dissenting in part. Justice Pashman’s opinion for the Court is a clear, explicit, and much-needed guide to the procedures to be employed with respect to automatic judicial review of continued restraints of UGI committees. It receives my enthusiastic endorsement with but two exceptions. Both points of difference go to the vexing question of burden of proof, an issue which has separated me from my colleagues since first we entered these troubled waters. See *315State v. Carter, 64 N. J. 382, 410, 423-27 (1974) (concurring and dissenting opinion) and State v. Krol, 68 N. J. 236, 267-77 (1975) .(dissenting opinion).

First, I disagree with the following from the opinion of the Court, ante at 299-300:

Just as the State bears the burden of persuasion by a preponderance of the evidence on the necessity for the initial imposition of such restraints, it must similarly reestablish its authority to restrict the liberty of the committee by showing that his present condition warrants their continuance.

[Emphasis added, footnote omitted.]

All the reasons for my favoring a “beyond a reasonable doubt” rule over one calling for only a “bare preponderance of the evidence” were set forth in my Krol dissent, supra. My manifestly inadequate supply of eloquence and the apparently limited persuasiveness of my arguments having proved unavailing there, I simply record here my continuing view that the State should be required to prove beyond any reasonable doubt the necessity to restrict one’s liberty on account of mental illness.

In the same vein I disagree further with so much of the majority opinion as allocates to the committee the burden of proof on the committee’s application, between periodic reviews, for a change in restrictions. In this limited regard I agree with Judge Conford that after the committee has met his burden of going forward with some evidence to support his challenge to the restrictions previously imposed on him, the ultimate burden of persuasion as to the extent to which any restrictions should be continued should rest with the State — and this beyond a reasonable doubt. I adopt the reasoning of that part of Judge Conford’s concurring opinion dealing with this issue.

In doing so I recognize a plain inconsistency with my position in State v. Carter, supra, wherein I advocated imposing on an applicant for conditional release the burden of demonstrating by a bare preponderance (as opposed to the ma*316jority’s then requirement of “clear and convincing” proof) his non-dangerousness, 64 N. J. at 424 — a position later adopted by the Court in State v. Krol, supra, .68 N. J. at 263, n. 13. I now think my position in Carter did not go far enough. Because of the emphasis I would place on a committee’s constitutionally protected liberty interest as set forth in Krol, supra, 68 N. J. at 271-77, I would require the State to bear its burden of demonstrating beyond a reasonable doubt the committee’s dangerousness to himself or others or property, at every stage of commitment or review thereof, and would in no instance impose on the committee the burden of proving the converse.

In all other respects I join fully in the opinion of the Court.

Confoed, P. J. A. D., concurring in the result.

For reversal and remandment — Chief Justice Hughes, Justices Sullivan, Pashman, Scheeibee and Handles and Judge Confoed — 6.

Dissenting in part — Justice Cliffoed — 1.

The opinion of the court was delivered by

In State v. Knot, 68 N. J. 236 (1975), this Court announced certain interim rules to govern the disposition of persons acquitted of criminal charges by reason of insanity pursuant to N. J. S. A. 2A:163-2 and 3. We authorized the automatic temporary commitment of an insanity-acquittee for up to sixty days for the purpose of a psychiatric evaluation of his current mental condition and propensity for future antisocial conduct. See 68 N. J. at 256. *289We further authorized the State to seek, within that period, tire indefinite imposition of restraints on the insanity-aequittee’s liberty because of his alleged mental illness and potential dangerousness to himself or to society. We required a judicial hearing on the State’s request at which the State must satisfy the court, by a preponderance of the evidence, that the insanity-acquittee is presently mentally ill (as that term is defined in N. J. S. A. 30:4-23) and is likely to pose a danger to himself or to society. If the State’s burden is met, the court must order “suitable restraints” placed upon the insanity-acquittee’s liberty in order to “protect the public and provide defendant with appropriate treatment.” 68 N. J. at 257. The suitability of the restraints is determined by the court’s assessment of the degree of restraint upon the liberty of the insanity-acquittee necessary to “reduce the risk of danger which he poses to an acceptable level.” 68 N. J. at 261. We provided for the modification or termination of orders imposing such restraints upon a showing, again by a preponderance of the evidence, by the party seeking to change the status quo that the degree of dangerousness presently posed by the committee warrants a different level of restraints or even none at all. 68 N. J. at 263 and n. 13.

Appellant Hetra Fields was indicted for the stabbing murder of her boyfriend and at her subsequent jury trial was acquitted by reason of her insanity at the time of the commission of the offense. Pursuant to the procedures set forth in Krol, the trial judge ordered her temporary confinement for observation in an appropriate psychiatric facility. Upon completion of the psychiatric evaluation, a hearing in conformity with Krol’s directives was held to determine what restraints were required under the Krol criteria. The court determined that Fields was presently mentally ill, and constituted a danger to herself and society. Accordingly the judge ordered her continued confinement at a psychiatric hospital until his further order. At a commitment review hearing conducted six months later the court made an iden*290tical determination and disposition, although, authorizing the grant of furloughs in the discretion of the hospital staff.

A second such hearing was held six months thereafter at which the court found that the committee’s mental illness .(schizophrenia, chronic, undifferentiated type) was incurable. The reviewing judge concluded that although Fields was in a state of remission, she remained a probable danger to herself and society and continued to be in need of “controlled supervision.” The court ordered her continued commitment under the same terms as provided in its previous order, subject to further review one year later, and directed that the hospital take certain steps aimed at her eventual conditional release. The committee appealed from this order and we directly certified the appeal on our own motion, B. 2:12-1, while it was pending unheard before the Appellate Division, 75 N. J. 588 (1977), in order to resolve certain important questions not settled by our decision in Krol and not yet resolved by legislative action.1

The main subject of controversy at the hearing below was the proper allocation of the burden of proof, both as to production of evidence and risk of non-persuasion, in court-*291ordered review proceedings. Counsel for Fields argued that under Krol, persons found not guilty by reason of insanity who are subsequently committed (hereinafter designated as “UGI committees”) must be treated the same as persons whose commitment did not result from their involvement in the criminal justice system, “civil” committees. Under B. 4:74r-7(e), (f) and (g), which govern the initial entry and automatic periodic review of orders for civil commitment and conditional release, that burden is placed upon the State, as the party seeking the continued imposition of restraints upon the liberty of the committee, at both stages in the civil context.2 Thus, counsel for Fields argued that *292the burden at the review stage of proceedings involving an NGI committee must remain where Knot had placed it at the initial commitment stage • — ■ upon the State. The court rejected this argument, ruling that neither party bore the burden of persuasion at a periodic review hearing held at the initiative of the court. He distinguished such a hearing, in terms of the proper allocation of the burden of persuasion, from review proceedings initiated by the committee or by the state at which, under Krol, the burden is placed upon the moving party. See ante at 288-290. The judge stated that the periodic review proceeding required proof “to the satisfaction of the court” that the committee’s mental illness and dangerousness continued. The court further ruled that the State was not required to produce fresh psychiatric testimony at each court-ordered periodic review hearing but could rely instead on the proofs previously adduced in support of the continuation of the commitment.

The only new evidence presented at the hearing was testimony on behalf of the committee by a senior psychiatrist at the hospital to which she had been committed. The psy*293chiatrist testified that the committee was in complete remission and did not require further hospitalization and indeed would be detrimentally affected by it. He stated that future violent psychotic episodes were unlikely so long as Eields was able to stay away from alcohol. The reviewing court was nevertheless concerned with the report of such an episode on one of the committee’s weekend furloughs where, while under the influence of alcohol, she had threatened one of her daughters with a knife. The judge attributed her state of remission to the structured institutional environment and feared that exposure to the stress of the non-institutional world coupled with availability of alcohol might lead to another violent reaction. He accordingly refused to approve her conditional release at that time.

The narrow question before us is the entitlement of NGI committees to automatic periodic judicial review of the validity of the continued restraints upon their liberty on the basis of their dangerousness to themselves or others by reason of mental illness. We shall also address the ancillary questions of the allocation of the burden of persuasion on that issue and the appropriate standard and type of proof to be required.

We hold that NGI committees possess the same right to automatic periodic review of the justification for their commitment ,(or lesser restraints, as the case may be) as that enjoyed by civil committees. We further hold that the State must bear the ultimate burden of proof in justifying any continued restrictions upon the liberty of NGI committees at each periodic review proceeding by establishing, by a preponderance of the evidence, that such restrictions currently meet the criteria set forth in Xrol for the initial imposition of restraints.

I

Entitlement to■ Periodic Review

Our decision in Xrol spoke broadly of the constitutional requirement of substantially identical treatment for *294all persons subject to a loss of liberty because of their probable dangerousness due to mental disability. In view of the comprehensive treatment this issue received in Krol, we need not retrace our steps- here. See 68 N. J. at 250-255. Simply stated, the fact that a mentally ill person has committed an act which would expose a mentally competent person to criminal sanction is a constitutionally unacceptable justification for granting him less procedural and substantive protection against involuntary commitment than that generally afforded all -other members of society. Jackson v. Indiana, 406 U. S. 715, 724, 92 S. Ct. 1845, 32 L. Ed. 2d 435 (1972); Baxstrom v. Herold, 383 U. S. 107, 86 S. Ct. 760, 15 L. Ed. 2d 620 (1966). Krol mandated such parallel treatment for prospective NGI committees at the- initial consideration of the appropriateness of restricting their liberty for the purpose of community or personal protection ■because of their predictive dangerousness. It also indicated that the same criteria should be dispositive of the propriety of modifying or terminating orders requiring institutionalization or lesser restraints at the request of the NGI committee or the -State, as- is the case where the release of civil committees is at issue. However, Krol was silent with respect to the question -of mandatory judicial review, on a regular periodic basis, of the current validity of the restraints imposed upon NGI committees. At the time Krol was decided, no comparable provision for periodic judicial reevaluation of the need for continuation of restrictions upon the liberty of civil committees existed. Shortly thereafter, B. 4:74-7(f) and (g) became effective. Those rules mandate such automatic review for civil committees without the necessity of a request therefor.

W-e- have concluded that the significant safeguards afforded civil committees by virtue of the periodic review provisions of B. 4:7U-7(f) and (g) must be extended to NGI committees as well. We shall accordingly refer the matter to our Criminal Practice Committee for the purpose of formulating an appropriate revision of B. 3 :19-2 *295which will guarantee NGI committees equivalent protection against the unwarranted continuance of state-imposed restrictions on their liberty.

Due process would seem to require a meaningful periodic review of the continued legitimacy of restraints on the liberty of all persons whose alleged dangerousness by reason of mental disability brought about these restrictions. Their present dangerousness to themselves and others must be assessed at reasonable intervals. The standard of proof and the burden of meeting it at each periodic review hearing must be identical to that required án the initial proceeding mandated by Krol — whether the committee actually meets the criteria of mental illness and dangerousness by reason thereof necessary to validate the total or partial deprivation of his liberty by the State. See 68 N. J. at 249, 257.

In Jackson v. Indiana, 406 U. S. 715, 92 S. Ct. 1845, 32 L. Ed. 2d 435 (1972), the United States Supreme Court ruled that there were certain substantive constitutional limitations on state power to confine persons found to be mentally ill. However, the Court chose to articulate only the limitation pertinent to the case before it:

At the least, due process requires that the nature and duration of commitment bear some reasonable relation to the purpose for which the individual is committed.

[406 U. S. at 738, 92 S. Ct. at 1858]

The logical corollary of the Court’s observation is that where that purpose is the protection of society or the individual himself from harm he may inflict, the deprivation of that person’s liberty can constitutionally continue only so long as the potential for that harm remains sufficiently great so that his confinement would be warranted were his initial commitability at issue. In O'Connor v. Donaldson, 422 U. S. 563, 95 S. Ct. 2486, 45 L. Ed. 2d 396 (1975), the Court, citing Jackson v. Indiana, supra, was more explicit on this point:

*296Nor is it enough that [the committee’s] original confinement was founded upon a constitutionally adequate basis, if in fact it was, because even if his involuntary confinement was initially permissible, it could not constitutionally continue after that basis no longer existed.

[422 U. S. at 574-75, 95 S. Ct. at 2493]

Simply stated, state power to continue restrictions on the liberty of a committee terminates when his condition no longer satisfies the legal standards for their initial imposition. The authority of the state to impose restraints on a committee’s liberty is thus contingent upon the continuance of the justification therefor as measured by that individual’s present mental condition and degree of dangerousness. Due process requires a review procedure adequate to effectuate those limitations. Cf. Fasulo v. Arafeh, 173 Conn. 473, 378 A. 2d 553, 555 (1977). In Fasulo, supra, the Supreme Court of Connecticut interpreted the due process clause of that state’s constitution to require that civil committees be granted periodic judicial review “of the propriety of their continued confinement.” 378 A. 2d at 556. The Court’s analysis of O’Connor v. Donaldson’s significance in terms of the demands of due process in this context is particularly instructive:

Since the state’s power to confine is measured by a legal standard, the expiration of the state's power can only be determined in a judicial proceeding which tests the patient’s present mental status against the legal standard for confinement. *■ * [S]inee the state’s power to confine is premised on the individual’s present mental status, the original involuntary commitment proceeding can only establish that the state may confine the individual at the time of the hearing and for the foreseeable period during which that status is unlikely to change. Upon the expiration of that period, the state's power to deprive the patient of his liberty lapses and any further confinement must be justified anew. * * * The state, therefore, must bear the burden of initiating recommitment proceedings.

[Id. at 556-57 (footnote omitted)]3

*297We are in substantial accord 'with this analysis. Similar reasoning was the underlying basis of our promulgation, in response to O'Connor v. Donaldson, of a civil practice rule mandating automatic periodic judicial reviews, at reasonable intervals, of orders for commitment and conditional release of civil committees. B. 4:74-7 (f) and (g). In light of the constitutional imperative of substantially equal treatment reflected in Baxstrom and Jackson, we discern no constitutionally satisfactory justification for denying comparable protection to NGI committees. The spirit, if not the express language, of Krol forecloses any contrary contention. Cf. State v. Clemons, 110 Ariz. 79, 415 P. 2d 324 (1973).

A defendant who has avoided criminal sanction by establishing his insanity at the time he committed the offense for which he was tried stands on essentially the same legal footing, in terms of his amenability to involuntary commitment, as any other member of society.4 Justification for imposing restraints upon his liberty must be found under legal criteria which do not “deviate substantially from those applied to civil commitments generally.” State v. Krol, supra, 68 N. J. at 251. Prospective NGI committees and prospective civil committees are constitutionally indistinguishable in terms of their entitlement to procedural and substantive due process rights. As we indicated in Krol, the “labels ‘criminal commitment5 and ‘civil commitment5 are *298of no constitutional significance” in this context. 68 N. J. at 351. Although the particular mechanisms for commitment to which such persons are subject may differ, the exercise of State power to deprive them of their liberty must be initially sanctioned and subsequently reaffirmed under the same substantive test — present dangerousness by reason of mental illness.5 To underscore the identity of these two forms of involuntary imposition of restrictions on an individual's liberty due to mental disability, we direct that henceforth commitment proceedings involving prospective NGI committees shall be captioned as civil actions and entitled “In the Matter of the Commitment of-.”

Our holding that NGI committees are entitled to periodic review of the continued validity of the restraints on their liberty does not affect K^ol’s recognition of the right of such committees to seek, by an appropriate motion, seie post at 303-304, modification or termination of those restraints on the ground of a demonstrable improvement in their condition. Nor does it prevent the State from seeking an appropriate tightening of the prevailing restraints whenever it deems such action necessary because of an increase in the danger a particular individual would pose to himself or the community without such a modification. Bee generally, 68 N. J. at 363 and n. 13. The fact that automatic periodic review is provided does not mean, either here or in the context of B. 4:74-7(f) and (g), that the committee or the State must await the next regularly scheduled hearing to seek such relief as an intervening change in circumstances might warrant. Only where neither the committee nor the *299State initiates any action will periodic review be the exclusive means of judicial examination of the propriety of the present level of restraints. Periodic review is provided as an automatic safeguard against the wrongful perpetuation of a deprivation of liberty. It does not supersede the committee’s right to challenge the legitimacy of the prevailing restraints upon his liberty.

II

The Nature of Periodic Review

The central inquiry at each periodic review hearing is the current existence of a legitimate justification for the continuance of the prevailing restraints on the liberty of the committee. Just as the State bears the burden of persuasion by a preponderance of the evidence6 on the necessity *300for the initial imposition of such restraints, it must similarly reestablish its authority to restrict the liberty of the committee by showing that his present condition warrants their continuance. The Supreme Court of Connecticut has eloquently stated the compelling reasons for such an allocation of the burden of proof: *301We accordingly hold that the State must renew its authority to continue to subject a committee to a partial or total deprivation of his liberty at each periodic review hearing by demonstrating that such a deprivation is warranted by the committee’s current condition.

*300The burden should not be placed on the civilly committed patient to justify his right to liberty. Freedom from involuntary confinement for those who have committed no crime is the natural state of individuals in this country. The burden must be placed on the state to prove the necessity of stripping the citizen of one of his most fundamental rights, and the risk of error must rest on the state. Since the state has no greater right to confine a patient after the validity of the original commitment has expired than it does to commit him in the first place, the state must bear the burden of proving the necessity of recommitment, just as it bears the burden of proving the necessity for commitment.

[Fasulo v. Arafeh, supra, 378 A. 2d at 557] 7

*301At the initial judicial proceeding where the appropriateness of the imposition of any restraints on the prospective committee’s liberty is first considered, the State is required to satisfy the court both as to the need for such restraints and the suitability of the extent of the restrictions it seeks in the particular case. As we said in Krol, the initial order

* * * should be molded so as to protect society’s very strong interest in public safety but to do so in a fashion that reasonably minimizes infringements upon [the NGI committee’s] liberty and autonomy and gives him the best opportunity to receive appropriate care and treatment. * * * [T]he object of the order is to impose that degree of restraint upon [the NGI committee] necessary to reduce the risk of danger which he poses to an acceptable level. Doubts must be resolved in favor of protecting the public, but the court should not, by its order, infringe upon [the NGI committee’s] liberty or autonomy any more than appears reasonably necessary to accomplish this goal. * * *

[68 N. J. at 257-58, 261-62 (footnote omitted) (emphasis added) ]

The purpose of the first periodic review hearing is the reevaluation ' of the current suitability of the restraints then prevailing pursuant to the initial order. The reviewing court must consider any improvement or deterioration in the committee’s condition since the initial hearing, which serves to increase or decrease the danger he would pose to himself or the community under the current level of restraints. If the State perceives a need for the tightening of those re*302straints because they are presently inadequate to serve their protective function, the State must, absent consent by the committee, persuade the reviewing court of the current justification for subjecting the committee to more severe restrictions upon his liberty. If the State is satisfied that continuance of the current restraints would be adequate, it must demonstrate only that there has been no material change in the committee’s condition and degree of potential dangerousness which would warrant a relaxation of the prevailing level of restraints upon his liberty. If the State determines that the committee’s current condition would permit some relaxation of those restraints, it may either stipulate with the committee to a relaxation to that extent8 or prove that while some relaxation is in order, the continuance of some restraints remains necessary. The inquiry will be essentially identical at each subsequent periodic review proceeding. The order ultimately entered after each periodic review establishes the validity of the restraints imposed therein until the next periodic hearing (unless challenged prior to that time by the committee, see infra.).

If at any periodic review proceeding the State is unable to meet its burden of justifying the continuance of the currently prevailing restraints upon the liberty of the committee, it becomes the task of the reviewing judge again to “mold” an appropriate order based upon his evaluation of the level of restraints dictated by the committee’s present condition. KroVs guidelines, quoted above, for the formulation of the initial order remain applicable. The mere failure of the State to prove the necessity of continuing the prevailing restraints does not entitle the committee to relaxation of those restraints to any extent he might desire. The new order should provide for the least restrictive restraints which *303are found by the judge to be consistent with the well-being of the community and the individual. The determination of the suitable level of restraint is a matter entrusted to the sound discretion of the reviewing judge based on his firsthand evaluation of the particular case and is one as to which he must be accorded a wide range of flexibility. However, even where the committee’s condition shows marked improvement, only the most extraordinary case would justify modification in any manner other than by a gradual deescalation of the restraints upon the committee’s liberty. Eor example, where the State is unable to justify the continuance of an order for restrictive confinement, the outright release of the committeee into the community without the use of any intermediate levels of restraint, see State v. Carter, 64 N. J. 382, 403-404 (1974), would normally constitute a manifestly mistaken exercise of the reviewing court’s discretion.

It cannot be emphasized too strongly that the relaxation of the restraints on the committee’s liberty must proceed in gradual stages. As the level of dangerousness posed by the committeee decreases, he should be afforded the opportunity to demonstrate his ability to cope responsibly with the stresses of normal everyday life with -diminishing degrees of supervision. Only after the committee has progressed to the point where he has proven that he can function in normal society with minimal supervision should consideration be given to unconditional release. This process of gradual deescalation will substantially minimize the risk of erroneous determinations of non-dangerousness and will thus protect the State’s compelling interest in maintaining the safety and security of its citizens.

At each periodic review proceeding the ultimate determination will set the level of restraints on the liberty of the committee which will prevail until the next review hearing, periodic or otherwise. The committee remains free to challenge the propriety of those restraints at any time in the *304interim between the scheduled periodic review hearings. However, no hearing on the committee’s request for review of the current validity of the 'prevailing restrictions on his liberty need be granted in the absence of a prima facie showing that a material change in circumstances has occurred since the last periodic review hearing. The requisite showing must allege an improvement in the committee’s condition and consequent diminution in his probable dangerousness. This allegation must be accompanied by supporting documentation from a source other than the committee and must be such as would, if credited, justify a relaxation of the prevailing restraints. At any such committee-initiated review proceeding, the burden of ’ proof is on the committee. As the moving party, the committee must demonstrate by a preponderance of the evidence that under the applicable criteria his request for the modification or termination of those restraints, as the case may be, should be granted.9 See State v. Knot, supra, 68 N. J. at 263 n. 13; State v. Hesse, 373 A. 2d 345 (N. H. Sup. Ct. 1977). The disposition of this request would of course determine the prevailing restraints which are to be the subject of judicial examination at the next periodic review hearing.

Since the committee’s condition and consequent dangerousness are subject to possible change, the State must, in order to meet its burden, provide the reviewing court with a reasonably contemporaneous psychiatric evaluation at each *305periodic hearing. At the periodic reviews within the first two years after the entry of the initial order, the State should ordinarily offer the testimony of a psychiatrist concerning the committee’s present mental condition and dangerousness attributable thereto. This two year period is the critical period within which progressive developments in the committee’s condition are likely to occur, a fact evidenced by the frequency with which periodic reviews pursuant to B. 4 :74-7 (f) are required to be held during that time. At the annual periodic review proceedings after the first two years, the necessity for presentation of testimony of a psychiatrist may be relaxed. The State may instead choose to utilize written reports of a psychiatrist who had testified in an earlier proceeding which reaffirmed or modified his earlier conclusions concerning a committee so long as those reports are based on a current reevaluation. However, whenever a committee intends to present his own psychiatric testimony at any hearing in support of a request that the prevailing restraints on his liberty should be either modified or terminated, he must notify the State in advance. This shall be required whether or not it is a periodic review hearing or one initiated by the committee. Where the State has been so notified and challenges that request, it should ordinarily support its position with psychiatric testimony. Moreover, whenever the State seeks a tightening of restraints, its request should ordinarily be justified through psychiatric testimony in addition to the other evidence upon which it will rely in support of its claim that the committee’s potential dangerousness has increased.

Our allowance of the limited use of written psychiatric reports by the State to justify the continuance of prevailing restraints is not intended to sanction any deprivation of the committee’s right to meaningful confrontation. Any factual evidence of the committee’s behavior bearing upon present dangerousness and controverted should be presented through competent evidence. Of course, where a justifiable reason exists for the witness not appearing, an exception may be *306made. We underscore, however, that the committee in these proceedings enjoys rights of procedural and substantive due process comparable to those available in any judicial proceeding where liberty is at state. We agree with a federal appeals court which, in addressing this issue in a different context, held that “[t]he conclusions in the papers must be sufficiently detailed and supported by facts to allow the court to perform its function * * * .” In re Barnard, 147 U. S. App. D. C. 302, 307, 455 F. 2d 1370, 1375 (D. C. Cir. 1971). However, we must distinguish the situation where the issues of mental illness and dangerousness have been fully and fairly litigated at earlier review hearings (either periodic or committee-initiated) so that the committee has been afforded the opportunity meaningfully to test the legal bases of the State’s imposition of restraints upon his liberty. We discern no unfairness in permitting written psychiatric reports, based on contemporaneous reevaluations, which confirm that the committee’s condition has not materially changed for the better, to be sufficient to support the continuance of the prevailing restraints in the absence of any contrary evidence adduced by the committee. Of course, where such an evaluation does reveal a significant improvement, the State will presumably consent to appropriate relaxation of the restrictions placed upon the committee.

With respect to the issue of determining the extent to which the committee would endanger himself or others in the absence of any restraints upon his liberty, the guidelines we set forth in Krol remain applicable in all review proceedings :

The standard is “dangerous to self or society.” Dangerous conduct is not identical with criminal conduct. Dangerous conduct involves not merely violation of social norms enforced by criminal sanctions, but significant physical or psychological injury to persons or substantial destruction of property. Persons are not to be indefinitely incarcerated because they present a risk of future conduct which is merely socially undesirable. Personal liberty and autonomy are of too great value to be sacrificed to protect society against the possibility of future behavior which some may find odd. *307disagreeable, or offensive, or even against the possibility of future non-dangerous acts which would be ground for criminal prosecution if actually committed. Unlike inanimate objects, people cannot be suppressed simply because they may become public nuisances.

Commitment requires that there be a substantial risk of dangerous conduct within the reasonably foreseeable future. Evaluation of the magnitude of the risk involves consideration both of the likelihood of dangerous conduct and the seriousness of the harm which may ensue if such conduct takes place. It is not sufficient that the state establish a possibility that defendant might commit some dangerous acts at some time in the indefinite future. The risk of danger, a product of the likelihood of such conduct and the degree of harm which may ensue, must be substantial within the reasonably foreseeable future. On the other hand, certainty of prediction is not required and cannot reasonably be expected.

A defendant may be dangerous in only certain types of situations or in connection with relationships with certain individuals. An evaluation of dangerousness in such cases must take into account the lihelihood that defendant will be exposed to such situations or come into contact with such individuals.

Determination of dangerousness involves prediction of defendant’s future conduct rather than mere characterization of his past conduct. Nonetheless, defendant’s past conduct is important evidence as to his probable future conduct. It is appropriate for the court to give substantial weight to the nature and seriousness of the crime committed by defendant and its relationship to his present mental condition.

It should be emphasized that while courts in determining dangerousness should take full advantage of expert testimony presented by-the State and by defendant, the decision is not one that can be left toholly to the technical expertise of the psychiatrists and psychologists. The determination of dangerousness involves a delicate balancing of society’s interest in protection from harmful conduct against: the individual’s interest in personal liberty and autonomy. This decision, while requiring the court to make use of the assistance which medical testimony may provide, is ultimately a legal one, not at medical one.

[68 N. J. at 259-261 (citations and footnotes omitted) (emphasis added)]

We emphasize the last point made in the quoted portion of Krol. Judges considering the committee’s likely dangerousness for the purpose of assessing the appropriate level of restraints upon his liberty to be initially imposed and subsequently modified or terminated too often accord *308undue deference to the presumed expertise underlying psychiatric opinion on that issue. While such psychiatric opinion certainly possesses probative significance, it is no more conclusive on the dangerousness issue than is evidence from lay sources concerning particular instances where the committee has manifested actual or potential harmful behavior. The final decision on the need for and appropriate extent of restrictions on the committee’s liberty is for the court, not the- psychiatrists.

Judges may appropriately consider both expert opinion evidence and lay factual evidence in making their evaluation of the committee’s potential dangerousness under the Krol guidelines and the consequent need for restraint. They must always be mindful that it is their responsibility to ensure that the restraints imposed on the committee will adequately promote the goals of societal and personal safety. The determination of the level of restraints suitable for a particular committee is a matter entrusted to their sound discretion. Their decisions should be based upon their evaluation of all of the relevant evidence before them concerning the committee — psychiatric or otherwise — according each type such weight as they see fit. The ultimate order should reflect a careful judicial weighing of the competing concerns in light of that evidence, explicitly stating the bases for the conclusion reached.

We underscore another teaching of Krol. While the fact that UGI committees have committed an act which would be criminal but for their successful plea of insanity will not justify according them substantially different legal treatment from that accorded other committees, nevertheless a legally cognizable distinction between the two groups does exist. With respect to an NGI committee, his propensity, by reason of his mental illness, to engage in serious antisocial conduct has on at least one occasion crystallized into the commission of what otherwise would constitute a criminal offense. The focus of course must always be upon the actual conduct of the committee, not merely upon its *309characterization as a “criminal” or anti-social conduct. The potential dangerousness of most civil committees ordinarily remains predictive and hypothetical, and may not have been so cogently demonstrated. In justifying the automatic commitment of persons acquitted by reason of insanity, the New York Court of Appeals has observed:

An individual who has committed an act of violence, and has thus demonstrated his dangerousness, and who has successfully asserted an insanity defense, may quite properly be treated somewhat differently from other individuals who, although they may in fact be potentially equally dangerous as a result of mental problems, have not yet so vehemently demonstrated their dangerousness by violent antisocial behavior.

[People ex rel. Henig v. Commissioner of Mental Hygiene, supra, 401 N. Y. S. 2d at 465]

In addition to justifying temporary observational confinement of a prospective N6I committee, see State v. Krol, 68 N. J. at 256, the fact that he has actually engaged in dangerous conduct otherwise criminal should weigh heavily in the court’s assessment of the need for the continued imposition of restraints upon his liberty. See State v. Krol, 68 N. J. at 261 and n. 12. Although his commission of an offense is, like any other instance of demonstrated dangerous behavior, of only evidentiary significance in the judge’s determination of his dangerousness, evidence pertaining to that offense is highly probative of the ultimate issue. This is so because the actual commission of an act which violates the rules governing conduct in the community is of a different order of magnitude than mere threats to engage in such conduct. It is also more telling than is the commission of antisocial acts of lesser gravity. In short, an NGI committee’s prior commission of an act for which he has been relieved of criminal responsibility is powerful evidence of his potential dangerousness and should be weighed accordingly in making that legal judgment.

*310Each review proceeding, periodic or otherwise, should be a comprehensive evaluation of all evidence pertaining to the committee’s mental illness and potential dangerousness by reason thereof. Although the purpose of each review hearing is the assessment of the committee’s current condition and the consequent need for restrictions upon his liberty as a matter of community or individual protection, all prior evidence, both factual and expert, pertaining to those issues remains relevant. See In re Stephenson, supra, 367 N. E. 2d at 1281. The reviewing judge must evaluate the current evidence submitted to him in light of all evidence adduced in earlier proceedings. The committee does not enter each review proceeding with a clean slate, although certainly the passage of time might diminish the relevancy of certain expert diagnostic evidence to the point where it may be insignificant. The accuracy of the reviewing judge’s evaluation of the committee’s probable dangerousness will be greatly enhanced by ensuring that he has the maximum amount of relevant information concerning the committee at his disposal. See “Developments • — • Civil Commitment of the Mentally Ill,” 87 Harv. L. Rev. 1190, 1224 (1974).

Finally, although not raised by either party, a question concerning the appealability of the order entered below and of orders entered in review proceedings generally is presented. The appellant herein filed a notice of appeal “from the whole of the final judgment” continuing her commitment. Orders entered after a review proceeding are “final” in the sense that they definitively set the restrictions on the committee’s liberty which will prevail until the next periodic review hearing unless the committee or the State interposes a request for review on the ground of a change in circumstances prior to that time. Nevertheless, it is at least arguable that such orders are interlocutory and not final judgments, since they are subject to mandatory review on a regular basis by the issuing court. However, the policy underlying the final judgment rule, the undesirability of premature and piecemeal appellate review, would not be *311frustrated by permitting appeals as of right from the judgments entered by reviewing courts since such orders are dispositive adjudications on the record before the court for the time being. We therefore hold such judgments final and appealable pursuant to B. 2:2-3 (a) (1). Nevertheless, the scope of appellate review of such judgments will be extremely narrow, with the utmost deference accorded the reviewing judge’s determination as to the appropriate accommodation of the competing interests of individual liberty and societal safety in the particular case. Such sensitive decisions will be subject to appellate modification only where the record reveals a clear mistake in the exercise of the reviewing judge’s broad discretion in evaluating the committee’s •present condition and formulating a suitable order.

Ill

The Effect of This Ruling

Pursuant to court order, Fields has been receiving periodic reviews of the validity of her commitment. However, there was a misalloeation of the burden of justifying the present restraints upon her liberty in the proceeding below. Thus, a remand is necessary so that the reviewing court can determine the validity of those restraints under the principles announced herein. The reviewing judge did, however, act within his broad discretion in refusing to accord dispositive weight to the psychiatric testimony concerning the committee’s alleged lack of dangerousness.

As was the case in Krol, see 68 N. J. at 266-67, we have determined that our ruling herein shall have retrospective application to all persons who have been acquitted by reason of insanity who are presently subject to any restraints upon their liberty. The interests of those persons presently subject to restraints upon their liberty in obtaining judicial review of the current validity of those restraints are compelling. Conversely, the burden on the State to *312hold bearings conforming to the guidelines announced herein is not that great, considering the small number of persons involved.10 Such persons are entitled to' receive periodic review hearings in accordance with the directives of this opinion within 60 days and thereafter at the intervals specified in B. 4:74-7(f) until the appropriate revision of B. 3:19 — 2 is promulgated.11

Eor the foregoing reasons, the decision below is reversed and the matter remanded to the reviewing court for further proceedings consistent with this opinion.

(temporarily assigned), concurring. I concur in the Court’s judgment of remand of appellant’s case for further review and in much of its comprehensive opinion concerning review procedures in cases of persons acquitted of crime on grounds of insanity (UGI) and consequently committed. However, I disagree with certain aspects of the opinion.

I am not in accord with the Court’s allocation of the burden of proof where a committee seeks a change in restrictions *313between periodic reviews. The Court’s opinion grants the committee the conditional right to seek alteration of restraints in between regular periodic review hearings but in such ease would reverse the burden of proof from the State, which bears the burden originally as well as on periodic review, to the committee, (pp. 303-304) I believe this position to be wrong in principle. I understand the Court’s basic approach to be that, generally speaking, the State bears the burden of persuasion on depriving an individual of his liberty, in whole or in part, for mental reasons. I agree. However, it can logically make no difference as to burden of persuasion that the committee has chosen to exercise his court-granted right to challenge the restrictions at a time other than on regular periodic review. The State should still bear the ultimate burden because the underlying issue remains the same — whether the individual should be kept under restriction as to his liberty. Nevertheless, a proper distinction can be made as to the burden of coming forward with some evidence to support his claim. That burden can appropriately be placed on the committee when he seeks a change in restrictions between periodic reviews. But if he meets that burden, the ultimate burden of persuasion on the disputed issue of the degree of restriction to be continued on the committee should remain with the State. Cf. State v. Abbott, 36 N. J. 63, 72 (1961) (as to the respective burdens of State and defendant with respect to the defense of self-defense).

I also see no justification for the Court’s setting a two year period after the initial order as a line of demarcation to determine the kind of psychiatric testimony which will be required of the State, (pp. 304-305) Each case stands on its own feet as to the degree of seriousness of the condition of the committee and as to the kinds of proofs which should move a court to determine the nature of the restrictions, if any, which should continue to be imposed on him. The decision has to rest in the sound judgment of the judge; so does, generally, the kind of proof he will properly exact *314from the State to meet its burden as of any particular hearing date.

I would agree, however, that a valid distinction can be made between the first hearing and all later ones as to the kind and nature of proof to be required of the State; At the first hearing I would mandate psychiatric expert proof on the part of the State and I would require in other respects that the rules of evidence be followed as in a criminal trial. At all later hearings, however, the case would be treated as a continuum in any phase of which all prior proofs would be deemed applicable and updated reports of physicians who had testified at earlier stages of the case would be admissible in the discretion of the judge. Subject to the foregoing general directions, I believe it more useful and practicable to vest broad discretion as to proof and procedural requirements in the hearing judge at such later hearings than to direct the manner of his exercise of discretion at every step of the way in an opinion of this Court.

In the foregoing vein, I particularly disagree with the Court’s direction that if the committee gives notice that he intends to offer psychiatric testimony at any subsequent hearing the State must offer rebutting psychiatric testimony (p. 305), I can easily conceive of situations where, in the light of the past history of the case and the medical condition and recent objective behavior of the committee, the judge in his discretion could' properly find the State to have met its burden without putting a physician on the stand.