dissenting.
There is always an unappealing aspect to the use of doctors and nurses, ministers of mercy, to obtain incriminating evidence against the supposed objects of their ministration— although here, it is correctly pointed out, the doctors and *92nurses were ministering not just to the mothers but also to the children whom their cooperation with the police was meant to protect. But whatever may be the correct social judgment concerning the desirability of what occurred here, that is not the issue in the present case. The Constitution does not resolve all difficult social questions, but leaves the vast majority of them to resolution by debate and the democratic process — which would produce a decision by the citizens of Charleston, through their elected representatives, to forbid or permit the police action at issue here. The question before us is a narrower one: whether, whatever the desirability of this police conduct, it violates the Fourth Amendment’s prohibition of unreasonable searches and seizures. In my view, it plainly does not.
I
The first step in Fourth Amendment analysis is to identify the search or seizure at issue. What petitioners, the Court, and to a lesser extent the concurrence really object to is not the urine testing, but the hospital’s reporting of positive drug-test results to police. But the latter is obviously not a search. At most it may be a “derivative use of the product of a past unlawful search,” which, of course, “work[s] no new Fourth Amendment wrong” and “presents a question, not of rights, but of remedies.” United States v. Calandra, 414 U. S. 338, 354 (1974). There is only one act that could conceivably be regarded as a search of petitioners in the present case: the taking of the urine sample. I suppose the testing of that urine for traces of unlawful drugs could be considered a search of sorts, but the Fourth Amendment protects only against searches of citizens’ “persons, houses, papers, and effects”; and it is entirely unrealistic to regard urine as one of the “effects” (i. e., part of the property) of the person who has passed and abandoned it. Cf. California v. Greenwood, 486 U. S. 35 (1988) (garbage left at curb is not property protected by the Fourth Amendment). Some would argue, *93I suppose, that testing of the urine is prohibited by some generalized privacy right “emanating” from the “penumbras” of the Constitution (a question that is not before us); but it is not even arguable that the testing of urine that has been lawfully obtained is a Fourth Amendment search. (I may add that, even if it were, the factors legitimizing the talcing of the sample, which I discuss below, would likewise legitimize the testing of it.)
It is rudimentary Fourth Amendment law that a search which has been consented to is not unreasonable. There is no contention in the present case that the urine samples were extracted forcibly. The only conceivable bases for saying that they were obtained without consent are the contentions (1) that the consent was coerced by the patients’ need for medical treatment, (2) that the consent was uninformed because the patients were not told that the tests would include testing for drugs, and (3) that the consent was uninformed because the patients were not told that the results of the tests would be provided to the police.1 (When the court below said that it was reserving the factual issue of consent, see 186 F. 8d 469, 476 (CA4 1999), it was referring at most to these three — and perhaps just to the last two.)
*94Under our established Fourth Amendment law, the last two contentions would not suffice, even without reference to the special-needs doctrine. The Court's analogizing of this case to Miranda v. Arizona, 384 U. S. 436 (1966), and its claim that “standards of knowing waiver" apply, ante, at 85, are flatly contradicted by our jurisprudence, which shows that using lawfully (but deceivingly) obtained material for purposes other than those represented, and giving that material or information derived from it to the police, is not unconstitutional. In Hoffa v. United States, 385 U. S. 293 (1966), “[t]he argument [was] that [the informant’s] failure to disclose his role as a government informant vitiated the consent that the petitioner gave” for the agent's access to evidence of criminal wrongdoing, id., at 300. We rejected that argument, because “the Fourth Amendment [does not protect] a wrongdoer’s misplaced belief that a person to whom he voluntarily confides his wrongdoing will not reveal it.” Id., at 302. Because the defendant had voluntarily provided access to the evidence, there was no reasonable expectation of privacy to invade. Abuse of trust is surely a sneaky and ungentlemanly thing, and perhaps there should be (as there are) laws against such conduct by the government. See, e. g., 50 U. S. C. §403-7 (1994 ed., Supp. IV) (prohibiting the “Intelligence Community['s]” use of journalists as agents). That, however, is immaterial for Fourth Amendment purposes, for “however strongly a defendant may trust an apparent colleague, his expectations in this respect are not protected by the Fourth Amendment when it turns out that the colleague is a government agent regularly communicating with the authorities.” United States v. White, 401U. S. 745, 749 (1971) (emphasis added). The Hoffa line of eases, I may note, does not distinguish between operations meant to catch a criminal in the act, and those meant only to gather evidence of prior wrongdoing. See, e.g., United States v. Miller, 425 U. S. 435, 440-443 (1976); cf. Illinois v. Perkins, 496 U. S. 292, 298 (1990) (relying on Hoffa in holding the *95Miranda rule did not require suppression of an inmate confession given an agent posing as a fellow prisoner).
Until today, we have never held — or even suggested — that material which a person voluntarily entrusts to someone else cannot be given by that person to the police, and used for whatever evidence it may contain.2 Without so much as discussing the point, the Court today opens a hole in our Fourth Amendment jurisprudence, the size and shape of which is entirely indeterminate. Today’s holding would be remarkable enough if the confidential relationship violated by the police conduct were at least one protected by state law. It would be surprising to learn, for example, that in a State which recognizes a spousal evidentiary privilege the police cannot use evidence obtained from a cooperating husband or wife. But today’s holding goes even beyond that, since there does not exist any physician-patient privilege in South Carolina. See, e. g., Peagler v. Atlantic Coast R. R. Co., 232 S. C. 274, 101 S. E. 2d 821 (1958). Since the Court declines even to discuss the issue, it leaves law enforcement officials entirely in the dark as to when they can use incriminating evidence obtained from “trusted” sources.3 Presumably the *96lines will be drawn in the ease-by-ease development of a whole new branch of Fourth Amendment jurisprudence, taking yet another social judgment (which confidential relationships ought not be invaded by the police) out of democratic control, and confiding it to the uncontrolled judgment of this Court — uncontrolled because there is no common-law precedent to guide it. I would adhere to our established law, which says that information obtained through violation of a relationship of trust is obtained eonsensually, and is hence not a search.4
*97There remains to be considered the first possible basis for invalidating this search, which is that the patients were coerced to produce their urine samples by their necessitous circumstances, to wit, their need for medieal treatment of their pregnancy. If that was coercion, it was not coercion applied by the government — and if such nongovernmental coercion sufficed, the police would never be permitted to use the ballistic evidence obtained from treatment of a patient with a bullet wound. And the Fourth Amendment would invalidate those many state laws that require physicians to report gunshot wounds,5 evidence of spousal abuse,6 and (like the South Carolina law relevant here, see S. C. Code Ann. §20-7-510 (2000)) evidence of child abuse.7
*98II
I think it clear, therefore, that there is no basis for saying that obtaining of the urine sample was unconstitutional. The special-needs doctrine is thus quite irrelevant, since it operates only to validate searches and seizures that are otherwise unlawful. In the ensuing discussion, however, I shall assume (contrary to legal precedent) that the taking of the urine sample was (either because of the patients’ necessitous circumstances, or because of failure to disclose that the urine would be tested for drugs, or because of failure to disclose that the results of the test would be given to the police) coerced. Indeed, I shall even assume (contrary to common sense) that the testing of the urine constituted an uneon-sented search of the patients’ effects. On those assumptions, the special-needs doctrine would become relevant; and, properly applied, would validate what was done here.
The conclusion of the Court that the special-needs doctrine is inapplicable rests upon its contention that respondents “undert[ook] to obtain [drug] evidence from their patients” not for any medical purpose, but “for the specific purpose of incriminating those patients.” Ante, at 85 (emphasis in original). In other words, the purported medical rationale was merely a pretext; there was no special need. See Skinner v. Railway Labor Executives’ Assn., 489 U. S. 602, 621, n. 5 (1989). This contention contradicts the District Court’s finding of fact that the goal of the testing policy “was not to arrest patients but to facilitate their treatment and protect both the mother and unborn child.” App. to Pet. for Cert. A-38.8 This finding is binding upon us unless clearly erro*99neous, see Fed. Rule Civ. Proe. 52(a). Not only do I find it supportable; I think any other finding would have to be overturned.
The cocaine tests started in April 1989, neither at police suggestion nor with police involvement. Expectant mothers who tested positive were referred by hospital staff for substance-abuse treatment, ante, at 70 (opinion of the Court) — an obvious health benefit to both mother and child. See App. 48 (testimony that a single use of cocaine can cause fetal damage). And, since “[ijnfants whose mothers abuse cocaine during pregnancy are born with a wide variety of physical and neurological abnormalities,” ante, at 89 (Kennedy, J., concurring in judgment), whieh require medical attention, see Brief in Opposition A76-A77, the tests were of additional medical benefit in predicting needed postnatal treatment for the child. Thus, in their origin — before the police were in any way involved — the tests had an immediate, not merely an "ultimate,” ante, at 82 (opinion of the Court), purpose of improving maternal and infant health. Several months after the testing had been initiated, a nurse discovered that local police were arresting pregnant users of eocaine for child abuse, the hospital’s general counsel wrote the county solicitor to ask "what, if anything, our Medical Center needs to do to assist you in this matter,” App. 499 (South Carolina law requires child abuse to be reported, see S. C. Code Ann. §20-7-510), the police suggested ways to avoid tainting evidence, and the hospital and police in conjunction used the testing program as a means of securing what the Court calls the "ultimate” health benefit of coercing drug-abusing mothers into drug treatment. See ante, at 70-73,82. Why would there be any reason to believe that, once *100this policy of using the drug tests for their "ultimate” health benefits had been adopted, use of them for their original, immediate, benefits somehow disappeared, and testing somehow became in its entirety nothing more than a “pretext” for obtaining grounds for arrest? On the face of it, this is incredible. The only evidence of the exclusively arrest-related purpose of the testing adduced by the Court is that the police-cooperation policy itself does not describe how to care for cocaine-exposed infants. See ante, at 73, 82. But of course it does not, since that policy, adopted months after the cocaine testing was initiated, had as its only health object the “ultimate” goal of inducing drug treatment through threat of arrest. Does the Court really believe (or even hope) that, once invalidation of the program challenged here has been decreed, drug testing will cease?
In sum, there can be no basis for the Court’s purported ability to “distinguislh.] this case from circumstances in which physicians or psychologists, in the course of ordinary medical procedures aimed at helping the patient herself, come across information that... is subject to reporting requirements,” ante, at 80-81, unless it is this: That the addition of a law-enforcement-related purpose to a legitimate medical purpose destroys applicability of the “special-needs” doctrine. But that is quite impossible, since the special-needs doctrine was developed, and is ordinarily employed, precisely to enable searches by law enforcement officials who, of course, ordinarily have a law enforcement objective. Thus, in Griffin v. Wisconsin, 483 U. S. 868 (1987), a probation officer received a tip from a detective that petitioner, a felon on probation, possessed a firearm. Accompanied by police, he conducted a warrantless search of petitioner’s home. The weapon was found and used as evidence in the probationer’s trial for unlawful possession of a firearm. See id., at 870-872. Affirming denial of a motion to suppress, we concluded that the “special need” of assuring compliance with terms of release *101justified a warrantless search of petitioner’s home. Notably, we observed that a probation officer is not
“the police officer who normally conducts searches against the ordinary citizen. He is an employee of the State Department of Health and Social Services who, while assuredly charged with protecting the public interest, is also supposed to have in mind the welfare of the probationer .... In such a setting, we think it reasonable to dispense with the warrant requirement.” Id., at 876-877.
Like the probation officer, the doctors here do not “ordinarily condue[t] searches against the ordinary citizen,” and they are “supposed to have in mind the welfare of the [mother and child].” That they have in mind in addition the provision of evidence to the police should make no difference. The Court suggests that if police involvement in this case was in some way incidental and after-the-fact, that would make a difference in the outcome. See ante, at 80-84. But in Griffin, even more than here, police were involved in the search from the very beginning; indeed, the initial tip about the gun came from a detective. Under the factors relied upon by the Court, the use of evidence approved in Griffin would have been permitted only if the parole officer had been untrained in ehain-of-eustody procedures, had not known of the possibility a gun was present, and had been unaccompanied by police when he simply happened upon the weapon. Why any or all of these is constitutionally significant is baffling.
Petitioners seek to distinguish Griffin by observing that probationers enjoy a lesser expectation of privacy than does the general public. That is irrelevant to the point I make here, which is that the presence of a law enforcement purpose does not render the special-needs doctrine inapplicable. In any event, I doubt whether Griffin’s reasonable expectation of privacy in his home was any less than petitioners’ reasonable expectation of privacy in their urine taken, *102or in the urine tests performed, in a hospital — especially in a State such as South Carolina, which recognizes no physician-patient testimonial privilege and requires the physician’s duty of confidentiality to yield to public policy, see McCormick v. England, 328 S. C. 627, 633, 640-642, 494 S. E. 2d 431, 434, 438-439 (App. 1997); and which requires medical conditions that indicate a violation of the law to be reported to authorities, see, e.g., S. C. Code Ann. §20-7-510 (2000) (child abuse). Cf. Whalen v. Roe, 429 U. S. 589, 597-598 (1977) (privacy interest does not forbid government to require hospitals to provide, for law enforcement purposes, names of patients receiving prescriptions of frequently abused drugs).
The concurrence makes essentially the same basic error as the Court, though it puts the point somewhat differently: “The special needs cases we have decided,” it says, “do not sustain the active use of law enforcement... as an integral part of a program which seeks to achieve legitimate, civil objectives.” Ante, at 88. Griffin shows that is not true. Indeed, Griffin shows that there is not even any truth in the more limited proposition that our cases do not support application of the special-needs exception where the “legitimate, civil objectives” are sought only through the use of law enforcement means. (Surely the parole officer in Griffin was using threat of reincareeration to assure compliance with parole.) But even if this latter proposition were true, it would invalidate what occurred here only if the drug testing sought exclusively the “ultimate” health benefits achieved by coercing the mothers into drug treatment through threat of prosecution. But in fact the drug testing sought, independently of law enforcement involvement, the “immediate” health benefits of identifying drug-impaired mother and child for necessary medical treatment. The concurrence concedes that if the testing is conducted for medical reasons, the fact that “prosecuting authorities then adopt legitimate procedures to discover this information and prosecution follows *103... ought not to invalidate the testing.” Ante, at 90 (emphasis added). But here the police involvement in each case did take place after the testing was conducted for independent reasons. Surely the concurrence cannot mean that no police-suggested procedures (such as preserving the chain of custody of the urine sample) can be applied until after the testing; or that the police-suggested procedures must have been designed after the testing. The facts in Griffin (and common sense) show that this cannot be so. It seems to me that the only real distinction between what the concurrence must reasonably be thought to be approving, and what we have here, is that here the police took the lesser step of initially threatening prosecution rather than bringing it.
* * *
As I indicated at the outset, it is not the function of this Court — at least not in Fourth Amendment cases — to weigh petitioners’ privacy interest against the State’s interest in meeting the crisis of “crack babies” that developed in the late 1980’s. I cannot refrain from observing, however, that the outcome of a wise weighing of those interests is by no means clear. The initial goal of the doctors and nurses who conducted cocaine testing in this case was to refer pregnant drug addicts to treatment centers, and to prepare for necessary treatment of their possibly affected children. When the doctors and nurses agreed to the program providing test results to the police, they did so because (in addition to the fact that child abuse was required by law to be reported) they wanted to use the sanction of arrest as a strong incentive for their addicted patients to undertake drug-addiction treatment. . And the police themselves used it for that benign purpose, as is shown by the fact that only 30 of 253 women testing positive for cocaine were ever arrested, and only 2 of those prosecuted. See App. 1125-1126. It would not be unreasonable to conclude that today’s judgment, authorizing the assessment of damages against the county *104solicitor and individual doctors and nurses who participated in the program, proves once again that no good deed goes unpunished.
But as far as the Fourth Amendment is concerned: There was no unconsented search in this case. And if there was, it would have been validated by the special-needs doctrine. For these reasons, I respectfully dissent.
The Court asserts that it is improper to “disaggregate the taking and testing of the urine sample from the reporting of the results to the police,” because “in our special needs cases, we have routinely treated urine screens taken by state agents as.searches within the meaning of the Fourth Amendment.” Ante, at 76, n. 9. But in all of those cases, the urine was obtained involuntarily. See Chandler v. Miller, 520 U. S. 305 (1997); Vernonia School Dist. 47J v. Acton, 515 U.S. 646 (1995); Skinner v. Railway Labor Executives' Assn., 489 U.S. 602 (1989); Treasury Employees v. Von Raab, 489 U. S. 656 (1989). Where the taking of the urine sample-is uneonsented (and thus a Fourth Amendment search), the subsequent testing and reporting of the results to the police are obviously part of (or infected by) the same search; but where, as here, the taking of the sample was not a Fourth Amendment search, it is necessary to consider separately whether the testing and reporting were.
Hoffa did say that the Fourth Amendment can be violated by "guileful as well as by forcible intrusions into a constitutionally protected area.” 385 U. S., at 301. The case it dted for that proposition, however, shows what it meant: Gouled. v. United States, 255 U. S. 298 (1921), found a Fourth Amendment violation where a Government agent who had obtained access to the defendant’s office on pretext of a social visit carried away private papers. “Guile” (rather than force) had been used to go beyond the scope of the consented access to evidence. Whereas the search in Gouled was invalidated, the search was approved in Lewis v. United States, 385 U. S. 206 (1966), where an equally guileful agent stayed within the bounds of the access to defendant’s home, carrying away only a package of drugs that had been voluntarily provided.
The Court contends that its opinion does not leave law enforcement officials in the dark as to when they can use incriminating evidence from trusted sources, since it “do[es] not address a case in which doctors independently complied with reporting requirements,” ante, at 85, n. 24. I find it hard to understand how not addressing that point fails to leave *96it enshrouded in darkness — unless the Court means that such reporting requirements are clearly bad. (H voluntary betrayal of a trust in mere cooperation with the police constitutes a Fourth Amendment search, surely betrayal of a trust at the direction of the legislature must be.) • But in any event, reporting requirements are an infinitesimal part of the problem. What about a doctor’s — or a spouse’s — voluntary provision of information to the police, without the compulsion of a statute?
The Court contends that I am “mischaracteriz[ing]” its opinion, since the Court is merely “assum[ing] for purposes of decision that the patients did not consent to the searches, and [leaves] the question of consent for the Court of Appeals to determine." Ibid. That is not responsive. The “question of consent” that the Court leaves open is whether the patients consented, not merely to the taking of the urine samples, but to the drug testing in particular, and to the provision of the results to the police. Consent to the taking of the samples alone — or even to the taking of the samples plus the drug testing — does not suffice. The Court’s contention that the question of the sufficiency of that more limited consent is not before us because respondents did not raise it, see ante, at 74, n. 6, is simply mistaken. Part II of respondents’ brief, entitled “The Petitioners consented to the searches,” argues that “Petitioners... freely and voluntarily ... provided the urine samples”; that “each of the Petitioners signed a consent to treatment form which authorized the MUSC medical staff to conduct all necessary tests of those urine samples — including drug tests”; and that “[t]here is no precedent in this Court’s Fourth Amendment search and seizure jurisprudence which imposes any . . . requirement that the searching agency inform the consenting party that the results of the search will he turned over to law enforcement.” Brief for Respondents 38-89. The brief specifically takes issue with the District Court’s charge to the jury — which the Court chooses to accept as an unexaminable “given,” see ante, at 74, n. 6 — that “the Respondents were required to *97show that the Petitioners consented to MUSG disclosing the information to law enforcement.” Brief for Respondents 39.
In sum, I think it dear that the Court’s disposition requires the holding that violation of a relationship of trust constitutes a search. The opinion itself implies that in its description of the issue left for the Court of Appeals on remand, see ante, at 77, n. 11: whether “the tests were performed without the informed consent of the patients,” ante, at 77 (emphasis added) — informed, that is, that the urine would be tested for drugs and that the results would be given to the police. I am happy, of course, to accept the Court’s illogical assurance that it intends no such holding, and urge the Court of Appeals on remand to do the same.
See, e. g., Cal. Penal Code Aim. §11160 (West Supp. 2001); N. Y. Penal Law §265.25 (McKinney 2000); S. C. Code Ann. §16-3-1072 (Supp. 2000).
See, e.g., Cal. Penal Code Ann. §11160 (West Supp. 2001); Colo. Rev. Stat. §12-36-135 (2000).
The Court contends that I “would have us . . . resolve the issue of consent in favor of respondents,” whereas the Court’s opinion “more prudent[ly) allowjs] [the Court of Appeals] to resolve the legal and factual issues in the first instance, and ... expresses] no view on those issues.” Ante, at 77, n. 11. That is not entirely so. The Court does not resolve the factual issue whether there was consent to the drug testing and to providing the results to the police; and neither do I. But the Court does resolve the legal issue whether that consent was necessary, see ante, at 77,84-85, and n. 24; and so do I. Since the Court concludes it was necessary, the factual inquiry is left for the Fourth Circuit on remand. Since I conclude it was not necessary (and since no one contends that the taking *98of the urine sample was unconsented), there is on my analysis no factual consent issue remaining.
The Court believes that this finding “must be read in light of” the District Court's comment that “‘these searches were not done by the medical university for independent purposes. . . . [T]he police came in and there was an agreement reached that the positive screens would be shared with the police. And then the screen is not done independent of police, *99it’s done in conjunction with the police and that implicates the Fourth Amendment.’” Ante, at 75, n. 8, quoting App. 1247-1249. But all this shows is that the explicit finding of medieal purpose was not a finding of exclusive medical purpose. As discussed later in text, the special-needs doetrine contains no such exclusivity requirement.