Nothing should be less controversial than protecting children from abuse. As the prevalence of the "battered child syndrome" came to popular attention in the 1960s, every state in the country required certain classes of professionals to report their suspicions of abuse to a child protection agency (1). In 1974 Congress passed the Child Abuse Prevention and Treatment Act, which provided financial assistance to states that met federal standards, and it established the National Center for Child Abuse and Neglect to further these efforts. But mandatory reporting of suspected child abuse has turned out not to be a simple matter, especially not for mental health professionals.

The best estimates suggest that the scope of the child abuse problem is substantial. About 1.6 million children a year are thought to be exposed to physical, sexual, or emotional abuse or to substantial neglect (2), with more than 1,000 deaths each year attributed to these causes (3). The incidence of serious abuse increased an estimated 67 percent between 1986 and 1993, the latest year for which comparable data are available (2). Sequelae of child abuse can include lifelong psychiatric and emotional problems, such as depression and posttraumatic stress disorder. Persons who were abused as children are at higher risk of later perpetrating abuse themselves or committing other acts of violence (4).

Because the alarm over battered children was first raised by physicians who saw the effects of abuse in emergency rooms and on pediatric units (5), the model adopted for combating the problem echoed the casefinding approach used for infectious diseases, gunshot wounds, and other public health problems. Statutes initially focused on requiring physicians and other medical personnel to report cases to a state agency that would investigate the complaints and initiate appropriate action.

The scope of the statutes was soon widened, however, to include a broad variety of people who might, in the course of their work, become aware of child abuse (1). Massachusetts is typical in including, among other categories, physicians, psychologists, dentists, nurses, public and private school teachers, educational administrators, guidance counselors, family counselors, probation officers, social workers, police officers, and even firefighters (6). Some states require anyone becoming aware of abuse to report the situation (7).

Although reporting statutes are superficially similar from state to state, they actually vary significantly (7). (An on-line compilation of all state child abuse reporting statutes can be found at the Web site of the National Clearinghouse on Child Abuse and Neglect Information at www.calib. com/nccanch.) The statutes generally require a report to be made when a mandated reporter has a "reasonable cause" (or some similar term) to suspect abuse, and they often impose time limits, such as 48 hours, within which contact with the agency must take place.

But some states' language limits the obligation to professionals who have actually seen the abused child—those "who have reason to believe … that a child coming before them is a victim of child abuse"—whereas others impose the duty regardless of how the professional learns of the suspected abuse (8). In addition, laws may focus on the effects of abuse, requiring reporting if the child "is suffering serious physical or emotional injury resulting from abuse" (6). Others may require disclosure whenever a suspicion is raised that abuse has occurred (7,9).

Based on these requirements, approximately 2 million reports of child abuse are filed in the United States each year (3). Most states impose sanctions, ranging from civil fines to criminal penalties, on mandated reporters who fail to act in accord with the statutes (7). Although prosecutions for failure to report are rare, they do occur, and criminal convictions have been handed down (10). One might therefore expect that most mental health professionals would be punctilious about reporting their suspicions of abuse in cases that come to their attention. In fact, this is far from what occurs.

By now a substantial body of studies agree that it is not uncommon for clinicians to ignore reporting requirements. In one review of 12 studies, an average of 40 percent of psychologists acknowledged that they had failed to report at least one case of suspected child abuse, with the figure ranging as high as 66 percent in one survey (11). There is no reason to believe that the numbers are different for the other mental health professions. Although ignorance of reporting statutes is surprisingly high in some studies, rates of nonreporting remain elevated even when clinicians are aware of the laws in their states. Overall, child protective agencies in 1993 are estimated to have investigated only 28 percent of cases of children harmed by abuse or neglect, with a substantial part of the remainder accounted for by cases that are never reported to the agencies (3).

Why are clinicians reluctant to report cases in which they suspect that abuse might have occurred? Studies have shown that abuse that is recent or ongoing, that is of greater severity, that involves younger victims, and about which the mental health professional feels more certain has a greater likelihood of being reported (11). On the other hand, cases are less likely to be reported if clinicians believe that an intervention will do more harm than good (12). Thus mental health professionals insist on exercising some degree of discretion in decisions to report, shying away in particular from cases in which some doubt exists about the occurrence of abuse.

These consistent findings cause no end of consternation among many experts on child abuse. Reporting statutes are designed specifically to require reporting whenever some reasonable suspicion of abuse is present. Clinicians, this argument goes, are poorly situated to perform investigations to verify whether abuse has in fact occurred—precisely the task of the child protective agencies. Legislators have made a judgment that a low threshold should apply to decisions to report, recognizing that some degree of overreporting is inevitable in these circumstances, and accepting that risk. What right do mental health professionals have to decide otherwise?

Although one cannot gainsay the validity of these contentions, more than three decades of resistance by mental health professionals to these arguments suggests that they are giving equal weight to other considerations. Many clinicians who have reported suspicions of child abuse have had disappointing experiences with the child protective agencies (1). Investigations may be highly intrusive, and decisions made abruptly or on the basis of limited evidence. Children removed from households do not always end up in placements that seem preferable, especially when it is not clear that abuse has occurred, or when abuse has taken place in the past but is not ongoing (13).

Having learned that child abuse reporting does not always lead to benign outcomes, it is hardly surprising that therapists are reluctant to file reports unless they have a high degree of certainty that a child is at risk and will benefit from the intervention. Even the threat of criminal penalties fails to overcome this adherence to what clinicians believe is their patients' best interests.

Without doubt, another major consideration for clinicians is the likelihood that a child abuse report will be disruptive to treatment. This concern is particularly relevant when the patient is the abuser, rather than the victim. When a person who has abused a child seeks treatment, the therapist will usually be required under the mandatory reporting law immediately to report any abuse that is disclosed. Many clinicians have a reasonable fear, supported by case reports and empirical data, that patients who have not been deterred from seeking treatment already, perhaps because they are ignorant of the law's requirements, will be forced from treatment if therapists report their behavior (9,14,15). The problem is compounded by the common requirement that reports be made within 48 hours, which limits the time available to discuss these issues with patients.

Understandably, clinicians feel torn between their obligation to do what is best for their patients and their duty to protect actual and potential victims of child abuse. A variety of suggestions have been made for ameliorating the situation (1,13,14). The least radical changes would involve narrowing the definitions of reportable circumstances to ensure that only the occurrence or threat of serious harm is covered. Less serious, perhaps arguable, instances of neglect or abuse, especially emotional abuse, might clearly be excluded under such a proposal (1).

Somewhat more dramatically, some commentators have suggested exempting all therapists who are treating perpetrators of abuse from mandatory reporting, while permitting them to report voluntarily if that seemed essential for the protection of a child (14). This approach would permit therapy to continue in circumstances that the clinician judged to be low risk, for example, if the abuser no longer has contact with children. Victims of abuse who are in treatment might be given some limited right to block therapists' reporting, out of privacy concerns or for other reasons, as long as no one was immediately in danger. Other persons who become aware of abuse would still be covered by the reporting statutes.

Perhaps the most innovative suggestion advocates adoption of the Dutch system, whereby confidential investigations take place when a report is made, without the immediate threat of coercive interventions (13). The primary goal of case workers is to provide treatment and support for families and children. Only if these interventions fail, or if parents are uncooperative, would consideration be given to removing the child from the home.

The data on reporting by mental health professionals, characterized as it is by frequent failure to comply with existing laws, suggest that little might be lost and a good deal gained by considering alternative approaches. Were, for example, therapists working with patients who had been abusive in the past given discretion about reporting, treatment would likely be facilitated. A positive outcome in such cases would be better for abuser and victim alike, compared with the intrusive and often ineffective techniques employed today. Moreover, the terribly overburdened child protective agencies, their caseloads reduced by elimination of some proportion of cases where the threat to children is less imminent, might be better able to concentrate on more serious cases of abuse.

When laws are ignored systematically, it is usually because they do not make sense to those who are subject to them. There is good evidence that this is the case with child abuse reporting statutes. The possibility that there is wisdom in the intuitive responses of so many clinicians ought to be considered.

Dr. Appelbaum, editor of this column, is A. F. Zeleznik professor and chair of the department of psychiatry of the University of Massachusetts Medical School. Address correspondence to him at the Department of Psychiatry, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, Massachusetts 01655.