WIEAND, Judge.

Gary D. Moore was tried by jury and was found guilty of homicide by vehicle while driving under the influence of alcohol, homicide by vehicle, driving while under the influence of alcohol, and related summary offenses. Post-trial motions were denied, and Moore was sentenced to serve a term of imprisonment for not less than three (3) years nor more than six (6) years. On direct appeal from the judgment of sentence, Moore challenges evidentiary rulings made by the trial court as well as the court's jury instructions.

The evidence at trial was that during the afternoon and evening of January 19, 1990, Moore had been drinking beer with his uncle, Dwayne Canipe, and his uncle's friend, Mark Pierson. Starting their drinking at Canipe's apartment, the men stopped at a local bar shortly after 4:00 p.m. and then returned to Canipe's apartment complex, where they spent several hours with friends who lived in other apartment units. Sometime after 8:00 p.m., Canipe took Moore's keys, and later suggested that Moore spend the night, because it had begun sleeting and Moore had been drinking. Canipe last saw Moore at approximately 1:00 a.m. on January 20.

Shortly after 4:00 a.m. that morning, Timothy Snowden came upon the scene of an automobile accident while travelling north on Route 41 in Chester County. He stopped his car and proceeded toward the accident on foot. Hearing cries for help, he saw Gary Moore lying partially inside a van, which was one of the vehicles which had been involved in the collision. According to Snowden's testimony at trial, when he helped Moore out of the van he detected an odor of alcohol. Snowden also examined the other vehicle, whose driver, Lawrence DeClue, had been killed in the collision. Snowden then stopped a passing truck, whose driver called for police assistance on his radio. While directing traffic around the accident scene, Snowden lost sight of Moore. Moore was next spotted by Christopher Wilson, whose car had been stopped in traffic heading south on Route 41. Wilson testified that Moore staggered to Wilson's car, with face and knee covered with blood, and asked Wilson for a ride. Wilson allowed Moore to get into his car and then drove north on Route 41 to a convenience store, where he waited for medical assistance. Wilson also testified that he had noticed an odor of alcohol from Moore while in the car. An ambulance arrived shortly thereafter and took Moore to Southern Chester County Medical Center.

State Police Trooper John W. Laufer came to the accident scene at approximately 4:30 a.m. He observed that the van and the other vehicle, a Cadillac, had been involved in a head-on collision. It appeared that the van had been proceeding north and had crossed the center line of the highway, striking the south-bound Cadillac. Trooper Laufer went to the hospital to question Moore, where he told Moore that he was not under arrest but that Laufer and Sergeant Jeffrey Miller wished to ask him questions regarding the accident. Moore was also advised of his Miranda rights. He stated that he had left the Kennett Square area at approximately 11:30 p.m. in his father's van, and that the next thing he had recalled was being picked up by Christopher Wilson. Trooper Laufer smelled alcohol on Moore's breath and observed that Moore's eyes were glassy and bloodshot. Moore told him that he had had three to four beers at his uncle's apartment, but had stopped drinking at approximately 10:00 p.m. Laufer administered a horizontal gaze nystagmus test, which Moore failed. Moore refused to give a blood sample, because blood had already been taken by hospital personnel and he did not wish to have more drawn. On January 22, after he had been released from the hospital, Moore was arrested. Police subsequently obtained and served a subpoena for the production of Moore's medical records at the preliminary hearing. At the same time, police requested and obtained copies of portions of the medical records from hospital personnel.

Appellant's first argument is that the police seizure of his hospital records was improper and that information derived therefrom should have been suppressed. In reviewing the denial of a defense motion to suppress, an appellate court will

consider only the evidence presented by the prosecution, and so much of the evidence for the defense as, fairly read in the context of the record as a whole, remains uncontradicted. Assuming that there is support in the record for the suppression court's findings, we are bound thereby and may reverse only if the legal conclusions drawn from those facts are in error.

Commonwealth v. Pestinikas, 421 Pa. Super. 371, 390, 617 A.2d 1339, 1349 (1992) (en banc).

Here, the suppression court found that a subpoena had been validly issued for appellant's medical records. It had been duly issued to compel hospital personnel to bring Moore's records to the preliminary healing. When it was served upon a secretary in the records department of the hospital, however, the state trooper serving the subpoena requested an immediate copy which he could take with him. The secretary thereupon prepared and delivered to the trooper copies of appellant's emergency room report, lab report and x-ray report. This enabled police to review the records prior to the preliminary hearing.

This is not a case in which a subpoena was used to obtain copies of appellant's hospital records before any charges had been filed. Compare: Commonwealth v. Jolly, 337 Pa. Super. 130, 486 A.2d 515 (1984) and Commonwealth v. Haynos, 363 Pa. Super. 1, 525 A.2d 394 (1987), allocatur denied, 517 Pa. 604, 536 A.2d 1329 (1987). This case involves a subpoena duces tecum validly issued to require the presence of hospital records for a preliminary hearing. The subpoena, moreover, was not issued as part of a "fishing expedition" but upon probable cause to believe that appellant's hospital records contained relevant information pertaining to appellant's physical condition at the time of the fatal accident. Based upon observations and information given by witnesses, police were aware that appellant had been involved in a serious accident in which his vehicle had crossed into the lane for oncoming traffic and that alcohol had been detected on his breath. This constituted probable cause to believe that a criminal offense had been committed. Commonwealth v. Haynos, supra; Commonwealth v. Pelkey, 349 Pa. Super. 373, 378-379, 503 A.2d 414, 416 (1985).

Under these circumstances, we conclude, as did the trial court, that the police use of a subpoena to compel the production of appellant's medical records for the preliminary hearing did not violate any constitutionally protected right of privacy which appellant possessed in his medical records. Cf. Commonwealth v. Hipp, 380 Pa. Super. 345, 551 A.2d 1086 (1988).

Appellant also contends that the trial court committed error when it allowed evidence of the results of blood serum tests to be received in evidence. An identical argument, however, was rejected by the Superior Court in Commonwealth v. Dagnon, 413 Pa. Super. 281, 605 A.2d 360 (1992).

At trial, a toxicologist testified that appellant's blood alcohol level at the time of testing was 0.14%. He also testified, however, that without knowing when appellant had last consumed alcohol he could not state conclusively the blood alcohol level of appellant at the time of the accident two hours earlier. Appellant now argues that his blood level at the time of testing was irrelevant and should not have been received in evidence. At trial, however, appellant's objection to the evidence was withdrawn; and in post-trial motions the issue was not raised. Under these circumstances the relevancy of the evidence as an issue was waived and has not been preserved for appellate review. Commonwealth v. Clair, 458 Pa. 418, 422-423, 326 A.2d 272, 274 (1974); Commonwealth v. Garofalo, 386 Pa. Super. 363, 367, 563 A.2d 109, 111 (1989). The waiver is not nullified merely because the trial court may have discussed the relevancy of the evidence in its opinion denying post-trial relief. Because the evidence had been received, moreover, the prosecuting attorney could properly refer to it during closing argument.

Appellant's principal argument and the most difficult issue in this appeal concerns the trial court's ruling which allowed the jury to hear and consider testimony that appellant had failed a horizontal gaze nystagmus test. The testimony was improper, it is argued, because the evidence failed to establish that the test had been generally accepted by the scientific community.

In Commonwealth v. Topa, 471 Pa. 223, 369 A.2d 1277 (1977), the Supreme Court adopted language from Frye v. United States, 54 App.D.C. 46, 293 F. 1013 (1923), and said:

Admissibility of the evidence depends upon the general acceptance of its validity by those scientists active in the field to which the evidence belongs[.]

"Just when a scientific principle of discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs." Frye v. United States, 54 App.D.C. 46, 293 F. 1013, 1014 (1923) (emphasis supplied).

Id. 471 Pa. at 231, 369 A.2d at 1281. See also: Commonwealth v. Dunkle, 529 Pa. 168, 602 A.2d 830 (1992); Commonwealth v. Middleton, 379 Pa. Super. 502, 550 A.2d 561 (1988).

HGN, according to the testimony in this case, "is a test which relates to the detection of an involuntary jerking of the eyes, which [is] accentuated by the introduction of alcohol or other central nervous system depressants into the body." "Horizontal gaze nystagmus is an involuntary, rapid oscillation of the eyes which occurs when a person looks to the side at an object, and is characterized by an involuntary pendular (back and forth) jerking movement of the eye. The field test is a measurement of the angle of onset of this jerking movement when the eye tracks a steadily moving object such as a pencil or pen-sized flashlight. . . . [T]he HGN test presumes that through measuring the point at which the subject's eyes begin jerking, a rough approximation of blood alcohol content can be determined." 4 Am Jur Proof of Facts 3d 439, 446.

In some jurisdictions, the results of an HGN test are admissible without the need for further scientific evidence. See: State v. Superior Court, 149 Ariz. 269, 718 P.2d 171 (1986) (HGN test satisfies Frye standard and is admissible to corroborate or attack, but not to quantify, the presence of blood alcohol); State v. Murphy, 451 N.W.2d 154 (Iowa 1990); State v. Clark, 234 Mont. 222, 762 P.2d 853 (1988); State v. Nagel, 30 Ohio App.3d 80, 506 N.E.2d 285 (1986); Finley v. State, 809 S.W.2d 909 (Tex.App. 1991). See also: State v. Garrett, 119 Idaho 878, 811 P.2d 488 (1991); State v. Armstrong, 561 So.2d 883 (La.App. 2 Cir. 1990). Other courts, however, have held that the results of HGN tests are not admissible without a showing of the validity of the underlying scientific principles. See: Ex parte Malone, 575 So.2d 106 (Ala. 1990); State v. Witte, 251 Kan. 313, 836 P.2d 1110 (1992); State v. Borchardt, 224 Neb. 47, 395 N.W.2d 551 (1986); State v. Reed, 83 Or.App. 451, 732 P.2d 66 (1987); State v. Wheeler, 764 S.W.2d 523 (Mo.App. 1989); State v. Barker, 179 W. Va. 194, 366 S.E.2d 642 (1988).

The issue has not yet been considered by the Supreme Court of Pennsylvania. On two occasions, however, the admissibility of HGN testing has been before the Superior Court, which has refused to allow the results of HGN testing without first establishing an adequate foundation. This foundation, the Court said in Commonwealth v. Miller, 367 Pa. Super. 359, 532 A.2d 1186 (1987), is not established without proof "that the scientific principles upon which the test is based are generally accepted by scientists active in the appropriate scientific communities." Id. at 365-366, 532 A.2d at 1189. Later, in Commonwealth v. Apollo, 412 Pa. Super. 453, 603 A.2d 1023 (1992), allocatur denied, 531 Pa. 650, 613 A.2d 556 (1992), the Court held that the trial court had not abused its discretion where, despite testimony of a behavioral optometrist, it concluded from other evidence that the reliability of the horizontal gaze nystagmus test was not a settled proposition within the scientific community and ruled the evidence inadmissible. The testimony of the optometrist that the test was reliable, the Court observed, reflected his personal views and not necessarily those of the general scientific community.

In the instant case, the Commonwealth relied on the testimony of Michael F. Wahmann, a county detective who had been certified by the Department of Transportation as an instructor in the horizontal gaze nystagmus test. He testified that the principles upon which the test is based are generally accepted in the optometric and police science communities. He conceded, however, that there were differences of opinion regarding the reliability of HGN test results among ophthalmologists and in the medical community. On the basis of this evidence, the trial court found that the test was "acceptable in the police and also some portions of the scientific community" and held the test results to be admissible.

It may be that the Supreme Court, if it were to consider the issue, would hold that HGN testing meets the standards of Frye and Topa. At the present time, however, the present state of the law is as stated by the Superior Court in Miller and Apollo. Under those decisions, the testimony of County Detective Wahmann did not provide an adequate basis for finding that HGN testing had gained general acceptance in the scientific community, particularly in the field of medical science represented by ophthalmology. In this case, therefore, the results of an HGN test should not have been received. To correct this error, a new trial is necessary.

The error was not harmless. Appellant was subjected to a blood test approximately two hours after the accident. The result of the test was a finding of a blood alcohol content of 0.14 percent. As we have observed, however, the toxicologist who testified as a Commonwealth witness conceded that he could not estimate the alcoholic content of appellant's blood at the time of the accident. Under these circumstances, we cannot determine beyond a reasonable doubt that the improperly received result of the HGN testing did not contribute to the jury's finding. See: Commonwealth v. McGrogan, 523 Pa. 614, 568 A.2d 924 (1990); Commonwealth v. McGinnis, 511 Pa. 520, 515 A.2d 847 (1986).

We find no merit in appellant's arguments that there were improprieties in the prosecuting attorney's closing argument and in the trial court's jury instructions. Because the case must be retried, moreover, the arguments of counsel and the instructions of the trial court may well be different at the next trial. Therefore, we find no reason to provide an extensive discussion of these issues.

Reversed and remanded for a new trial. Jurisdiction is not retained meanwhile.