(dissenting):
In 1967 Lewis Ecker, suffering from mental illness, brutally raped and murdered a young woman. He was tried and found not guilty by reason of insanity. After a Bolton hearing1 he was committed to St. Elizabeth’s Hospital. Although his illness was serious and his condition chronic, he has improved significantly during the course of his years at the hospital. Tr. 22, 123, 137-138, 150, 172, 197-198. This is reflected both in IQ tests (his IQ rose 15-25 points by 1972) and in other psychological tests administered most recently in June 1974. They show continued inner turmoil and fantasies, but they demonstrate an improved ability to deal with problems. These tests also indicate that Ecker now regards his fantasies as “quite ego-alien,” Tr. 23-27, App. 61, that he can step aside and deal with them at a distance, Tr. 203, avoiding the necessity of “acting out,” App. 63, and that he now has a better sense of the “time spectrum,” appreciating “that there is not immediate relief to any sense of tension.” Tr. 113; see also Tr. 83-84. And since 1972 he has been virtually without medication. Tr. 13; contrast United States v. Ecker (Ecker I), 156 U.S.App.D.C. 223, 225, 227, 479 F.2d 1206, 1208, 1210 (1973).
Ecker has, as well, taken an active volunteer role in the life of the hospital community, showing movies to elderly patients and participating in editing and publishing a hospital newsletter. Tr. 10, 17-18, 97. He has had grounds privileges (unsupervised access to the hospital grounds) since 1971. Tr. 11, 173. Perhaps most important, he has participated in vocational therapy, first with the Behavioral Studies Division and then in the Sheltered Workshop at the hospital. Tr. 14-17. He has shown interest in and aptitude for electronics work, and he has been given some supervisory responsibilities at the Sheltered Workshop. Tr. 15-17. He has, in fact, advanced through substantially all of the training opportunities open to him at the hospital; the logical next step of his treatment is electronics training at a vocational training institution outside the hospital grounds. Tr. 56.
In order to permit him to take this next step the hospital, after review, certified Ecker for conditional release under 24 D.C. Code § 301(e) (1973),2 thereby indicating *202that, in the hospital’s judgment, Ecker would not be “dangerous to himself or others” under the proposed conditions. He was to be released from the hospital only for sufficient time to travel to and from the school and attend classes; he would always spend the night and evening at the hospital. If his progress was satisfactory he could have further time for library and lab work. Finally he might gain permission to spend certain weekends or holidays with his family. All these measures were seen as the logical next steps in his treatment.
The Government does not challenge the hospital’s right to seek conditional release for purposes of treatment. Indeed it could not; treatment must be regarded as the most compelling reason for conditional release of a patient not completely recovered. See generally Ashe v. Robinson, 146 U.S. App.D.C. 220, 222, 450 F.2d 681, 683 (1971); Covington v. Harris, 136 U.S.App.D.C. 35, 43, 419 F.2d 617, 625 (1969); Tribby v. Cameron, 126 U.S.App.D.C. 327, 328, 379 F.2d 104, 105 (1967); Rouse v. Cameron, 125 U.S.App.D.C. 366, 373 F.2d 451 (1966); Hough v. United States, 106 U.S.App.D.C. 192, 196, 271 F.2d 458, 462 (1959) (“temporary freedom is often an essential part of the therapeutic process”); 21 D.C.Code § 562 (1973). The Government does assert, though, that Mr. Ecker’s situation does not meet the “dangerousness” standard for conditional release under Section 301(e). And it asserts at the threshold that, contrary to Ecker’s argument, the Government does not bear the burden of proof on the dangerousness issue when it opposes release, despite the hospital’s certification.3
I
On the burden of proof question the statute is hardly a model of clarity.4 I agree with Judge Lumbard that on this record we need not decide this dispute. We should, moreover, be especially cautious in finally deciding the question in view of the difficulty of the issue. Judge Wilkey, for example, says in Part IV of his opinion that in conditional release proceedings “there is no assignable burden of proof as we would know it in a criminal or civil case.” Yet in Part II of his opinion he seems clearly to place the burden of persuasion on the patient or perhaps the hospital.5 Judge Lum*203bard, although agreeing that the question need not be reached, offers his view that the burden is on the hospital. But this puts the United States Attorney’s office in the anomalous position (unless the patient hires counsel to represent the hospital-patient position) of potentially being called upon to represent both sides — choosing on its own to oppose the release, but being called upon to serve as lawyer for the government hospital.6 In any event, hospital certification amounts to significant support for the patient in his effort to secure conditional release; it must mean that the hurdles are not as high as if he sought conditional release on his own motion. This is particularly true where, as here, the conditional release is for treatment.
II
I dissent from the judgment affirming the District Court because that court’s opinion does not rest upon findings adequate to support the outcome reached or to provide for meaningful judicial review. United States v. McNeil, 140 U.S.App.D.C. 228, 229, 434 F.2d 502, 503 (1970) (per curiam);7 cf. Millard v. Cameron, 125 U.S.App.D.C. 383, 386, 373 F.2d 468, 471 (1966), reversed after remand, sub nom. Millard v. Harris, 132 U.S.App.D.C. 146, 406 F.2d 964 (1968); Whittaker v. Overholser, 112 U.S.App.D.C. 66, 68, 299 F.2d 447, 449 (1962).
The most glaring deficiency lies in the District Court’s failure to indicate what other conditions for release were considered and — obviously — rejected. Both the statute and our cases make clear that if the court is dissatisfied with the conditions suggested by the hospital — if, that is, it feels the danger is too high under those conditions— then it is the court’s , duty to consider and impose such other conditions as will make conditional release acceptably safe. United States v. McNeil, supra, 140 U.S.App.D.C. at 239, 434 F.2d at 513 (Bazelon, C. J., concurring); Dixon v. Jacobs, 138 U.S.App. D.C. 319, 325, 427 F.2d 589, 595 (1970); Hough v. United States, supra, 106 U.S.App.D.C. at 195, 271 F.2d at 461; cf. Lake v. Cameron, 124 U.S.App.D.C. 264, 268, 364 F.2d 657, 661 (1966) (en banc). The record indicates the District Court’s probably justified displeasure with the laxity of some hospital-proposed conditions. Tr. 210, memorandum op. 18. But more narrow — and still reasonable — conditions which would remedy at least some of the problems quickly come to mind.8 Yet the District Court has given us no indication of other conditions considered or of the reasons for reject*204ing them. The only mention of other possible conditions comes in the concluding paragraph of the memorandum opinion: Mr. Ecker’s release would not Joe safe “under the proposed conditions of release (or any conditions which could reasonably be imposed) * * Memorandum op. 19. This conclusory statement is inadequate. I would remand for reconsideration and for a more careful articulation, on the record, of alternative conditions considered and the reasons for rejecting them.
Ill
The District Court’s findings are inadequate in another important respect. There is a certain momentum in cases of this sort, it seems to me, pulling strongly toward reading the conditional release provisions out of the statute. The initial episode which sent Ecker to St. Elizabeth’s was so brutal that there is a tendency to seek proof of complete recovery and assurance of safety under any conceivable conditions before granting even a limited release. If courts yield to this entirely understandable urge, then patients like Ecker will be entitled either to unconditional release or to no release at all.
Congress cannot have been unaware of these concerns. Yet in Section 301(e) Congress has explicitly chosen to permit, in appropriate circumstances, conditional release of patients not completely recovered. And the only patients who come under Section 301(e) are those who were initially hospitalized after committing a criminal act. Courts must take care to honor the congressional judgment. Congress has provided, I emphasize, that a person committed following an insanity acquittal may be entitled to conditional release without proving that he has recovered completely, without proving that, unmonitored, he would never be dangerous, and certainly without proving that medical authorities have always been 100 percent accurate in past judgments of his condition.9
The question for the court in this case is whether “the individual has recovered sufficiently so that under the proposed conditions — or under conditions which the statute empowers the court to impose ‘as [it] shall see fit,’ — ‘such person will not in the reasonable future be dangerous to himself or others.’ ” Hough v. United States, supra, 106 U.S.App.D.C. at 195, 271 F.2d at 461 (footnote omitted). And the court is of course concerned not with finding the mere possibility of danger but with assessing its likelihood. See Dixon v. Jacobs, supra, 138 U.S.App.D.C. at 325 n.17, 427 F.2d at 595 n.17; Millard v. Harris, 132 U.S.App.D.C. 146, 159, 406 F.2d 964, 977 (1968).10
*205The District Court’s findings here are inadequate for us to determine whether this was in fact the question it asked and answered. The only truly useful explanation of the court’s decision comes in the following passage from Paragraph 9 of the memorandum opinion:
This conclusion [that conditional release should be denied] is made with particular reference to [Ecker’s] mental illness, its chronic nature, the continuing existence of fantasies in his mind and his uncertain ability to deal with them, the continuing turmoil he is experiencing, and the incidents of his misbehavior in 1974 with respect to female patients at Saint Elizabeths Hospital.11
Most of these factors show only that Eek-er remains mentally ill — a fact not contested. His illness is undeniably chronic, and he does experience fantasies and turmoil. None of this necessarily bars him from conditional release. Apart from the two incidents of “misbehavior,” which are exceedingly minor in nature, the only listed factor with any real bearing on dangerousness is Ecker’s “uncertain ability to deal” with his fantasies. I agree that we cannot be certain about his ability to deal with them. But the real question is whether that uncertainty is so great that Ecker must be considered “dangerous” within the meaning of the statute if granted a limited conditional release.12 The District Court must do more *206to explain why it answered yes to that question, particularly in the face of the hospital’s certification and a record suggesting the contrary. All four doctors who testified, including the one called by the Government, concurred in the hospital’s recommendation for conditional release and the judgment that Ecker would not be dangerous under the proposed conditions. Tr. 19, 123, 163, 171, 205. There is in the record, moreover, an impressive range of evidence indicating that the vocational therapy to be provided under the proposed conditional release would yield continued improvement in Ecker’s condition because it would provide the needed next step in his treatment. If his condition were to deteriorate — and the witnesses thought this unlikely because of his improved “controls”— the deterioration would come slowly. With Ecker returning to the hospital each afternoon after class, there would be numerous “points of intervention” where the hospital could take note of the developments and initiate appropriate action.13 Tr. 19, 54-60, 83-84, 113, 118, 123, 128, 137-140, 150, 173, 197, 203-205.
I do not suggest that the District Court was bound to accept all of this testimony.14 It may in fact be that Ecker is too dangerous for limited release under any reasonable conditions. But if so, this court, and Ecker, and indeed the public (for the public has an important interest in the liberty of fellow citizens and in treatment to restore their *207health, along with its obvious interest in safety), are entitled to a more complete statement indicating why the likelihood of danger was adjudged sufficiently great to bar conditional release.
Nothing in the court’s holding today would prevent the District Court from supplying a more complete statement if Ecker should apply again at a later date for conditional release. And, of course, nothing in today’s holding precludes a District Court from ordering conditional release later should it determine at that time that the statutory tests are met.15
I respectfully dissent.
On Suggestion for Rehearing En Banc.
Before BAZELON, Chief Judge, and WRIGHT, McGOWAN, TAMM, LEVEN-THAL, ROBINSON, MacKINNON, ROBB and WILKEY, Circuit Judges.ORDER
PER CURIAM.The suggestion of appellant Lewis C. Ecker, II for rehearing en banc having been transmitted to the’ full court and there not being a majority of the judges in regular active service in favor of having this case reheard en banc, it is
ORDERED by the court en banc that the aforesaid suggestion for rehearing en banc is denied.
Statement of Circuit Judge LEVEN-THAL Concerning His Vote To Deny Rehearing En Banc LEVENTHAL, Circuit Judge:While the suggestion for rehearing en banc raises some interesting questions, I would find it difficult to consider them because in my view the issues have not been sufficiently focused by the record.
What I think is a more central difficulty is the problem, glimpsed in part III of Judge Wright’s dissenting opinion, but not focused in a claim of error, that the trial judge has in this case disagreed with the conclusions of the only experts who gave testimony without indicating why.
This is a proceeding seeking release (or conditional release) of a person detained following a verdict that he is guilty of a serious crime but for mental illness. The trial judge has the responsibility, and he may decline relief even though all the testifying experts recommend release. But he must indicate why he comes to a differenct conclusion. As Judge Tamm said in another connection, the requirement of a statement of reasons — “more than a simple conclusion” — “is not onerous if the matter was dealt with in a conscientious manner in passing on the merits.” Davis v. Clark, 131 U.S.App.D.C. 379, 380-81, 404 F.2d 1356, 1357-58 (1968).
Although this case is structured as a habeas corpus hearing with expert testimony, what is also involved is a decision by a unique agency — a hospital (here, St. Elizabeth’s Hospital) holding someone detained under court order. The predicate of a hearing is that the hospital superintendent has issued a certificate, for filing with the clerk of the court, of his opinion that the person confined is entitled to release. 24 D.C. Code § 301(e). This is not a situation where the court must defer to the agency, if it is supported by substantial evidence. The hearing is before the court, the court has authority and responsibility to make the decision, and the hospital agency is only advisory. But if the machinery of government is to make sense the agency must be informed of why it is that the advice of its staff and officials has been rejected.
The forecasting of dangerousness is itself dangerous, and difficult, but the dangers and difficulties will only be increased unless the approach of the court is indicated, so that the experts of the hospital agency can adhere to that approach, or so that they can better present the situation when it is re*208viewed a year hence, if they believe that the judge’s approach reflects a misunderstanding of their views or the underlying situation.*
Statement of Circuit Judge J. SKELLY WRIGHT, in Which Chief Judge BAZELON and Circuit Judge SPOTTSWOOD W. ROBINSON, III, Join, of Reasons for Voting for Rehearing En Bane.Appellant Ecker was committed to St. Elizabeth’s Hospital in 1968 after being acquitted by reason of insanity of the murder of a young woman. 24 D.C. Code § 301(d) (1973). In 1974 the hospital certified Ecker for conditional release under id. § 301(e) (1973) so that he could attend a vocational training school by day, returning to the hospital each night. Such training was seen as an essential step in Ecker’s treatment program, since he had basically exhausted all the training opportunities available at the hospital. Testimony at the hearing firmly supported the conclusion that limited release of this kind is needed for Ecker’s continued improvement. It also provided a strong showing that, as the hospital determined, Ecker is not likely to be dangerous to himself or others on conditional release. Nonetheless the District Court rejected the hospital’s determination and denied release. A panel of this court affirmed.
Our cases have consistently underscored the importance of honoring a patient’s right to treatment. See, e. g., Ashe v. Robinson, 146 U.S.App.D.C. 220, 222, 450 F.2d 681, 683 (1971); Covington v. Harris, 136 U.S.App.D.C. 35, 41, 419 F.2d 617, 623 (1969); Tribby v. Cameron, 126 U.S.App.D.C. 327, 328, 379 F.2d 104,105 (1967); Rouse v. Cameron, 125 U.S.App.D.C. 366, 373 F.2d 451 (1966). Where there is a strong showing that conditional release is needed for treatment purposes, it is especially incumbent upon the District Court to make a careful, detailed, and explicit statement of its findings and reasons for denying release, both to assure that the need for treatment is properly respected and, more importantly, to aid the hospital in planning its future course. The hospital ought to know where to concentrate its future efforts so that, barring deterioration in Ecker’s condition, it may some day devise an acceptable release proposal.
The conclusory findings presented by the District Court in this case give the hospital no real guidance. This failure to give adequate reasons in the face of the hospital’s certificate recommending conditional release is a significant departure from the statutory scheme, which places initial responsibility for this judgment on the hospital. The statute indicates that hospital and court should work cooperatively in carrying out its purpose.
For this reason, and for the other reasons stated in my dissent from the panel opinion, I vote for rehearing en banc.
. Bolton v. Harris, 130 U.S.App.D.C. 1, 395 F.2d 642 (1968).
. Section 301(e) is set forth in its entirety in footnote 12 of Judge Wilkey’s opinion. The statute makes it clear that the patient may be released on the strength of the hospital’s certificate alone unless the Government or the court sua sponte moves for a hearing. In that event the court weighs the evidence and decides for itself whether the statutory requisites are met. This provision for court review of hospital decisions to grant release applies only to those committed under § 301(d) — that is, those ac*202quitted of an offense by reason of insanity. Patients civilly committed may be released when the hospital itself determines that the statutory requisites are met, without court review. See, e. g., 21 D.C.Code § 548 (1973).
. Note that the certificate is sufficient of itself to authorize release of the patient. See note 2 supra.
. The statute provides four methods by which the insanity-acquitted may gain release. Under two of them it expressly places the burden of proof on the patient. 24 D.C.Code §§ 301(d)(2) & (k)(3). Section 301(g) preserves the patient’s right to petition for habeas corpus. It gives no indication of the burden of proof in such a proceeding, but this court has construed the provision to place the burden on the patient, principally because of the usual allocation of the burden in habeas corpus cases. Bolton v. Harris, supra note 1, 130 U.S.App.D.C. at 12, 395 F.2d at 653. Section 301(e) is likewise silent on burden of proof, but here there is no centuries-old tradition compelling the conclusion that the burden is on the patient — or on the hospital, which is really the moving party.
. Judge Wilkey writes, quoting from Judge Leventhal’s concurrence in Dixon v. Jacobs, 138 U.S.App.D.C. 319, 332, 427 F.2d 589, 602 (1970), that continued confinement is appropriate “unless the district court can make an ‘affirmative finding that it is at least more probable than not that [the patient] will not be violently dangerous in the future.’ ” Wilkey op. at 18. A preponderance standard for the burden of persuasion is meant to operate only in that very limited class of cases where the fact finder, after weighing all the evidence, finds himself in a position of equilibrium, not persuaded by either position. In such an instance, he is to find against the party bearing the burden of proof. The quoted passage from Judge Wilkey’s opinion therefore amounts, unavoidably, to placing the burden of persuasion on the patient (or hospital) seeking release.
The discussion in Part II of Judge Wilkey’s opinion is clouded by the reference to “a substantial problem of danger in the reasonable future” — a phrase never defined. Wilkey op. at 18. Whatever this may mean, it can neither add to nor detract from the statutory test, as authoritatively construed in Hough v. United States, 106 U.S.App.D.C. 192, 195, 271 F.2d 458, 461 (1959): determining by a preponderance standard whether the patient will or “will not in the reasonable future be dangerous to himself or others” under the proposed condi*203tions or any others which may reasonably be imposed.
I do wish to state my agreement with the major thrust of Part IV of Judge Wilkey’s opinion. Conditional release proceedings are “not strictly adversary proceedings.” Lake v. Cameron, 124 U.S.App.D.C. 264, 268, 364 F.2d 657, 661 (1966) (en banc). I take this to mean, at a minimum, that the burden of going forward does not fall rigidly upon either side. Rather the judge has an affirmative duty to take an active and vigorous role to see that all the facts are developed to the fullest extent possible at the hearing. He may direct any order of presentation that seems most likely to effectuate that goal.
. St. Elizabeth’s is clearly a government hospital. See 24 U.S.C. § 161 et seq. (1970 and Supp. IV 1974); 32 D.C.Code § 401 et seq. (1973).
. Judge Wilkey’s attempt to distinguish McNeil, in footnote 40 of his opinion, strikes me as unconvincing. The brief per curiam opinion in that case was not directed only to the vice of denying conditional releases with no explanation whatever. See Hough v. United States, supra note 5, 106 U.S.App.D.C. at 195, 271 F.2d at 461. The same concerns that led to remand there warrant remand where, as here, the District Court’s explanation sheds insufficient light on the crucial determinations.
. The court expressed a concern, for example, that the hospital’s proposal “still does not provide adequate controls over Mr. Ecker during the periods he would be away from the hospital.” Memorandum op. 18. This may have reference — we are not further enlightened — to the testimony that the hospital would not learn of any truancy episodes until a week following any absence. Tr. 60. If this is the problem, it would be relatively simple to impose a condition that the school telephone the hospital each morning reporting Mr. Ecker’s arrival or absence. Or the court could require that the hospital, at least at first, transport him to and from the school, rather than permitting him to ride unaccompanied on public transportation. See Tr. 210.
. Judge Wilkey suggests obliquely that the expert testimony here — unanimously in favor of conditional release — is entitled to especial skepticism because other experts, year| ago, failed to foresee that Ecker’s illness might lead to the murder he committed when they released him from Shepard-Pratt Hospital in 1965. Wilkey op. at 25. This strikes me as deeply unfair to Ecker and to the medical profession. It is well settled that the question before the court in a conditional release proceeding is the patient’s present condition. Rouse v. Cameron, 125 U.S.App.D.C. 366, 376 n.43, 373 F.2d 451, 461 n.43 (1966). Past diagnoses may have proven tragically in error, but this does not by any means indicate that current medical opinion is necessarily wrong.
Moreover, the suggestion yields to no limiting principles. There will always be in Ecker’s case history the erroneous 1965 diagnosis. Does this mean that he can never secure release, no matter what the psychiatrists and psychologists may say from the stand? Furthermore, there is no inherent reason why the past error should apply only to Ecker. Logically it might as well apply to any hospital recommendation for any patient’s conditional release: hospitals have been badly wrong in the past; they might be wrong here. Everyone knows that psychiatric predictions fall well short of perfect accuracy — a fact candidly acknowledged by the experts who testified in this case, e. g., Tr. 56, 119 — but no one seriously believes that every hospital recommendation must therefore be rejected.
I would suggest that Judge Wilkey’s reliance on this factor does not constitute a part of the holding of this court.
. Assessing the likelihood of danger is, beyond doubt, a peculiarly difficult determination. In Covington v. Harris, 136 U.S.App.D.C. 35, 45-46, 419 F.2d 617, 627-628 (1969), the court explored these difficulties with particular eloquence. The issue there was whether the hospital’s decision to keep the patient in the maximum-security John Howard Pavilion was valid. In the course of its consideration this *205court posed questions which could as well be asked about a court’s decision in cases like the one before us:
“[D]angerousness” is a many splendored thing. Unless muzzled by discriminating analysis, it is likely to weigh against nominally competing considerations the way a wolf weighs against a sheep in the same scales: even if the sheep is heavier when weighed separately, somehow the wolf always prevails when the two are weighed together. Keeping dangerousness on a taut leash is especially difficult where there is danger of murder, since the danger is admittedly grave and since its improbability, which theoretically discounts its gravity, is exceedingly difficult to quantify.
Moreover, once a man has shown himself to be dangerous, it is all but impossible for him to prove the negative that he is no longer a menace. The specters of the murder appellant committed 35 years ago (expiated by a long jail sentence) and the murder he may have committed more than 10 years ago obviously haunt the hospital at the very thought of granting him the least measure of freedom[.] * * * Their concern is understandable and may well be fully justified. But for all that appears, the murders and the unpredictable consequences will still be there after twenty years or after fifty. Appellant was not convicted of the second murder, and his hospitalization is not to be tacitly converted into a life sentence to John Howard. In these circumstances it is fair to ask the hospital how appellant can ever demonstrate his readiness for a less pervasive confinement: What evidence of improvement are they looking for? What is the prospect that they will ever find it? * * *
. Judge Wilkey sets forth in full the other paragraphs of the memorandum opinion which have some bearing on the District Court’s conelusion. Wilkey op. at 20-21. They contain certain details of the incidents of misbehavior and they amplify the diagnosis of Ecker’s condition. Beyond showing that Ecker is mentally ill, however, they shed very little light on the court’s crucial dangerousness determination.
. Judge Wilkey refers to “ambiguities and uncertainties” and to “real doubts about the sufficiency of the investigation of Ecker’s condition” which he implies are just as great as when we considered the hospital’s first certification for Ecker’s conditional release, United States v. Ecker (Ecker I), 156 U.S.App.D.C. 223, 479 F.2d 1206 (1973). Wilkey op. at 23-24. This misconstrues our holding in Ecker I. The major uncertainties mentioned there have been resolved: Ecker has been off medication for over two years, without untoward effects; and an elopement incident, recent at the time of Ecker I, has faded into the past without apparent repetition. The court in 1973 expressed lesser reservations about a third factor, the hasty testing program to which he was subjected and which yielded doubts about his control over the turmoil reflected in the tests. But it was clear after Ecker I what the hospital could do to provide an adequate investigation — and it has done so. The 1974 testing was not hastily undertaken to meet an artificial deadline. Though it revealed some evidence of a continued fantasy life, the uncertainty about Ecker’s controls has clearly been reduced. Judge Wilkey supplies the hospital with no indication whatever of how the remaining ambiguities and uncertainties may be cleared up. If this is because they never can be cleared up, given the irreducible imprecision of psychological judgments, see, e. g., Developments in the Law—Civil Commitment, 87 Harv.L.Rev. 1190, 1240-1244 (1974), then Ecker is being held to a standard far stricter than Congress has chosen to impose.
. If the court was concerned that, despite this testimony, the hospital would not carry out faithfully its monitoring and supervisory functions as well as necessary corrective action should some deterioration appear, the court could have imposed further conditions. For example, it might have required detailed reporting to the court in such a fashion that careful monitoring and supervision would be assured.
. This court has wisely held that the trier of fact is not bound by expert opinions expressed during the hearing. United States v. McNeil, 140 U.S.App.D.C. 228, 230, 434 F.2d 502, 504 (1970). But at the same time the trier cannot simply ride roughshod over what the experts have said. Our cases are “not authority for disregarding expert testimony. It must be considered with the other evidence, not arbitrarily rejected.” Douglas v. United States, 99 U.S.App.D.C. 232, 239, 239 F.2d 52, 59 (1956).
That the trial court is charged with the important duty of superintending the hospital’s application of the legal standards for release to the particular patient does not mean that the court may substitute its own opinions for uncontradicted expert testimony with regard to factual matters within the specialized competence of psychiatrists and psychologists.
sk * * * * *
Where uncontradicted expert testimony is provided with an adequate foundation, the trier of fact may not simply pick and choose among the experts’ opinions, accepting some and rejecting others according to his own personal views. * * * The Supreme Court has recently reiterated the “elementary” requirement of due process that “the decision maker’s conclusion * * * must rest solely on the legal rules and evidence adduced at the hearing.” * * *
United States v. McNeil, supra, 140 U.S.App.D.C. at 240-241, 434 F.2d at 514-515 (Bazelon, C. J., concurring) (footnotes omitted) (emphasis in original), quoting from Goldberg v. Kelly, 397 U.S. 254, 271, 90 S.Ct. 1011, 1022, 25 L.Ed.2d 287, 300 (1970).
In this light, I am not at all convinced that Judge Wilkey’s characterization is correct when he says that the trial judge here did nothing more than reject “the experts’ opinions on questions of law.” Wilkey op. at 22. “The likelihood of future misconduct, the type of misconduct to be expected, and its probable frequency” are questions of fact. Dixon v. Jacobs, supra note 5, 138 U.S.App.D.C. at 325 n.17, 427 F.2d at 595 n.17. Only after the level of factual dangerousness is thus established— on the basis of the record before the court — is the court called upon to make a determination of law: whether the established likely harm is sufficient to warrant continued confinement. Id.
Although it is undeniably difficult to keep the two determinations distinct, it is incumbent upon us to insist that the District Court do so to the fullest extent possible. We would then have a proper basis for determining whether the District Court’s factual conclusion is or is not “clearly erroneous” when viewed against the record. If the factual finding is acceptable, we may then review whether the legal standard was properly applied. Again, I do not minimize the difficulty involved. We cannot pretend to anything like mathematical application of fixed numerical scales. But we can certainly approach more closely to analytical clarity in the twofold determination demanded by proceedings for conditional release. See Ecker I, supra note 12, 156 U.S.App.D.C. at 226 n.9, 479 F.2d at 1209 n.9.
. The question before the court at such time will be Ecker’s present mental state, Rouse v. Cameron, supra note 9, and the outcome of this proceeding by no means predetermines the decision in future proceedings.
The law, 24 D.C. Code § 301(e), specifically contemplates that at the hearing there will be submission of evidence as to mental condition of the person confined “including the testimony of one or more psychiatrists from said hospital.” The law then provides that the “court shall weigh the evidence.” The statute directs the court to order release if the court finds that the person has recovered his sanity and will not in the reasonable future be dangerous to himself or others. If the court does not so find, it is directed to order the person returned to the hospital.
Even though an order of return need not be accompanied by a “finding,” but rather a statement that the court does not make the “finding” necessary for release, the process is a reasoned one, with a requirement of weighing the evidence, and the reason for failure to make the finding fulfills the concept of accountability underlying government under law.