dissenting:
I join in Part II of the opinion of the court which sets forth the legal standards applicable by virtue of the fourteenth amendment to the involuntary imposition of medical treatments by persons acting under color of state law. I dissent, however, from the judgment of the court insofar as it modifies the preliminary injunction which was entered in the trial court, because the findings supporting that injunction were not clearly erroneous and the scope of pendente lite relief which was ordered was well within the range of permissible district court discretion. Part III of the opinion of the court does not satisfactorily explain why, in reviewing that preliminary injunction, we are free in this case to disregard those findings or to substitute our discretion for that of the trial judge.
The majority opinion concludes that Administrative Bulletin 78-3 sets forth procedures which, if followed, satisfy fourteenth amendment due process standards for the protection of the liberty interest it recognizes. In the abstract that may be so. The district court was not, however, dealing with abstractions. It was presented with a record of ongoing substantive violations of that liberty interest by those state officials having custody of the unfortunate class members whom Rennie represents. The record also shows that Bulletin 78-3 was adopted only after this lawsuit was commenced.
This represented the Division’s first formal attempt to deal with the problem of patients who refuse medication, and was in response to the issues raised in this litigation.
Rennie v. Klein, 476 F.Supp. 1294, 1303 (D.N.J.1979). It is a fair inference that the Bulletin is nothing more than a response to the stimulus of a threatened court order. The implementation of the Bulletin was hardly enthusiastic. Although after March, 1979 the central office of the Division of Mental Health was supposed to review the *866case of any patient compelled to take medication, a representative of the office could confirm only two such reviews having taken place by late summer of that year, although 40 to 50 patients a month were recorded as refusing drugs at Marlboro Hospital alone. The witness added that a Dr. Rotov:
might have reviewed “several” other cases [besides Rennie’s], and examined some patients personally.... The Court finds that this level of review is hardly a regular procedure, and is not quickly or reliably available to patients.
476 F.Supp. at 1303. The court found, moreover, that the Bulletin was not rapidly implemented at the various hospitals. It was ignored in two hospitals for nine months, and was not even called to the attention of the medical director of Marlboro Hospital for eight months. 476 F.Supp. at 1303. Given the evident lack of enthusiasm in the Division of Mental Health for addressing the problem in the first place, and for implementing the Bulletin after it was devised as a response to the lawsuit, the court’s skepticism about relying on it as a substitute for an injunction is understandable.
I find it unnecessary to determine whether the Bulletin sets forth the process that is due. The opinion of the court does not refute plaintiff’s evidence establishing that defendants have largely failed to comply with the procedures which it specifies. My principal disagreement with the majority is over the implicit suggestion that those procedures, though not implemented, nonetheless limit the remedial authority of the court to cure a substantive violation, because in the abstract or on their face they satisfy the minimum requirements of due process. The Bulletin cannot limit the scope of pendente lite injunctive relief. It must be recalled that the trial court has not yet acted finally. It concluded, however, based upon its evaluation of all the evidence and the testimony, that plaintiff class members were suffering irreparable harm and that, at least until it received further information, more protection against forced medication was, in the New Jersey institutional climate, required. To support that conclusion the court made detailed findings in light of which the preliminary injunctive relief it ordered was demonstrably not an abuse of discretion. Oburn v. Shapp, 521 F.2d 142 (3d Cir. 1975); see also Doran v. Salem Inn, 422 U.S. 922, 95 S.Ct. 2561, 45 L.Ed.2d 648 (1976).
For example, the court found, significantly, that refusals had been underreported since the Bulletin’s adoption:
A major problem in implementation of Bulletin 78-3 is identifying those patients who refuse drugs. At Marlboro Hospital about 40 to 50 patients a month have been recorded as refusing. Tr. XXIX, 80, 87-88; Ex. D-34, pp. 45-48. Trenton Hospital has reported about 15 to 40 a month. Ex. D-38, pp. 46-49. Ancora Hospital has invoked the Bulletin for only Mr. Rennie ■sand the 23 year old woman discussed above. No other patient has been reported to have refused since the beginning of the year. Ex. P-33, p. 77; Ex. D-35, pp. 24-26. Greystone has only reported three patients refusing since the beginning of the year, Ex. P-33, pp. 80-81; Ex. P — 47, pp. 29-31, and the Gardner Geriatric Center has only acknowledged one refusal. Ex. D — 47, pp. 26-28; Ex. J-6, p. 24.
The record also reflects the fact that in a ward of 70 patients at Boston State Hospital, which has been prohibited from forcing medication by federal court order, 10 to 12 patients refuse medication at any one time. Tr. XXVII, 35-36. A psychiatrist with experience at state hospitals in New York testified that five percent of patients refuse medication at a given time. Tr. XXVIII, 12.
The patient populations at Trenton, Marlboro and Boston State are similar to those at Greystone and Ancora. The court attributes the discrepancy in statistics to a substantial failure by staff at the latter hospitals to report refusals. Also, certain hospitals overuse the exception in the Bulletin for emergency situations.
*867476 F.Supp. at 1303-04. The court’s factual findings that staff members have failed to report refusals and have abused the “emergency” exception in the Bulletin are not clearly erroneous. The court also found that many patients are too intimidated to refuse medication and thereby trigger the review procedures in the Bulletin:
Even if all hospitals accurately reported the number of overt refusals by patients, many patients are too intimidated to attempt to refuse medication and would still be ignored. Certainly there has been extensive use of • forced injections in the hospitals when both voluntary and involuntary patients refused to take medication orally. Tr. XVIII, 71, 75, 88, 115; Tr. XX, 8-20; Tr. XXII, 18-20, 104; Tr. XXIII, 74-76; Tr. XXIX, 112-19; Tr. XXXI, 95-107; Tr. XXXII, 65; Ex. P-28, p. 11; Ex. P — 35, p. 6; Ex. P-60; Ex. P-61; Ex. P-65; Ex. P-75. The Marlboro medical director candidly admitted that drugs are still systematically forced on patients. Tr. XXIX, 137. Often forced injections are doses of long-acting prolixin which not only has a much longer effect than other psychotropics-but usually has mote immediate adverse side effects. Tr. XXII, 104; Tr. XXXI, 108; Ex. D-23, 70-71; Zander, Prolixin Decanoate: A Review of the Research, in 2 Mental Disability L.Rep. 37-39 (1977). Therefore it would be quite rational for patients to conclude that resistance to drugs would result in their receiving a more unpleasant medication.
476 F.Supp. at 1304. This finding of intimidation is not clearly erroneous. Intimidation, if it exists, makes the Bulletin meaningless since it only applies to refusals of medication. The court also found that although a review is initiated only after a refusal to accept medication, there is little evidence that the hospitals have taken affirmative steps to inform patients of their right to refuse. 476 F.Supp. at 1304. Thus the court found that a combination of underreporting, intimidation, and induced ignorance seriously undermined confidence that the review procedures of the Bulletin would ever become readily available to members of the plaintiff class.
For those instances in which refusals of medication actually occur, the court’s findings cast doubt upon the effectiveness in actual operation of the Bulletin’s procedures: .
Much of the evidence concerning use of Bulletin 78-3 focused on review by the medical director of refusals by involuntary patients. Few of the recorded refusals even reach that level, partly as a result of legitimate and desirable agreement by patients with their doctors or treatment team to take some medication, and partly the result of wrongful coercion of patients. The Marlboro medical director overruled three of the ten cases which reached his desk since adoption of the Bulletin. Tr. XXIX, 21-22; Ex. D-34, pp. 45-47. The Ancora medical director upheld the treating physician in the two cases referred to him. Tr. XXXIII, 71; Ex. D-35, pp. 24-26. The same statistic applies to Greystone. Tr. XXXI, 14; Ex. P — 47, pp. 29-31. At Trenton, the medical director rescinded half of the six forced medication orders she reviewed. Ex. D-38, pp. 46-49. There is no evidence of any decision reaching the medical director at Gardner. Ex. D-36, pp. 45-47.
The court does not find these statistics themselves indicative of the capability or independence of the medical directors. But these officials can be faulted for their failure to insure that the Bulletin is fully implemented, that patients’ refusals are acknowledged and counted, and that force and threat, of force or punishment is not used to administer medication.
The medical directors asserted in their testimony that they indeed can exercise independent and conscientious reviews. Tr. XXIX, 21-22; Tr. XXXI, 13; Tr. XXXIII, 40. But some of the individual cases described above rebut those assertions. The Greystone medical director even delegated his responsibility to every staff psychiatrist under a clause in the Bulletin which allows another doctor to act for the medical director when he is *868away; however, this interpretation was later overruled by Dr. Rotov. Tr. XXXI, 48^49. The Ancora medical director testified that, in his opinion, an emergency justifying forced medication could last over 30 days. Ex. P-29, p. 31. In fact, the medical directors have not even acknowledged a constitutional right to refuse treatment or followed the court’s guidelines on the law in deciding refusal cases. Tr. XXIX, 21; Tr. XXXIII, 40.
The medical directors’ actions demonstrate a lack of independence and objectivity when reviewing the actions of their staffs. The court believes this stems largely from their responsibilities; they must have the support of their personnel, whose jobs are made easier when patients are subdued by medication. Unfortunately the rights and health of patients are sometimes ignored.
476 F.Supp. at 1305 (emphasis supplied). These findings, not clearly erroneous, certainly support the court’s conclusion that something more than reliance on the Bulletin was appropriate.
The majority disregards these findings and substitutes its own intuitive judgment that the Bulletin will “if carefully followed, pose only a minor risk of erroneous deprivation.” P. 850. It bears repeating, however, that a pattern of patient abuse, mistaken diagnoses, and inadequate record-keeping has been documented. The medical directors have denied the very existence of an underlying right to refuse treatment and the appellants still contest its existence on appeal. Whatever may be said about the contours of procedural due process in the abstract, concretely the court had the obligation to attempt to fashion a temporary injunction which would be effective in bringing the pattern of substantive violations to an end. Plainly it cannot be considered an abuse of discretion for the court to have concluded that a more independent review was required than that proposed as a litigation response by the reluctant defendants. Nor did the court misconstrue Supreme Court precedents or apply erroneous legal standards.1 With respect to the *869need for greater independence, the court found:
Certain testimony concerned plaintiffs’ request for independent review by some type of hearing board or examiner of decisions to forcibly medicate. The court accepts the opinions of those experts who indicated that someone outside the hospital structure would provide a fairer, more accurate review of patients’ refusal rights. Tr. XVII, 85; Tr. XXIV, 108-11; Tr. XXVI, 58; Tr. XXXII, 160-63.
The court also believes that if an independent check on forced medication is established, doctors would learn the legal limits of involuntary medication and the number of attempts to force psychotropics would eventually decrease. This would also mean that the number of cases reaching an independent decision-maker would eventually drop to a relatively small number of disputes. The court cannot specifically gauge the cost of an independent review system, but it rejects the $3 million per year figure provided by defendants, Tr. XXX, 42 — 46, as based on overestimates of the number of reviews and the number of hours required for each hearing.
More sensitivity on the part of treating physicians would also engender better patient-doctor relations. Less use of psychotropics and more attention to patient’s feelings about their treatment would likely improve patient-staff relations and foster patients’ individual dignity. Tr. XXIX, 19, 78-79, 89-90. While the authority granted an outside review officer will initially cause resentment among many doctors and staff members of the hospital, Tr. XXVII, 49; Tr. XXXIII, 43, the court believes that a carefully structured review system would eventually be accepted by most personnel as review under the Bulletin has been accepted. Tr. XXIX, 19; Tr. XXXI, 18; Ex. P-29, p. 47.
476 F.Supp. at 1306. These findings, which involve credibility determinations, are not clearly erroneous. They suggest the need to take steps to educate the reluctant defendant doctors in becoming sensitive to patients’ constitutional rights, the existence of which, even today, they deny.
Moreover the court in the quoted findings rejects the defendants’ overblown estimates of cost of the independent review system it ordered. The majority, while not expressly reversing this credibility judgment, notes its acute awareness of finite state resources (P. 851). However, aside from the exaggerated estimate which the trial court refused to accept, there is no record evidence from which it can be concluded that the system which the trial court ordered will involve any net cost. There is abundant evidence of excessive reliance upon psychotropic drugs. No evidence was presented as to the cost of those drugs, but it is quite possible that the effect of enforcing patients’ rights in this area would be to reduce the hospital’s overall medication budget.2 The majority’s discussion of finite state resources is rank speculation.
Even less defensible is the majority’s decision to vacate that part of the preliminary injunction which directs the defendants to submit to the court monthly reports describing in reasonable detail their implementation of the Administrative Bulletin and the court’s order pending entry of a final order. 476 F.Supp. at 1315. This elementary precaution, designed to enlighten the court as to the appropriate contours of final injunctive relief, has been vacated without discussion, despite the fact that the district court retains jurisdiction to consider at a later date the need for measures to enforce compliance with the procedures contained in Bulletin 78-3. Respectfully, I *870ask my colleagues how that provision can, under any definition of the term, be deemed to be an abuse of discretion. Certainly the judge is entitled to know whether the steps which have been taken as a result of the lawsuit have actually been effective in bringing to an end the pattern of violations of constitutional rights which, he found, were committed by health professionals. Judge Brotman was hardly donning a white coat, as the majority opinion suggests, when insisting that information essential for the formation of an informed judgment at final hearing be preserved and brought to his attention.
Part III of the majority opinion contains at least three fundamental errors. First, it disregards the trial court’s factual findings. Second, it departs from the appropriate standard of review of district court discretion with respect to pendente lite relief. Third, it confuses the standard for injunctive relief from an established pattern of illegal conduct with the distinct question of what, in the abstract and absent such an established pattern, would be an appropriate procedural rule for safeguarding a substantive right. The last error is the most egregious. The case before us is not one in which the defendants admit the existence of the substantive right and plaintiffs direct their challenge solely to the sufficiency of the state procedures on their face. Rather the record demonstrates that in practice these procedures have failed to protect plaintiffs’ substantive liberty interest, thus entitling them to the protection of the court’s broad remedial powers. We are dealing then with more than a dispute over what minimum procedural due process the fourteenth amendment requires for the protection of an acknowledged right. Cf. Matthews v. Eldridge, 424 U.S. 319, 96 S.Ct. 893, 47 L.Ed.2d 18 (1976). Clearly a court, faced with defendants who dispute the very existence of the constitutional right at stake, may supplement such minimum safeguards if, within its sound discretion, it concludes that this is necessary to the fashioning of effective pendente lite relief. See, e. g., Todaro v. Ward, 565 F.2d 48, 52-53 (2d Cir. 1977); Morris v. Travisono, 509 F.2d 1358 (1st Cir. 1975); cf. 28 U.S.C. § 1651. The emotional negative response which our recognition of basic human rights for the involuntarily committed evokes in some quarters should not lead the majority of the court to announce an opinion which in this and other contexts will have unfortunately inhibiting consequences upon the ability of trial judges to fashion effective injunctive relief.
Thus I dissent from Part III of the opinion of the court and from the judgment. I would affirm the district court order in all respects.
. Indeed, the Supreme Court has not foreclosed the possibility that institutional pressures or biases may disqualify a decision-maker; it has required impartiality and detachment and said that a staff physician will suffice “so long as he or she is free to evaluate independently” the evidence. Parham v. J.R., 442 U.S. 589, 607, 99 S.Ct. 2493, 2506, 61 L.Ed.2d 101 (1979); see also Wolff v. McDonnell, 418 U.S. 539, 559, 94 S.Ct. 2963, 2976, 41 L.Ed.2d 935 (1974); Morrissey v. Brewer, 408 U.S. 471, 489, 92 S.Ct. 2593, 2604, 33 L.Ed.2d 484 (1972). In each of these cases, a district court finding that intramural administrators were fit to review decisions made by their subordinates was affirmed by the Court expressly on the record and not as a matter of law. See, e. g., Parham v. J.R., supra, 442 U.S. at 618, 99 S.Ct. at 2512; Wolff v. McDonnell, supra, 418 U.S. at 571, 94 S.Ct. at 2989. Accordingly, even were the injunction before us a permanent one, the question whether medical directors in New Jersey mental hospitals lacked the objectivity necessary to review the actions of their staff should be subject to the clearly erroneous standard of review.
Likewise, the Supreme Court has recognized that some provision for patient advocacy, such as the district court ordered in this case, may be the unavoidable price of due process when those persons whose liberty interest is threatened are particularly vulnerable or helpless. See Vitek v. Jones, 445 U.S. 480, 100 S.Ct. 1254, 63 L.Ed.2d 552 (1980). Frowning on the adversariness implicit in such a decree, the opinion of the court vacates it in reliance upon Parham v. J.R.. The analogy is misplaced, however, since the doctors there served as a disinterested check on the decisions of parents affecting their children. In the context of involuntary drugging, however, the doctors are called upon to review the decisions of colleagues or subordinates within their institution.
In the comparable context of Vitek v. Jones, the Court found the risk of error in determining a prisoner’s need for compelled behavior modification treatment to be “substantial.” Acknowledging that the inquiry involved was essentially medical, and that officials with medical, rather than legal training should make the decisions, the Court nonetheless responded that “[i]t is precisely ‘[t]he subtleties and nuances of psychiatric diagnoses’ that justify the requirement of adversary hearings.” 445 U.S. at 495, 100 S.Ct. at 1264, quoting Addington v. Texas, 441 U.S. 418, 430, 99 S.Ct. 1804, 1811, 60 L.Ed.2d 323 (1979). A majority of the justices agreed that due process under such circumstances required the provision of “independent assistance” to an inmate, so that his interests would be effectively advocated at the hearing. 445 U.S. at 498, 100 S.Ct. at 1266 (Powell, J., concurring).
By discounting the need for independent decision-makers and patient advocates without regard to the risk of error demonstrated in the *869record, the opinion of this court seriously blunts the significance of the foregoing Supreme Court precedents.
. Ancora Hospital reported that because of changes in response to the Rennie litigation, $100,000 was saved on medications in one year. Indeed, comparable savings in other hospitals might offset the cost of independent review and patient advocacy. See A.D. Brooks, The Constitutional Right to Refuse Antipsychotic Medications, 8 Bull.Am.Acad.Psychiat. & L. 179, 202 (1981).