Jimmie Louise GRAY, individually and as personal representative of the estate of Charles Gray; Charles Gray, deceased; Teresa Leeper; David Gray; and Tim Gray, Plaintiffs–Appellants, v. UNIVERSITY OF COLORADO HOSPITAL AUTHORITY, a body corporate and political subdivision of the State of Colorado; University Of Colorado Hospital; Mark Spitz, Archana Shrestha, M.D., Christy Barbee–Young, M.D., Mollie Startzer, R.N., Beverly Solas–Fajardo, C.N.A., Jason Booe, and Jacqueline Funk, in their individual and official capacities, Defendants–Appellees.
Decedent Charles Gray sought treatment for epilepsy at Defendant University of Colorado Hospital. In the course of his withdrawal from medication, hospital staff left decedent unattended and he died after suffering a seizure. Plaintiffs, decedent's estate and family members, filed this civil rights suit pursuant to 42 U.S.C. § 1983. In their complaint, Plaintiffs alleged among other things, that Defendant hospital, and affiliated doctors, nurses, and staff acting in their capacity as “employees and/or agents” of the hospital, deprived decedent of life without due process of law in violation of the Fourteenth Amendment. The district court granted Defendants' motion to dismiss the complaint pursuant to Fed.R.Civ.P. 12(b)(6) for failure to state a constitutional claim.1 Plaintiffs appeal. Our jurisdiction arises under 28 U.S.C. § 1291. We review the district court's dismissal of Plaintiffs' complaint de novo, accepting the pleading's factual allegations as true. See Smith v.. United States, 561 F.3d 1090, 1098 (10th Cir.2009). Applying the appropriate legal standards, we affirm, but for reasons somewhat different than those proffered by the district court.
Plaintiffs make the following factual allegations in their complaint. At the outset, we note most of these allegations refer generally to Defendant hospital and unspecified doctors, nurses, and staff. In other words, the complaint's material allegations largely fail to specify exactly who allegedly did what. Decedent Charles Gray was admitted to the University of Colorado Hospital's Epilepsy Monitoring Unit (EMU). Defendants arranged to wean decedent off his anti-seizure medication while monitoring him to determine if he would benefit from ameliorative surgery. Defendants represented to decedent and his family that he would receive continuous, 24–hour–per–day intensive care monitoring during his stay in the EMU. Defendants provided decedent with an information sheet that said a neuro-diagnostic technologist would be in the monitoring booth at all times to maintain the equipment and gather data. Defendants knew uninterrupted monitoring of decedent was necessary for his protection, especially during periods of sleep. The hospital's protocol, however, allowed EMU staff to leave patients unattended and unobserved.
On the first full day off his medications, decedent experienced two complex partial seizures. Because these seizures were difficult to localize, Defendants continued to withhold anti-seizure medication from decedent in order to capture data sufficient to determine whether surgery was advisable. Shortly before midnight that same day, an attending technician, unidentified in the complaint, left decedent to “troubleshoot another ICU patient's electrodes.”2 About twenty minutes later, around 12:20 a.m. the following day, decedent suffered a generalized seizure requiring immediate medical attention. At 12:22 a.m., decedent stopped breathing. Around 1:00 a.m., the technician returned to the EMU and discovered decedent was not breathing. Efforts to resuscitate him were unsuccessful. Decedent was pronounced dead at 1:37 a.m. The hospital's Vice President for Patient Safety acknowledged that Defendants made false representations to decedent and his family. The administrator admitted that if the hospital had required constant monitoring in the EMU, decedent likely would have survived.
Based on these facts, Plaintiffs alleged three federal claims on behalf of decedent's estate. Plaintiffs labeled their first claim for relief “Failure to Provide Medical Care and Treatment.” This claim alleged Defendants acted with “deliberate indifference” to decedent's due process “right not to be denied necessary medical care and treatment.” Plaintiffs alleged Defendants exhibited deliberate indifference to decedent's well-being by failing to monitor his condition, despite their knowledge of his serious medical needs. Plaintiffs labeled their second claim for relief “Supervisory Liability for Failure to Train and Supervise.” Plaintiffs' second claim cursorily alleged certain Defendants' failure to adequately train and supervise hospital personnel was the cause of decedent's constitutional deprivation. Plaintiffs labeled their third claim for relief “Substantive Due Process/Danger Creation.” This claim alleged Defendants acted, pursuant to policy and custom, with “reckless disregard” for decedent's right “not to be subjected to serious dangers created by and under the control of the Defendants.” Plaintiffs alleged Defendants knew their repeated assurances were contrary to hospital protocol permitting staff to leave decedent unattended in the EMU for extended periods.
Defendants moved to dismiss Plaintiffs' § 1983 claims for failure to state a cause of action. The district court granted Defendants' motion. The court first reasoned that “where non-prisoners voluntarily seek medical care from state actors, negligent and even willfully indifferent treatment does not amount to a violation of the U.S. Constitution.” Gray v. Univ. of Colo. Hosp. Auth., 2010 WL 3430785, at *2 (D.Colo.2010) (unpublished). The district court wrote: “Plaintiffs do not appear to dispute that [Tenth Circuit] cases ․ are dispositive of their simple constitutional claims, but they contend that these cases do not address the substantive due process claim premised on ‘danger creation.’ “ Id. Considering the context in which the district court referenced “simple constitutional claims,” the court presumably was referring to Plaintiffs' first two claims for relief, i.e., (1) for failure to provide medical care and treatment based on a theory of personal liability, and (2) for failure to train and supervise based on a theory of supervisory liability. See Brown v. Montoya, 662 F.3d 1152, 1163–64 (10th Cir.2011) (distinguishing between § 1983 claims based on personal liability and supervisory liability). As to Plaintiffs' third claim, the court expressed doubt regarding Defendants' argument that the “ ‘danger creation’ doctrine is limited to circumstances where violence by a [private] third party is the cause of the victim's injury.” Gray, 2010 WL 3430785, at *3. But the court concluded it need not decide that question because, assuming the danger creation theory applied, the alleged facts failed to demonstrate Defendants' conduct was “conscience-shocking” as required by our precedent. According to the court, Defendants' conduct was at most negligent. The court reasoned that “elevating such careless conduct to the level of a constitutional deprivation would radically broaden the scope of constitutional protection, essentially allowing it to replace ordinary tort law.” Id. at *4.
Because Plaintiffs precisely identify neither the claim or claims for relief—one, two and/or three—nor the theory or theories of liability—personal, supervisory, and/or danger creation—they wish to press upon us, we must determine ourselves what exactly Plaintiffs are appealing. In the issue portion of their opening brief, Plaintiffs characterize their appeal as raising three issues. First, Plaintiffs ask us to decide whether their complaint “failed to establish a cognizable claim for relief” under § 1983. Such a broadly worded issue tells us absolutely nothing about the precise nature of the legal questions Plaintiffs would have us resolve, and is particularly unhelpful. The second issue Plaintiffs raise on appeal is more telling because it refers to their third claim, i.e ., the claim based on a danger creation theory. Plaintiffs phrase this issue as whether the district court “applied an incorrect legal standard or multi-criteria test ․ as the basis for its dismissal of Plaintiffs' claim for deprivation of substantive due process, under the danger creation exception.” Plaintiffs' third issue also speaks of danger creation and thus also appears to address their third claim. Plaintiffs say the third issue is whether their complaint “sufficiently alleged that Defendants' actions or inactions contributed to the danger that gave rise to a constitutional duty to protect the deceased ․ and Defendants acted recklessly and with deliberate indifference to the known risk of harm.”
Unfortunately, the argument portion of Plaintiffs' opening brief does not track their statement of the issues, so we need briefly summarize its main points in our attempt to ascertain where to commence. Plaintiffs begin in Part I by telling us that this case challenges action taken by hospital employees pursuant to Defendants' policies and customs. Plaintiffs say state action taken pursuant to a governmental policy or custom may give rise to supervisory liability under § 1983. But they fail to provide us with developed argumentation as to why Defendants' policies and customs in this case translate into constitutional liability under the danger creation theory identified in their statement of the issues. Plaintiffs next seek to persuade us in Part II that decedent's hospitalization was akin to a custodial setting because he was involuntarily restrained as a result of, among other things, tainted consent. But Plaintiffs do not explain how decedent's supposed “confinement” affects the applicable legal standards under their danger creation theory. Part III of Plaintiffs' argument is somewhat more focused. In Part III, Plaintiffs evoke their third issue statement by arguing that “Defendants created or enhanced [decedent's] vulnerability to harm.” Here, Plaintiffs run through an incomplete list of factors our precedents require to establish Defendants' liability under their danger creation theory. Finally in Part IV, Plaintiffs appear to address their second issue statement. Plaintiffs argue the district court erred in holding the facts alleged did not “shock the conscience”—a factor necessary to establish substantive due process liability absent a custodial relationship between the victim and the State—because the court “failed to appreciate that the Defendants' conduct was the product of their deliberate indifference to the danger [decedent] faced as a result of their conduct.”
So where does this morass suggest we begin? Based on the issues and arguments Plaintiffs raise on appeal, as well as our review of the briefs and record, we construe Plaintiffs' appeal as challenging only the district court's dismissal of their third claim for relief, that is, the claim labeled “Substantive Due Process/Danger Creation .” And that claim, as best we can tell, is this: Defendants' customary misrepresentations, coupled with their implementation of a hospital policy, gave rise to a constitutional duty to protect decedent from the danger that came to fruition when the technician left him unattended. In other words, Defendants jeopardized decedent's right to substantive due process, i.e., the right to be free from arbitrary government action, by affirmatively acting to create the danger that precipitated his death. Plaintiffs' emphasis on Defendants' policy that allowed the attending technician to abandon decedent in the EMU despite their contrary assurances leads us to this conclusion. Plaintiffs' statement of the issues and the overwhelming weight of their arguments speak to this claim alone.
Buttressing our understanding of this appeal is the fact that Plaintiffs never have refuted the district court's observation that their “simple constitutional claims,” i.e., those claims apart from Plaintiffs' “substantive due process claim premised on ‘danger creation,’ “ are meritless. Gray, 2010 WL 3430785, at *2. Nor have Plaintiffs refuted Defendants' characterization of Plaintiffs' response to their motion to dismiss in the district court. Defendants state in their appellate brief that Plaintiffs “responded to the motion to dismiss, but did not present any argument related to the first claim for relief ․ or the second claim for relief․ Instead, the only theory upon which [Plaintiffs] proceeded was a due process violation under the ‘danger creation’ theory.” Our review of Plaintiffs' response confirms the accuracy of Defendants' characterization.
Because Plaintiffs have failed to contest in any meaningful way the district court's dismissal of their first two federal claims, they have voluntarily foregone any right to further adjudication of those claims. See, e.g., Carpenter v. Boeing Co., 456 F.3d 1183, 1198 n.2 (10th Cir.2006) (noting we generally will not consider arguments appellants failed to raise in the district court); United States v. Wooten, 377 F.3d 1134, 1145 (10th Cir.2004) (noting we generally will not consider issues appellants advert to in their opening brief only in a perfunctory manner without developed argumentation); Mattioda v. White, 323 F.3d 1288, 1291 n.2 (10th Cir.2003) (noting we generally will not consider issues appellants fail to raise on appeal). Accordingly, we now proceed with a discussion of what we refer to as the state-created danger theory of constitutional liability as that dicey theory has evolved within the Tenth Circuit. With a proper understanding of the law in place, we then explain why that theory, in light of the facts alleged, does not support the complaint's third claim for relief. Bear with us, because in their efforts to invoke the state-created danger theory and obtain a federal remedy for an apparent state tort claim, Plaintiffs have made this case much more difficult than it need be.
The story of the state-created danger theory goes at least as far back as DeShaney v. Winnebago Cnty. Dept. of Soc. Serv., 489 U.S. 189 (1989). In DeShaney, a child alleged that a state social service agency and its employees deprived him of liberty without due process of law “by failing to intervene to protect him against a risk of violence at his father's hands of which they knew or should have known.” Id. at 193. The Court was unmoved and established the general rule that the State's failure to protect an individual “against private violence simply does not constitute a violation of the Due Process Clause.” Id. at 197. The Due Process Clause, the Court explained, “does not transform every tort committed by a state actor into a constitutional violation.” Id. at 202. That the child claimed “the State ․ specifically proclaimed, by word and by deed, its intention to protect him against that danger” made no difference. Id. at 197. A constitutional duty to protect on the part of the State does not arise “from the State's knowledge of the individual's predicament or from its expressions of intent to help him.” Id. at 200. As an exception to the general rule, the Court stated that “when the State takes a person into custody and holds him there against his will, the Constitution imposes upon it a corresponding duty to assume some responsibility for his safety and general well-being.” Id. at 199–200. The Court explained:
In the substantive due process analysis, it is the State's affirmative act of restraining the individual's freedom to act on his own behalf—through incarceration, institutionalization, or other similar restraint of personal liberty—which is the “deprivation of liberty” triggering the protections of the Due Process Clause, not its failure to act to protect his liberty interest against harms inflicted by other means.
Id. at 200 (emphasis added).
We subsequently recognized the state-created danger theory of constitutional liability for the first time in Graham v. Indep. Sch. Dist. No. I–89, 22 F.3d 991 (10th Cir.1994). We did so on the basis of the Court's statement in DeShaney that “[w]hile the State may have been aware of the dangers that [the child] faced in the free world, it played no part in their creation, nor did it do anything to render him any more vulnerable to them.” DeShaney, 489 U.S. at 201. In Graham, two mothers alleged public school officials breached a constitutional duty to protect their sons from the violent acts of other students. The mothers based their claims on the officials' failure to take action designed to ensure their sons' safety in the face of known dangers. We observed that, by negative implication, DeShaney “leaves the door open for liability” where the State creates a dangerous situation for citizens in the free world or renders them more vulnerable to danger.3 Graham, 22 F.3d at 995 (internal quotations omitted). We walked through that door only to uphold the dismissal of the complaint because it failed to allege “affirmative actions by the defendants that created or increased the danger to the victims.” Id.
At this early stage, we placed an important limitation on the state-created danger theory consistent with DeShaney. We established “[t]his state-created danger doctrine necessarily involves affirmative conduct on the part of the state in placing the plaintiff in danger.” Id. (internal quotations omitted) (emphasis added). We expressly “h[e]ld foreseeability cannot create an affirmative duty to protect when plaintiff remains unable to allege a custodial relationship.” Id. at 994. “[I]naction by the state in the face of a known danger is not enough to trigger the obligation” unless the State has “limited in some way the liberty of a citizen to act on his own behalf.”4 Id. at 995 (internal quotations omitted). We also recognized the unremarkable proposition that a complaint seeking recovery under § 1983 based on the state-created danger theory must allege a sufficient causal link between the danger created by an affirmative act of the State and the harm inflicted upon the victim by a private party:
[A]ny danger to the victims was too remote a consequence of defendants' action [of enrolling the aggressors in public school] to hold them responsible under the federal civil rights law. In most every circuit court decision imposing § 1983 liability because the State affirmatively created or enhanced a danger, the immediate threat of harm has a limited range and duration unlike the indefinite risk created by enrolling the aggressor[s] in public school.[5 ]
Id. (internal brackets and quotations omitted).
Our second foray into the state-created danger theory came in Uhlrig v. Harder, 64 F.3d 567 (10th Cir.1995). In that case, state mental health administrators terminated a special unit in a mental hospital reserved for criminally insane inmates. A former resident of that unit, housed with the general population, murdered an activity therapist. The decedent's estate sued, alleging the administrators' were “liable under § 1983 for violating [decedent's] substantive due process rights by recklessly creating the danger that led to her death.” Id. at 569. We rejected the allegation in the context of summary judgment, holding the facts did not establish that the administrators acted recklessly. As part of our analysis, we described the state-created danger theory as a means by which state actors may be held constitutionally liable for acts of private violence under prescribed circumstances. The theory, we explained, constitutes a second exception to the DeShaney rule that “state actors are generally only liable under the Due Process Clause for their own acts and not for private violence.” Id. at 572. Because “the conduct complained of ․ was committed by a private third party (i.e. [the inmate] ) rather than by a state actor,” we required plaintiff to “demonstrate ․ that the state recklessly created the danger that caused the constitutional violation.” Id. at 571–72. But not just any danger would do. Instead, we required plaintiff to show that the State's reckless conduct created a “constitutionally cognizable danger:”
[M]any state activities have the potential for creating some danger ․ but not all such activities constitute a “special” danger giving rise to § 1983 liability. For the state to be liable under § 1983 for creating a special danger (i.e. where a third party other than a state actor causes the complained of injury), a plaintiff must allege a constitutionally cognizable danger. That is, the danger creation theory must ultimately rest on the specifics of a substantive due process claim—i.e. a claim predicated on reckless or intentionally injury-causing state action which “shocks the conscience.”
Id. at 572.
The third case relevant to our discussion is Armijo ex rel. Chavez v. Wagon Mound Pub. Sch., 159 F.3d 1253 (10th Cir.1998). Like DeShaney, Graham, and Uhlrig, Armijo addressed an act of private violence allegedly precipitated by state actors. But unlike those cases, a third party did not inflict the violence in Armijo. Rather, a special education student at a public school shot and killed himself after being suspended and driven home without parental notification in violation of the school's disciplinary policy. Addressing the parents' claim that the school had violated their son's right to substantive due process, we rejected the argument that the school had a constitutional duty to protect the decedent based on a custodial relationship: “Banning a student from school grounds does not rise to the same level of involuntary restraint as arresting, incarcerating, or institutionalizing an individual.” Id. at 1261. We held, however, that fact issues under the parents' danger creation theory of liability precluded summary judgment on the basis of qualified immunity for school officials. Our holding effectively but unremarkably extended application of the state-created danger theory to instances of suicide, undeniably another form of private violence .6
Next we decided Sutton v. Utah State Sch. for the Deaf and Blind, 173 F.3d 1226 (10th Cir.1999), another case involving private violence, but in some respects presenting a factual scenario similar to our own. The child victim suffered from cerebral palsy, mental retardation, blindness, and an inability to speak. The child communicated to his mother that he had been sexually assaulted by a larger boy while using the bathroom at the state school. Thereafter, the principal assured the mother that her child “would be supervised at all times while in the bathroom.” Id. at 1230. But the older boy sexually assaulted the child a second time when a teacher's aide abandoned her post outside the bathroom to answer the phone. The district court dismissed the mother's danger creation claim for failure to allege the violation of a constitutional right. On appeal, we rejected the mother's argument that the principal created the danger to her son “by directly participating in placing him in harm's way.” Id. at 1238. Citing the “strict standards of substantive due process” we had employed in Graham and Uhlrig “for succeeding on a danger creation claim,” we held the complaint did not sufficiently plead “affirmative acts on the part of [the principal] which give rise to a claim for a deprivation of [the child's] constitutional rights.”7 Id. at 1239.
At this point, our discussion would be incomplete and intellectually dishonest absent reference to DeAnzona v. City and Cnty. of Denver, 222 F.3d 1229 (10th Cir.2000), a case illustrative of the state-created danger theory's osmotic, ill-considered tendency to invade the province of both common law negligence and state tort law. See Schroeder v. City of Chicago, 927 F.2d 957, 961 (7th Cir.1991) (Posner, J.) (noting that substantive due process's “tendency to formlessness has been blamed for a host of usurpative decisions”). A child participating in a city day-camp drowned when busy camp counselors failed to notice him at the lake's edge. His parents sued the manager of the city's parks and recreation department, among others, alleging that “seven failures” caused their child's death, including the failure to train counselors and the failure to institute precautionary measures against the risk that befell their son. DeAnzona, 222 F.3d at 1234. These failures, according to the parents, gave rise to the state-created danger of day-camp drownings which resulted in a violation of their child's substantive due process right. We made nary a mention that the harm to the child in no sense involved (at least on a fair reading of the opinion) affirmative conduct on the part of state actors or private violence—factors that Graham and Uhlrig established as preconditions to the state-created danger theory's application in the Tenth Circuit. Instead of inquiring into the presence of these necessary preconditions as an initial matter, we proceeded directly to a six-factor test originating with Uhlrig and revised in Armijo:
When a plaintiff alleges a danger was created by the defendant, the plaintiff must demonstrate that 1) plaintiff was a member of a limited and specifically definable group; 2) defendant's conduct put plaintiff at substantial risk of serious, immediate, and proximate harm; 3) the risk was obvious or known; 4) defendant acted recklessly in conscious disregard of that risk; ․ 5) such conduct when viewed in total, is conscience shocking[;] and 6) defendant ․ created the danger or increased the plaintiff's vulnerability to the danger in some way.
Id. at 1235 (internal brackets, ellipses, and quotations omitted). We jumped to the fifth factor and concluded the manager was entitled to summary judgment because the requirement that her conduct “be conscience shocking, is impossible to meet in the current case.”8 Id.
We subsequently returned to course in Ruiz v. McDonnell, 299 F.3d 1173 (10th Cir.2002). Specifically, we reaffirmed Graham's ruling that “this state-created danger doctrine necessarily involves affirmative conduct on the part of the state in placing the plaintiff in danger.” Id. at 1183 (quoting Graham, 22 F.3d at 995) (internal brackets and quotations omitted). In Ruiz, a mother enrolled her child in a state-licensed home daycare. The operator of the daycare literally abused the child to death. The mother brought suit against the state human services department and its director alleging defendants' act of licensing the provider, who failed to meet state requirements for licensure, violated her deceased son's right to substantive due process. We held that the mother failed to allege “any pertinent affirmative conduct” on the part of defendants and upheld the dismissal of her § 1983 claim. Id. In reaching our decision, we focused on the second of the six factors noted above as it bears upon affirmative conduct, i. e ., the requirement that defendants' act of licensing the daycare place the child “at substantial risk of serious, immediate, and proximate harm.” Id. (emphasis added). As in Graham, we reasoned that the threat of harm must be of “limited range and duration,” rather than generally applicable to a broader populace:
Affirmative conduct for purposes of § 1983 should typically involve conduct that imposes an immediate threat of harm, which by its nature has a limited range and duration․ Moreover, the conduct should be directed at a discrete plaintiff rather than at the public at large․
[W]e do not view the mere licensure of [the day care] as constituting the requisite affirmative conduct necessary to state a viable § 1983 claim. Specifically, the improper licensure did not impose an immediate threat of harm. Rather, it presented a threat of an indefinite range and duration. Moreover, the licensure affected the public at large; it was not aimed at [the child] or [his mother] directly․ [T]he mere licensure of [the daycare] was not an act directed at [the child] which, in and of itself, placed [the child] in danger.
Id. at 1183.
We continued on course in Moore v. Guthrie, 438 F.3d 1036 (10th Cir.2006). This time we reaffirmed Uhlrig's observation that the state-created danger theory is an exception to the rule that state actors generally are not liable for acts of private violence. In Moore, a city police officer was injured during a training exercise when a plastic bullet flew up beneath his standardized riot helmet and struck him in the right eye. The police chief had not authorized the purchase of protective head and body gear recommended by the bullet's manufacturer. The officer sued the city and its police chief alleging a due process violation of his right to bodily integrity. Among other things, the officer argued his complaint sufficiently pled “a violation of his right to bodily integrity under the ‘danger creation’ theory.” Id. at 1042. The district court dismissed the complaint for failure to state a constitutional claim and we affirmed. We described the state-created danger theory as “a narrow exception, which applies only when a state actor affirmatively acts to create, or increase[ ] a plaintiff's vulnerability to, danger from private violence. It does not apply when the injury occurs due to the action of another state actor.” Id. (internal citations and quotations omitted) (emphasis added). We held that because plaintiff “was injured by a ․ bullet fired by a fellow police officer and not a private ․ party, the danger creation doctrine is inapplicable.” Id.
We conclude our history lesson with Robbins v. Oklahoma, 519 F.3d 1242 (10th Cir.2008), a case returning us full circle to where we began—with DeShaney. In Robbins, an infant enrolled in state-subsidized daycare suffered fatal blunt force trauma to her head. Her parents filed suit against the state human services department and a number of its employees alleging defendants instructed them to place their child “in a specific daycare,” and that this particular daycare was the only one “which [their infant] could attend due to financial considerations.” Id. at 1250. The parents also alleged defendants “lulled [them] into a false sense of security about [their infant's] welfare” and “failed to correct the misimpressions that the [department's] report of available daycare facilities engendered.” Id. The parents claimed these and other allegations, accepted as true, established a violation of their infant's substantive due process right under the state-created danger theory. We disagreed and directed the district court to dismiss the complaint for failure to state a constitutional claim.9
We explained that DeShaney established the Due Process Clause does not require a State to “protect ‘the interests of life, liberty, and property of its citizens against invasion by private actors.’ “ Id. at 1251 (quoting DeShaney, 489 U.S. at 195). We recognized that only two of plaintiffs' claims could “escape the sweep of DeShaney ”: “(1) that the defendants are liable for danger creation for continuing to license [the daycare] facility, and (2) that defendants provided affirmative representations and assurances as to the quality of care that [the infant] would receive at the ․ daycare.” Id. at 1251. Ruiz's decision that “negligence in licensing was not a sufficiently affirmative act under the standard set by DeShaney because it did not pose an immediate threat of harm and was directed at the public in general,” disposed of the parents' first claim. Id. As for the parents' claim that defendants' misrepresentations “lulled” them “into a false sense of security,” we explained that on the facts alleged we need not decide whether to extend the state-created danger theory to affirmative misrepresentations because “plaintiffs do not allege defendants made any affirmative statements regarding the quality of the ․ daycare. The allegations of ‘lulling’ and ‘doing nothing’ do not give rise to a reasonable expectation of relief, given that DeShaney requires an affirmative act before imposing liability.” Id. at 1252.
At this point, the astute reader understands that, for any number of reasons, attempting to apply the state-created danger theory to the facts alleged in Plaintiffs' complaint is like trying to fit a square peg in a round hole. We begin our analysis of Plaintiffs' attempt with what is an unremarkable proposition in the Tenth Circuit: The state-created danger theory is a means by which a state actor might be held liable for an act of private violence absent a custodial relationship between the victim and the State, under narrowly prescribed circumstances bearing upon conduct, causation, and state of mind, provided the danger the state actor created, or rendered the victim more vulnerable to, precipitated a deprivation of life, liberty, or property in the constitutional sense. From that proposition, we proceed.
We first address Plaintiffs' argument that the state-created danger theory is a viable theory of recovery in this case because decedent was in the custody of Defendants at the time of his death. Given that application of the state-created danger theory depends upon the absence of a custodial relationship between the victim and the State, however, we are at a loss to understand Plaintiffs' extended efforts at the outset to convince us decedent was in Defendants' custody at the time of his death. Uhlrig explained, as have many other cases, that where a private party inflicts harm upon the victim, the State incurs an antecedent constitutional duty to protect the victim only if plaintiff demonstrates “either ” (1) the existence of a “special custodial relationship” between the two; “or ” (2) the State recklessly created the danger that precipitated the constitutional deprivation. Uhlrig, 64 F.3d at 571–72 (emphasis added). DeShaney stressed the first or “special relationship” exception to the general rule that state actors are not responsible for acts of private violence. DeShaney, 489 U.S. at 197–200. This first exception applies only in the presence of a custodial relationship between the victim and the State. Graham provided us the second or “danger creation” exception to that rule. Graham, 22 F.3d at 995. This second exception applies only in the absence of a custodial relationship between the victim and the State.10
In any event, decedent most assuredly was not in the custody of Defendants at the time of his death because Defendants did not affirmatively act to place him there. DeShaney listed arrest, incarceration, and institutionalization as examples where the State's “affirmative act” of exercising physical control and dominion over the person constitutes a deprivation of liberty triggering substantive due process protection. DeShaney, 489 U.S. at 197–200 (citing Revere v. Massachusetts Gen. Hosp., 463 U.S. 239 (1983) (detained suspect); Estelle v. Gamble, 429 U.S. 97 (1976) (convicted prisoner); Youngberg v. Romero, 457 U.S. 307 (1982) (involuntarily committed mental patient)). We subsequently recognized a State's affirmative act of placing a child in involuntary foster care as a similar restraint of liberty.11 DeAnzona, 222 F.3d at 1234. But these examples do not help Plaintiffs. Rather, they demonstrate that the restraint of liberty necessary to invoke substantive due process protection under the special relationship exception requires state action involving force, the threat of force, or a show of authority, with the intent of exercising dominion and control over the person. See DeShaney 489 U.S. at 200 (explaining the State's affirmative duty to protect arises from the limitation it imposes on the freedom to act); see also id. at 206 (Brennan J., dissenting) (“[T]o the Court, the only fact that seems to count as an ‘affirmative act of restraining an individual's freedom to act on his own behalf’ is direct physical control.”). The complaint's allegations do not satisfy this demanding standard. Apart from the fact that the complaint does not allege the special relationship upon which the custody inquiry depends, Defendants did not restrain decedent's liberty or freedom to act through a show of force or authority. See Ye v. United States, 484 F.3d 634, 641 (3d Cir.2007) (holding a doctor's assurances that his patient had “nothing to worry about” and was “fine” did not constitute a restraint of liberty regardless of the patient's reliance on those assurances to forego medical assistance). Defendants did not force decedent against his will to become dependent upon them. Decedent voluntarily checked himself into Defendants' hospital for medical observation and testing. Defendants did not prohibit decedent from seeking alternative sources of assistance for his condition. Although Defendants may have played “some causal role” in decedent's death, they did so only because decedent “voluntarily availed himself” of their services.12 Monahan v. Dorchester Counseling Ctr., 961 F.2d 987, 993 (1st Cir.1992).
Plaintiffs beg to differ, repeatedly reminding us Defendants recklessly informed decedent on more than one occasion that the EMU would provide him with uninterrupted monitoring, when they knew or should have known those representations were false. Undoubtedly, Defendants subjected decedent to an increased risk when he relied on their representations, a risk he might have foregone given accurate information. But we fail to see how Plaintiffs' allegation that decedent trusted Defendants to do what they said they would do differs from DeShaney, where the State “specifically proclaimed, by word and by deed, its intention to protect [the child] against [the] danger” his father posed. DeShaney, 489 U .S. at 197. The four year old child in DeShaney was no more able to protect himself from his father's abuse than decedent was able to protect himself from his fatal seizure. But the State's false assurances—even if in some way responsible for the tragic result—did not render the child in custody there and did not render the decedent in custody here.13
With decedent's status as a free man resolved, we turn to the question of whether Plaintiffs' complaint alleges affirmative conduct on the part of Defendants sufficient to sustain application of the state-created danger theory. As we have seen, “ ‘affirmative conduct’ “ is a necessary precondition to such application. Graham, 22 F.3d at 995. We begin by considering whether Defendants' untruthful assurances to decedent and his family constitute “affirmative conduct on the part of the state in placing [decedent] in danger.” Id. (internal quotations omitted). We conclude those assurances do not support Plaintiffs' substantive due process claim based on danger creation. The reason those false assurances do not constitute an affirmative act rendering decedent vulnerable to danger within the meaning of the danger creation exception is the same reason those assurances do not constitute an affirmative act in restraint of decedent's liberty within the meaning of the special relationship exception—DeShaney tells us so.
DeShaney's facts stalwartly suggest assurances of protection from the State do not constitute affirmative conduct sufficient to invoke the state-created danger theory of constitutional liability. In DeShaney, the State “specifically proclaimed, by word and by deed,” its intention to protect the child from the danger his father posed. DeShaney, 489 U.S. at 197. Nonetheless, the Court observed that the State played no part in creating the danger the child faced from his father, “nor did it do anything to render him any more vulnerable to [it].” Id. at 201. The Court rejected the child's argument that as a result of the State's assurances of protection, “the State acquired an affirmative ‘duty,’ enforceable through the Due Process Clause, to [protect him] in a reasonably competent fashion.” Id. at 197. We reached the same conclusion in Sutton, where we rejected the argument that the principal directly participated in creating the danger of sexual assault upon the child based on assurances to the child's mother that he “would be supervised at all time while in the bathroom.” Sutton, 173 F.3d at 1230; see also Rivera v. Rhode Island, 402 F.3d 27, 37 (1st Cir.2005) (“[M]erely rendering a person more vulnerable to risk does not create a constitutional duty to protect.”).14 The allegation in this case that decedent consented to enter the EMU because he believed Defendants would do what they said they would do does not materially differ from the allegations in DeShaney and Sutton. There, like here, state actors were aware of the risk, expressly promised to eliminate the risk, and failed to do so, with no constitutional implications.
Plaintiffs' contrary insistence notwithstanding, the fact that Defendants' reckless behavior was attributable to their established policies and customs counsels against application of the state-created danger theory, rather than for it. In their opening brief, Plaintiffs remind us that Defendants' false assurances to decedent “were committed ․ as a matter of long-standing official policy and custom.” Plaintiffs tell us “[t]he false promises the Defendants made to [decedent] and his family were the same promises they had made to other patients and other family on many different occasions in the past.” Similarly, Plaintiffs repeatedly point out “Defendants' policy had, for years, permitted its staff to leave patients experiencing ongoing, life-threatening seizure activity in the EMU alone and unattended.” According to Plaintiffs, “[t]he factual allegations in the complaint ․ show that both the misrepresentations concerning patient observation and the lack of observation afforded patients experiencing seizure activity were part of the Defendants' official policy and custom.”
But like the States' licensures of the daycare facilities in Ruiz and Robbins, a State's adoption of generally-applicable policies and customs does not foist upon anyone an “immediate threat of harm” having “a limited range and duration.” Graham, 22 F.3d at 995 (internal quotations omitted). The act of establishing such policies and customs itself does not put any particular individual “at substantial risk of serious, immediate, and proximate harm.” Rost ex rel. K.C. v. Steamboat Springs RE–2 Sch. Dist., 511 F.3d 1114, 1126 (10th Cir.2008) (reaffirming the six-factor test applicable to danger creation claims). And because the act of establishing such policies and customs does not pose a direct threat to any one particular individual but affects a broader populace, we deem such act too remote to establish the necessary causal link between the danger to the victim and the resulting harm. See Ruiz, 299 F.3d at 1183; see also Kaucher v. Cnty. of Bucks, 455 F.3d 418, 432 (3d Cir.2006) (opining a “direct causal relationship” must exist between the State's affirmative act and plaintiff's harm); Jones v. Reynolds, 438 F.3d 685, 697 (6th Cir .2006) (explaining that where the victim is not identifiable at the time of the alleged state action, § 1983 will not support a claim under the state-created danger theory). In other words, the affirmative conduct required to support a danger creation claim “should be directed at a discrete plaintiff.” Ruiz, 299 F.3d at 1183 (emphasis added). In Uhlrig, we explained “many state activities have the potential for creating some danger ․ but not all such activities constitute a ‘special’ danger giving rise to § 1983 liability.” Uhlrig, 64 F.3d at 572. The Sixth Circuit explained:
[B]ecause many state activities have the potential to increase an individual's risk of harm, we require plaintiffs alleging a constitutional tort under § 1983 to show ‘special danger’ in the absence of a special relationship․ The victim faces ‘special danger’ where the state's actions place the victim specifically at risk, as distinguished from a risk that affects the public at large.
Kallstrom v. City of Columbus, 136 F.3d 1055, 1066 (6th Cir.1998).
At the time Defendants adopted the alleged policies and customs about which Plaintiffs complain, decedent was not an identifiable victim. Plaintiffs acknowledge these policies and customs were “long-standing” and had been in effect “for years.” Thus, they were not aimed at decedent directly, and did not pose a threat of immediate and proximate harm to him. Rather they “presented a threat of an indefinite range and duration.” Ruiz, 299 F.3d at 1183. To be sure, Defendants' policies and customs did not increase the danger to the public at large in any real sense, but rather to a defined group, namely, patients in the EMU. See Kallstrom, 136 F .3d at 1067 (holding a city's policy of freely releasing information from undercover officers' personnel files created a constitutionally cognizable special danger). But this does not undermine our analysis. The restrictive six-factor test we apply to danger creation claims requires not only that plaintiff be a “member of a limited and specifically definable group,” but also that defendant's conduct specifically “put plaintiff at substantial risk of serious, immediate, and proximate harm.” Rost, 511 F.3d at 1126. In Ruiz and Robbins, we held the States' wrongful licensures of the daycare facilities did not constitute affirmative conduct sufficient to warrant application of the state-created danger theory, despite that in reality those licensures only enhanced the danger of abuse to the unfortunate children enrolled in those daycare facilities. The “public at large” or “the public in general” was unaffected despite our lax use of these phrases in those cases. Ruiz, 299 F.3d at 1183; Robbins, 519 F.3d at 1251. Given Ruiz and Robbins, and our requirement that a defendant's conduct put the victim at substantial risk of immediate and proximate harm, we conclude Defendants' adoption of policies and customs generally applicable to all EMU patients, even if done in reckless disregard of a generalized risk, did not constitute affirmative conduct sufficient to impose § 1983 liability on Defendants under the state-created danger theory.
We conclude our analysis of Plaintiffs' danger creation claim by pointing out its most glaring defect. We have observed throughout this opinion that a precondition to our application of the state-created danger theory is an act of “private violence.” Quite simply, the complaint does not allege this indispensable precondition. Instead, the complaint alleges that the immediate or direct cause of decedent's death was negligence on the part of state actors:
16. At all relevant times, the physicians, nurses, technicians, and staff of University Hospital whose acts and omissions are referenced herein were the employees and/or agents of Defendant[ ] University Hospital ․, acting within the course and scope of their employment, under color of state law ․
17. John and Jane Does are persons who, at times relevant to this complaint, were currently unidentified staff or personnel in the EMU ․ responsible for negligently causing the death of Charles L. Gray․
The Due Process Clause of the Fourteenth Amendment by its plain language applies only to state action: “[N]or shall any State deprive any person of life, liberty, or property, without due process of law.” U.S. Const. amend. XIV. The state-created danger theory indulges the legal fiction that an act of private violence may deprive the victim of this constitutional guarantee. Before the fiction may operate, however, a state actor must create the danger or render the victim more vulnerable to the danger that occasions the deprivation of life, liberty, or property. The danger that the state actor creates or enhances must be differentiated from the harm that the private party inflicts. Under the state-created danger theory, a constitutional deprivation is dependent on a private act that deprives the victim of life, liberty, or property, even though private action itself is never cognizable under § 1983. The state actor's affirmative act creating the danger or rendering the victim more vulnerable to it does not constitute a constitutional deprivation. The Due Process Clause does not provide an individual the right to be free from state-created dangers in a vacuum. This is because “an increased risk is not itself a deprivation of life, liberty, or property.” Rivera, 402 F.3d at 37–38 (emphasis added). A “would-be” victim who averts the danger, and thus the harm, may not claim the denial of a constitutional right to be free from state-created dangers. That would be nonsensical. Although the State may incur a constitutional duty to protect the victim from harm its conduct has rendered more likely, the victim has suffered no constitutional deprivation if the victim is not harmed.
Courts simply need not indulge this legal fiction where a state actor, rather than a private individual, is directly responsible for causing the harm. This is because the state actor directly responsible for the deprivation of life, liberty, or property may be held personally liable under § 1983. Whether other state actors further down the chain of causation also may be liable poses separate questions of personal and/or supervisory liability. See Brown, 662 F.3d 1152, 1163 (explaining that “[a] § 1983 defendant may be subject to personal and/or supervisory liability” in an individual capacity). The answers to these questions in no way depend on a legal fiction that declares an act of a private party necessary to effectuate a constitutional deprivation. For this reason, a private act must directly cause the victim's harm before we even so much as consider the state-created danger theory. See Moore, 438 F.3d at 1042 (explaining the state-created danger theory “does not apply when the injury occurs due to the action of another state actor” rather than a private party).
But not just any private act will suffice. The private act must be a violent one. Black's defines violence as, among other things, “physical force unlawfully exercised with the intent to harm.” Black's Law Dictionary 1705 (9th ed.2009). At the very least, the term “violence” in its legal sense typically connotes an act involving some degree of deliberateness. The view that a private party must act with some degree of deliberateness before a victim's harm is actionable under the state-created danger theory is sound. This is because the harm associated with a negligent act is never constitutionally cognizable under the Due Process Clause.16 Daniels v. Williams, 474 U.S. 327, 328 (1986).
Rather, due process guarantees historically have applied only to “deliberate decisions.” Id. at 331. Those guarantees are “not implicated by a negligent act of an official causing unintended loss of injury to life, liberty, or property.” Id. at 328; see also Parratt v. Taylor, 451 U.S. 527, 548 (1981) (Powell, J., concurring), overruled by Daniels, 474 U.S. at 328 (foretelling Daniels by opining that a negligent act causing “unintended loss” never “works a deprivation in the constitutional sense”). In other words, regardless of the circumstances preceding the act, a negligent act that is directly responsible for causing harm to the victim never constitutes a substantive due process violation because such an act never constitutes a constitutional deprivation of life, liberty, or property.17 Reason dictates that if state actors are not answerable under § 1983 for their own negligent acts, they are not answerable under § 1983 where a private party's underlying negligent act is directly responsible for the harm. The rationale is simple: The victim has not suffered a constitutional deprivation in either case.18
Plaintiffs' complaint plainly alleges that those individuals in the EMU responsible for monitoring decedent were “employees and/or agents” of Defendant hospital acting “under color of state law.” Plaintiffs' complaint also plainly alleges those individuals are responsible for “negligently causing” decedent's death. A precondition to our application of the state-created danger theory is “private violence.” The conduct Plaintiffs allege to be directly responsible for decedent's death is neither private nor violent.
Accordingly, because the state-created danger theory of constitutional liability has no role to play in a proper resolution of Plaintiffs' grievance, the judgment of the district court is AFFIRMED.
I join in affirming the dismissal of Plaintiffs' claims. I specifically join Sections I, II and subpart C only of Section IV of the opinion.
1. The district court declined to exercise supplemental jurisdiction over Plaintiffs' pendent state law claims. See 28 U.S.C. § 1367(c)(3).
2. While Plaintiffs' complaint names numerous hospital personnel as Defendants, nowhere does it allege that any named Defendant was the technician responsible for monitoring decedent.
3. No one can say for sure that the outcome in DeShaney would have differed if the State had created the danger to the child or rendered him more vulnerable to it. See Dorothy J. v. Little Rock Sch. Dist., 7 F.3d 729, 734 (8th Cir.1993) (“We cannot agree with those who have suggested that one comment toward the end of the DeShaney opinion ․ signals the Supreme Court's approval of § 1983 liability whenever a state actor has increased the risk of harm from private sources.”). The Court's historical reluctance to expand the concept of substantive due process “because guideposts for responsible decisionmaking in this unchartered area are scarce and open-ended,” leaves us to wonder. Collins v. City of Harker Heights, 503 U.S. 115, 125 (1992).
4. Graham did not present a situation where a “special” or custodial relationship gave rise to a constitutional duty on the part of the school to protect the student victims. We had previously held in Maldonado v. Josey, 975 F.2d 727, 732 (10th Cir.1992), that “compulsory attendance laws do not create an affirmative constitutional duty to protect students from the private actions of third parties while they attend school.”
5. Consider here the Supreme Court's decision in Martinez v. California, 444 U.S. 277 (1980), where the Court addressed a claim based on danger creation solely in terms of causation. Therein the Court rejected a claim that state officials' decision five months earlier to parole an inmate subjected his murder victim to a deprivation of life without due process of law. The Court reasoned that because the officials were not aware that decedent, as distinguished from the public at large, faced any “special danger,” her death was “too remote a consequence” of the officials' action to hold them responsible for the murder under § 1983. Id. at 285.
6. We note here that many of our cases after Armijo inaccurately refer to “danger creation” in terms of private third-party violence rather than simply private violence. See, e.g., Rost ex rel. K.C. v. Steamboat Springs RE–2 Sch. Dist., 511 F.3d 1114, 1125 (10th Cir.2008). Even Armijo and a later decision applying the theory in the context of a suicide alluded to third-party violence. Armijo, 159 F.3d at 1260; Christiansen v. City of Tulsa, 332 F.3d 1270, 1279 (10th Cir.2003) (applying the state-created danger theory where decedent shot himself during a police standoff). As both these suicide cases demonstrate, however, we do not strictly require third-party violence as a precondition to invoking the state-created danger theory, but rather only private violence.
7. As an alternative to our holding in Sutton that the principal on the facts alleged could not be liable for his “direct participation in enhancing the danger” to the child, we held the principal could be liable on those same facts for his inaction in failing “to adequately train school employees or adopt or implement a policy to prevent sexual assaults like those against [the child].” Sutton, 173 F.3d at 1238–39. This approach to the principal's accountability might well have succeeded under a theory of supervisory liability, if the mother first was able to establish an underlying constitutional violation on the part of the teacher's aide or other state actor. See, e.g., City of Los Angeles v. Heller, 475 U.S. 796, 799 (1986) (per curiam) (holding that if a police officer inflicted no constitutional injury on a suspect, “it is inconceivable” that the police commissioners could be liable to the suspect); Morris v. Lanpher, 563 F.3d 399, 403 (8th Cir.2009) (recognizing that without an underlying constitutional violation, a § 1983 claim for failure to supervise “necessarily fails”); see also Currier v. Doran, 242 F.3d 905, 922–23 (10th Cir.2001) (relying on Sutton to uphold a claim for failure to supervise against a child welfare supervisor as part of a complaint that also adequately alleged an underlying constitutional violation against subordinate social workers based on danger creation); cf. Martinez v. Beggs, 563 F.3d 1082, 1092 (10th Cir .2009) (holding a sheriff could not be liable for implementing county policies where no underlying violation of decedent's rights had occurred). We believe Sutton recognized the requirement that supervisory liability be premised on an underlying constitutional violation. We take this view of Sutton for two reasons. First, Sutton exclusively relied on cases involving claims of municipal or supervisory liability arising out of a subordinate's allegedly unconstitutional conduct. See Sutton, 173 F.3d at 1239–41 (discussing City of Canton v. Harris, 489 U.S. 378 (1989); Green v. Branson, 108 F.3d 1296 (10th Cir.1997); Meade v. Grubbs, 841 F.2d 1512 (10th Cir.1988); Greason v. Kemp, 891 F.2d 829 (11th Cir.1990)). Second, Sutton made plain not once, not twice, but three times, that in reaching its decision regarding the principal's alleged failure to supervise, it was construing the complaint pursuant to the now defunct and “best forgotten” Conley standard. Conley v. Gibson, 355 U.S. 41, 45–46 (1957), overruled by Bell Atl. Corp. v. Twombly, 550 U.S. 544, 562–63 (2007). That standard purported to prohibit a complaint's dismissal for failure to state a claim “unless it appears beyond doubt that the plaintiff can prove no set of facts in support of his claim which would entitle him to relief.” Sutton, 173 F.3d at 1239 (quoting Conley 355 U.S. at 45–46); see also id. at 1236, 1241. To be sure, confusion may arise from Sutton's reference to the mother's failure to supervise theory as a “variation[ ] of the ‘danger creation’ doctrine.” Id. at 1237–38. But if that was literally true, Sutton would be difficult if not impossible to square with the precedents by which it was bound, in particular DeShaney and Graham. DeShaney told us that absent a custodial relationship, a State's failure to protect a child from private violence does not violate the Due Process Clause. DeShaney, 489 U.S. at 197. And Graham told us that the “state-created danger doctrine necessarily involves affirmative conduct on the part of the state in placing the plaintiff in danger,” and that “inaction by the state in the face of a known danger is not enough” to invoke the protections of the Due Process Clause. Graham, 22 F.3d at 995 (internal quotations omitted). To the extent Sutton may be said to conflict with these decisions, Sutton would have no precedential effect. See Utah Envtl. Cong. v. Troyer, 479 F.3d 1269, 1292 (10th Cir.2007) (stating the rule that where Tenth Circuit panel decisions conflict, the earliest decision controls). We need not definitively resolve the tension between Sutton and our prior precedents addressing the state-created danger theory, however, because Plaintiffs in this case do not raise Sutton's alternative theory of relief as part of their danger creation claim. In fact, Plaintiffs do not cite Sutton in either of their appellate briefs and when questioned at oral argument about its alternative theory of relief, knew little, if anything, about it. See Carpenter v. Boeing Co., 456 F.3d 1183, 1198 n.2 (10th Cir.2006) (recognizing we will not consider theories on appeal alluded to in a vague or ambiguous manner).
8. As a postscript to DeAnzona, we note the six-factor test we have employed since Uhlrig to evaluate state-created danger claims is proper as far as it goes. As DeAnzona illustrates, however, our test is prone to generate oversight on the part of courts and counsel alike because it fails to expressly incorporate those preconditions necessary in this Circuit—affirmative conduct and private violence—to invoking the state-created danger theory.
9. In Robbins, we explained as an initial matter that the complaint's substantive due process count did not provide adequate notice to the defendants concerning the nature of the claims against them individually because the pleading failed “to isolate the allegedly unconstitutional acts of each defendant.” Robbins, 519 F.3d at 1250. The same might be said for Plaintiffs' complaint in this instance. See supra at 2–3. A federal complaint must allege both (1) a claim for relief and (2) facts supporting such claim. A “plain statement” of the claim showing a plaintiff is entitled to relief is necessary to provide defendants “fair notice of what the claim is and the grounds upon which it rests.” Twombly, 550 U.S. at 555 (internal ellipses and quotations omitted) (citing Fed.R.Civ.P. 8(a)). Because in § 1983 actions, named defendants often include a governmental entity and numerous state actors, “it is particularly important in such circumstances that the complaint make clear exactly who is alleged to have done what to whom, to provide each individual with fair notice as to the basis of the claims against him or her, as distinguished from collective allegations against the state.”Robbins, 519 F.3d at 1250. To provide adequate notice as to the nature of multiple claims against multiple defendants, a complaint must isolate the allegedly unlawful acts of “each defendant.” Id. Otherwise, defendants have no way of knowing “what particular unconstitutional acts they are alleged to have committed.” Id. Because in this case Plaintiffs' danger creation claim and the facts alleged to support that claim use either the collective term “Defendants” or “University Hospital” and do not specify what acts are attributable to whom, the individual Defendants cannot be sure what unconstitutional acts they are alleged to have committed. But because Defendants have not raised this particular objection to Plaintiffs' complaint, we leave it at that.
10. By definition, the special relationship theory necessarily only applies where a “custodial relationship” exists between the victim and the State. DeShaney tells us such a relationship exists in instances, such as incarceration or institutionalization, where “the State takes a person into custody and holds him there against his will.” DeShaney, 489 U.S. at 199–200. In other words, the State's exercise of control must “so restrain [ ] an individual's liberty that it renders him unable to care for himself, and at the same time fails to provide for his basic human needs.” Id. at 200. A careful reading of DeShaney further informs us that application of the state-created danger theory is unnecessary and unwarranted where a custodial or special relationship exists between the victim and the State because in that case, the relationship itself gives rise to the State's corresponding duty to protect the victim. The idea that the doctrine of substantive due process encompasses state-created dangers, which, in turn, may give rise to a duty on the part of the State to protect the victim apart from any special relationship, arises from DeShaney: “While the State may have been aware of the dangers that [the child] faced in the free world, it played no part in their creation, nor did it do anything to render him more vulnerable to them.” Id. at 201 (emphasis added). “[T]he free world” does not embody a special or custodial relationship.
11. No doubt our recognition that involuntary foster care gives rise to a special relationship between the victim and the State was based on DeShaney's view that “[h]ad the State by the affirmative exercise of its power removed [the child] from free society and placed him in a foster home operated by its agents, we might have a situation sufficiently analogous to incarceration or institutionalization to give rise to an affirmative duty to protect.” DeShaney, 489 U.S. at 201 n.9 (emphasis added).
12. We acknowledge that at the time of his fatal seizure, decedent may have been functionally dependent on his caretakers. We reject, however, Plaintiffs' suggestion that decedent was in custody because he was connected to his bed by medical devices and was not allowed to leave his room without authorization. Virtually every patient in the intensive care unit of a public hospital, and many other patients as well, would be in the custody of the State if Plaintiffs had their way. Decedent was no more “restrained” than many ill persons who, like decedent, voluntarily check into public hospitals.
13. Even assuming for the moment that decedent was in state custody at the time of his death, the complaint's failure to allege private violence-a failure about which we will have more to say subsequently—would render the special relationship exception inapposite. In that case, the questions simply would be those posed by the complaint's first claim for relief, namely, whether Defendants exhibited deliberate indifference to decedent's serious medical needs and whether that indifference caused a constitutional deprivation. See, e.g., DeShaney, 489 U.S. at 198–199 n.5 (suggesting deliberate indifference is necessary to establish a substantive due process violation in a custodial setting); Sawyer v. County of Creek, 908 F.2d 663, 666 (10th Cir.1990) (recognizing a state doctor may violate a detainee's right to substantive due process “by exhibiting deliberate indifference to [his] serious medical needs”), overruled on other grounds by Leatherman v. Tarrant Cnty. Narcotics Intelligence and Coordination Unit, 507 U.S. 163 (1993).
14. In Rivera, a teenage girl under subpoena was shot dead to prevent her from testifying at a murder trial. In response to the girl's prior pleas, the police promised to protect her but failed to do so. The First Circuit concluded her mother's claims, though based on the danger creation exception rather than the special relationship exception, were “indistinguishable from those in DeShaney.” Rivera, 402 F.3d at 38.
15. The phrase “special danger” appears to originate in Martinez, where the Court held that a murder victim faced no “special danger” from state officials' prior decision to parole the assailant. Martinez, 444 U.S. at 285; see, supra n.5.
16. Save the case of the child's accidental drowning in DeAnzona—an opinion we have necessarily discounted as an incomplete application of Graham and Uhlrig—our case law is entirely consistent with the principle that an act of private violence is a necessary precondition to our application of the state-created danger theory. See Graham, 22 F.3d at 993 (stabbing; shooting/murder); Uhlrig, 64 F.3d at 571 (murder); Liebson v. New Mexico Corr. Dept., 73 F.3d 274, 275 (10th Cir.1996) (kidnapping /sexual assault); Seamons v. Snow, 84 F.3d 1226, 1230 (10th Cir.1996) (hazing); Radecki v. Barela, 146 F.3d 1227, 1228 (10th Cir.1998) (murder); Armijo, 159 F.3d at 1257 (self-inflicted gunshot/suicide); Sutton, 173 F.3d at 1230–31 (sexual molestation); Currier, 242 F.3d at 909–10 (child abuse/homicide); Martinez v. Uphoff, 265 F.3d 1130, 1132 (10th Cir.2001) (murder); Ruiz, 299 F.3d at 1178 (child abuse); Christiansen, 332 F.3d at 1277 (self-inflicted gunshot/suicide); Johnson ex rel. Estate of Cano v. Holmes, 455 F.3d 1133, 1135 (10th Cir.2006) (child abuse/homicide); Rost, 511 F.3d at 1117–18 (sexual molestation); Robbins, 519 F.3d at 1245–46 (child abuse/homicide).
17. Defendants suggest state medical care gone awry, if voluntarily undertaken by an unconfined patient, can never support a substantive due process claim regardless of the responsible party's conduct or state of mind. In Johnson ex rel. Johnson v. Thompson, 971 F.2d 1487 (10th Cir.1992), a number of parents alleged a substantive due process violation on behalf of their infants born with spina bifida. We reasoned that because “substantive due process” does not encompass a “right to treatment,” the parents' assertion that the infants' deaths arose from state medical providers' intentionally discriminatory decision to deny them treatment made no difference. Id. at 1495–96. The providers' decision to provide some medical services to the infants did not alter our conclusion. We cursorily opined that the providers' actions, “while possibly negligent or even willfully indifferent or reckless,” did not implicate the Constitution. Id . at 1496 (quoting Monahan, 661 F.2d at 993). Johnson's holding that the Due Process Clause does not require a state hospital to provide its patients with competent medical services is analogous to the Supreme Court's holding in Collins v. City of Harker Heights, 503 U.S. 115 (1992), that the Due Process Clause does not require a city to provide its employees with a safe work environment. Id. at 127. In Collins, a city employee suffocated after entering a manhole to unstop a sewer line. Like Johnson, Collins addressed a § 1983 personal liability claim based on an alleged substantive due process violation outside a custodial setting. Collins differs from Johnson, however, in an important respect. In Collins, the employee's widow alleged not only that the city acted with deliberate indifference in breaching a constitutional duty to provide her husband with a safe work environment, but also that the city's deliberate indifference to her husband's safety was “arbitrary government action” that “shock[ed] the conscience of federal judges.” Id. at 126 (internal quotations omitted). The Court implicitly recognized a constitutional cause of action based on the widow's allegations of “arbitrary government action”—allegations apparently lacking in Johnson—when it “refus[ed] to characterize the city's alleged [failure to train or warn] in this case as arbitrary in a constitutional sense.” Id. at 128 (emphasis added). We have no reason to believe the Court in Collins limited such cause of action to the employment context. In fact, the Court noted that the employment relationship was “not of controlling significance” and that its analysis would apply to “pedestrians” harmed by a city's deliberate indifference to his or her safety. Id. at 119; see also County of Sacramento v. Lewis, 523 U.S. 833, 836 (1998) (holding a police chase of a fleeing suspect that resulted in a fatal accident may support a substantive due process claim based on the theory that the officer's conduct constituted arbitrary government action shocking the judicial conscience). Today, however, we need not address the breadth of Johnson's holding or how that holding might apply in the context of a danger creation claim because Plaintiffs allege Defendants' misrepresentations coupled with their policy of permitting EMU staff to leave patients unattended precipitated only a negligent act that caused decedent's death. Sufficient is our holding that Plaintiffs' complaint fails to allege a constitutional deprivation because “the Due Process Clause is simply not implicated by a[n] [underlying] negligent act.” Daniels, 474 U.S. at 328.
18. Importantly, our conclusion is consistent with the well-established rule that state officials may not be held constitutionally accountable for the negligent acts of their subordinates. Supervisory liability, like that alleged in the complaint's second claim for relief, simply will not lie absent an underlying constitutional deprivation. See Heller, 475 U.S. at 799; see also supra n.7.
BALDOCK, Circuit Judge.