COLUMBIA JOURNAL OF RACE AND LAW VOL. 11 JULY 2021 NO. 3 ARTICLE MUTUAL DEFERENCE BETWEEN HOSPITALS AND COURTS: HOW MANDATED REPORTING FROM MEDICAL PROVIDERS HARMS FAMILIES Clara Presler This Article explores the phenomenon of “mutual deference” between the medical and legal systems to show that placing mandated reporting responsibilities on clinicians results in lasting harm for families. On the medical side, clinicians are obligated to defer any “reasonable suspicion” that a child may be at risk to the legal system; their concern may be mild or severe, medical or non- medical in nature. But the legal system, comprised of lay-people in the field of medicine, is ill- equipped to evaluate a medical concern, and so defers back to the clinician’s report when making critical decisions around family integrity. This deference often functions to elevate a clinician’s “reasonable suspicion” to a finding of “imminent risk,” justifying needless and prolonged separation of families. More systemically, mutual deference creates and reinforces medical and legal associations between low-income communities of color and notions of child maltreatment. Mutual deference insulates the medical reporter and the  Supervising Attorney, Family Defense Practice, The Bronx Defenders; J.D., The University of Michigan Law School, 2012; B.A., Carleton College, 2004. I thank Emma Ketteringham for providing critical comments on a previous draft of this Article, and also for establishing links to local hospitals that have become an ongoing avenue for advocacy on behalf of parents affected by the family regulation system. I thank Miriam Mack and Mara Fleder for providing important edits; Hymavati Kavuri, M.D., and Edith Presler, N.P., for insights from a medical perspective; Rebecca Suldan and Ester Johannson- Lebron for contributing legal research; and the editorial staff of CJRL for their support in the publication process. Finally, I honor and thank the parents who agreed to have their story included in this Article. 734 COLUM. J. RACE & L. [Vol. 11:733 legal system from liability while imposing tremendous harm on the families caught in the middle. That mandated reporting laws discourage clinicians from considering this harm when deciding whether to report a family reflects the extent to which the family regulation system has prioritized prosecution over supporting families. Efforts to re-envision how society’s support for and protection of families can move away from state- sanctioned violence and towards strengthening families within their communities must begin with removing mandated reporter responsibilities from medical providers. 2021] MUTUAL DEFERENCE 735 I. Introduction ............................................................................. 736 II. Mutual Deference: Why the Theory of Mandated Reporting Fails in Medical Cases ..................................... 740 A. Mandated Reporting Laws Require and Incentivize Reporters to Defer Their Suspicions to the System, Promising a Process of Checks and Balances ....................................................................... 741 B. The System Is Unable or Unwilling to Provide a Check on a Medical “Reasonable Suspicion” and Instead Defers to the Clinician’s Concern ................. 745 III. Lack of Systemic Guidance Around “Reasonable Suspicion” Means a Clinician’s Concern May Reflect Non-Medical Factors, Including Race and Class Biases .................................................................................. 749 A. A Clinician’s “Reasonable Suspicion” Is an Undescriptive Metric ................................................... 749 B. Hospitals’ Use of Social Risk Factors and Screening Tools to Detect Maltreatment Embed Race and Class Disparities into Medical Opinions ... 751 IV. How Mutual Deference Harms Families in Practice ......... 755 A. A Clinician May Refer a Family to the Family Regulation System for Supportive Intentions; the Court Interprets the Report as Imminent Risk Pending a Hearing ....................................................... 757 B. An Injury Is Severe, but the Doctor’s True Concern is Non-Medical .............................................. 759 C. Doctors Can Make Mistakes, and Critical Legal Decisions Are Based on Those Mistakes; Litigation to Resolve Them Results in Prolonged Separation of Families .................................................................... 761 V. Concluding Remarks: Resisting Mutual Deference ........ 764 736 COLUM. J. RACE & L. [Vol. 11:733 I. INTRODUCTION Race disparities pervade the foster system: families forcibly separated by the state are primarily families of color; Black and Brown children spend more time in the foster system than white children.1 Interrogation of the system that enforces this separation—historically referred to generally as “child protective services” and more recently as the “family regulation system”2—requires that we examine the mechanisms by which families come to the attention of the system in the first place. The hospital setting is one critical juncture,3 and families’ experiences there diverge along race and class lines. Many parents of color must weigh a child’s need for medical attention against the real possibility that their decision to seek care will trigger an investigation and that they will leave the hospital without their child. A parent brings a child to the hospital for medical care or advice. Something about the child’s condition, the clinical history, or the parent’s demeanor sparks a clinician’s concern about the child’s safety. A child may have a physical injury and 1 See Tanya A. Cooper, Racial Bias in American Foster Care: The National Debate, 97 MARQ. L. REV. 215, 258 (2013) (“Studies repeatedly show that ‘children of color are overrepresented at all decision points of the child welfare system: reporting, investigation, substantiation, placement, and exit from [foster] care.’”) (citing ALLIANCE FOR RACIAL EQUITY IN CHILD WELFARE, POLICY ACTIONS TO REDUCE RACIAL DISPROPORTIONALITY AND DISPARITIES IN CHILD WELFARE: A SCAN OF ELEVEN STATES (2009), http://www.antiracistalliance.com/PolicyActionstoReduceRacialDisproportional ityandDisparitiesinChildWelfare.pdf [https://perma.cc/8DJ4-345L]). For a robust review of literature citing data on race disparities, see Tina Lee, Processes of Racialization in New York City’s Child Welfare System, 28 CITY & SOC. 276 (2016). 2 Dorothy Roberts, Abolishing Policing also Means Abolishing Family Regulation, IMPRINT (June 16, 2020, 5:26 AM), https://imprintnews.org/child- welfare-2/abolishing-policing-also-means-abolishing-family-regulation/44480 [https://perma.cc/N7F8-XU6M]. 3 2018 data show that reports of suspected child maltreatment from medical personnel comprised 10.5% of those that were screened in for investigation. U.S. DEP’T HEALTH & HUM. SERVS., CHILD.’S BUREAU, CHILD MALTREATMENT 2018, at 9 exhibit 2-D (2020), https://www.acf.hhs.gov/sites/ default/files/cb/cm2018.pdf [https://perma.cc/4EF4-JSTB]. Physician reports of suspected maltreatment of children have been shown to be the most likely to be supported be subsequent child welfare investigation. See, e.g., Jody E. Warner & David J. Hansen, The Identification and Reporting of Physical Abuse by Physicians: A Review and Implications for Research, 18 CHILD ABUSE & NEGLECT 11 (1994). 2021] MUTUAL DEFERENCE 737 a parent does not know how it was caused or the hospital does not believe the explanation;4 a child may have a medical condition and the parent has missed doctor’s visits;5 a newborn or their parent may test positive for an illegal substance at birth;6 or a parent may disagree with the hospital’s course of treatment for their child’s medical condition.7 The treating clinician may be concerned about the risks caused by the myriad challenges that financial and housing instability pose for a family.8 The concern may be mild or severe, medical or non-medical in nature. To be on the safe side, or because the clinician is a mandated reporter of suspected child maltreatment, or because the clinician assumes that a child protective team will connect the family to supportive programs, the clinician reports this concern to the state. What happens next is unimaginable for parents who have experienced hospitals primarily as safe and reassuring places: a caseworker, and possibly the police, interview the family at the hospital. These officials defer to the doctor’s intuition and 4 See, e.g., Jacqueline Kuruppu et al., Tipping the Scales: Factors Influencing the Decision to Report Child Maltreatment in Primary Care, 21 TRAUMA, VIOLENCE, & ABUSE 427 (2020). See also Jessica Horan-Block, A Child Bumps Her Head. What Happens Next Depends on Race., N.Y. TIMES (Aug. 24, 2019), https://www.nytimes.com/2019/08/24/opinion/sunday/child-injuries- race.html [https://perma.cc/M44D-J49J]. 5 See, e.g., Kristine Fortin, When Child Neglect Is an Emergency, 21 CLINICAL PEDIATRIC EMERGENCY MED. 100784 (2020). 6 See, for example, Comprehensive Addiction and Recovery Act (CARA), Pub. L. No. 114-198, 130 Stat 695 (codified as amended in scattered sections of 42 U.S.C.), Child Abuse Prevention and Treatment Act (CAPTA), 42 U.S.C. §§ 5101–5116i, and 42 U.S.C. § 5106, requiring states to implement policies to “notify” child welfare agencies of babies who fall into one of the three categories: being “affected by substance abuse,” affected by “withdrawal symptoms resulting from prenatal drug exposure,” or having Fetal Alcohol Spectrum Disorder, which has led to hospitals implementing testing policies for birthing women. See also Emma S. Ketteringham et al., Healthy Mothers, Healthy Babies: A Reproductive Justice Response to the “Womb-to-Foster-Care Pipeline”, 20 CUNY L. REV. 77 (2016). 7 See, e.g., Maxine Eichner, Bad Medicine: Parents, the State, and the Charge of “Medical Child Abuse”, 50 U.C. DAVIS L. REV. 205 (2016). 8 Effrosyni D. Kokaliari et al., African American Perspectives on Racial Disparities in Child Removals, 90 CHILD ABUSE & NEGLECT 139, 140 (2019) (“A corollary to the disproportionately high poverty rate among African American children, is the greater likelihood poor parents will face charges of neglect and possible child removal based on conditions related to their precarious financial standing such as poor food quality or lack of medical supervision—factors with which affluent parents are not confronted.”). 738 COLUM. J. RACE & L. [Vol. 11:733 medical knowledge. Relying on the clinician’s report, the caseworker files a case in Family Court alleging the child is neglected or abused. The judge, needing to make an emergency decision, reluctant to weigh in on a medical condition, and trusting the word of a doctor over the parents, removes the child from the care of their parent. Unless the parent contests the removal, the clinician may never be consulted and may never know the effect of their call. Contesting the removal requires navigating hospital bureaucracies, competing schedules of clinicians, and over-clogged court systems. This can take weeks or months. During this time, the child is separated from their parents. This common scenario represents a phenomenon that this Article names and will refer to as “mutual deference.” Current mandated reporting laws require that certain professionals, including medical professionals,9 defer any “reasonable suspicion” to the family regulation system. This low burden reflects the aspiration that a system of checks and balances will follow. But a clinician’s concern cannot be effectively investigated and evaluated on an emergency basis because it is—or is perceived to be—based on specialized medical knowledge. Instead, the family regulation system and the court system (collectively, the “legal system”)—comprised of lay people in the field of medicine—overly defers to the clinician’s concern, making critical decisions affecting family integrity without a full medical context. While mutual deference insulates each part of the system from liability, it devastates the families in the middle. Mutual deference is particularly harmful for Black and Brown families given studies showing the disproportional reports and investigations of children from low-income families of color from hospitals.10 And there is an ominous circularity to it: individual 9 In New York, the original mandated reporting statute of 1964 required only physician and surgeons to report an incident of suspected abuse to a specified agency because they were considered to be reluctant to interfere with family affairs. Iris Ann Albstein, Note, Child Abuse and Maltreatment: The Development of New York’s Child Protection Laws, 5 FORDHAM URB. L.J. 533, 536 (1977). Notably, it was enacted as part of New York Penal Law, but is now contained in New York Social Services Law. Id. See also N.Y. SOC. SERV. LAW § 491 (McKinney 2021). 10 See generally Kathryn S. Krase, Differences in Racially Disproportionate Reporting of Child Maltreatment Across Report Sources, 7 J. 2021] MUTUAL DEFERENCE 739 and collective biases influence mandated reporting and these biases are reinforced by the legal system. Data around which families are caught in the family regulation system then influence how medical institutions screen for potentially at-risk children.11 This Article argues that mandated reporting for medical providers, instead of protecting children, perpetuates the disregard for the bonds of Black and Brown families that characterizes the family regulation system as a whole. Parts II and III examine mutual deference on a systemic level. Part II traces the origins of “mutual deference” to statute and case law, revealing tensions between reporters’ obligations on the one hand and the deference to medical concerns by the legal system on the other. Part III explains why mutual deference is particularly harmful for low-income families of color. Non-medical factors, including clinicians’ individual biases and perceived social risk factors, have been shown to influence clinicians’ reports, yet receive the deference of a medical diagnosis. Part IV illustrates how mutual deference harms families in practice. It describes the experience of three parents in the Bronx who were separated from their children after seeking medical care at a hospital. Concluding remarks propose that removing mandated reporting responsibilities from clinicians is a critical step towards re-envisioning support for families away from the family regulation system entirely. Further, eliminating mandated reporting would restore the primacy of the physician-patient relationship and permit a critical analysis of how child maltreatment has been diagnosed and adjudicated. I offer this Article into the discourse about fundamental challenges to the family regulation system in my personal capacity. But, the experiences that give rise to this Article are rooted entirely in my role as a Family Defense Attorney in the Bronx. In that capacity, I represent parents charged with abuse and neglect of their children in Family Court. I have also delivered trainings at New York City hospitals on mandated reporting and have spent hours speaking with hospital staff— PUB. CHILD WELFARE 351 (2013); Daniel Hirschman & Emily Adlin Bosk, Standardizing Biases: Selection Devices and the Quantification of Race, 6 SOCIO. RACE & ETHNICITY 348 (2020). 11 See infra Part III.B (discussing studies showing racial disparities in reporting patterns among clinicians with more specificity). 740 COLUM. J. RACE & L. [Vol. 11:733 residents, doctors, social workers—about the harmful effects of mandated reporting on families. This Article describes what I have seen. II. MUTUAL DEFERENCE: WHY THE THEORY OF MANDATED REPORTING FAILS IN MEDICAL CASES The theory of mandated reporting depends on a balance of power between the reporter, the investigatory branch of the government, and the court system.12 Statutes and case law13 instruct mandated reporters to defer investigation to the system under a theory of checks and balances. The system promises that caseworkers will investigate the concern and, where necessary, seek judicial review. Critics of mandated reporting have cited its ineffectiveness and unintended consequences.14 This section 12 For a robust history of the emergence and development of mandated reporting, see, for example, Albstein, supra note 9; Monrad Paulsen et al., Child Abuse Reporting Laws—Some Legislative History, 34 GEO. WASH. L. REV. 482 (1965); Leonard G. Brown III & Kevin Gallagher, Mandatory Reporting of Abuse: A Historical Perspective on the Evolution of States’ Current Mandatory Reporting Laws with a Review of the Laws in the Commonwealth of Pennsylvania, 59 VILL. L. REV. TOLLE LEGE 37 (2013). 13 This Article examines primarily New York law around mandated reporting, but the concepts are transferrable to other states as well. Although by 1974, all states had some sort of mandatory reporting law, passage of the federal CAPTA fueled the expansion of state-wide systems. CAPTA aimed to systematize and strengthen existing programs by “provid[ing] financial assistance for a demonstration program for the prevention, identification, and treatment of child abuse and neglect” to establish a National Center on Child Abuse and Neglect, “and for other purposes.” See, Child Abuse Prevention Act of 1973: Hearings Before the Subcomm. on Children and Youth of the Comm. on Labor and Public Welfare, 93d Cong. 137 (1973) [hereinafter CAPTA Hearings] (statement of Sen. Walter Mondale, Chairman, Subcomm. on Child. & Youth). 14 For critiques of mandated reporting, see, for example, Richard Wexler, Mandatory Child Abuse Reporting Belongs in Dustbin, New Research Shows, YOUTH TODAY (Feb. 28, 2020), https://youthtoday.org/2020/02/ mandatory-child-abuse-reporting-belongs-in-dustbin-new-research-makes- clear/ [https://perma.cc/K53G-CG5H]; Richard Wexler, Increasing Mandated Reporting of Alleged Child Abuse and Neglect Will Hurt Children, NAT’L COAL. FOR CHILD PROT. REFORM, https://nccpr.org/the-nccpr-evidence-base-brief- analyses-and-commentaries/ [https://perma.cc/NPB9-9LBG]; Mical Raz, Unintended Consequences of Expanded Mandatory Reporting Laws, PEDIATRICS Apr. 2017, at 1; Jill R. McTavish et al., Mandated Reporters’ Experiences with Reporting Child Maltreatment: A Meta-Synthesis of Qualitative Studies, BMJ OPEN, July 2017, at 1; Mical Raz, More Mandatory Reporting Won’t Keep Children Safe from Predators, WASH. POST (May 1, 2018, 7:00 PM), 2021] MUTUAL DEFERENCE 741 shows how mutual deference in medical cases makes mandated reporting particularly problematic: when the issue is or appears to be medical, the court system does not function as the objective check the system envisioned it to be. Instead, the courts defer to the report absent a countering medical opinion—for practical reasons, such an opinion is unavailable at the time a call is made and often still unavailable when a child is removed from their parent. Deference obscures opportunities for the court to issue orders designed to keep children in their parents’ care,15 rendering the legal system both impotent and complicit in the resulting harm. A. Mandated Reporting Laws Require and Incentivize Reporters to Defer Their Suspicions to the System, Promising a Process of Checks and Balances The resounding message to New York’s mandated reporters is to defer any suspicion a child may be at risk to the family regulation system. Passed in 1973, New York’s Child Protective Services Act addressed the concern that child abuse was going undetected and acted on a legislative intent to increase reporting of suspected child maltreatment to the state.16 The Act https://www.washingtonpost.com/news/made-by-history/wp/2018/05/01/more- mandatory-reporting-wont-keep-children-safe-from-predators/ [https://perma.cc/3JQ9-3W5J]. 15 Pursuant to New York’s Family Court Act section 1028, prior to removing a child from a parent, a judge must consider whether any orders would mitigate the risk of harm. N.Y. FAM. CT. ACT § 1028 (McKinney 2021) See also Nicholson v. Scoppetta, 3 N.Y.3d 357, 378 (2004) (“The court must do more than identify the existence of a risk of serious harm. Rather, a court must weigh, in the factual setting before it, whether the imminent risk to the child can be mitigated by reasonable efforts to avoid removal.”). 16 N.Y. SOC. SERV. LAW § 411 (New York’s Child Protective Services Act was designed “to encourage more complete reporting of suspected child abuse and maltreatment”). See also Diana G-D ex rel. Ann D. v. Bedford Cent. Sch. Dist., 33 Misc. 3d 970, 982 (N.Y. Sup. Ct. 2011), aff’d, 104 A.D.3d 805 (N.Y. App. Div. 2013) (reviews the legislative history of N.Y. SOC. SERV. LAW § 413 and states that “[a]ccording to a June 4, 1973 memorandum from the Department of Social Services in support of Assembly Bill 6514A, which includes enactment of Social Services Law § 413, the Department of Social Services believed that the law is intended to address the issue of the difficulty in obtaining an accurate measure of the [child abuse] problem. It believed there were more instances of child abuse than reported. The objective of the new legislation was to accurately report such abuse.” (internal quotation removed)). See also Satler v. Larsen, 131 A.D.2d 125, 129 (N.Y. App. Div. 1987) (“The importance of rapidly detecting and addressing instances of an evil as pernicious as child abuse cannot be overstated.”). 742 COLUM. J. RACE & L. [Vol. 11:733 instructs medical professionals, teachers, counselors, social service workers, and many others to “report or cause a report to be made” whenever they “have reasonable cause to suspect that a child coming before them in their professional or official capacity is an abused or mistreated child.”17 The system promises to investigate any concern and address a family’s needs in a way that prioritizes keeping families together.18 “Reasonable suspicion” is a low standard, emphasizing that reporters are not meant to investigate or achieve a particular quantum of evidence before making a report. Instead, statutory and case law endorse reporting if a “reasonable person” could be concerned and even when maltreatment is just one of many possible explanations.19 Nor should the reporter delay their reporting: the statute specifies that reports of suspected child abuse or maltreatment under the statute must be made “immediately.”20 The regulations under the statute reassure reporters that their suspicion will be investigated: “There may be times when you have very little information on which to base your suspicion of abuse or maltreatment, but this should not prevent you from calling the SCR. A trained specialist at the SCR will help to determine if the information you are providing can be registered as a report.”21 17 N.Y. SOC. SERV. LAW § 413(1)(a). 18 See, e.g., OFF. CHILD. & FAM. SERVS., CHILD PROTECTIVE SERVICES MANUAL, ch. 6, § H (2020) [hereinafter OFF. CHILD. & FAM. SERVS., MANUAL], https://ocfs.ny.gov/programs/cps/manual/2020/2020-CPS-Manual.pdf [https://perma.cc/B594-RGHA] (“when a child has been assessed to be in imminent danger (i.e. unsafe), CPS should also consider a broad range of safety oriented responses other than removal.”). See also N.Y. COMP. CODES R. & REGS. tit. 18, § 423.3; N.Y. COMP. CODES R. & REGS. tit. 18, § 430.9 (2021); and N.Y. SOC. SERV. LAW § 409-a (2019) (mandating that core preventative services must be made available to a child and the family when there is a danger that the child may be separated from the family and services may prevent such removal or separation). 19 Isabelle v. City of New York, 541 N.Y.S.2d 809 (App. Div. 1989) (finding that required reporters were immune from civil liability for reporting a suspicion of child sexual abuse if there is no willful misconduct or gross negligence, even though the tests for venereal disease came back negative two days later, and commenting, “[m]andated reporters need not await conclusive evidence of abuse or maltreatment but must act on their reasonable suspicions and the law allows them a degree of latitude to err on the side of protecting children who may be suffering from abuse”). 20 N.Y. SOC. SERV. LAW § 415. 21 OFF. CHILD. & FAM. SERVS., SUMMARY GUIDE FOR MANDATED REPORTERS IN NEW YORK STATE (2019) (emphasis added), 2021] MUTUAL DEFERENCE 743 That the role of investigating the report is meant for the state officials and not the reporter is evident in the relatively sparse information a reporter is asked to provide in the report. The regulations request basic identifying information and the basis for concern.22 Notably absent is any instruction that the source include alternative possible causes or mitigating factors for the investigating specialist to consider—for example, information about a child’s special needs or a family’s strengths that would encourage prioritization of family unity despite the reporter’s concerns. Also absent from this list is information that would distinguish poverty or other financial instability from neglect.23 This implies that such information—much of which is required information once a case comes to Family Court24—is within the realm of investigation, while the report is intended to provide the agency only the most basic information needed to begin an investigation. New York incentivizes the reporting of any reasonable suspicion, no matter how minor, by attaching legal and financial penalties to a mandated reporter’s failure to report25 and https://ocfs.ny.gov/publications/Pub1159/OCFS-Pub1159.pdf [https://perma.cc/MS3S-YEY8]. 22 OFF. CHILD. & FAM. SERVS., MANUAL, supra note 18, at ch. 2, § A-3. 23 Section 1012(f)(i)(A) of the Family Court Act distinguishes poverty from neglect by defining a neglected child as one whose “physical, mental or emotional condition has been impaired or is in imminent danger of becoming impaired as a result of the failure of his parent . . . to exercise a minimum degree of care in supplying the child with adequate food, clothing, shelter or education . . . , or medical, dental, optometrical or surgical care, though financially able to do so.” N.Y. FAM. CT. ACT § 1012(f)(i)(A) (emphasis added). 24 OFF. CHILD. & FAM. SERVS., MANUAL, supra note 18, at ch. 6, § H (obligating caseworkers to consider in-home safety measures before executing a removal of a child); see also N.Y. FAM. CT. ACT § 1028 (requiring the court to consider any orders that could ensure the safety of a child in order to avoid a removal). 25 See N.Y. SOC. SERV. LAW § 420; N.Y. COMP. CODES R. & REGS. tit. 18, § 432.8. Most states identify a failure to report as a misdemeanor; some states have raised the penalties to a felony in certain circumstances, for example for a second failure or if the alleged offense is a criminal act. In New York, a mandated reporter’s willful failure to report is considered a Class A misdemeanor, punishable by up to a year in jail or a fine of up to $1,000. For an extended discussion of penalties attached to failure of mandated reporters to report suspicions, see Brown & Gallagher, supra note 12, at 37, 63, 79 (providing state by state list of penalties for a mandated reporter’s failure to report). See generally CHILD.’S BUREAU, CHILD WELFARE INFO. GATEWAY, PENALTIES FOR FAILURE TO REPORT AND FALSE REPORTING OF CHILD ABUSE AND NEGLECT 744 COLUM. J. RACE & L. [Vol. 11:733 immunity for reporters who are later sued.26 Reporters are presumed to be acting in good faith and any future liability for reports that turn out to be unfounded are predicated on a showing of actual malice.27 Indeed, as long as they are acting on a reasonable cause to suspect maltreatment and in good faith, immunity attaches.28 In the face of statutory instructions to report immediately, civil and criminal penalties for failure to do so, and immunity for reports that turn out to be unfounded, “objectivity” emerges as the main check on “reasonable suspicion.” In considering what reasonable suspicion means substantively, courts have commented that, “[w]hether reasonable cause exists to suspect child abuse is an objective question that must be answered in light of the information available to the reporter at the time of her report.”29 Invoking the “reasonable cause” standard in criminal law, courts have looked to what the “ordinarily prudent and cautious [person] under the (2018), https://www.childwelfare.gov/topics/systemwide/laws- policies/statutes/report/ [https://perma.cc/HA5Y-LVSS]. 26 For example, in order to bring a libel case for a false report, a plaintiff has to show that the report was motivated by malice. See, e.g., Dunajewski v. Bellmore-Merrick Cent. High Sch. Dist., 526 N.Y.S.2d 139 (App. Div. 1988). In order to bring a negligence suit, the plaintiff has to show willful misconduct or gross negligence. See, e.g., Ervin v. Bronx Lebanon Hosp. Ctr., 794 N.Y.S.2d 41, 41 (App. Div. 2005); Estiverne v. Esernio-Jenssen, 581 F. Supp. 2d 335 (E.D.N.Y. 2008). See generally CHILD.’S BUREAU, CHILD WELFARE INFO. GATEWAY, IMMUNITY FOR REPORTERS OF CHILD ABUSE AND NEGLECT (2018) https://www.childwelfare.gov/topics/systemwide/laws- policies/statutes/immunity/ [https://perma.cc/8GPW-DEUV]. 27 N.Y. SOC. SERV. LAW § 419 (2019). In a libel suit, the plaintiff bears the burden of proving that the statement was motivated by malice. A plaintiff bringing a negligence suit must overcome qualified immunity and show that the reporter engaged in willful misconduct or were grossly negligent in making the disputed report in order to overcome qualified immunity. Finally, due process claims are contingent on whether the plaintiff can show that the mandated reporter was acting as a state actor. Caselaw has indicated that a hospital complying with the Social Services Law and communicating with Child Protective Services is not sufficient to prove that the reporter acted under the color of state law. See Thomas v. Beth Israel Hospital, Inc., 710 F. Supp. 935, 940 (S.D.N.Y.1989); Estiverne v. Esernio-Jenssen, 581 F. Supp. 2d 335, 345 (E.D.N.Y. 2008). 28 Thomsen v. Kefalas, No. 15-CV-2668 (BCM), 2018 WL 1508735, at *17 (S.D.N.Y. Mar. 26, 2018) (“[E]vidence that the report was unfounded . . . does not–standing alone–undercut the existence of ‘reasonable cause,’ nor rebut the presumption of good faith.” (quoting JC v. Mark Country Day Sch., No. 03-CV- 1414 (DLI) (WDW), 2007 WL 201163, at *7 (E.D.N.Y. Jan. 23, 2007))). 29 Thomsen, 2018 WL 1508735, at *15. 2021] MUTUAL DEFERENCE 745 circumstances” would consider suspicious.30 To establish objectivity, the reviewing court must parse “mere ‘hunch’ or ‘gut reaction’” from objective knowledge that has “at least some demonstrable roots.”31 Therefore, in order for the system of checks and balances to work, the system—here, the investigating agency and the court system—needs the ability and the information necessary to act as an “objective” observer. B. The System Is Unable or Unwilling to Provide a Check on a Medical “Reasonable Suspicion” and Instead Defers to the Clinician’s Concern When a reporter is a medical professional, the family regulation system fails to be the objective check on the low threshold of “reasonable suspicion.” This is evident when comparing medical cases with cases coming from schools, another significant source of reports. The clinician’s report carries the weight of a medical opinion rooted in specialized information; as such, courts’ opinions are comparatively cursory, presuming the reporter’s medical training provides the basis for concern. To analyze “reasonable suspicion,” courts ask whether a reporter acted in good faith when reporting a “reasonable suspicion.”32 In school cases, the facts are easily accessible—a child reports feeling uncomfortable at home or has excessive absences—so courts are able to engage with the information known at the time and consider what the reasonable person would have done.33 For example, in Vacchio v. St. Paul’s United 30 Vacchio v. St. Paul’s United Methodist Nursery Sch., No. 001332/95, 1995 WL 17959412, at *5–6 (N.Y. Sup. Ct. Aug. 18, 1995) (the term “reasonable cause” is defined, as follows: “‘Reasonable cause to believe that a person has committed an offense’ exists when evidence or information which appears reliable discloses facts or circumstances which are collectively of such weight and persuasiveness as to convince a person of ordinary intelligence, judgment and experience that it reasonably likely that such offense was committed and that such person committed it.” (citing N.Y. CRIM. PROC. § 70.10)). 31 People v. Sobotker, 43 N.Y.2d 559, 564 (1978). 32 These cases are primarily liability or negligence cases against a reporter by a family. While this is not the posture that affects the removal of a child from their parent, the analysis in these cases illustrates courts’ limitation in medical cases. Because decisions to remove a child from their parent are often emergency decisions made in summary or oral decisions, there is not a body of Family Court case law documenting judges’ rationale. 33 What is useful about this comparison is not the ultimate decision reached—the presumption that the reporter acts in good faith means that courts act largely as a rubber stamp in all cases—but the degree to which the court engages in a fact-specific analysis. 746 COLUM. J. RACE & L. [Vol. 11:733 Methodist Nursery School, the court considered whether it was reasonable for a school to suspect a student had been abused when the child appeared with a black eye.34 The court denied the school summary judgment, holding that the decision to report a black eye without conducting a preliminary inquiry might support a finding of gross negligence.35 In doing so, the court weighed the presentation of the child with the paucity of other information available to the teacher at the time.36 Even when the court ultimately dismisses the plaintiff’s case, courts do so after considering the underlying facts of the case. In Cox v. Warwick Valley Cent. Sch. Dist., the court considered whether the behavior of a student was sufficiently concerning for the school to make a report of suspected neglect. In affirming summary judgment for the school district, the Second Circuit examined the information available to the school at the time of the report, detailing that the child’s journal entries, misbehavior, and expressions of suicidal thoughts were— objectively—cause for concern and the school was not acting with actual malice.37 Reports by medical professionals receive far more deference and less analysis. Courts presume reports made by medical professionals are grounded in their professional 34 Vacchio, 1995 WL 17959412, at *7 (“[D]oes the presentation of a child with a blackened eye, without more, give rise to a reasonable suspicion of child abuse, or may such conclusion more appropriately be characterized as within the ambit of the term “hunch”?”) 35 Id. at *8–9. 36 Id. See also Thomsen 2018 WL 1508735, at *14, where the court considered whether a teacher had reasonable suspicion to think a child may have been sexually abused by another teacher. In denying summary judgment to the teacher, the court considered facts that undermined the likelihood the abuse took place, such as presence of other adults on the day in question, as well as the defendant’s history of making reports and possible motivations for making a false report. 37 Cox v. Warwick Valley Cent. Sch. Dist., 654 F.3d 267, 276 (2d Cir. 2011) (finding that where the court affirms a reporter’s decision not to report, the inquiry is similarly fact-specific). See Diana G-D ex rel. Ann D. v. Bedford Cent. Sch. Dist., 33 Misc. 3d 970 (N.Y. Sup. Ct. 2011), aff’d, 104 A.D.3d 805 (N.Y. App. Div. 2013), where the court dismissed a negligence claim against a school that did not report sexual abuse allegations concerning a student. The court examined the actions the school took when it became aware of the possibility of abuse and the information that was available to the teachers and administrators. The court considered the child’s behavior, the answers she gave the teachers when she was questioned, and the content of a proximate parent- teacher conference. 2021] MUTUAL DEFERENCE 747 expertise and therefore reasonable. They do not review the medical basis for the concern as they did in the school cases cited above; rather a professional’s concern is the medical basis. This circular reasoning is apparent in Storck v. Suffolk County Department of Social Services, where the court, in granting “good faith immunity” to doctors who had suspected a parent of neglect, commented, “Clearly, when the doctors reported their suspicions of abuse, they were acting ‘in the discharge of their duties and within the scope of their employment.’”38 In Bowes v. Noone, the Appellate Division even implied in dicta that the court should have no role in evaluating whether a doctor was acting within their professional judgment: To determine whether the act of a medical professional deviates from accepted medical standards, it must first be determined whether the act involves the exercise of professional judgment . . . . Here, the medical experts . . . testified that the issue whether a medical professional should report suspected child abuse to the central register involves the exercise of professional judgment.39 This deference leads to cursory reviews of a clinician’s concern. For example, in Kempster v. Child Protective Services, the court found that a report by a hospital based on a baby’s swollen nose was reasonable, citing broadly the “medical data and other available information.” The court deferred to the hospital’s assertion that the injuries were concerning and the mother’s explanation did not explain them.40 In Miriam P., the court presumed the hospital acted in good faith when it reported that a child had a fractured leg the mother was unable to explain. This common theory—that a parent’s inability to adequately explain the cause of an injury is a reasonable basis to suspect 38 Storck v. Suffolk Cnty. Dep’t of Soc. Servs., 62 F. Supp. 2d 927, 946 (E.D.N.Y. 1999). 39 Bowes v. Noone, 748 N.Y.S.2d 440 (App. Div. 2002). 40 Kempster v. Child Protective Servs. of Dep’t of Soc. Servs. of Suffolk Cty., 130 A.D.2d 623, 625 (N.Y. App. Div. 1987). See also Isabelle V. v. City of New York, 541 N.Y.S.2d 809 (App. Div. 1989) (The court examined a hospital’s report of suspected sexual abuse based on the vaginal discharge of two children, despite their denial that any abuse had taken place, then when the cultures came back negative for any venereal disease, and the parents later sued the hospital, the court deferred to the hospital’s concern about the symptoms.). 748 COLUM. J. RACE & L. [Vol. 11:733 abuse—itself reflects deference to the medical profession’s opinion about which explanations are adequate and which are suspicious.41 Similar to courts’ deference to the professional judgment of clinicians is their instruction that caseworkers should rely on and defer to medical professionals’ suspicions. In V.S. v. Muhammad, when plaintiffs argued that the caseworkers’ reliance on a pediatrician known to give unreliable diagnoses in the field of child abuse was sufficiently unreasonable to remove qualified immunity, the Second Circuit disagreed. It commented, “to impose on an [Administration for Children’s Services (ACS)]42 caseworker the obligation in such circumstances of assessing the reliability of a qualified doctor’s past and present diagnosis would impose a wholly unreasonable burden of the very kind qualified immunity is designed to remove.”43 This is perhaps the most explicit acknowledgement of the family regulation system’s inability to provide a check on what counts as “reasonable suspicion” when the reporter is a medical professional.44 41 Miriam v. City of New York, 163 A.D.2d 39, 43–44, 558 N.Y.S.2d 506, 509 (1990). See also Jessica Horan-Block & Elizabeth Tuttle Newman, Accidents Happen: Exposing Fallacies in Child Protection Abuse Cases and Reuniting Families Through Aggressive Litigation, 22 CUNY L. REV. 382, 418 (2019) (discussing how a parent’s lack of explanation for a child’s condition is often erroneously transformed into a basis for an abuse charge). 42 In New York City, the child protective agency is called the Administration for Children’s Services (ACS). 43 V.S. v. Muhammad, 595 F.3d 426 (2d Cir. 2010). See also Estate of Keenan v. Hoffman-Rosenfeld, No. 16-cv-0149 (SFJ) (AYS), 2019 WL 3416374, at *21 (E.D.N.Y. July 29, 2019), aff’d, No. 19-2730-cv, 2020 WL 6494881 (2d Cir. Nov. 5, 2020). 44 Notably, when a lawsuit against a medical professional or hospital survives a motion to dismiss based on immunity, it tends to be for reasons other than the court looking at the basis for the clinician’s reasonable suspicion. For example, in Ying Li v. City of New York, the court allowed discovery to proceed in a civil rights suit based on the fact that the doctor may have gone beyond reporting and instead taken an active role in the investigation and prosecution of the plaintiff. Ying Li v. City of New York, 246 F.Supp.3d 578 (2d Cir. 2017). In Estiverne v. Esernio-Jenssen, the plaintiffs survived a motion to dismiss a civil rights case when they pled information beyond the basis for the individual clinician’s concern, including that the doctor knew the diagnosis to be false and that the clinician disregarded the contrary diagnosis of a colleague. Estiverne v. Esernio-Jenssen, 833 F. Supp. 2d 356 (E.D.N.Y. 2011). In other words, it was only when the court was presented with information that undermined the court’s ability to defer to the professional judgment of the clinician that it found a triable issue as to a presumption of good faith. See id. 2021] MUTUAL DEFERENCE 749 III. LACK OF SYSTEMIC GUIDANCE AROUND “REASONABLE SUSPICION” MEANS A CLINICIAN’S CONCERN MAY REFLECT NON-MEDICAL FACTORS, INCLUDING RACE AND CLASS BIASES Certainly, the legal system cannot make decisions around medical issues without medical evidence. But the extent to which the legal system defers to the initial report from a clinician presumes that a clinician’s “reasonable suspicion” is probative of imminent risk. In fact, studies show that it is largely an undescriptive metric. Clinicians report for a host of reasons that may provide little guidance to a court. The severity of a clinician’s concern may be mild or severe. The possibility of maltreatment may be the leading diagnosis, or one of many possibilities. Further, studies show that a clinician’s reasonable suspicion may be influenced by a range of non-medical factors, including race and class biases, that are invisible to—or shared by—the system that investigates and adjudicates. These biases gain the status of medical opinions and therefore define the course of a family’s experience in the legal system.45 A. A Clinician’s “Reasonable Suspicion” Is an Undescriptive Metric A report of reasonable suspicion provides little description about the clinician’s level of concern. One study found that a report may represent that the reporter perceives abuse to be “very likely” or simply “likely.”46 Additionally, when a clinician 45 See, e.g., Cooper, supra note 1, at 252. Cooper refers to “critical junctures” in the foster care system, “where incentives reinforce interconnections or dynamics between players.” Id. at 251. The many critical junctures described by Cooper, id. at 257–58, are consolidated into reporting of abuse, investigation, substantiation, placement, and exit from foster care. A medical professional’s concern, therefore, dominates four of those critical junctures: the clinician is incentivized to defer any suspicion to the family regulation system, without examination of whether that suspicion is rooted in race or class-based assumptions. A court is incentivized to defer to the medical professional’s suspicion, particularly in the absence of contrary information. The willingness to defer, which often means presuming a child is at risk in their parent’s care, can be traced to the judicial system’s own associations between maltreatment, race, and class. Mutual deference, therefore, is one reason why “children of color are overrepresented at all decision points of the child welfare system.” Id. at 258 (citations omitted). 46 Benjamin H. Levi & Georgia Brown, Reasonable Suspicion: A Study of Pennsylvania Pediatricians Regarding Child Abuse, 116 PEDIATRICS e5 (2005) 750 COLUM. J. RACE & L. [Vol. 11:733 makes a list of differential diagnoses—the list of possible causes of a condition—a report of suspected abuse may indicate that it is thought to be the leading cause, or it may rank as low as tenth on a list of differential diagnoses.47 In terms of the probability of abuse, the study found that twice as many clinicians thought that a report would represent a ten to thirty-five percent probability of abuse than a seventy-five percent probability.48 Further, studies show that a clinician’s decision to report a reasonable suspicion can be influenced by non-medical factors. For example, “familiarity with the patient or family, including any previous involvement of the family with CPS[;] . . . reference to elements of the case history[;] . . . use of available resources; and . . . clinicians’ perceptions of anticipated outcomes of CPS intervention” were significant factors in one study.49 The study also found that clinicians were less likely to report when they had a significant relationship with the family,50 and a clinician’s (indicating that the responding clinicians reported seventy-three percent of the children they considered likely or very likely to be abused; and twenty-four percent of the children they considered possibly abused). The influence of probabilistic language such as this in child abuse diagnoses is problematized by another scholar. See Steven C. Gabaeff, Recognizing the Misuse of Probabilistic Language and False Certainty in False Accusations of Child Abuse, J. RSCH. PHIL. & HIST., Dec. 2020, at 1. 47 Levi & Brown, supra note 46, at e7 (finding that twelve percent of clinicians responded that abuse would have to rank first or second on the list of differential diagnoses before it would be considered reportable; forty-one percent indicated a rank of third or fourth; forty-seven percent reported a rank from fifth to tenth on the list of differential diagnoses). 48 Id. (Thirty-five percent of pediatricians responded that, to report a suspicion, the probability of abuse would need to be ten to thirty-five percent; by contrast, fifteen percent required a probability of more than seventy-five percent. Further, any one individual pediatrician was not necessarily internally consistent in the level of certainty they required: the average pediatrician required fifty to sixty percent probability that abuse occurred, but responded that child abuse could rank as low as fourth or fifth on the differential diagnosis list and still merit a report.). See also Kuruppu et al., supra note 4, at 430 (finding that “each clinician seem[s] to have their own personal threshold of suspicion that would activate their reporting duty”). 49 Risé Jones et al., Clinicians’ Description of Factors Influencing Their Reporting of Suspected Child Abuse: Report of the Child Abuse Reporting Experience Study Research Group, 122 PEDIATRICS 259, 261 (2008). See also Kuruppu et al., supra note 4, at 430 (citing “personal threshold of suspicion, knowing the family, having little faith in the system, and education” and training as significant non-medical factors influencing primary care physicians’ decisions to report). 50 This in itself can lead to disparate reports for low-income and Black families who are less likely to have a primary care provider and more likely to 2021] MUTUAL DEFERENCE 751 decision not to report was influenced by doubt that it would benefit the family.51 Further, past experiences of the clinician with the family regulation system lowered the likelihood that a suspicion resulted in a report.52 B. Hospitals’ Use of Social Risk Factors and Screening Tools to Detect Maltreatment Embed Race and Class Disparities into Medical Opinions The low standard of “reasonable suspicion” is vulnerable to personal biases influencing a clinician’s decision to report. This leads to race and class disparities in reporting and has ripple effects on systemic views of child maltreatment. In turn, data about which children are reported by clinicians to be “neglected” or “abused”—whether or not they have been adjudicated legally as such—inform how clinicians are trained to look for signs of maltreatment, which in turn influence subsequent decisions to report. Tracking of the use of skeletal surveys, full body x-rays that are often conducted when an injury is deemed suspicious of abuse, offers a clear example of racial disparities in medical investigations. Studies have shown that non-white children presenting with head injuries are more likely to receive skeletal surveys,53 as are those who are uninsured or on public resort to an emergency room for medical care. Rick Hong et al., The Emergency Department for Routine Healthcare: Race/Ethnicity, Socioeconomic Status, and Perceptual Factors, 32 J. EMERGENCY MED. 149, 155 (2006) (finding that Black and Hispanic patients were approximately twice as likely as white patients to be routine ED users, probably because of coinciding socioeconomic factors, primarily lack of insurance). 51 Jones et al., supra note 49, at 264. 52 Id. See also Vernoica L. Gunn et al., Factors Affecting Pediatricians’ Reporting of Suspected Child Maltreatment, 5 AMBULATORY PEDIATRICS 96 (2005) (finding that a decision not to report was independently linked to the following factors: men who have been in practice longer, have been deposed or testified in a related matter, or had been threatened with a lawsuit); McTavish et al., supra note 14 (providing qualitative feedback from mandated reporters, including clinicians, who cite negative experiences from reporting). 53 Kent P. Hymel et al., Racial and Ethnic Disparities and Bias in the Evaluation and Reporting of Abusive Head Trauma, 198 J. PEDIATRICS 137, 138 (2018) (finding skeletal surveys to be twice as likely to be ordered for non-white patients under three years old who presented with head injuries as white/non-Hispanic patients; here, “evaluated” referred to radiologic skeletal survey and/or retinal examination by an ophthalmologist). See also Wendy G. Lane, Racial Differences in the Evaluation of Pediatric Fractures for Physical 752 COLUM. J. RACE & L. [Vol. 11:733 insurance.54 Even though socioeconomic status is a significant factor,55 the disparity in skeletal surveys between public and privately insured white patients has been shown to be greater than for Black and Latinx patients, who were more likely to receive skeletal surveys across the board.56 This disparity did not correlate with a positive diagnosis for abuse.57 One way that clinicians have attempted to structure the diagnosis of child neglect or abuse is evaluation of so-called “risk factors.” Risk factors refer to conditions that are considered to be correlated with abuse or neglect. The factors span medical and non-medical concerns: poverty; past history of social services involvement, housing instability, unemployment, and drug use; maternal smoking; and being born to an unwed mother, to name a few.58 But because race is associated “to a shameful degree” Abuse, 288 J. AM. MED. ASS’N 1603 (2002) (finding that the effect of race on ordering of skeletal surveys and reporting to CPS remains significant). 54 Christine W. Paine & Joanne N. Wood, Skeletal Surveys in Young, Injured Children: A Systematic Review, 76 CHILD ABUSE & NEGLECT 237, 242 (2018). 55 Antoinette L. Laskey et al., Influence of Race and Socioeconomic Status on the Diagnosis of Child Abuse: A Randomized Study, 160 J. PEDIATRICS 1003, 1003 (2012) (finding greater likelihood that physician would label a fracture as abuse in patients with low socio-economic status (SES) and remain unsure about the etiology in patients with high SES, but not finding an independent effect for race). See also Emalee G. Flaherty et al., From Suspicion of Physical Child Abuse to Reporting: Primary Care Clinician Decision-Making, 122 PEDIATRICS 611 (2008) (reviewing studies finding no racial differences in reporting when families did not have private insurance but finding also that having private insurance can protect white children from being reported). 56 Joanne N. Wood et al., Disparities in the Evaluation and Diagnosis of Abuse Among Infants with Traumatic Brain Injury, 126 PEDIATRICS 408, 408 (2010) (The difference in skeletal survey performance for infants with public or no insurance versus private insurance was greater among white (82% vs. 53%) infants than among Black (85% vs. 75%) or Hispanic (72% vs. 55%) infants.). 57 See id. (the probability that the survey would lead to a diagnosis of abuse among white infants was higher (61%) than among Black (51%) or Hispanic (53%) infants.); Paine & Wood, supra note 54, at 246 (noting that, although Black children and children with public or no insurance were evaluated with skeletal surveys more often than white infants and infants with private insurance, Black infants had similar likelihood of having a positive skeletal survey compared to white infants). 58 Risk Factors That Contribute to Child Abuse and Neglect, CHILD WELFARE INFO. GATEWAY, CHILD.’S BUREAU, https://www.childwelfare.gov/ topics/can/factors [https://perma.cc/S882-L74B] (last visited May 22, 2021). See also Caitlin A. Farrell, Community Poverty and Child Abuse Fatalities in the United States, 139 PEDIATRICS (2017); Hillary W. Petska & Lynn K. Sheets, Sentinel Injuries: Subtle Findings of Physical Abuse, 61 PEDIATRIC CLINIC N. 2021] MUTUAL DEFERENCE 753 with clinicians’ perceptions of social risk factors, this leads to over-reporting and over-investigation of Black and Brown communities.59 Reporting families exhibiting these risk factors at a higher rate perpetuates stereotypes around risk,60 without providing an accurate assessment of risk. Moreover, one’s approach to risk factors is itself subjective; broad use of risk factors to diagnose maltreatment can lead to significantly varying results.61 To reduce the role of personal bias and alleviate race disparities in reporting, some medical institutions use screening tools, or questionnaires, that aim to standardize identification of AM. 923 (2014) (citing young parental age, mental health disorders, exposure to domestic violence as risk factors for child physical abuse); Cindy W. Christian, The Evaluation of Suspected Child Physical Abuse, 135 PEDIATRICS 1337, 1339 (2015) (citing literature claiming that “[r]isk factors for infant abuse include maternal smoking, the presence of more than 2 siblings, low infant birth weight, and being born to an unmarried mother. Children with disabilities are at high risk for physical, sexual, and emotional abuse. Young, abused children who live in households with unrelated adults are at exceptionally high risk of fatal abuse, and children previously reported to CPS are at significantly higher risk of both abusive and preventable accidental death compared with peers with similar sociodemographic characteristics.” (citations omitted)) 59 Daniel M. Lindberg, Bias and Objectivity when Evaluating Social Risk Factors for Physical Abuse: Of Babies and Bathwater, 198 J. PEDIATRICS 13, 13 (2018). 60 Id. (arguing that clinicians are “ill-equipped” to apply social risk factors and instead “use their intuition to estimate social risk”). 61 Heather T. Keenan et al., Social Intuition and Social Information in Physical Child Abuse Evaluation and Diagnosis, PEDIATRICS, November 2017, at 1. This study bears mentioning for illustrating how use of risk factors and personal intuition lead to varying results in diagnoses. Keenan et al. used three scenarios to demonstrate how the diagnoses of child abuse pediatricians (CAPs) vary depending on the type of information that is available about a family. First, CAPs received all the information a clinician would get from meeting the family; the study labeled this the “gut reaction”: social intuition, social information, risk indicators, and social cues, as well as a full medical report. Second, the CAP received the social information and the medical history, but none of the perceptions from meeting the family or information about race. Third, the CAP received only the medical history. The study found that the more information the CAP had about the family, the more diagnostic certainty CAPs reported. But agreement among CAPs dropped when social information was present. In one out of five diagnoses, knowledge of social information reversed the diagnosis when all other information held constant. Further, CAPs who met the family performed a less complete evaluation than the other two categories, suggesting that “meeting the family encourages an intuitive thinking pathway (‘gut feeling’).” Id. at 5–6. As the Article points out, if a CAP’s intuition is based on social risk factors that are correlated with but not causative of child abuse, it leads to over-reporting for certain groups. 754 COLUM. J. RACE & L. [Vol. 11:733 risk factors.62 But, similar to risk factor evaluations, these tools can easily perpetuate the precise racial and class biases that they are designed to dampen.63 For example, one study found that a screening tool used before administering drug tests to birthing parents inadvertently reinforced “the process of identifying more Black than white women.”64 The protocol mandated that women would be ordered to test based on factors that were more common among Black parents, “including no prenatal care, an earlier positive toxicology test during the pregnancy, current intoxication or signs of placental abruption[,] . . . limited/late prenatal care, having children out of care, past drug or alcohol problems, and previous negative birth outcomes.65 These 62 See, e.g., Lindberg, supra note 59; Mauricio A. Escobar et al., Development of a Systematic Protocol to Identify Victims of Non-Accidental Trauma, 32 PEDIATRIC SURGICAL INT. 377 (2016); Eveline C.F.M. Louwers et al., Effects of Systematic Screening and Detection of Child Abuse in Emergency Departments, 130 PEDIATRICS 457 (2012). 63 Hirschman & Bosk, supra note 10, at 352 (“Because racial inequalities are best diagnosed as reflecting structural racism not (just) individual bias, efforts to reduce racial inequality through standardizing gatekeeping decision making may have little effect.”); Flaherty et al., supra note 55, at 612, 617 (discussing “injury encounter cards” that clinicians would fill out when they diagnosed an injury). Among questions about the type of injury and its seriousness, are questions around social factors and questions that ask for a practitioner’s individual opinion: parents appear to “have little social support,” parents have a “history of drug or alcohol use,” parents are a “victim of [child/spousal] abuse,” parent/child interactions cause concern, prior involvement with CPS. Id. See also Louwers et al., supra note 62, at 458 (evaluating the effectiveness of a checklist, the list—labeled the “Escape Form” to be used in Emergency Departments—included questions that rely on the practitioner’s intuition: “Is the behavior of the child/the carers and the interaction appropriate?” and “Are there any other signals that make you doubt the safety of the child or other family members?”). 64 S. C. M. Roberts et al., Does Adopting a Prenatal Substance Abuse Use Protocol Reduce Racial Disparities in CPS Reporting Related to Maternal Drug Use? A California Case Study, 35 J. PERINATOLOGY 146, 149 (2015). See also Maternal Mortality and Morbidity in New York City: Hearing Before the N.Y. City Council’s Comms. on Hospitals, Health, and Women & Gender Equity, Dec. 7, 2020 (joint written testimony of Ancient Song Doula Services, The Bronx Defenders, Movement for Family Power, National Advocates for Pregnant Women, and the New York Civil Liberties Union) (providing additional background on the issue of testing of pregnant women leading to disproportionate outcomes). 65 Roberts et al., supra note 64, at 147. Even with universal screening, Black women have been shown to be four times more likely to be reported for suspected maltreatment than white women, despite the fact that all women were screened. Sarah C.M. Roberts & Amani Nuru-Jeter, Universal Screening for 2021] MUTUAL DEFERENCE 755 indicators are more common among Black parents, the last being a salient example of how measures like screening tools can perpetuate the cycle of the foster system.66 The law asks clinicians to make reports of suspected child maltreatment to a system that is unable—or unwilling—to decipher a clinician’s reasonable suspicion at the pace necessary to avoid catastrophic harm to a family. The obligation to report and the instruction to defer all investigation to the system obscures critical information, such as the gravity of the concern; whether it is a medical diagnosis or a personal concern; the role of screening tools or hospital policies that triggered the report rather than an acute safety concern. Instead, all these issues receive the deference given to a medical opinion. The report may appear to the system as a medical diagnosis, but any diagnostic error that results from this report is not examined.67 Instead, the process of reporting enhances the heuristic associations between abuse, neglect, race, and class. IV. HOW MUTUAL DEFERENCE HARMS FAMILIES IN PRACTICE Formally, parents’ due process rights are strongest when facing the possible removal of their child from their care:68 in New York, the state must prove that the child would be at “imminent risk” of harm in their parents’ care, and the parent has a right to an emergency hearing to contest a removal.69 Whether to a remove a child is first examined at the Alcohol and Drug Use and Racial Disparities in Child Protective Services Reporting, 39 J. BEHAV. HEALTH SERV. & RSCH. 3, 3 (2012). 66 See Ketteringham et al., supra note 6. 67 See NAT’L ACADS. OF SCIENCES, ENGINEERING, & MED., IMPROVING DIAGNOSIS IN HEALTH CARE 56 (2015) (“Prolonged learning in a regular and predictable environment increases the successfulness of heuristics, whereas uncertain and unpredictable environments are a chief cause of heuristic failure. There are many heuristics and biases that affect clinical reasoning and decision making.” (citations omitted)). See also Ruth Gilbert et al., Recognizing and Responding to Child Maltreatment, 373 LANCET 167 (2009) (commenting on the difficulty of understanding the meaning of a “substantiated” report, in that it can be a reflection of an agency’s determination of risk of future harm rather than confirmation of the reporter’s concern). 68 See, e.g., Stanley v. Illinois, 405 U.S. 645 (1972) (outlining parents’ fundamental right to make decisions regarding the care and custody of their children). 69 N.Y. FAM. CT. ACT § 1028 (establishing the state’s burden and the right of a parent facing possible removal of a child to an emergency hearing to contest a removal or request the return of their child from foster care). 756 COLUM. J. RACE & L. [Vol. 11:733 arraignment—the initial appearance in court when a judge makes a determination about a child’s placement by considering the reporter’s narrative, the caseworker’s investigation, the parent’s counter-narrative, and possibly the child’s position. If a child is removed and a parent contests it, an emergency hearing commences in which the question of imminent risk is reviewed in more detail.70 But when faced with a report from a medical setting, judges routinely remove a child based on a clinician’s “reasonable suspicion” alone. The investigation by the caseworker likely reiterates the clinician’s report; the parent’s defense alone is unlikely to nullify any medical concern. There are no medical records available yet and certainly no live testimony from the doctor to provide context to the report. In these cases, the removal effectively transforms the provider’s “reasonable suspicion” into a finding of “imminent risk.”71 This section illustrates mutual deference in practice, using three case examples from Bronx Family Court. In each case, the clinician’s decision to report, concededly, falls squarely within the purview of “reasonable suspicion” contemplated by the statute. But in each case, the system interpreted the clinician’s report in the most severe light possible, presuming the worst of the parents, all of whom are parents of color. The need, or perceived need, for countering medical information obscured the legal system’s ability, or justified its unwillingness, to issue orders that would keep the family intact. Instead, these families were separated based on the initial report alone. 70 See Nicholson v. Scoppetta, 3 N.Y.3d 357, 378 (2004) (“The court must do more than identify the existence of a risk of serious harm. Rather, a court must weigh, in the factual setting before it, whether the imminent risk to the child can be mitigated by reasonable efforts to avoid removal. It must balance that risk against the harm removal might bring, and it must determine factually which course is in the child’s best interests.” (emphasis added)). 71 Even in cases where the parent’s attorney does have additional medical information at the arraignment, the judge must hear that evidence in the context of a formal hearing. That hearing will be scheduled on a different day and may take days, weeks, or months to complete, depending upon the congestion of the particular courthouse. Where medical concerns are at stake, these hearings are more prolonged than other hearings because of the delay of obtaining expert medical opinions. 2021] MUTUAL DEFERENCE 757 A. A Clinician May Refer a Family to the Family Regulation System for Supportive Intentions; the Court Interprets the Report as Imminent Risk Pending a Hearing Anthony,72 a seven-year-old child with sickle-cell anemia, missed seven appointments for his condition in two months. The hospital made a report of suspected neglect, stating that Anthony needed to be monitored closely because, if he developed a fever, he would need immediate medical attention. The report stated that Anthony had a fever the previous week, and his mother, Ms. Oros, did not bring him to the doctor. Based on this report, ACS removed Anthony from his mother on an emergency basis and filed a neglect petition against her in Family Court. The petition alleged medical neglect for the missed appointments and stated that Ms. Oros “admitted” to missing the appointments and that she found them “overwhelming.” The petition alleged that, although Ms. Oros knew she should bring her son to the hospital immediately if he developed a fever, she did not do so for a week. At the arraignment, ACS asked the judge to continue the removal. The judge deliberated: “This is a difficult case because . . . it really depends on how . . . this child is doing medically.” She further commented, “[h]ere’s the problem I have. I don’t have enough medical information.”73 The judge continued the removal of Anthony from his mother, and Ms. Oros requested a hearing. In the meantime, Anthony lived with his maternal grandmother while the court awaited further medical information. This arrangement remained in effect for almost two months. Two months after the filing of the petition and over twenty court appearances later, Anthony’s 72 The parent gave permission for these facts to be shared, but names have been changed to protect privacy. 73 Transcript of Arraignment, Matter of A. (Bronx Cnty. Fam. Ct. July 6, 2017) (docket number withheld to protect client confidentiality). 758 COLUM. J. RACE & L. [Vol. 11:733 treating physician testified in Family Court about her concerns when making the report.74 The testimony revealed that, although the hospital contacted ACS out of a concern that Anthony had a high fever, his doctor did not intend for the child to be taken from his mother’s care. Instead, the doctor testified that one motivation for calling ACS was the hope they could help with “case management”: assistance to the mother with the appointments and having in-home services put in place to help with medication management. The same day that the judge heard from the doctor, she found that there was no imminent risk to Anthony in his mother’s care. He was sent home with an order that the child be taken immediately to the hospital if his temperature exceeded 101 degrees. Notably, these were the same orders the attorneys for Anthony‘s mother had requested at the arraignment. *** Anthony’s two-month separation from his mother, which was contrary to any medical advice, is a direct effect of mutual deference. That the clinician’s concern resulted in a call of suspected child maltreatment in the first place is a response to the incentives on and guidance to mandated reporters to defer quickly to the state. That the court system received incomplete information from the treating physician reflects the failure of the family regulation system to adequately investigate medical concerns. The court’s paralysis when faced with a medical report reflects the knee-jerk deference of the legal system to the clinical opinion. Finally, that Ms. Oros is a single Black mother cannot be overlooked when understanding the legal system’s prioritization of prosecution over supportive interventions in this case.75 74 Given the challenge of coordinating schedules between the court and the physicians, and the delay of obtaining medical records for all parties to review before a physician testifies, this amount of delay is typical. 75 One month into this hearing, ACS revealed that it had yet to make a single referral to a supportive service to Ms. Oros, and did so only upon order of the court. That referral was for in-home medical preventive services, which had a significant waitlist. 2021] MUTUAL DEFERENCE 759 Embedded in mutual deference is a willingness to disrupt the parent-child relationship pending investigation and the presumption that the state could address Anthony’s condition better than his mother, a presumption that applies to parents of color at a far greater rate than white families.76 B. An Injury Is Severe, but the Doctor’s True Concern is Non-Medical Rysheen Summers77 brought his four-month-old daughter to the hospital with severe burns on her legs. Five days earlier, Mr. Summers had left the bathroom briefly while his daughter was in the bathtub with the water running and drain unplugged. The water temperature in his homeless shelter spiked and she was badly burned. Scared to go to the hospital for fear of losing their children, he and his girlfriend treated the burns themselves. When they did seek medical treatment, the hospital notified the police and ACS. The parents were arrested and charged with felonies; their baby and older child were removed from their care. The parents were charged with abuse— intentionally burning their baby—in Family Court. With no other information than the report from a hospital of the burn and ACS’s allegations of abuse, the judge continued the removals and placed the children in kinship foster care. While the family remained separated, the attorneys for ACS maintained that they would call an expert witness at trial to prove abuse.78 76 See Cooper, supra note 1 at 258. 77 Mr. Summers gave permission for these facts to be shared, and requested that his name be included. 78 Discovery was not expedited because Mr. Summers did not exercise his right to an emergency hearing to request the return of his children, as is his right pursuant to N.Y. FAM. CT. ACT § 1028. This is because he faced felony charges in Criminal Court; any hearing in Family Court would necessarily require him to testify; defendants are often advised not to testify in open court right after serious criminal charges are filed. As a result, ACS was not immediately required to produce the medical records in discovery. Mr. Summers, through his counsel, obtained them, but for bureaucratic reasons this took 760 COLUM. J. RACE & L. [Vol. 11:733 But the medical records later revealed that the Child Abuse Pediatrician (CAP) who examined the baby at the hospital and who made the initial report believed the burns were accidental, as the parents had described. Instead, the basis of the call was her concern about the parent’s judgment in leaving the baby unattended in the bathtub and declining to seek medical care earlier. Based on these medical records, ACS ultimately agreed to settle the case with a neglect finding based on parental judgment, withdrawing the allegation that the injury was intentionally inflicted. Once the court posture reflected the reality of the medical concern, the parents were able to move forward in their case, expanding their time with their children towards reunification. *** Here, the family regulation system assumed that Mr. Summers, a Black man in his twenties, had intentionally burned his baby. The system rushed to remove the children and file abuse allegations without speaking to the source of the report about the true basis of her concern. The court took severe measures under the assumption that a doctor would testify to abuse. Ultimately, the basis for the report was parental judgment—leaving a baby unattended in a bathtub, particularly with unpredictable temperatures, and delaying medical care. These are not medical issues. At worst, they reflect lack of foresight about potential dangers that come with leaving a baby unattended, even if the drain was unplugged, which it was. More accurately, they reflect inequity in housing safety and acute familiarity with the power of the state to remove children which discourages many from seeking prompt medical attention. While arguably these issues did not have to be prosecuted at all, once in front of a judge, orders could have addressed these issues. The conditions allowing for the children to be returned to their parents could have been issued from the start, including taking several months. This is another harm of over-inflation of charges in these cases: parents must navigate multiple court cases with conflicting incentives. 2021] MUTUAL DEFERENCE 761 a first aid class, an order to seek timely and regular medical attention for future concerns, and a parenting class in the form of parent-child therapy. Here, the clinician’s report was not only elevated to imminent risk, but also inflated to abuse, and this resulted in prolonged separation.79 C. Doctors Can Make Mistakes, and Critical Legal Decisions Are Based on Those Mistakes; Litigation to Resolve Them Results in Prolonged Separation of Families Ms. Tolbert80 received a call from her partner that their four-month-old daughter, Beatrice, had rolled off their bed when he stepped away to take a work call. Beatrice had a bad bruise on her eye, and the parents rushed her to the hospital. At the hospital, Ms. Tolbert was asked to agree to a CT scan and then a full skeletal survey of her baby; she agreed, assuming it was for medical purposes. Subsequently, she learned Beatrice was being held for “investigation.” When Ms. Tolbert learned from the CAP that Beatrice had a healing skull fracture and two healing rib fractures, she realized the hospital did not believe her daughter had fallen. They thought she had been abused—repeatedly. Once Beatrice was ready for discharge, ACS filed an abuse petition in Family Court requesting that she be removed from her parents’ care and put in foster care.81 79 This case illustrates the ongoing harm of hospitals’ close relationship with the police and the family regulation system. First, families that are familiar with the family regulation system avoid seeking medical care out of fear that they will lose their children. Second, the disproportionate legal response stifles adjudication. Faced with felony charges in criminal court, it was not advised for Mr. Summers to request an immediate hearing in Family Court, in which he would have to testify, before discovery had taken place in either forum. This resulted in unfortunate delay in obtaining the medical records that ultimately brought the true concern to light. 80 Ms. Tolbert gave permission for these facts to be shared, but names have been changed to protect privacy. 81 The application of ACS was for Beatrice to be in stranger foster care. But because Ms. Tolbert’s mother was able to move to the Bronx from out of state, Beatrice was able to stay with her grandmother with her mother in the home. However, pending litigation, Ms. Tolbert was not allowed to be alone with 762 COLUM. J. RACE & L. [Vol. 11:733 Ms. Tolbert’s lawyers provided the records and radiology to a neurosurgeon and a radiologist from different hospitals, who confirmed that the pediatrician’s diagnosis was incorrect and in fact, the skull fracture did not exist. The radiologist also confirmed that the location and nature and appearance of the two healing rib fractures suggested that they had been caused accidentally, and that it was likely that they had been asymptomatic. Based on this alternative medical opinion, Ms. Tolbert and her partner asked for an emergency hearing for the return of their baby. They were ultimately successful in that hearing, but more than two months passed from the date of filing to the time they were reunified with their child.82 *** Ms. Tolbert’s case illustrates three aspects of mutual deference. First, the system is unequipped to investigate a physician’s reasonable suspicion. Here, ACS relied on the opinion of one CAP83 who suspected that this child had been abused based on erroneous interpretations of radiology. Even though the court ultimately deemed the CAP’s opinion a “rush to judgment,” 84 her suspicion functioned as the basis of “imminent risk” for two and a half months while the case was litigated.85 Beatrice, and Beatrice’s father was not allowed to be in the home at all, except for during scheduled visits. 82 Because this hearing happened during the early stages of the COVID-19 pandemic and all doctors were able to testify virtually, two months was likely less time than the litigation would have taken in person. 83 Child Abuse Pediatrics is a sub-specialty of pediatrics that emerged in 2009; in New York City, CAPs lead Child Advocacy Centers, which are institutes within hospitals that collaborate with law enforcement in child abuse investigations. Child Advocacy Centers were established by law in 2006 under New York Social Services Law sections 423 and 423-a. 84 Matter of B.D. (Bronx Cnty. Fam. Ct. Aug. 7, 2020) (Passidomo, J.) (docket number withheld to protect client confidentiality). 85 Although not the precise focus of this Article, this highlights the role of Child Abuse Pediatricians and the harm that flows from the deference that they receive. The ethical concerns flowing from CAPs are myriad. For a comprehensive review of the ethical concerns relating to the role of Child Abuse Pediatricians, see GEORGE J. BARRY & DIANE L. REDLEAF, MEDICAL ETHICS CONCERNS IN PHYSICAL CHILD ABUSE INVESTIGATIONS: A CRITICAL 2021] MUTUAL DEFERENCE 763 Second, the distrust with which Ms. Tolbert, a Black mother, was treated at the hospital led to tangible medical harm.86 Because the hospital considered the story of her baby rolling off the bed unlikely, Beatrice was subjected to unnecessary radiation exposure and two nights in the hospital at the height of the COVID-19 pandemic. These tests revealed findings that reinforced the hospital’s suspicion and exacerbated the legal intervention, even though the findings turned out to be benign. Third, Ms. Tolbert’s only tool to challenge one doctor’s suspicion was litigation. Even if a parent is assigned an attorney with the resources to retain experts, the cost of litigation is delay.87 The over-clogged court system and doctors’ demanding schedules result in significant scheduling challenges. Delay in these cases means children remain separated from their parents, resulting in lasting harm.88 *** Unjustified elevation of a clinician’s concern to evidence of imminent risk is the harm of mutual deference. At best, mutual deference fails to ensure that the players in the system— clinicians, caseworkers, lawyers, judges—have the information necessary to perform their job and make informed decisions at each stage of the case. At worst, mutual deference provides PERSPECTIVE (2014). What is most relevant and representative in Ms. Tolbert’s case is that CAPs are afforded deference even when rendering opinions outside their training that can only be made reliably by radiologists, neurologists, orthopedists, and other medical specialists because “[t]he idea that the child abuse pediatrician’s has greater expertise than other subspecialists has been more broadly accepted than is justified, especially if the child abuse pediatrician fails to fully consult with subspecialists in forming her abuse conclusions.” Id. at 4. See also Rachel Blustain, Doctors Say They Shook Their Baby. They Didn’t, DAILY BEAST (Apr. 13, 2017, 2:37 PM), https://www.thedailybeast.com/doctors- said-they-shook-their-baby-to-death-they-didnt. 86 In public testimony delivered by Ms. Tolbert in an out-of-court setting referring to her experience at the hospital with Beatrice, she commented, “I have never felt more Black[].” 87 See, e.g., Horan-Block & Newman, supra note 41, at 410 (2019) (showing that even where early and aggressive litigation of suspected physical abuse results in reunification of parents and children, the delay is considerable). In 2019, litigation of cases of serious abuse shortened the length of separation from an average of 595 days without a hearing to 226 days with a hearing. Id. 88 Vivek S. Sankaran & Christopher Church, Easy Come, Easy Go: The Plight of Children who Spend Less than Thirty Days in Foster Care, 19 U. PA. J.L. & SOC. CHANGE 207, 210–13 (2016). 764 COLUM. J. RACE & L. [Vol. 11:733 insulation to the players in the system who decline to examine the biases that inform their role in a traumatic intervention: why an investigation was triggered, why a removal was conducted, and why it was legally sanctioned by the courts.89 V. CONCLUDING REMARKS: RESISTING MUTUAL DEFERENCE The relationship between the family regulation system and medical providers is historical; in fact, when mandated reporting was first established in law, many states labeled physicians as the only mandated reporters.90 Since its inception, therefore, the family regulation system has depended on medical professionals to provide a pool of families to investigate and surveil. But if society values the therapeutic relationship, why would it delegate surveillance efforts to clinicians when that surveillance disrupts the patient-doctor relationship so fundamentally?91 A report of suspected child maltreatment carries immediate side effects and grave risks for the family and greater community. There is the medical harm—possible radiation, testing, and stress; the loss of trust from the family, the loss of confidentiality, and the loss of the patient’s continuity of care. There is long-term harm, too—the lasting trauma of removal for 89 See, e.g., Annette R. Appell & Bruce A. Boyer, Parental Rights vs. Best Interests of the Child: A False Dichotomy in the Context of Adoption, 2 DUKE J. GENDER L. & POL’Y 63, 66 (1995) (“Judges must be careful to distinguish cultural or value based differences in child-rearing practices from parental conduct that falls beneath minimally acceptable parenting standards and raises a legitimate concern about the health, safety, or welfare of the child.”). 90 Monrad G. Paulsen, Child Abuse Reporting Laws: The Shape of the Legislation, 67 COLUM. L. REV. 1, 3 (1967) (citing the reasons the Children’s Bureau placed the primary duty to report on physicians for three reasons: physicians were in a unique position of having access to information about abuse when a caretaker would seek medical attention for a child; the special skill and training of the physician to detect instances of child abuse; reluctance of physicians to report for fear of “meddling” or violating “professional confidence”). 91 The American College of Obstetrics and Gynecology comments: “Legally mandated testing and reporting puts the therapeutic relationship between the obstetrician-gynecologist and the patient at risk, potentially placing the physician in an adversarial relationship with the patient.” Substance Abuse Reporting and Pregnancy: The Role of the Obstetrician-Gynecologist, Committee Opinion No. 473, COMM. ON HEALTH CARE FOR UNDERSERVED WOMEN (January 2011) https://www.acog.org/clinical/clinical-guidance/committee-opinion/ articles/2011/01/substance-abuse-reporting-and-pregnancy-the-role-of-the- obstetrician-gynecologist [https://perma.cc/SZ3A-G24N]. 2021] MUTUAL DEFERENCE 765 the child no matter how short the separation; as well as housing, medical, and educational disruption. More broadly, the fear of legal intervention at hospitals undermines the public policy interest in encouraging prompt medical attention for children. The systemic harm of mutual deference is the reinforcement of links between medical and legal terms that divert Black and Brown families into the family regulation system at disproportionately high rates. With mandated reporting, these considerations are irrelevant to the instruction to report a suspicion. A clinician, tasked with the obligation to “do no harm,” is forbidden from considering the potential harms of initiating this course of action. In this way, the clinician’s obligation to the state supersedes its obligation to the patient. This tension can only be resolved by presuming that making the report does protect the patient, but there is no mechanism to ensure that the balance of harms weighs in favor of the patient. Mutual deference shows it often does not. It is only through removing clinicians’ reporting obligations to the state that clinicians can be empowered to reconceptualize a report to the family regulation system as an invasive treatment—one with risks and harmful side effects that often dissuade clinicians from choosing a particular course of treatment. Once seen as a dangerous intervention, mandated reporting can receive the critical examination that other diagnoses—and diagnostic errors—receive. Removing liability around reporting creates space for interrogation of the class- and race-based associations that medical and legal institutions have made between neglect, abuse, and the challenges endemic to low- income communities of color. More broadly, it can trigger—or even require—diversion of resources to therapeutic rather than prosecutorial methods of addressing the root causes of perceived and real challenges.92 This strengthens community programs, rather than state agencies, towards the dissolution of the family regulation system entirely. Efforts to re-envision how support for and protection of families can move away from state-sanctioned violence and 92 See Cooper, supra note 1,at 251 (“[C]hanging the players or elements has the least effect on the system, but changing dynamics between elements and especially the ultimate purpose of the system has the greatest effect.”). 766 COLUM. J. RACE & L. [Vol. 11:733 towards strengthening families within their communities93 must critically examine the role of hospitals in establishing the current system. Hospitals hold tremendous potential to support families by redirecting their resources and expertise back into the community and away from state surveillance. This begins with removing mandated reporting responsibilities. 93 Dorothy Roberts & Lisa Sangol, Black Families Matter: How the Child Welfare System Punishes Poor Families of Color, APPEAL (Mar. 26, 2018), https://theappeal.org/black-families-matter-how-the-child-welfare-system- punishes-poor-families-of-color-33ad20e2882e/ [https://perma.cc/WLR4-EG29]; Erin Miles Cloud, Toward the Abolition of the Foster System, S&F ONLINE, http://sfonline.barnard.edu/unraveling-criminalizing-webs-building-police-free- futures/toward-the-abolition-of-the-foster-system/ [https://perma.cc/NGY6- VY7J].