Defendant Dr. Bruce Johnson brings this appeal seeking reversal of the lower court's denial of his motion for summary judgment. Plaintiff Walter Lojuk filed this lawsuit seeking damages allegedly suffered from his receipt of electro-convulsive therapy ("ECT", also known as electro-shock therapy) at the Veterans Administration ("VA") Medical Center, North Chicago, Illinois, in March 1979. For purposes of this appeal, Dr. Johnson, the treating psychiatrist, agrees that Mr. Lojuk received this treatment without his consent and over his family's express objections.1 The parties have stipulated that Dr. Johnson was acting within the scope of his employment as a VA physician at all relevant times.

This case is now before us for the second time. In our previous consideration of this lawsuit, Lojuk v. Quandt, 706 F.2d 1456 (7th Cir.1983) ("Lojuk I "), we held among other things that plaintiff was a de facto involuntary patient and that Dr. Johnson's decision to administer ECT without plaintiff's consent infringed his liberty interest under the Fifth Amendment. Plaintiff alleges both a constitutional tort claim due to this infringement of his liberty interest and a battery claim under Illinois common law. Dr. Johnson argues in defense before us, as he did below, that he has absolute immunity from the common-law claim and qualified immunity from the constitutional claim. The district court disagreed and therefore denied defendant's motion for summary judgment. For reasons to be discussed herein, we affirm the judgment of the district court in part and reverse in part. But before reaching the merits of this appeal, we must first determine whether jurisdiction is properly laid in this Court.

* This Court has "jurisdiction of appeals from all final decisions of the district courts of the United States * * * except where a direct review may be had in the Supreme Court." 28 U.S.C. Sec. 1291. Finality means that the district court has nothing left to do " 'but execute the judgment.' " Coopers & Lybrand v. Livesay, 437 U.S. 463, 467, 98 S. Ct. 2454, 2457, 57 L. Ed. 2d 351 (quoting Catlin v. United States, 324 U.S. 229, 233, 65 S. Ct. 631, 633, 89 L. Ed. 911). Cohen v. Beneficial Industrial Loan Corp., 337 U.S. 541, 546, 69 S. Ct. 1221, 1225, 93 L. Ed. 1528, articulated one very narrow exception to this strict finality rule. The collateral order doctrine of Cohen applies to interlocutory orders that "conclusively determine the disputed question, resolve an important issue completely separate from the merits of the action, and [are] effectively unreviewable on appeal from a final judgment." Coopers & Lybrand, 437 U.S. at 468, 98 S.Ct. at 2458.

We hold, as have other courts before us, see Kenyatta v. Moore, 744 F.2d 1179, 1183 n. 10 (5th Cir.1984) (citing cases), that the denial of absolute immunity is immediately appealable. All three requirements enunciated in Coopers & Lybrand, supra, are met, as they must be for an order to be appealable under the collateral order doctrine. The right to be free from suit, regardless of culpability, is an important issue completely separate from the merits of the action. The district court's denial of absolute immunity conclusively determines the question, because absolute immunity protects the right to be free from trial, rather than just the right to avoid liability. See Powers v. Lightner, 752 F.2d 1251, 1255 (7th Cir.1985); Kenyatta, 744 F.2d at 1183. Consequently, the only effective appellate review is immediate appellate review. Hence the district court's denial of absolute immunity on the common-law claim is reviewable under 28 U.S.C. Sec. 1291.

Whether the claim of qualified immunity is also appealable is a separate question and one that the Supreme Court has recently addressed. In Mitchell v. Forsyth, --- U.S. ----, 105 S. Ct. 2806, 86 L. Ed. 2d 411, the Supreme Court held "that a district court's denial of a claim of qualified immunity, to the extent that it turns on an issue of law, is an appealable 'final decision' within the meaning of 28 U.S.C. Sec. 1291 notwithstanding the absence of a final judgment." Id. at ----, 105 S.Ct. at 2817. Crucial to the Supreme Court's analysis was its determination that qualified immunity, like absolute immunity, constituted "immunity from suit rather than a mere defense to liability." Id. at ----, 105 S.Ct. at 2816 (emphasis in original). Indeed officials often argue the two bases of liability in the alternative--that they are absolutely immune or, if not absolutely immune, qualifiedly immune. See Mitchell (Attorney General argued that he should have absolute immunity for performing his national security functions delegated by the President or at least qualified immunity); Harlow v. Fitzgerald, 457 U.S. 800, 813, 102 S. Ct. 2727, 2735, 73 L. Ed. 2d 396 (senior Presidential aides argued that they were absolutely immune or otherwise had qualified immunity).

Dr. Johnson has stipulated that, as Mr. Lojuk alleges, he ordered the ECT treatment without Mr. Lojuk's consent or the consent of his family. The Supreme Court in Mitchell expressly contemplated immediate appeals from a district court's ruling that "if the facts are as asserted by the plaintiff, the defendant is not immune." --- U.S. at ----, 105 S. Ct. at 2816. Since that language describes precisely the instant situation, Mitchell controls and we have jurisdiction over this appeal as an appealable collateral order.2

We have jurisdiction over the district court's denial of both absolute immunity and qualified immunity. We turn first to the question of absolute immunity.

Absolute immunity is a doctrine largely of judicial origin and evolution. Barr v. Matteo, 360 U.S. 564, 569, 79 S. Ct. 1335, 1338, 3 L. Ed. 2d 1434. Its purpose is to advance "the fearless, vigorous, and effective administration of policies of government," id. at 571, 79 S.Ct. at 1339, by insulating governmental officials from lawsuits based on acts done in the furtherance of their official duties. Although doing so means that "[t]here may be occasional instances of actual injustice which will go unredressed," id. at 576, 79 S.Ct. at 1342, that price is the one paid to preserve innocent officials from the threat and the burden of baseless litigation. It is also the price paid to assure that capable individuals do not shun government service due to the threat of liability.

The basic inquiry is whether the lawsuit is based on state common law, and whether the federal official executed the action complained of within the outer perimeter of his delegated duty. Id. at 574-575, 79 S.Ct. at 1341-1342; Oyler v. National Guard Association of the United States, 743 F.2d 545, 551-553 (7th Cir.1984). Nonetheless, the process is one of balancing the advantages derived from compensating individuals for damage inflicted by government officials against the harm to the public interest incurred by the effect of vindictive or ill-founded damage suits on public officials. Barr, 360 U.S. at 564-565, 79 S.Ct. at 1335-1336. Consequently, the doctrine should not be extended unthinkingly, and each situation should be evaluated to ascertain whether extending immunity would harmonize with the policies underlying the absolute immunity doctrine. Doe v. McMillan, 412 U.S. 306, 320, 93 S. Ct. 2018, 2028, 36 L. Ed. 2d 912. Cf. Mitchell v. Forsyth, --- U.S. at ----, 105 S. Ct. at 2813-2814 (functional approach used to deny blanket immunization of Attorney General's alleged violation of constitutional rights; qualified immunity adopted instead); Ferri v. Ackerman, 444 U.S. 193, 202-205, 100 S. Ct. 402, 408-410, 62 L. Ed. 2d 355 (refusing to afford absolute immunity to a court-appointed lawyer sued by his former client for malpractice). Doing so is especially important when the immunity extended has not been mandated by Congress but is of judicial origin alone.

Although we suggested in Lojuk I, supra, that absolute immunity might be unwarranted in this case, we did not consider the question due to the limited briefing on the issue at that time and because consideration of the issue would not result in disposal of the entire suit. Instead, we remanded to the district court for a fuller consideration. 706 F.2d at 1469. We now hold that absolute immunity should not be accorded to Dr. Johnson.

Understanding our reasons for doing so requires a reexamination of our holding in Lojuk I. There we characterized Mr. Lojuk's medical malpractice claim against Dr. Johnson as one for battery, because he alleged a total lack of consent to ECT. Consequently, his receipt of the treatment constituted an intentional unauthorized touching of his person. Id. at 1460. As a result, he was foreclosed from proceeding against the United States under the Federal Tort Claims Act ("FTCA"), due to its exception for intentional tort cases. Id. at 1462. Had he alleged lack of informed consent, rather than no consent, the complaint would have sounded in negligence, id. at 1460, and a FTCA action would have been available.

We then considered whether the individual VA defendants were immunized from suit by 38 U.S.C. Sec. 4116(a), which applies to suits requesting damages for personal injury "allegedly arising from malpractice or negligence of a physician." Our examination of the statute convinced us that it immunizes VA medical personnel only when the injured person has available a remedy against the United States, such as through the FTCA. Because plaintiff in this case had no such alternative remedy, the immunity from suit in Section 4116(a) did not apply. Lojuk I, 706 F.2d at 1464.

Strongly supporting this determination was Subsection (e), added to Section 4116 in 1973 and providing in relevant part:

The [VA] Administrator may, to the extent the Administrator deems appropriate, hold harmless or provide liability insurance for any person to whom the immunity provisions of this section apply * * *, for damage for personal injury * * *, negligently caused by such person while furnishing medical care or treatment * * * in the exercise of such person's duties * * *, if such person * * * is acting under * * * circumstances which would preclude the remedies of an injured third person against the United States * * * for such damage or injury.

The House Report explaining this provision stated that the statute would "extend this protection [for the defense of negligence and malpractice claims] to cases where Federal Tort claims actions would not lie, but actions could still be brought against the VA employee personally for actions arising in the exercise of his duties." H.R.Rep. No. 368, 93d Cong., 1st Sess., reprinted in 1973 U.S.Code Cong. & Ad.News 1688, 1710. Thus the House identified Subsection (e) as applying to VA doctors in Dr. Johnson's situation, who are sued by individuals without an alternative remedy directly against the United States because their medical malpractice claim sounds in intentional tort. The VA Administrator at the time reinforced this conclusion by explaining that the provision would "provide a means of protecting * * * medical personnel * * * who are sued for assault and battery, * * * in connection with the performance of their assigned duties." S.Rep. No. 776, 92d Cong., 2d Sess. 51 (1972) (reprinted in Lojuk I, 706 F.2d at 1463). Because absolute immunity would attach solely to acts done within the outer perimeter of a government official's duties, the protection provided by Section 4116(e) would have no meaning unless Congress and the head of the agency directly affected by the statute were right in believing that absolute immunity should not and does not extend to this situation.

Giving full effect to Congressional intent in enacting Section 4116(e) requires that we affirm the district court's denial of absolute immunity. Although one could argue that the statute represents mere overkill, this explanation is overly glib in view of the inexplicable duplication. Any other interpretation would render Subsection (e) superfluous, and "[w]e should not and do not suppose that Congress intended to enact unnecessary statutes." Jackson v. Kelly, 557 F.2d 735, 740 (10th Cir.1977) (en banc ) (interpreting 10 U.S.C. Sec. 1089(f), immunizing military medical employees, parallel statute to Section 4116(e)).

Without question a primary Congressional purpose underlying the legislation of which Section 4116(e) was a part was "to improve recruitment and retention of career personnel" in the VA. H.R.Rep. No. 368, supra, reprinted in 1973 U.S.Code Cong. & Ad.News 1688. Dr. Johnson vigorously argues that absolute immunity is essential to effectuate this purpose and to promote the government's policies in providing medical care to veterans. He stressed at oral argument that the threat of liability without medical malpractice insurance, coupled to the lower salaries VA doctors receive, would seriously affect the government's ability to recruit medical personnel. But the contention that insurance coverage is lacking is simply not true, given the protection Congress purposefully provided in 1973 by enacting Section 4116(e).3 We noted in Lojuk I Congress' purpose in immunizing medical employees only if an alternative remedy is available against the United States. Lojuk I, 706 F.2d at 1463 (discussing three similar statutes, 42 U.S.C. Sec. 233 (Public Health Service); 42 U.S.C. Sec. 2458a (NASA); 10 U.S.C. Sec. 1089 (Armed Forces)). The Tenth Circuit explained in Jackson, supra, that Congress intended by enacting 10 U.S.C. Sec. 1089 "to protect military medical personnel from the ever-present danger of personal liability and relieve them of the expense of supplying private malpractice insurance, while preserving a means for compensating malpractice victims for their injuries." 557 F.2d at 740. We believe that Congress had the same twin aims in mind when it enacted Subsection (e).

The reasoning adopted by the Tenth Circuit in Jackson is very persuasive. The court there noted that Congress' decision to provide medical personnel who are sued for malpractice either indemnity or insurance, rather than absolute immunity, indicated an intent to ensure a remedy to victims of malpractice while protecting military employees at the same time. Id. at 740-741. The same reasoning applies with equal force to the instant situation. Congress could have immunized VA medical employees without incurring the added expense of indemnity or insurance. That Congress chose to incur this expense indicates a firm purpose to provide compensation to those who are injured. Congress' choosing to do so avoids the societal cost implicated in denying recompense to injured individuals.

That Section 4116(e) provides the VA Administrator with discretionary, rather than mandatory, authority to provide indemnity or insurance does not change the result. Congress has consistently used such language in enacting such provisions. See 10 U.S.C. Sec. 1089(f) (Armed Forces); 42 U.S.C. Sec. 233(f) (Public Health Service); 42 U.S.C. Sec. 2458a(f) (NASA). Yet previous courts considering the question have assumed insurance or indemnity would be generally available. See, e.g., Jackson, supra. Congress itself places little weight on such discretionary language, assuming that protection will be provided in most instances. See S.Rep. 1264, 94th Cong., 2d Sess. 1, 10, reprinted in 1976 U.S.Code Cong. & Ad.News 4443, 4451-4452 (10 U.S.C. Sec. 1089(f)) (most extensive discussion). Congress has exhibited a strong preference for protecting its medical employees without depriving potential victims of malpractice of a remedy. That Congress has also decided to allow agency heads in some circumstances to deny complete relief, so that a small possibility exists that Dr. Johnson might not be completely indemnified, does not alter the Congressional decision to protect Dr. Johnson and those in his position through statutory remedies rather than through the judicial doctrine of absolute immunity. We recognize that the authority to provide indemnity or insurance was a power that the VA sought to strengthen its ability to recruit the most qualified medical personnel, and that the chances that the VA Administrator would refuse to exercise this power in appropriate circumstances are quite small indeed. In any case, Congress' enacting legislation establishing a preferred scheme of compensation represents a balancing of interests the courts ought not to overset. The argument that Dr. Johnson might still require protection is little more than a red herring.

Also a red herring is the argument that Section 4116(e) is not applicable to malpractice claims like Mr. Lojuk's that sound in intentional tort. Subsection (e) does refer to injuries that are "negligently caused." The House Report, however, notes that "[v]irtually identical provisions were added to the Public Health Service Act by PL 91-623 in 1970 and were included." H.R.Rep. No. 368, 93d Cong., 1st Sess., reprinted in 1973 U.S.Code Cong. & Ad.News 1688, 1710. The referenced provision in the Public Health Service Act is 42 U.S.C. Sec. 233. Section 233 is indeed quite similar to Section 4116, except it is in six subparts and not five. Its last two subsections cover separately what Congress lumped together in Section 4116(e). Section 233(e) waives the FTCA intentional tort exemption for claims of "assault or battery arising out of negligence in the performance of medical * * * functions," while Section 233(f) authorizes the provision of insurance or indemnity to Public Health Service employees detailed abroad, since such injuries are also exempted from the FTCA. Thus patients administered treatment in Illinois without their consent by Public Health Service doctors can bring a FTCA action, while medical employees detailed abroad may rely on the government to provide indemnity or insurance should they be sued. Congress adopted this second method of protecting VA medical personnel, sued for assault or battery arising out of negligence, whether employed inside or outside of the United States.

Although Section 4116(e) does not explicitly provide protection for malpractice claims that technically sound in intentional tort, it must cover such claims. We have recently reiterated that Congress intended in enacting Section 4116 to grant broad immunity protection to VA medical personnel. Quilico v. Kaplan, 749 F.2d 480, 486-487 (1984). In fact we accepted there our earlier intimation that Section 4116(e) provides protection for assault and battery claims arising out of negligence. Id. at 487 n. 13. That Congress chose to provide indemnity or insurance to those sued for assault or battery claims arising out of negligence, rather than waiving the FTCA exemption from such claims as Section 233 and other similar statutes did, does not detract from the clear intent of the legislation. Rather, Congress' indication that Section 233 was its model, which explicitly encompasses assault and battery claims arising out of negligence, infers that such coverage is implicit in Section 4116(e) and that Congress understood that assault and battery claims arising out of medical malpractice were within the statute's reach. The House Report implies as much, and the VA Director at the time, who cooperated closely with Congress in getting the legislation passed, explicitly said so.

The driving force behind Section 4116, as behind analogous statutory provisions, is protecting medical personnel without depriving those who are injured of a remedy. Section 4116(a) authorizes a FTCA claim against the government directly for medical malpractice or negligence claims; Section 4116(e) operates to provide insurance or indemnity only if some exception in the FTCA bars direct action against the government under Subsection (a) for an employee's malpractice. Certainly courts interpreting statutes analogous to Section 4116 have understood Section 4116 to have the same scope, as well as the same Congressional purposes, as the statutes waiving the FTCA's exemption for assault and battery claims that might arise out of medical malpractice. See, e.g., Mendez v. Belton, 739 F.2d 15, 19 (1st Cir.1984) (42 U.S.C. Sec. 233). A strictly literal reading of the word "negligently" to exclude intentional torts arising out of medical malpractice would contradict the broad protection, to both employees and those who might be victims of malpractice, that Congress authorized in Section 4116 and similar statutes.

Because Section 4116(e) provides for indemnification and insurance, Dr. Johnson's argument that absolute immunity is necessary fails. If he were afforded the immunity he requests, we would be imposing a significant cost on Mr. Lojuk with no counterbalancing advancement of governmental policies. Doing so runs counter to the reasoning in Barr, supra, justifying the doctrine of absolute immunity. Doing so would also contravene the discriminating application of absolute immunity necessary to ensure that its application does not exceed its justification. Doe, supra. The existence of Section 4116(e) sets Mr. Lojuk's claim apart from other cases in which we have considered the doctrine of absolute immunity. See Oyler v. National Guard Association of the United States, supra; Scherer v. Morrow, 401 F.2d 204 (7th Cir.1968), certiorari denied, 393 U.S. 1084, 89 S. Ct. 868, 21 L. Ed. 2d 777; Skolnick v. Campbell, 398 F.2d 23 (7th Cir.1968); Scherer v. Brennan, 379 F.2d 609 (7th Cir.1967), certiorari denied, 389 U.S. 1021, 88 S. Ct. 592, 19 L. Ed. 2d 666; LeBurkien v. Notti, 365 F.2d 143 (7th Cir.1966).4 In this line of precedent, absolute immunity clearly promoted significant and important governmental policies, as for example in Scherer involving the Secret Service or in Oyler involving the National Guard. Congressional enactment of Section 4116(e), however, emphasizes the importance to Congress not only of protecting VA medical personnel but also of according recompense to the victims of medical malpractice or negligence--so that the improvement of the VA's medical capabilities does not come at the expense of the general public. Had Mr. Lojuk alleged some other kind of injury--defamation, for instance--then the protection provided by Section 4116(e) would be unavailable. Cf. Mendez v. Belton, 739 F.2d at 19-20 (42 U.S.C. Sec. 233). In that situation the policy reasons that bolstered our earlier decisions providing absolute immunity would operate to preclude liability.

Our decision today accords with the holdings of at least three other circuits. Davis v. Knud-Hansen Memorial Hospital, 635 F.2d 179 (3d Cir.1980); Jackson v. Kelly, 557 F.2d 735 (10th Cir.1977) (en banc ); Henderson v. Bluemink, 511 F.2d 399 (D.C.Cir.1974).5 These three cases distinguish between governmental policy decisions and medical treatment decisions and argue that absolute immunity applies only to the former.6 This distinction is fruitful insofar as it requires a court to examine whether absolute immunity in a given situation will accord with its policy underpinnings. We do not, however, expressly rely on the rationale because our examination of the statutory immunity question leaves little doubt that Congress intended VA medical personnel to rely on the protection offered by Section 4116(e) rather than judicially created absolute immunity.

These circuits have also imported a ministerial-discretionary distinction into their analysis of the absolute immunity question, a distinction that is unfounded and which this Circuit has never adopted. Cf. Oyler, 743 F.2d at 554 (noting that Illinois state immunity is not as broad as federal immunity, in part because Illinois provides immunity "for discretionary acts taken within the scope of [state officials'] duty," inferring that federal immunity extends to both discretionary and mandatory acts).7 The adoption of this distinction apparently derives from references by Justice Harlan in his plurality opinion in Barr, supra, that petitioner had appropriately exercised his discretion. 360 U.S. at 574-575, 79 S. Ct. at 1341-1342. Justice Harlan also emphasized that a person's rank was not material to a determination of immunity, although higher-ranking officials would probably receive more protection because such people are afforded a wider range of discretion, making more actions fall within the outer perimeter of their duties. Id. at 573, 79 S.Ct. at 1340. But the Court apparently meant that an official could receive immunity for discretionary acts, as well as for mandatory or ministerial ones. In justifying affording immunity for acts not required by law but undertaken in the individual's discretion, Justice Harlan wrote,

That petitioner was not required by law or by direction of his superiors to speak out cannot be controlling in the case of an official of policy-making rank, for the same considerations which underlie the recognition of the privilege as to acts done in connection with a mandatory duty apply with equal force to discretionary acts at those levels of government where the concept of duty encompasses the sound exercise of discretionary authority.

Id. at 575, 79 S.Ct. at 1341 (emphasis in original) (footnote omitted). In other words, Justice Harlan was explaining how immunity for discretionary decisions could be justified as being as much within an official's duty as his mandatory duties. The clear inference is that the discretionary-ministerial distinction is unfounded.8

As suggested above, the result reached by the Third, Tenth, and D.C. Circuits can be justified without resort to the ministerial-discretionary distinction, by arguing that absolute immunity is appropriate for governmental policy decisions alone and a medical treatment decision simply has no relation to the effective administration of government. This view has some force, especially when a statute such as Section 4116(e) has been enacted that protects the employee who is sued. With such a statute in force the defendant cannot persuasively argue that absolute immunity would undermine the government's recruitment efforts or dampen an individual's effectiveness in carrying out his assigned duties. To the extent Dr. Johnson would be distracted from his duties with time-consuming litigation, that result is one Congress intended when it chose to protect him by providing him with insurance or indemnity, thereby allowing a trial to proceed with the requisite testimony of Dr. Johnson. Even a direct remedy against the United States such as a FTCA action would require Dr. Johnson's testimony in court. Congress evidently believes that involving medical employees in trial has few deleterious effects, as long as they are otherwise protected from personal liability on damages that may be awarded, and that compensating victims for their injuries is far more important. Our quarreling with this Congressional decision would be inappropriate.

Indeed, the Third, Tenth, and D.C. Circuits all made reference either to Section 4116(e) or to a closely parallel statute to bolster the result reached. Davis, 635 F.2d at 187 n. 7 (Virgin Islands statute, V.I.Code Ann. tit. 27, Sec. 166e); Jackson, 557 F.2d at 740-741 (10 U.S.C. Sec. 1089(f)); Henderson, 511 F.2d at 404 nn. 28, 30 and accompanying text (Section 4116(e)). See also Burchfield, 516 F. Supp. at 1304 (10 U.S.C. Sec. 1089(f)). As noted above, Jackson paid great attention to the significance of Congressional enactment of an immunity statute like Section 4116(e), reliance that was surely appropriate. The proper focus is on the practical effects of allowing an official to be sued, for only then can a court assure itself that the justifications underlying absolute immunity are being promoted and society best served. The process is indeed one of balancing the individual's interests against those of the government, see Barr, 360 U.S. at 571-572, 79 S.Ct. at 1339-1340; id. at 578, 79 S.Ct. at 1343 (Warren, C.J., dissenting), a procedure that cannot be ignored if we are to ensure that the doctrine of absolute immunity serves its purpose.

Case law Dr. Johnson cites in support of his position is singularly unpersuasive. The majority of cases relied on are quite old, predating Congress' enactment of Section 4116(e). Because congressional enactment obviates the concern expressed in these earlier cases for the impact of liability on an individual's willingness to serve in the government, the reasoning expressed in such cases loses all force. The more recent cases cited are similarly unconvincing. Harris v. Tennessee Valley Authority, 507 F. Supp. 318 (E.D.Tenn.1980), does not involve a VA physician and follows Sixth Circuit precedent expressed in Estate of Burks v. Ross, 438 F.2d 230 (6th Cir.1971)--but without any discussion of possible statutory immunity or the impact that the Congressional enactment of Section 4116(e) in 1973 might have on the continued force of Burks. Claus v. Gyorkey, 674 F.2d 427 (5th Cir.1982), did involve the VA but not medical malpractice liability. The same is true of Vest v. Waring, 565 F. Supp. 674 (N.D.Ga.1983). We need only mention that insofar as cases such as Claus involve employment decisions, which are obviously central to the smooth, effective functioning of government agencies but can often give birth to discord and conflict, absolute immunity is appropriate and nothing in today's opinion infers otherwise. See also supra p. 626.

Having found that the district court correctly denied absolute immunity to Dr. Johnson on Mr. Lojuk's malpractice claim of battery, we next consider the district court's denial of qualified immunity to Dr. Johnson on Mr. Lojuk's constitutional tort claim.

In Lojuk I we recognized that the plaintiff had "a constitutional liberty interest in avoiding the unwanted administration of ECT that must be protected under the Due Process Clause" of the Fifth Amendment. Lojuk I, 706 F.2d at 1465. On remand Dr. Johnson raised the defense of qualified immunity. He can prevail on this defense if he can establish that his conduct did "not violate clearly established statutory or constitutional rights of which a reasonable person would have known." Harlow v. Fitzgerald, 457 U.S. 800, 818, 102 S. Ct. 2727, 2738, 73 L. Ed. 2d 396. The Supreme Court adopted this objective test to accommodate better the competing interests between the injured individual's need for compensation and the governmental and societal costs entailed in subjecting government officials to the threat of lawsuits and the demands of litigation. Id. at 814, 102 S.Ct. at 2736. Officials who violate laws that are not clearly established at the time, regardless of later evolution in the law, are not answerable in damages; others are.

In the case at bar, Mr. Lojuk is unable to demonstrate that the liberty interest at stake was clearly established in March 1979. To meet this standard we have "required caselaw which clearly and consistently recognized the constitutional right." Coleman v. Frantz, 754 F.2d 719, 730 n. 15 (1985). Specifically, that we were able to define the liberty interest at stake and find that it was protected by the Constitution is insufficient to determine whether the right was clearly established at the pertinent time. A reasonable government official cannot necessarily be expected to recognize the significance of a few scattered cases from disparate areas of the law for a right that is just evolving. Courts have given force to the qualified immunity concept to protect officials from liability imposed for conduct that from the vantage point of 1985 appears obviously to infringe constitutional rights but six years ago did not so clearly do so. Case law may exist that legitimates our recognition of the protected liberty interest, but it does not necessarily mean that the right was also clearly established. While cases involving the exact fact pattern at bar are unnecessary, case law in a closely analogous area is crucial to permit us to conclude that reasonably diligent government officials would have known of the case law, related it to the situation at hand, and molded their conduct accordingly.

In attempting to prove that the right to avoid unwanted administration of ECT was clearly established, Mr. Lojuk has pointed to several statutes, both federal and state. Last year the Supreme Court clarified the role statutes play in determining whether a constitutional right is clearly established. See Davis v. Scherer, --- U.S. ----, 104 S. Ct. 3012, 82 L. Ed. 2d 139. Plaintiff argued there that Florida officials had violated his clearly established constitutional rights, but the Court rejected out of hand the contention that the law had been clearly established. Id. at ----, 104 S.Ct. at 3018-3019. Plaintiff argued in the alternative that even if defendants had not violated clearly established constitutional rights, they should be denied qualified immunity because they had violated a state administrative regulation. The Court also rejected this contention, holding firm to the "clearly established" analysis laid out in Harlow, supra. Id. at ----, 104 S.Ct. at 3019-3021.

In so holding the Court did not reject referring to statutes to prove the existence of clearly established constitutional rights, provided the statute in question has some bearing on the constitutional right, as when the statute creates the substantive right that the Due Process Clause then protects. Id. at ---- n. 11, 104 S. Ct. at 3019, n. 11. In the case at bar the right Mr. Lojuk asserts rises not from a statute but directly out of the Fifth Amendment's concept of liberty, rendering any reference to applicable statutes tenuous at best. Most probative is 38 U.S.C. Sec. 4131, which requires the VA Administrator to assure that "to the maximum extent practicable, all patient care furnished under this title shall be carried out only with the full and informed consent of the patient or subject or, in appropriate cases, a representative thereof." Unquestionably Dr. Johnson violated this statute, as he also violated a hospital memorandum in effect at the time that required the family's consent before administering ECT to a patient unable himself to give consent. But statutes most frequently establish rights and procedural protections not constitutionally required, so that the mere existence of the statute is not adequate proof that a constitutional right is clearly established. The Supreme Court warned in Davis that holding officials liable "for violation of any constitutional right--one that was not clearly defined or perhaps not even foreshadowed at the time of the alleged violation--merely because their official conduct also violated some statute or regulation," --- U.S. at ----, 104 S. Ct. at 3020, violates the policy and rule established in Harlow. The federal statute at issue does not itself generate the liberty interest implicated. The violation of a solitary statute in circumstances in which relevant case law is scattered and almost nonexistent does not point to the existence of a clearly established constitutional right.9

The major portion of the case law cited by plaintiff is inapposite to determining whether the right to refuse ECT treatment was clearly established. In Lojuk I we expressly held that Eighth Amendment interests were not implicated in the case at bar. Lojuk I, 706 F.2d at 1464-1465. Consequently cases identifying the right recognized as arising in part from the Eighth Amendment are irrelevant for determining whether Dr. Johnson should have realized that the Fifth Amendment itself required him to obtain either the patient's consent or the consent of his family. Mr. Lojuk's reference to such cases is not probative. See Mackey v. Procunier, 477 F.2d 877 (9th Cir.1973); Wheeler v. Glass, 473 F.2d 983 (7th Cir.1973).

Likewise Mr. Lojuk's citation of Winters v. Miller, 446 F.2d 65 (2d Cir.1971), certiorari denied, 404 U.S. 985, 92 S. Ct. 450, 30 L. Ed. 2d 369, is unconvincing. Winters involved the forced medication of a Christian Scientist, whose religious beliefs precluded resort to drugs. The First Amendment right to practice one's religion free of state interference is quite distinct from the First Amendment interest in being able to think and communicate freely that we recognized as implicated by Mr. Lojuk's claim. Lojuk I, 706 F.2d at 1465.

Closer on point is our decision in Nelson v. Heyne, 491 F.2d 352 (7th Cir.1974), recognizing the rights of juveniles incarcerated in a medium security state correctional institution. Two arguments examined there relate to the case at bar. One involves the forced intramuscular administration of tranquilizing drugs; the other relates to the juveniles' claimed right to rehabilitation and treatment. Because our analysis of the first argument turned on the Eighth Amendment, id. at 357, that aspect of Nelson is inapposite to the instant situation. Insofar as our discussion of the second argument concerned the right to treatment, rather than mistreatment, it is also irrelevant. We agree with the defendants that Mr. Lojuk's attempt to state a claim for a right to treatment is really a restatement of his claim of mistreatment. The plaintiff does not claim that he was given no treatment, but that he was administered the wrong treatment, and that against his wishes.

Nonetheless, our discussion in Nelson regarding the adequacy of treatment as a subset of the right to rehabilitative treatment bears on the instant situation. In fact we stated there "that the juvenile process has elements of both the criminal and mental health processes." Id. at 360. But a close reading of the opinion reveals its inapplicability to the case at bar. Our discussion revolved around the juvenile justice system, so that the leap is fairly far from what is required for juveniles because they are imprisoned to what liberty interests regarding treatment decisions are retained by an involuntary mental patient. More importantly, the discussion still centered around the youths' right to individual treatment to avoid being "warehoused" rather than rehabilitated. Id. The issue of consent, therefore, was not involved in our discussion, and the case certainly cannot stand as clear proof of the existence of an established liberty interest of the type infringed in Mr. Lojuk's case.

While Ingraham v. Wright, 430 U.S. 651, 97 S. Ct. 1401, 51 L. Ed. 2d 711, involving the corporal punishment of schoolchildren, did reemphasize the importance of a person's freedom from bodily restraint and punishment, see id. at 673-674, 97 S.Ct. at 1413-1414, a right Mr. Lojuk's claim implicates, see Lojuk I, 706 F.2d at 1465, Ingraham itself discusses that right in a context fairly far removed from that of the involuntarily committed mental patients. Runnels v. Rosendale, 499 F.2d 733 (9th Cir.1974), discussed this right in the prison context, a case we recognize as being closely related to the situation at bar. Likewise, Rennie v. Klein, 462 F. Supp. 1131 (D.N.J.1978), affirmed, 653 F.2d 836 (3d Cir.1981) (en banc ), vacated in light of Youngberg v. Romeo, 458 U.S. 1119, 102 S. Ct. 3506, 73 L. Ed. 2d 1381 (1982), on remand, 720 F.2d 266 (3rd Cir.1983) (en banc ), involved the forced medication of involuntarily committed mental patients.10 The district court in that case easily found a due process right to refuse treatment in 1978, which would suggest a clearly established right. But the force of this inference is undercut by the decision of a sister district court in the same circuit at roughly the same time that the Eighth Amendment alone served as the source of rights under the Fourteenth Amendment for the involuntarily committed. Romeo v. Youngberg, 644 F.2d 147, 156 (3d Cir.1980) (en banc ), affirmed, 457 U.S. 307, 102 S. Ct. 2452, 73 L. Ed. 2d 28. There the Third Circuit held that defendants would have a qualified immunity if they were not aware or had no reason to be aware of plaintiff prisoners' constitutional rights. 644 F.2d at 171.

In summary, the most that plaintiff can point to is one district court case (Rennie ), one circuit court case (Runnels ), and several other decisions that are distantly related, at best.11 Such a showing is insufficient to prove that the right to refuse treatment was clearly established, so that we are barred from holding Dr. Johnson accountable for his violation of Mr. Lojuk's constitutional liberty interest. The district court's denial of qualified immunity is reversed, but its denial of absolute immunity is affirmed, and the case is remanded for trial. Each party will bear his own costs on appeal.