Opinion
CORRIGAN, J.
The question in this case is whether patients suffering from Alzheimer’s disease are liable for injuries they inflict on health care [1000]*1000workers hired to care for them at home. Because agitation and physical aggression are common late-stage symptoms of the disease, injuries to caregivers are not unusual. California and other jurisdictions have established the rule that Alzheimer’s patients are not liable for injuries to caregivers in institutional settings. We conclude that the same rule applies to in-home caregivers who, like their institutional counterparts, are employed specifically to assist these disabled persons. It is a settled principle that those hired to manage a hazardous condition may not sue their clients for injuries caused by the very risks they were retained to confront.
This conclusion is consistent with the strong public policy against confining the disabled in institutions. If liability were imposed for caregiver injuries in private homes, but not in hospitals or nursing homes, the incentive for families to institutionalize Alzheimer’s sufferers would increase. Our holding does not preclude liability in situations where caregivers are not warned of a known risk, where defendants otherwise increase the level of risk beyond that inherent in providing care, or where the cause of injury is unrelated to the symptoms of the disease.
We encourage the Legislature to focus its attention on the problems associated with Alzheimer’s caregiving. The number of Californians afflicted with this disease can only be expected to grow in coming years. Training requirements and enhanced insurance benefits for caregivers exposed to the risk of injury are among the subjects worthy of legislative investigation.
BACKGROUND
The relevant facts are undisputed. In 2005, defendant Bernard Cott contracted with a home health care agency to assist with his 85-year-old wife and codefendant Lorraine, who had long suffered from Alzheimer’s disease. The agency assigned plaintiff Carolyn Gregory to work in the Cotts’ home.
Gregory was trained to care for Alzheimer’s patients, and had done so in other assignments. She knew they could be violent. Bernard told her Lorraine was combative and would bite, kick, scratch, and flail. Gregory’s duties included supervising, bathing, dressing, and transporting Lorraine, as well as some housekeeping. In September 2008, Gregory was washing dishes while Lorraine sat at the kitchen table. Bernard was not at home. As Gregory was washing a large knife, Lorraine approached her from behind, bumped into her, and reached toward the sink. When Gregory attempted to restrain Lorraine, she dropped the knife, which struck her wrist. As a result, Gregory lost feeling in several fingers and experienced recurring pain.
Gregory has received workers’ compensation. She also sued the Cotts for negligence and premises liability, with a claim against Lorraine for battery. [1001]*1001The trial court granted a defense motion for summary judgment. A divided Court of Appeal affirmed, holding that Gregory’s claims were barred by the primary assumption of risk doctrine. We affirm the judgment.
DISCUSSION
Since its reformulation in Knight v. Jewett (1992) 3 Cal.4th 296 [11 Cal.Rptr.2d 2, 834 P.2d 696] (Knight), California’s assumption of risk doctrine has taken two quite different forms. Primary assumption of risk is a complete bar to recovery. It applies when, as a matter of law, the defendant owes no duty to guard against a particular risk of harm. Secondary assumption of risk applies when the defendant does owe a duty, but the plaintiff has knowingly encountered a risk of injury caused by the defendant’s breach. Liability in such cases is adjudicated under the rules of comparative negligence. (Cheong v. Antablin (1997) 16 Cal.4th 1063, 1067-1068 [68 Cal.Rptr.2d 859, 946 P.2d 817]; Knight, at pp. 314-315.)
The general duty to avoid injuring others extends to persons “of unsound mind.” (Civ. Code, §§ 41, 1714.)1 Accordingly, Lorraine Cott’s Alzheimer’s disease does not, per se, diminish the duty she owed to Gregory. To shield themselves from liability, the Cotts rely on the primary assumption of risk doctrine, which operates as an exception to the general duty of care.
Primary assumption of risk cases often involve recreational activity, but the doctrine also governs claims arising from inherent occupational hazards. (Nalwa v. Cedar Fair, L.P. (2012) 55 Cal.4th 1148, 1155, fn. 1 [150 Cal.Rptr.3d 551, 290 P.3d 1158]; Knight, supra, 3 Cal.4th at pp. 309-310, fn. 5.) The bar against recovery in that context first developed as the “firefighter’s rule,” which precludes firefighters and police officers from suing members of the public for the conduct that makes their employment necessary. (Neighbarger v. Irwin Industries, Inc. (1994) 8 Cal.4th 532, 538-540 [34 Cal.Rptr.2d 630, 882 P.2d 347] (Neighbarger); see 6 Witkin, Summary of Cal. Law (10th ed. 2005) Torts, § 845 et seq., p. 65 et seq.) After Knight, we have viewed the firefighter’s rule “not ... as a separate concept,” but as a variant of primary assumption of risk, “an illustration of when it is appropriate to find that the defendant owes no duty of care.” (Neighbarger, at p. 538.) Whether a duty of care is owed in a particular context depends on [1002]*1002considerations of public policy, viewed in light of the nature of the activity and the relationship of the parties to the activity. (Neighbarger, at p. 541; Knight, at pp. 314-315.)
We have noted that the duty to avoid injuring others “normally extends to those engaged in hazardous work.” (Neighbarger, supra, 8 Cal.4th at p. 536.) “We have never held that the doctrine of assumption of risk relieves all persons of a duty of care to workers engaged in a hazardous occupation.” (Id. at p. 538.) However, the doctrine does apply in favor of those who hire workers to handle a dangerous situation, in both the public and the private sectors. Such a worker, “as a matter of fairness, should not be heard to complain of the negligence that is the cause of his or her employment. [Citations.] In effect, we have said it is unfair to charge the defendant with a duty of care to prevent injury to the plaintiff arising from the very condition or hazard the defendant has contracted with the plaintiff to remedy or confront.” (Id. at p. 542.) This rule encourages the remediation of dangerous conditions, an important public policy. Those who hire workers to manage a hazardous situation are sheltered from liability for injuries that result from the risks that necessitated the employment.
In Neighbarger, the plaintiffs were safety supervisors at an oil company. The company had hired an outside maintenance contractor, whose employees negligently injured the plaintiffs. We held that the assumption of risk doctrine did not apply, because there was no contractual relationship between the plaintiffs and the maintenance contractor. “When [a] firefighter is publicly employed, the public . . . stands in the shoes of the person who hires a contractor to cure a dangerous condition. In effect, the public has purchased exoneration from the duty of care and should not have to pay twice, through taxation and through individual liability, for that service.” (Neighbarger, supra, 8 Cal.4th at pp. 542-543.) A privately employed safety employee, however, has no such relationship with a third party contractor. The contractor has not “paid in any way to be relieved of the duty of care .... Having no relationship with the employee, and not having contracted for his or her services, it would not be unfair to charge the [contractor] with the usual duty of care towards the private safety employee.” (Id. at p. 543.)
The defendant in Neighbarger relied on “ ‘veterinarian’s rule’ ” cases, in which veterinarians or their assistants were held to have assumed the risk of being bitten by dogs during treatment. (Neighbarger, supra, 8 Cal.4th at pp. 544-545, citing Cohen v. McIntyre (1993) 16 Cal.App.4th 650 [20 Cal.Rptr.2d 143], Wittenberg v. Superior Court (1986) 185 Cal.App.3d 185 [229 Cal.Rptr. 625], and Nelson v. Hall (1985) 165 Cal.App.3d 709 [211 Cal.Rptr. 668].) We noted, however, that the veterinarian’s rule does not support applying assumption of risk “when the defendant is a third party who [1003]*1003has not secured the services of the plaintiff or otherwise entered into any relationship with the plaintiff.” (Neighbarger, at p. 545.)
We took up the veterinarian’s rule in Priebe v. Nelson (2006) 39 Cal.4th 1112 [47 Cal.Rptr.3d 553, 140 P.3d 848] (Priebe). There, a worker in a veterinary kennel sued the owner of a dog that bit her. We noted that the veterinarian’s rule was recognized in Neighbarger and Knight as “yet another application of the doctrine of primary assumption of risk.” (Priebe, at p. 1122, citing Neighbarger, supra, 8 Cal.4th at pp. 544-546, and Knight, supra, 3 Cal.4th at p. 309, fn. 5.) We concluded that the plaintiff, “by virtue of the nature of her occupation as a kennel worker, assumed the risk of being bitten or otherwise injured by the dogs under her care and control . . . .” (Priebe, at p. 1132.)
The Priebe opinion identifies several policy rationales for the veterinarian’s rule. The most fundamental is rooted in the very nature of the profession. When an owner entrusts a dog to the care of trained professionals, the owner is no longer in charge. The professional determines how best to manage the animal, and is in the best position to take protective measures against being bitten. (Priebe, supra, 39 Cal.4th at pp. 1129-1130.) A second basis for the rule is the contractual relationship between the parties. The defendant has retained the plaintiff for services that necessarily include the safe handling of the dog. (Id. at pp. 1130-1131.) A third reason, and one that justified extending the veterinarian’s rule to kennel workers, is the social utility of allowing owners to place their dogs in kennels without the risk of liability. “Encouraging the use of secure kennel boarding facilities . . . serves the salut[a]ry purpose behind the dog bite statute — that of protecting members of the public from harm or injury by dogs not properly under their owners’ control . . . .” (Id. at p. 1131.)
The case most closely on point here is Herrle v. Estate of Marshall (1996) 45 Cal.App.4th 1761, 1770-1772 [53 Cal.Rptr.2d 713] (Herrle), decided after Neighbarger and before Priebe. The plaintiff, a certified nurse’s aide at a convalescent hospital, was struck and injured by an Alzheimer’s patient while moving the patient from a chair to a bed. The parties agreed that violent behavior is a common symptom of Alzheimer’s disease. The hospital had many such patients, and the plaintiff was trained to work with them. (Herrle, at p. 1764.) The Court of Appeal observed that the situation precisely matched the contours of primary assumption of risk, as outlined in Knight: “ ‘the nature of the activity’ was the protection of the patient from doing harm to herself or others; ‘the parties’ relationship to the activity’ was plaintiff’s professional responsibility to provide this protection, [and] the ‘particular risk of harm that caused the injury’ was the very risk plaintiff and her employer were hired to prevent.” (Herrle, at p. 1765, quoting Knight, supra, 3 Cal.4th at pp. 314-315.)
[1004]*1004The plaintiff contended that assumption of risk was not an available defense because mentally incompetent persons are liable for their torts under Civil Code section 41. The court disagreed, reasoning that “section 41 is intended to place the incompetent person in the same posture as the competent person, not in a legally worse position. Where no duty exists in the first place, section 41 does not create one. Competent persons can avail themselves of the doctrine of primary assumption of risk. Likewise the defense is available to the incompetent. Here, plaintiff, by the very nature of her profession, placed herself in a position where she assumed the duty to take care of patients who were potentially violent and to protect such patients from committing acts which might injure others. The danger of violence to the plaintiff was rooted in the ‘ “ ‘ “ ‘very occasion for [her] engagement.’ ” ’ ” ’ [Citations.] This is not a case of a person suffering from senile dementia who gets in a car and causes an accident.” (Herrle, supra, 45 Cal.App.4th at p. 1766.)
Herrle distinguished an early case, Mullen v. Bruce (1959) 168 Cal.App.2d 494 [335 P.2d 945], in which a patient hospitalized with delirium tremens was found liable for injuring a nurse. (Id. at pp. 495-496.) The Mullen court declined to hold as a matter of law that the plaintiff had assumed the risk of her injuries. Her actual knowledge of the danger and reasonable opportunity to safely avoid it were questions of fact, decided in her favor at trial. (Id. at p. 498.) Herrle pointed out that Mullen was decided under an outdated theory of assumption of risk. Under the modem doctrine, as explained in Knight and Neighbarger, the defendant may be held to owe no duty of care in the first place, as “a legal conclusion based on the relationship between the parties.” (Herrle, supra, 45 Cal.App.4th at p. 1767.)
The Herrle court reviewed decisions from Florida and Wisconsin concluding that institutionalized mental patients were not liable for injuries inflicted on their caretakers. (Herrle, supra, 45 Cal.App.4th at pp. 1768-1770, discussing Anicet v. Gant (Fla.Dist.Ct.App. 1991) 580 So.2d 273, 277 [insane patient], Mujica v. Turner (Fla.Dist.Ct.App. 1991) 582 So.2d 24, 25 [Alzheimer’s patient], and Gould v. American Family Mutual Ins. Co. (1996) 198 Wis.2d 450 [543 N.W.2d 282, 287] [Alzheimer’s patient].) The analysis in Gould was particularly consistent with California’s primary assumption of risk doctrine. The Wisconsin Supreme Court reasoned that the injured nurse who sought recovery “was not an innocent member of the public,” but a person “employed as a caretaker specifically for dementia patients.” (Gould, at p. 286.) The patient’s “disorientation and potential for violence [was] the very reason he was institutionalized and needed the aid of employed caretakers.” (Id. at p. 287) Although “ordinarily a mentally disabled person is responsible [1005]*1005for his or her torts,” the relationship of such a person with an “employed caretaker” justifies an exception to the rule. (Ibid.; see Herrle, at p. 1769.) The Herrle court deemed Gould “remarkably congruent” with the analysis in Neighbarger, because it found that “the basic relationship between the Alzheimer’s patient and his employed caretaker ‘justifies exonerating’ the patient ‘from the usual duty of care.’ ” (Herrle, at p. 1770, quoting Neighbarger, supra, 8 Cal.4th at p. 543.)
Herrle concluded that public policy favors exempting patients from liability to health care providers “for injuries inherent in the very condition for which treatment was sought.” (Herrle, supra, 45 Cal.App.4th at p. 1770.) “When the relationship between health care providers and health care recipients is considered, the idea that a patient should be liable for ‘conduct’ part and parcel of the very disease which prompted the patient (or, as here, the patient’s family) to seek professional help in the first place becomes untenable. It is the health care provider, not the patient, who is in the best position to protect against the risks to the provider rooted in the very reason for the treatment. Were we to reach a contrary conclusion, . . . risks most efficiently allocable to and traditionally borne by the health care industry would be shifted to individual patients and their families.” (Id. at pp. 1770-1771.)
The plaintiff in Herrle relied on Neighbarger to argue that the firefighter’s rule should not be extended to bar recovery by a private employee. (Herrle, supra, 45 Cal.App.4th at p. 1771.) The court rejected the argument. Neighbarger recognized that a private contractor “ ‘hired to remedy a dangerous situation . . . , as a matter of fairness, should not be heard to complain of the negligence that is the cause of his or her employment.’ ” (Herrle, at p. 1772, quoting Neighbarger, supra, 8 Cal.4th at p. 542.)2 The defendant in Herrle, through her relatives, had arranged for the plaintiff’s services. Thus, unlike the Neighbarger defendant, she had “paid to be relieved of a duty of care.” (Herrle, at p. 1772.) Given the relationship of caregiver and patient, “it would be unfair to . . . impose on defendant the very duty of care which she had contracted for plaintiff to supply.” (Ibid.)
As the Herrle court recognized, primary assumption of risk in its occupational aspect is readily applicable to the relationship between hired caregivers and Alzheimer’s patients. It was stipulated in Herrle that violent behavior is a common symptom of the disease, and that proposition is well [1006]*1006supported by medical texts,3 legal commentary,4 and the facts of reported cases.5 It follows that the risk of violent injury is inherent in the occupation of caring for Alzheimer’s patients.6 While many such patients never become violent, it is equally true that not all fires injure firefighters, and not all dogs bite veterinarians. Nevertheless, because the risk of injury from those causes is inherent in the occupations of firefighters and veterinarians, it is settled that no duty is owed to protect them from the very dangers they are hired to confront. (Priebe, supra, 39 Cal.4th at p. 1122; Neighbarger, supra, 8 Cal.4th at p. 542.) Herrle’s conclusion that Alzheimer’s patients owe no duty of care to protect hired caregivers from the risk of injury has found support, and no disagreement, in other jurisdictions. (Colman v. Notre Dame Convalescent Home, Inc., supra, 968 F. Supp. at p. 813; Creasy v. Rusk, supra, 730 N.E.2d at p. 667; Berberian v. Lynn, supra, 845 A.2d at p. 129; see 1 Dobbs et al., The Law of Torts (2d ed. 2011) § 237, p. 854.)
[1007]*1007Gregory does not claim that Herrle was wrongly decided. She urges instead that its rationale should not be applied to Alzheimer’s caregivers employed in private homes. She contends the home environment lacks the specialized equipment and trained health care professionals found in institutions. Thus, she argues, in-home caregivers cannot be said to be “in the best position to protect against the risks to the provider rooted in the very reason for the treatment.” (Herrle, supra, 45 Cal.App.4th at p. 1770.) Gregory notes that she was not a certified health care professional, and asserts that Lorraine was not her “patient.” She points out that unlike the plaintiff in Herrle, she was not caring for her client at the time of her injury, but instead was engaged in housekeeping. Accordingly, Gregory maintains that primary assumption of risk should not bar her suit, and her claims should instead be analyzed under the secondary assumption of risk doctrine.
Secondary assumption of risk, however, is predicated on the existence of a duty. “The first question is whether the defendant has breached a duty to the plaintiff. The duty analysis depends on the nature of the activity . . . and the parties’ relationship to it. (Knight, [supra, 3 Cal.4th] at p. 308.)” (Shin v. Ahn (2007) 42 Cal.4th 482, 498 [64 Cal.Rptr.3d 803, 165 P.3d 581].) It is only when “it has been established that a duty has been breached, however that duty is appropriately defined under the circumstances of the case, [that] the general principles of comparative fault are applied to assign liability in proportion to the parties’ respective fault.” (Id. at pp. 49S-499.)7 On the fundamental question of duty, Gregory’s attempts to distinguish home health care workers from those employed in institutions are not persuasive. In each setting, caring for patients with Alzheimer’s dementia is the “nature of the activity.” Caregivers are hired to protect the patients from harming themselves or others. If a patient injures a caregiver by engaging in the combative behavior symptomatic of Alzheimer’s disease, the “particular risk of harm that caused the injury” was among the very risks the caregiver was hired to prevent. (Knight, at pp. 314-315; see Herrle, supra, 45 Cal.App.4th at p. 1765.)
Gregory’s claim that caregivers in private homes face higher risks and have fewer risk management tools than institutional caregivers is entirely speculative. It might be that institutions, which house larger numbers of potentially dangerous patients, are riskier than private homes, where the caregiver may develop familiarity with one patient’s proclivities. The record here sheds no [1008]*1008light on the question. The reported cases reflect many more instances of injury to employees caring for Alzheimer’s patients in institutions than to in-home caregivers. Institutional health care workers were the plaintiffs in Herrle, supra, 45 Cal.App.4th at page 1764, Colman v. Notre Dame Convalescent Home, Inc., supra, 968 F. Supp. at page 810, Mujica v. Turner, supra, 582 So. 2d at page 24, Creasy v. Rusk, supra, 730 N.E.2d at page 661, Berberian v. Lynn, supra, 845 A.2d at page 123, and Gould v. American Family Mutual Ins. Co., supra, 543 N.W.2d at page 283. Only one case other than Gregory’s has involved injury in a private home, and that injury was not the result of violence. In Vinccinelli v. Musso (La.Ct.App. 2002) 818 So.2d 163, the patient spilled ice cream, on which the caregiver slipped and fell. The court drew no distinction because the injury occurred in a home. It relied on cases arising in institutional settings to conclude that the patient owed no duty to the employee whose contractual duty it was to care for her. (Id. at pp. 166-167.) Gregory fails to establish that in-home caregivers face appreciably higher risks than those employed in institutions.
Gregory contends that primary assumption of risk should not bar recovery by an in-home worker who is not a certified health care professional. She notes that the plaintiff in Herrle was a certified nurse’s aide. However, Gregory does not explain how certification affects the legal question of duty. Primary assumption of risk is analyzed in terms of function, not formality. Volunteer firefighters assume the risk of injury just like their officially employed counterparts. (Neighbarger, supra, 8 Cal.4th at p. 544; Baker v. Superior Court (1982) 129 Cal.App.3d 710, 717-718 [181 Cal.Rptr. 311].) The same is true of veterinarian’s assistants and their credentialed employers. (Priebe, supra, 39 Cal.4th at pp. 1129-1130; Nelson v. Hall, supra, 165 Cal.App.3d at pp. 714-715.)
The duties of the plaintiff in Herrle were quite similar to those performed by Gregory. The dissent noted that the plaintiff’s “duties, for which she was paid not much more than minimum wage, included changing bedpans, helping the elderly to and from their beds, and assisting them in feeding and dressing themselves.” (Herrle, supra, 45 Cal.App.4th at p. 1773 (dis. opn. of Wallin, J.).) If we were to apply a different rule to uncertified in-home caregivers, employees attending to multiple institutionalized Alzheimer’s patients would be deemed to have assumed the risk of injury, while those performing the same duties for single patients in private homes would not. The logic and the fairness of that outcome are not evident.
We acknowledge that Gregory is not a doctor or a nurse. However, it is her occupation to care for Alzheimer’s patients. We do not hold that anyone who helps with such patients assumes the risk of injury. The rule we adopt is limited to professional home health care workers who are trained and [1009]*1009employed by an agency.8 Although Gregory now claims the training she received from her employer was insufficient, she expressed no reservations about the adequacy of her training in her deposition testimony. In any event, the important consideration is that Bernard Cott contracted with an agency that promised to provide him an aide trained to manage his wife’s condition. By doing so, he paid to be relieved of a duty to protect the aide from the very risks she was retained to encounter. (Neighbarger, supra, 8 Cal.4th at pp. 542-543; Herrle, supra, 45 Cal.App.4th at p. 1772.) Certification and minimum training requirements for such workers are subjects suitable for legislative consideration.
Gregory does not contend that the level of her compensation is relevant to the primary assumption of risk analysis. However, as amici curiae, unions representing home health care workers argue that their low rate of pay makes it inequitable to apply the firefighter’s rule to their occupation. They point out that firefighters are rewarded for their dangerous work with special pay and benefits. The dissent below, like the Herrle dissent, makes the same point. However, this factor has never been determinative, as cases applying the firefighter’s and veterinarian’s rules demonstrate.
“Although cases often cite the special benefits and compensation that firefighters and police officers receive as one reason underlying the firefighter’s rule, no case has held that receipt of or eligibility for those benefits is a requirement for application of the firefighter’s rule. On the contrary, cases have concluded receipt of special compensation or benefits is not a requirement for application of the rule. (Hodges v. Yarian (1997) 53 Cal.App.4th 973, 983 [62 Cal.Rptr.2d 130]; Neighbarger, supra, 8 Cal.4th at p. 544; Baker v. Superior Court[, supra,] 129 Cal.App.3d 710, 717-718.) In Baker, volunteer firefighters who were paid a mere $5 per call for firefighting were barred from maintaining personal injury actions because the firefighter’s rule applied. (Baker, supra, at pp. 717-718; Neighbarger, supra, at p. 544.)” (City of Oceanside v. Superior Court (2000) 81 Cal.App.4th 269, 284 [96 Cal.Rptr.2d 621].) “It was not the amount of public compensation that was determinative in Baker . . . but the relationship between the public and the firefighters who serve it.” (Neighbarger, at p. 544; see City of Oceanside, at p. 285 [inadequate compensation did not preclude application of firefighter’s rule to lifeguards].)
In Priebe, supra, 39 Cal.4th 1112, the plaintiff was a “ ‘kennel technician’ ” who had been working “for about four weeks.” (Id. at pp. 1117-1118.) Her duties included “ ‘feeding, walking, cleaning, laundry, helping hold [1010]*1010animals, [and] assisting the veterinarians and the technicians holding animals.’ ” (Id. at p. 1117.) There was no mention of special compensation. Nevertheless, we held that the plaintiff assumed the risks associated with caring for dogs. (See Nelson v. Hall, supra, 165 Cal.App.3d 709, where the plaintiff was a veterinarian’s assistant.)
Gregory suggests she was as much a housekeeper as a caregiver, and emphasizes that she was injured while washing dishes, not directly attending to Lorraine. If Gregory had been retained as a housekeeper, primary assumption of risk would not bar her action because she would not have been hired to manage the risks posed by Lorraine’s dementia. But Gregory worked for a home health care agency, not a housekeeping service. The circumstance that her duties included some housekeeping does not alter the central reason for her employment: Lorraine’s inability to care for herself due to Alzheimer’s disease. This fact establishes their relationship as caregiver and patient, and supports the application of primary assumption of risk. It is undisputed that Gregory’s duties included constant supervision of Lorraine, to protect not only Lorraine but also Bernard and Gregory herself.
Gregory argues that Bernard should be held liable for failing to install restraining devices and mirrors to facilitate observation and prevent Lorraine from catching others by surprise. However, courts are ill equipped to prescribe safety standards. Were we to allow recovery on this ground, families who retain caregivers for Alzheimer’s patients would have little guidance as to which devices and modifications might be sufficient to avoid liability in their particular situation. This is not to say that Bernard owed no duty of care to Gregory relating to conditions in the home. Under the firefighter’s rule, recovery is not barred when the injury was caused by factors independent of the activity that required the plaintiff’s presence.9 Accordingly, the Cotts as homeowners would be liable to Gregory for torts unrelated to Lorraine’s Alzheimer’s disease, including those involving a dangerous condition of their property.
In general, primary assumption of risk does not bar recovery when the defendant’s actions have unreasonably increased the risks of injury beyond those inherent in the activity. (Nalwa v. Cedar Fair, L.P., supra, 55 Cal.4th at p. 1162, citing cases.) If Bernard had done or failed to do something that elevated Gregory’s risk of injury, this limitation on the [1011]*1011doctrine would apply. But, having hired Gregory to care for Lorraine, Bernard owed Gregory no duty to protect her from the ordinary risks that arose in the course of that employment. Gregory claims she was not specifically warned that Lorraine might approach her from behind while she was washing dishes. It is true that a defendant who misrepresents or hides a hazardous condition is subject to liability. (Priebe, supra, 39 Cal.4th at p. 1115; Lipson v. Superior Court (1982) 31 Cal.3d 362, 371 [182 Cal.Rptr. 629, 644 P.2d 822].) Gregory, however, claims no deception by Bernard. She concedes that he informed her of Lorraine’s combative tendencies. No advisement can reasonably be required to anticipate every variation of circumstance in which a disclosed risk might develop.
Gregory also contends that intentional conduct does not come within the scope of primary assumption of risk, so that her battery claim against Lorraine should survive. Determining Lorraine’s intent when Gregory was injured, or indeed the intentions of any late-stage Alzheimer’s patient, is an uncertain enterprise. In any event, whether “intentional” or not, violent conduct by such patients is an inherent aspect of the caregiving function, and therefore within the scope of the assumed risk. (Cf. Avila v. Citrus Community College Dist. (2006) 38 Cal.4th 148, 164-166 [41 Cal.Rptr.3d 299, 131 P.3d 383] [being intentionally hit by pitch is inherent risk in baseball].)
The assaultive conduct in Herrle, and the other cases involving institutionalized Alzheimer’s patients, could be characterized as “intentional.” Nevertheless, the courts held that the patients owed no duty to protect the plaintiffs from the behavior that was the reason for their employment.10 Absence of [1012]*1012duty bars recovery for intentional torts as well as for negligence. “A tort, whether intentional or negligent, involves a violation of a legal duty, imposed by statute, contract, or otherwise, owed by the defendant to the person injured.” (5 Witkin, Summary of Cal. Law, supra, Torts, § 6, pp. 48-49; see Cedars-Sinai Medical Center v. Superior Court (1998) 18 Cal.4th 1, 8 [74 Cal.Rptr.2d 248, 954 P.2d 511].)11
There is an argument, though Gregory does not explicitly make it, that liability should be imposed to encourage the institutionalization of patients who develop violent tendencies. Public policy, according to this view, is served by isolating the dangerously demented to minimize the threat they pose. We note, first, that the incentive to institutionalize is not entirely removed by the rule we adopt. As a general matter, Alzheimer’s patients and their families are liable under Civil Code section 41 for the injuries they inflict. Our holding bars recovery by only one class of plaintiffs: those employed to care for the patient. We also note that institutionalization is not an effective solution to the problem of injury to caregivers in general; as we have seen, it is not uncommon for Alzheimer’s patients to injure employees in institutions.12
Most importantly, however, the idea that tort liability should be imposed to encourage placing the mentally disabled in institutions is inconsistent with the modern policy preference for keeping these patients in their homes whenever possible. (See Note, Rejecting the Logic of Confinement: Care Relationships and the Mentally Disabled Under Tort Law (1999) 109 Yale L.J. 381.) Support for institutionalization can be found in older cases. But the public [1013]*1013policy disfavoring institutional confinement of the mentally disabled has gained strength in recent years, and legislatures have taken measures aimed at keeping patients in their homes.
The case law reflects these developments. The prevailing view in the older cases was that hired caregivers did not assume the risk of injury by their insane patients, whether at home or in institutions. (Mullen v. Bruce, supra, 168 Cal.App.2d at pp. 497-498 [assumption of risk by nurse in convalescent hospital resolved against patient at trial]; Burrows v. Hawaiian Trust Co. (1966) 49 Haw. 351 [417 P.2d 816, 821-822] [nurse assumed risk of injury in home only if it was preventable by exercise of due care, or if hazard was unreasonably accepted]; McGuire v. Almy (1937) 297 Mass. 323 [8 N.E.2d 760, 763] [nurse did not assume risk of assault by insane person during violent episode in home]; Van Vooren v. Cook (N.Y.App.Div. 1947) 273 A.D. 88 [75 N.Y.S.2d 362, 366] [attendant in mental hospital did not invite assault by patient].)
The law permitting recovery by caregivers began to change in 1991, when the Florida District Court of Appeal held that a mental patient owed no duty to the hospital attendant he injured. (Anicet v. Gant, supra, 580 So.2d at p. 276.) The court relied primarily on the considerations underlying the firefighter’s rule, but also reasoned that patients’ families should be exempt from liability because they had done as much as they could to protect the public from harm by confining the patient in an institution. (Ibid.) The latter rationale was mentioned in two subsequent cases, Gould v. American Family Mutual Ins. Co., supra, 543 N.W.2d at page 287, and Colman v. Notre Dame Convalescent Home, Inc., supra, 968 F.Supp. at page 813.
In 2000, the Indiana Supreme Court recognized the conflict between modem attitudes toward disability and the idea that confinement brings protection against liability. In Creasy v. Rusk, supra, 730 N.E.2d 659, a nursing assistant in a health care facility was injured by an Alzheimer’s patient. The court acknowledged that those with mental disabilities are generally held to the usual standard of care to avoid injuring others, but concluded that public policy and the nature of the relationship between a professional caregiver and a mentally disabled patient justifies an exception to the ordinary mie. (Id. at pp. 667-668.) However, the court expressly refrained from “endorsing the incentives for confinement” that arise from outdated views of the benefits of institutionalization. (Id. at p. 668.)
The Creasy court observed: “Since the 1970’s, Indiana law has strongly reflected policies to deinstitutionalize people with disabilities and integrate them into the least restrictive environment. National policy changes have led the way for some of Indiana’s enactments in that several federal acts either [1014]*1014guarantee the civil rights of people with disabilities or condition state aid upon state compliance with desegregation and integrationist practices. [Citations.]” (Creasy v. Rusk, supra, 730 N.E.2d at pp. 664-665, fn. omitted, citing Indiana statutes.) “It is clear . . . that contemporary public policy has rejected institutionalization and confinement for a ‘strong professional consensus in favor of . . . community treatment . . . and integration into the least restrictive . . . environment.’ Indeed, scholarly commentary has noted that ‘new statutes and case law . . . have transformed the areas of commitment, guardianship, confidentiality, consent to treatment, and institutional conditions.’ ” (Creasy v. Rusk, supra, 730 N.E.2d at p. 666, fns. omitted, citing, inter alla, Note, Rejecting the Logic of Confinement: Care Relationships and the Mentally Disabled Under Tort Law, supra, 109 Yale L.J. at p. 390.)
California law also strongly disfavors institutionalizing those with mental disabilities, including the elderly. The Legislature has provided for the licensure of “home health agencies” to provide residential services (Health & Saf. Code, § 1727 et seq.), with an eye toward “preventing, postponing, and limiting the need for unnecessary institutionalization.” (Health & Saf. Code, § 1727.7, subd. (a)(1).)13 “It is the intent of the Legislature to ensure that the department licenses and certifies home health agencies in a reasonable and timely manner to ensure that Californians have access to critical home- and community-based services. . . . Home health agencies help the state protect against the unnecessary institutionalization of individuals and are integral in ensuring the state’s compliance with the United States Supreme Court decision in Olmstead v. L.C. (1999) 527 U.S. 581 [144 L.Ed.2d 540, 119 S.Ct. 2176], which requires public agencies to provide services in the most integrated setting appropriate to the needs of qualified individuals with disabilities.” (Health & Saf. Code, § 1728.8, subd. (a).)14
The Legislature also expressed a policy preference for minimizing the institutionalization of the elderly and disabled when it passed the California Adult Day Health Care Act. (Health & Saf. Code, § 1570 et seq.) “The Legislature hereby finds and declares that there exists a pattern of overutilization of long-term institutional care for elderly persons or adults with disabilities, and that there is an urgent need to establish and to continue a community-based system of quality adult day health care which will enable elderly persons or adults with disabilities to maintain maximum independence. While recognizing that there continues to be a substantial need for [1015]*1015facilities providing custodial care, overreliance on this type of care has proven to be a costly panacea in both financial and human terms, often traumatic, and destructive of continuing family relationships and the capacity for independent living.
“It is, therefore, the intent of the Legislature in enacting this chapter and related provisions to provide for the development of policies and programs that will accomplish the following:
“(a) Ensure that elderly persons and adults with disabilities are not institutionalized inappropriately or prematurely.
“(b) Provide a viable alternative to institutionalization for those elderly persons and adults with disabilities who are capable of living at home with the aid of appropriate health care or rehabilitative and social services.” (Health & Saf. Code, § 1570.2.)
Like the Indiana Supreme Court, we are reluctant to subscribe to a rationale that would encourage the confinement of Alzheimer’s patients in institutions. California public policy clearly favors alternative arrangements in which these patients are assisted to remain at home. The contemporary view of institutionalization as a last resort counsels in favor of a rule that encourages families to retain trained home health care workers to supervise and assist late-stage Alzheimer’s patients. If families were protected from liability to caregivers only if they place the patient in an institution, the opposite incentive would operate.
After weighing the public policies involved, we agree with those sister-state jurisdictions which have concluded that workers’ compensation, rather than tort recovery, is the appropriate means of compensating hired caregivers for injuries caused by Alzheimer’s patients. (Creasy v. Risk, supra, 730 N.E.2d at p. 668; Berberian v. Lynn, supra, 845 A.2d at p. 129; cf. Anicet v. Gant, supra, 580 So.2d at p. 276.) The nature and extent of workers’ compensation benefits or other insurance requirements for these caregivers are questions beyond our purview. However, considering the importance of the services they provide, whether in institutions or in private homes, the Legislature may wish to consider those questions.
DISPOSITION
We affirm the Court of Appeal’s judgment.
Cantil-Sakauye, C. L, Baxter, L, and Chin, L, concurred.