concurring.
I write separately on the admissibility of the testimony of the hypnotized witnesses to emphasize why reliance on this evidence presents such difficult problems in this case and to suggest possibilities for avoiding some of these difficulties in the future. Although I think the question is exceedingly close and difficult, I agree with the majority that the admission of the testimony of the hypnotized witnesses does not require reversal. Nevertheless, the district court was presented with an unusually taxing problem, and I think we should take this occasion to propose standards that may be applied in determining whether to admit the testimony of witnesses who have been hypnotized before trial.
The majority correctly identifies the very serious dangers of using hypnotically induced testimony. The sources cited by the majority provide a thorough documentation of those hazards and I shall not attempt to explore them here in detail. Perhaps, the most pervasive (though not necessarily the most acute) danger of hypnosis is to make witnesses confident of their recall of matters they might otherwise feel themselves to be guessing at or even be sure they could not recall.
A persuasive case can be made for a rule of per se inadmissibility, and this is the route a number of states have taken. On the other hand, the Eighth Circuit has made the important point that a per se rule could be overly sweeping and might result in the exclusion of valuable and reliable evidence in some cases. See Sprynczynatyk v. General Motors Corp., 771 F.2d 1112, 1122 (8th Cir.1985), cert. denied, — U.S. -, 106 S.Ct. 1263, 89 L.Ed.2d 572 (1986). As a matter of fact, all of the federal courts of appeals that have so far considered the issue have declined to adopt a rule of per se inadmissibility. See Wicker v. McCotter, 783 F.2d 487, 492 (5th Cir.), cert. denied, — U.S. -, 106 S.Ct. 3310, 92 L.Ed.2d 723 (1986); Clay v. Vose, 771 F.2d 1 (1st Cir.1985), cert. denied, — U.S. -, 106 S.Ct. 1212, 89 L.Ed.2d 324 (1986); United States v. Adams, 581 F.2d 193, 198 (9th Cir.), cert. denied, 439 U.S. 1006, 99 S.Ct. 621, 58 L.Ed.2d 683 (1978). But at the very least, we should follow generally the path of the Eighth Circuit and adopt a rule that requires the district court, in cases in which hypnosis has been used, to conduct pretrial hearings on the procedures followed during the hypnotic session in question. The district court should assess the effect of hypnosis upon the reliability of the testimony before deciding on admissibility. Sprynczynatyk, 771 F.2d at 1122. The proponent of the evidence bears the burden of persuading the district court that the proposed testimony is sufficiently reliable and that its probative value outweighs its prejudicial effect. Id. at 1123. The district court must exercise great care to *255ensure that any statements made after hypnosis that are admitted into evidence are the product of the subject’s own recollections, rather than confabulation or recall tainted by suggestions received while under hypnosis.1 See United States v. Solomon, 753 F.2d 1522, 1525 (9th Cir.1985); United States v. Valdez, 722 F.2d 1196, 1200-04 (5th Cir.1984); United States v. Adams, 581 F.2d 193, 198-99 (9th Cir.), cert. denied, 439 U.S. 1006, 99 S.Ct. 621, 58 L.Ed.2d 683 (1978). Unless a showing to this effect can be made, the statements must be excluded.
In making its determination of admissibility, the district court should at least consider: 1) the appropriateness of using hypnosis for the kind of memory loss involved; see Sprynczynatyk, 771 F.2d at 1123; 2) the extent to which information concerning the events in question may be communicated intentionally or inadvertently to the subject; 3) whether there is any evidence to corroborate the hypnotically enhanced testimony, see Sprynczynatyk, 771 F.2d at 1123. Further, the district court should consider the extent to which the safeguards articulated in State v. Hurd, 86 N.J. 525, 545-47, 432 A.2d 86, 96-97 (1981) and Sprynczynatyk, 771 F.2d at 1122-23, were followed.2
One almost indispensable safeguard is to require that the hypnotist first elicit a free narrative report from the subject before questioning that subject more specifically. This procedure is important because information elicited in a free narrative is less likely to be confabulated.
It may not be crucial that testimony always be excluded if the Hurd and Sprync-zynatyk safeguards were not employed. Neither does it necessarily follow that testimony should always be admitted if these safeguards were observed. I believe, however, that the safeguards provide some minimal guidelines upon which the district court, faced with evaluating these elusive problems, may rely. Of course, the district court need not limit itself to the consideration of the factors identified in the guidelines but should consider all circumstances that may affect the reliability of the testimony in light of current scientific knowledge about the effectiveness and hazards of hypnosis.
In the present case, although there was, of course, a full pretrial hearing, few, if any, of the requirements of the guidelines had been followed in connection with the hypnosis sessions. It is therefore almost impossible for us to evaluate — certainly on a cold record — the reliability of the testimony affected by hypnosis. It is almost equally difficult to know with any assurance what role this testimony played in the thinking of the jury. Nonetheless, the majority has made as searching an inquiry as seems possible under the circumstances and I concur in its ultimate conclusion.
There are other substantial problems in this trial including showing the jury mug shots of the defendant taken in connection *256with another charge. The jury also learned in one way or another about Kimberlin’s connection with a variety of other “bad acts” like possession of the uniform of a Defense Department security officer, participation in a drug conspiracy and possession of weapons. Many of these matters, taken individually, may not be in themselves of crucial significance. But the cumulative impression on the jury raises real questions of prejudice. Although these and a number of other problems are very troubling, I think the majority has addressed them conscientiously and in a fashion that sustains the result.
. In this connection, the courts must, of course, exercise the highest degree of caution in criminal cases in order to protect the rights of the accused. See State v. Hurd, 86 N.J. 525, 547 n. 6, 432 A.2d 86, 97 n. 6 (1981).
. These are recited in the majority opinion but restated here in a slightly more detailed form:
(1) The hypnotic session should be conducted by a licensed psychiatrist or psychologist trained in the use of hypnosis and aware of its possible effects on memory so as to be able to aid in the prevention of improper suggestions and confabulation.
(2) The qualified professional conducting the hypnotic session should be independent of either party and should have little investment in the ultimate disposition of the case. The qualified professional should have minimal preconceptions about the case.
(3) Any information given to the hypnotist by either party should be noted in writing so that subsequently the extent of information that the subject received from the hypnotist may be determined.
(4) Before hypnosis, the hypnotist should obtain a detailed description of the facts from the subject, avoiding adding new elements to the subject’s description.
(5) The session should be recorded, and preferably videotaped, so that a permanent record is available to ensure against suggestive procedures.
(6) Only the hypnotist and the subject should be present during any phase of the hypnotic session.