Health Care Providers' Experience Reporting Child Abuse in the Primary Care Setting FREE

FEW REPORTS of suspected child abuse come from office-based physicians. In 1990 in Illinois, 103,000 children were reported as abused or neglected to child protective services (CPS). Although medical personnel reported 18% of children, only 4% of the total reports came from office-based physicians.¹ Only 1% of all reports to the Ohio Central Registry for Child Abuse in 1981 were from private physicians' offices.² Three explanations for the paucity of reports from office-based physicians are (1) low rate of presentation of abused children to primary care settings, (2) nonrecognition of abused children who present to primary care settings, and (3) nonreporting of recognized cases.

Little information is available about how frequently office-based physicians identify child abuse injuries in their patients or about their response to suspected child maltreatment. A 1983 survey of physicians in Virginia reported that they saw few cases of abuse or neglect in their offices; 26% had encountered no abuse or neglect in the previous year.³ In 1990, an American Academy of Pediatrics survey⁴ found that 45% of primary care pediatricians had treated or consulted on patients who were victims of child abuse during the previous 8 weeks. These pediatricians reported treating an average of 2.8 patients who were victims of physical abuse during this period. A 1998 American Academy of Pediatrics survey⁵ found that 57% of 608 pediatricians reported that they had treated injuries from child abuse in the previous year.

No information is available, to our knowledge, on how often health care providers report cases of abuse that they recognize. We studied the experiences that primary care providers have with suspected child abuse and child abuse reporting.

We hypothesized that (1) providers suspect more cases than they report and (2) provider characteristics, including previous experience with and education about child abuse and previous experience with CPS, affect their reporting behavior.

This research is part of a larger study prospectively examining the health care provider's management of injuries in pediatric practices.

Seventeen practices, all members of the Pediatric Practice Research Group, were sent letters inviting them to participate in a study, which included a provider questionnaire and prospective data collection on office visits for child injury over a 4-week period in the summer and fall of 1998 (each office completed the study over a 4-week period, but this period occurred at different dates for different offices).⁶ Practices that responded affirmatively to a single invitation completed the study. Health care providers (physicians, nurse practitioners, and physician assistants) completed a mailed questionnaire as a first step in project participation. The completed questionnaires are the subject of this article.

Providers were assured that any information they gave was confidential; information collected under an Agency for Health Care Policy and Research grant is protected by a strong confidentiality statute, namely, 42 USC §299a-1(c) of the Public Health Services Act. This statute does not allow the disclosure of information collected other than for the purpose for which it was supplied. This assurance was important because state child abuse reporting laws provide for disciplinary action and legal consequences if a health care provider willfully fails to report child abuse. The hospital's institutional review board approved this study.

The provider questionnaire included 3 sections: demographic characteristics of the provider and practice, the provider's career experience with child abuse, and the provider's previous year's experience identifying and reporting child maltreatment.

All respondents were asked about their career experience with suspected child abuse, about their access to other professionals who could help them make decisions about abuse, and how many hours of continuing education they had received about child abuse in the past 5 years. Providers were also asked to describe any adverse consequences they may have experienced as a result of reporting suspected child abuse and how many hours they had spent preparing and providing testimony about children they had reported for suspected abuse.

Respondents who had encountered child abuse within the previous year were asked to list the total number of children they suspected had experienced either physical or sexual abuse and the number of those children reported to CPS. Using a 4-point Likert scale, the survey asked them to evaluate the quality of their most recent experience with CPS. Providers were asked about the professionalism of the CPS response, their perception of any benefit to the child and to the family, and whether the experience made them more willing to report suspected child abuse in the future. Providers not reporting all children suspected of maltreatment were asked their reasons for not reporting.

Survey data were entered and analyzed using SPSS statistical software (SPSS Inc, Chicago, Ill). Statistical significance was set at P<.05.

Because analyses requested data concerning patients seen and CPS experience while in a practice setting during the past year, analyses were limited to respondents who had completed their medical training at least 1 year before the beginning of the study. Also, providers were excluded who included in their responses children referred specifically for evaluation of suspected child abuse. Providers who said they (1) had not reported some patients to CPS because the child had already been reported or (2) had referred the child to the hospital CPS team were considered as having reported all suspected abuse when their reporting information was analyzed.

χ² Tests for categorical data and Mann-Whitney U tests for nonnormally distributed numeric data were used to compare the characteristics of providers who saw a patient suspected of maltreatment vs providers who identified no abuse. These tests were also used to examine the characteristics of providers reporting all vs those not reporting all child abuse to CPS. For both analyses, the variables that achieved significance were entered into a forward stepwise logistic regression model.

Seventeen (39%) of 43 practices, including 89 providers, responded affirmatively to the initial request for participation. Because the total number of health care providers in these practices exceeded our initial goal of 50, we did not recruit further. All 89 providers at the 17 participating practices completed the survey forms. Two physicians were excluded from survey analysis because they had completed their residency training in the past year, and 2 were excluded because they included in their sample children referred specifically because of suspected child abuse. The responses of 85 providers (95%) were included in the analysis.

Fifty-one (60%) of respondents were women; 74 (87%) were white, 5 (6%) were African American, 4 (5%) were Asian, and 3 (4%) were Hispanic. Seventy-six respondents (89%) were physicians, 8 (9%) were nurse practitioners, and 1 (1%) was a physician assistant. Eighty-three providers (98%) listed pediatrics as their specialty, and 2 nurse practitioners listed family practice. Of the physician providers, only 1 was not board certified. Thirty-eight (49%) of the providers (78 providers responded) had finished their training in the past 10 years (range, 1-47 years since end of residency). Two practitioners (2%) reported that a parent or a caregiver had physically abused them during childhood.

Eight (47%) of 17 practices were located in the city, including 5 (29%) that served an inner-city population; the remaining were suburban practices. Only one solo practitioner participated; the remaining provided pediatric care in group practice settings ranging from 2 to 13 providers.

Most respondents reported experiences with child abuse outside their primary care practices. Thirty providers (35%) said that in addition to their primary care patients, children had been referred to them for further evaluation of suspected abuse. Although 25 (29%) of providers reported that they had received no continuing education in child abuse in the past 5 years, the other providers reported a median of 2 hours (range, 1-250 hours) of educational credit.

Seventy-four providers (87%) identified a link with other professionals who could help them decide whether an injury was caused by physical abuse. The most commonly mentioned professionals included other primary care physicians (44%), multidisciplinary child abuse teams (40%), and mental health professionals (38%). When providers were asked if they had experienced any adverse consequences in their career as a result of reporting suspected abuse, 28 (33%) replied affirmatively. The most common consequences were losing patients (n=17) and spending time in court or other legal proceedings (n=13). One practitioner wrote, "I felt threatened by the mother that she would abduct or harm my child." Individual practitioners reported that other patients had heard about the report and left the practice, the family was disrupted without benefit to the child, and the provider had spent a lot of time on the telephone related to the suspected child abuse case. None of the participants had been sued for malpractice as a result of reporting suspected abuse, but one was threatened with a lawsuit and one was reported to the state licensing board. Twelve providers (14%) said that in the previous year they had spent 1 to 22 hours (median, 5 hours) preparing to testify and providing testimony in court about children they had reported.

Forty-eight providers (56%) reported treating children they suspected had been abused in the previous 12 months. These providers said that they saw a total of 152 children who were possibly abused. Provider experience in treating child abuse is shown below.

Providers who identified 1 or more children with abuse were compared with providers who said they had seen none. These groups did not significantly differ on sex, race, ethnicity, years since finished training, positive vs negative perception of previous experience with CPS, any vs no adverse consequences experienced as a result of reporting, additional experience evaluating abused children, and their ability to identify other professionals who could provide support or consultation if abuse was suspected.

Providers who had seen child abuse in the previous 12 months were characterized according to their reporting practices: (1) reporting all children they suspected were abused; (2) not reporting all children because some were previously reported by someone else and some they referred to child abuse teams for further evaluation; or (3) not reporting for other reasons. Thirty-seven providers reported all children (n=114) they suspected had been abused. Four providers did not report 31 children, but listed as a reason that the child had already been reported or that the child was referred to a child abuse team. Because the providers were not asked to specify the reason each individual child was not reported, some of these children may have been neither reported nor referred.

Seven providers did not report suspected child abuse for other reasons. Each did not report 1 child. Therefore, about 5% of children with suspected abuse injuries were not reported. The most common reason given for not reporting these children was that the provider was not certain of the diagnosis (n=4). Other reasons were that they helped the families themselves (n=2), they experienced previous dissatisfaction with CPS (n=2), they did not want to hurt the relation with the family (n=1), a breakdown occurred between CPS and the prosecutor (n=1), social service on site was better (n=1), and they referred the child to a social worker for counseling (n=1).

The 41 providers who reported all suspected abuse differed from the 7 providers who did not in their educational experience. Those who reported all abuse were 10 times more likely to have received some formal education about child abuse in the past 5 years (odds ratio, 10.3; 95% confidence interval, 1.7-63.1; P=.01 by Fisher exact test). The groups did not significantly differ on sex, race, ethnicity, years since finished training, positive vs negative perception of previous experience with CPS, any vs no adverse consequences experienced in the past as a result of reporting, additional experience evaluating abused children, and their ability to identify other professionals who could provide support or consultation if abuse was suspected.

The 48 providers who had reported suspected abuse in the past year then described their most recent experience reporting to CPS (Table 1). Most providers said CPS workers responded promptly and professionally. Twenty-three providers reported that the families had benefited, but only 17 indicated that the children had benefited from CPS intervention. However, the difference was not statistically significant. Only 17 providers indicated that the agency kept them informed of the progress and disposition of its investigation. Only 22 providers indicated that their past experiences with CPS made them more willing to report child abuse to CPS in the future.

This survey of pediatric providers gives new insight on child abuse reporting in primary care settings. The major findings were that providers reported 95% of children with suspected abuse, recent education about child abuse improved reporting, and negative experiences with CPS made providers more hesitant to report. Results of this study confirm those of some previous research^3- 5,7 on reporting and provide insight as to why providers may not report all suspected abuse.

Forty-eight (56%) of 85 providers surveyed had seen suspected child abuse in the past year. These results are almost identical to those of the most recent American Academy of Pediatrics survey⁵ on this topic, which found that 57% of 608 pediatricians reported treating injuries from child abuse during the previous year. In another survey,³ 74% of practicing physicians reported that they had encountered abuse or neglect in the previous year. These results differ because physicians were asked to include physical neglect, emotional abuse, and medical neglect in their responses, whereas in our survey only the provider's experience with physical and sexual abuse was included.

Primary care providers surveyed reported almost all suspected child abuse to CPS. Practitioners in our sample indicated a 95% one-year reporting proportion, which is markedly higher than the 75% lifetime reporting proportion of pediatricians and 60% lifetime reporting proportion of physician assistants reported previously.⁸

Seven providers (8%) in our survey said they had not reported a child who they suspected was maltreated; these results are comparable to those of the study by Saulsbury and Campbell.³ Uncertainty as to whether a child was abused was the most common reason children were not reported. Providers listed this reason for not reporting in 2 previous physician surveys.^3,7

Although state statutes require that health care providers report all suspected abuse, some providers might at times be reluctant to report unless they possess the same degree of diagnostic certainty they have when treating other medical problems. We speculate that providers wish to spare children and families from unnecessary distress and need to make a more definitive diagnosis before reporting.

Although only one third of the participants said that they had been informed by CPS of the progress and disposition of the reported case, nearly two thirds believed that CPS intervention had not benefited the child. This study did not address the basis on which this conclusion was reached, but clearly providers need CPS feedback.

In deciding whether to report a case of suspected child abuse, our respondents indicated that they consider (1) the probability of the abuse, (2) the likely effect of CPS intervention if the child was not abused, (3) the likely effect of CPS intervention if the child was abused, and (4) the consequences for the provider. Although the first factor is most important, our findings suggest that the other 3 factors contribute to not reporting abuse based on previous career experience.

This study suggests 3 changes that would better enhance primary care provider reporting of suspected abuse. Past efforts at improving child abuse reporting have focused on improving providers' identification of abuse. Results of our study show that education makes a difference in reporting. It is not clear whether education increases identification or increases the provider's comfort level with the reporting process. Efforts must be made to ensure that all primary care providers receive continuing medical education about child abuse. Several states have mandated that all professionals who are legally required to report suspected child abuse must receive education about child abuse as a prerequisite to licensing.^9- 10

Second, the results of this study demonstrate that improving the outcome for children who have suspected abuse injuries is important for medical care provider reporting compliance. Health care providers must be convinced that their intervention will help the child and the family. Just more than one third of our sample believed that CPS intervention helped the child. McDonald and Reece¹¹ postulated that one barrier to physician reporting is the practitioner's belief that reporting child abuse accomplishes little for the child. After conducting an extensive literature review on variables that affect the identification and reporting of abuse, Warner and Hansen¹² speculated that improving positive outcomes for families and physicians would improve reporting.

Third, communication between CPS and providers should be improved, as our providers showed concern about inadequate feedback from CPS. Child protective services must keep medical providers informed about the progress and resolution of their investigation. We also conjecture that including some field experience with CPS during pediatric training would demystify the CPS process, improve communication, improve understanding, and, hence, improve compliance.

These survey results must be interpreted cautiously because providers were asked to supply retrospective information. If child abuse injuries present infrequently, an encounter with abuse may stand out in a provider's memory, making it difficult to place in timed recall. Also, because a provider may spend more time in following up these children, this extensive time commitment may be remembered as more children seen with abuse than actually occurred. On the other hand, some providers may have been reluctant to disclose that they had not reported abuse despite the assurance that the information was confidential.

Because 49% of the sample (38 providers) completed their training in the past 10 years, these recently trained providers may be more comfortable reporting suspected abuse and working with CPS than are their more seasoned counterparts. However, there was no statistical significance between the reporting habits of the 2 groups.

This study does not address how accurate providers are in identifying abuse. Providers who reported that they saw no abuse during the previous year may have failed to identify abuse in their patients.

In conclusion, the results of this survey of primary care providers confirm a high rate of reporting abuse (95% of suspected cases) and provide further insight into the reasons for not reporting, which center on provider diagnostic uncertainty and past experience with child abuse reporting. Steps to address providers' concerns should be made through education of health care providers and improvement of CPS communication and response. Health care providers may need to improve their recognition and diagnostic skills and their desire to partner with CPS to ultimately result in better intervention, protection, and care for abused children.

Accepted for publication November 4, 1999.

This work was supported by grant R03 HS09811 from the Agency for Health Care Policy and Research, Rockville, Md.

Portions of these data were presented at the San Diego Conference on Child Maltreatment, San Diego, Calif, January 26, 1999, and at the Pediatric Academic Society Meeting, San Francisco, Calif, May 1, 1999.

We thank Joseph Flaherty, MD, Robert Reece, MD, and Emily Wood for their support and the following members of the Pediatric Practice Research Group who participated in this study: Associated Pediatricians, Valparaiso, Ind; Child Life Center, Homewood, Ill; Winfield-Moody Health Center, Chicago, Ill; Child and Adolescent Health Associates, Chicago; Traisman, Benuck, Traisman, & Merens, Evanston, Ill; Lake Forest Pediatrics, Lake Forest, Ill; DuPage Pediatrics, Darien, Ill; Glenbrook Pediatrics, Glenview, Ill; Pediatric Specialists, Barrington, Ill; North Arlington Pediatrics, Arlington Heights, Ill; Northwestern University Medical Faculty Foundation, Chicago; Infant Welfare Society, Chicago; Arnulfo Delgado, MD, Chicago; Children's Healthcare, Chicago; Children's Primary Care Services, Chicago; and Pediatric Associates of Barrington SC, Barrington.

Corresponding author: Emalee Gottbrath Flaherty, MD, Children's Memorial Hospital, 2300 Children's Plaza, Box 16, Chicago, IL 60614.