's finding that the unit was tested in compliance with the Smith & Wesson advisory when the advisory was received. The test records indicate no RFI effects at the frequencies commonly used in the police department. It would be better practice to test the central transmitter at all police broadcasting frequencies, but we cannot say that the failure to test the effect on the breathalyzer of the rarely used channels vitiated the testing procedure.

Although the record suggests the possibility that the instrument's RFI sensitivity may have been altered by the repair and calibration performed between the date of the defendant's examination and the January, 1983, RFI testing procedure, the

Page 21

judge was entitled to rely on the independent RFI testing conducted during the calibration session preceding the date of the defendant's examination. That independent procedure was limited to checking the effects of the central transmitter on the breathalyzer, but, in light of testimony that police cruiser transmitters generally are not used in the immediate vicinity of the station, and given the dearth of evidence that any walkie-talkies were in operation within twenty-five yards of the breathalyzer at the time of the defendant's test (1:12 A.M.), the judge may justifiably have treated as negligible the likelihood that the defendant's test result was affected by active RFI. Nor was it clear error for the judge, giving consideration to the potential shielding effect of the instrument's location, [Note 21] to discount the possibility raised by the defendant's expert that a transient "background" frequency deflected the null meter at the precise moment of the defendant's examination but disappeared prior to the simulator test. [Note 22] Although the Commonwealth did not conclusively establish that the defendant's test result was not influenced by RFI, we do not require a demonstration of infalibility as a precondition to consideration of scientific evidence by a trier of fact. Commonwealth v. Lykus, 367 Mass. 191, 198 (1975). We think the likelihood of an inaccuracy caused by RFI was sufficiently diluted by the evidence at the hearing to warrant the judge's conclusion that exclusion of the result as a matter of law was not required.

Page 22

4. Reliability of single breathalyzer test. The defendant asserts that all parties would be better served if the Commonwealth abandoned its single-test approach in favor of a two-test system that would yield a second blood alcohol content reading for comparison with the result of the initial test. The Commonwealth concedes that the administration of two breath tests would be better practice. We have indicated above that a corroborative second test would provide a foundation for the admission in evidence of results obtained from breathalyzers subject to attack on the basis of RFI susceptibility, and agree with the Commonwealth that a two-test system is better practice. At this time, however, G. L. c. 90, Section 24 (1) (e), does not condition the admissibility of breathalyzer results on the availability of a corroborative test.

The routine performance of a simulator test after each breath test, and the statutory right to an independent test, provide safeguards against the conviction of a driving under the influence suspect on the basis of an inaccurate result. We also note that, unlike statutes in other jurisdictions that confer conclusive effect on a breathalyzer result, under our statute the result is "merely evidence of the defendant's being under the influence of liquor," Commonwealth v. Moreira, 385 Mass. 792, 795 (1982), to be considered in conjunction with other evidence. Compare Cal. Vehicle Code Section 23152(b) (West Supp. 1984), conclusively establishing guilt, without showing of impairment, if the test results equals or exceeds 0.10. People v. Trombetta, 142 Cal. App. 3d 138, 141 (1983). We conclude that the use of the defendant's test result in this case did not deprive the defendant of a fair trial. [Note 23]

Judgment affirmed.