Reporting child abuse in Rwanda: attitudes and practices of physicians-a cross-sectional survey

¹Department of Pediatrics, University of Rwanda, Kigali, Rwanda, ²Department of Pediatrics, University Teaching Hospital of Kigali, Kigali, Rwanda, ³Department of Pediatrics, Kacyiru Police Hospital, Kigali, Rwanda, ⁴Department of Pediatrics, Oakland University William Beaumont School of Medicine, Oakland, USA, ⁵Department of Emergency Medicine, Yale University, New Haven, Connecticut, USA, ⁶Rwanda Human Resources for Health (HRH) Program, Ministry of Health, Kigali, Rwanda

^&Corresponding author
Peter Cartledge, Department of Emergency Medicine, Yale University, New Haven, Connecticut, USA

Introduction    

Child abuse is a global problem with life-long severe consequences [1,2]. The immediate injuries and effects can be devastating; children's long-term physical, mental and emotional health is also impacted [3]. The landmark Adverse Childhood Experiences study discovered a strong relationship between experiences with abuse during childhood and many of the principal causes of death in adulthood [4]. Policymakers are increasingly recognizing the importance of awareness, education, comprehensive local systems, and robust governance to combat childhood violence in its many forms. The United Nations Sustainable Development Goal (SDG 16.2) specifically aims to “end abuse, exploitation, trafficking and all forms of violence and torture against children” [5]. Despite this SDG, many countries lack even baseline data on the prevalence of child abuse which is compounded by the factor that the number of children on the African continent who are abused is underestimated [6]. Physicians and other healthcare professionals (HCPs) often have insufficient education and support necessary to recognize abuse and adequately report it in order to protect a child. Child abuse is under-researched in Africa with most of the medical literature on child abuse coming from North America and Europe. The majority of African child abuse literature is focused on sexual abuse [7]. In Rwanda, much of the child abuse literature focuses on street children and transactional sex, as well as intimate partner violence [8]. There is no clear data on the overall prevalence of child abuse in Rwanda. Despite the lack of data, Rwanda has done much to support child rights. In 1991, Rwanda was one of the first countries to ratify the United Nations Convention on the Rights of the Child (CRC). In 2001, Rwanda approved The African Charter on the Rights and Welfare of the Child [9] and in 2011 specific legislation was implemented “relating to the rights and the protection of the child” [10]. In 2007, Rwanda developed the Child Right's Observatory, which was charged with monitoring and protecting the rights of children [11].

Many governments, including, the USA, Canada, Australia and a number of European countries have enacted mandatory reporting laws to facilitate early discovery of cases of child abuse and neglect [12]. In Rwanda, professionals, HCPs and all adults are legally required to report suspected child abuse. Reporting mandates are crucial for the protection of children. Victims typically do not self-report [13]. If victims do report abuse, it is usually not promptly after having suffered abuse, and therefore professionals play a vital role in identifying and reporting abuse and potentially reducing the negative consequences for the child [14]. Reporting abuse is a key step in safeguarding children. Physicians who work with children must be given evidence-based training, in not only the concepts and recognition of child abuse but also in how to report abuse and protect children [6,15]. Despite laws to mandate the reporting of abuse, it is known that many professionals, all over the world, fail to do so [16,17]. Many factors influence a HCPs willingness to report abuse. Barriers can be divided into four categories; i. individual barriers (such as the knowledge of healthcare professionals, their attitudes and beliefs), ii. Interpersonal barriers (fear of disconnecting therapeutic relationships and violation of privacy), iii. Organizational barriers (such as poor communication and weak legal processes for reporting), and iv. Situational barriers (such as victims´ characteristics and available evidence) [16,18-20]. If enacting mandatory reporting into law, then legal protections must also be provided to HCPs who report possible abuse in good faith [12]. Historically, there has been disagreement, amongst HCPs, about what constitutes abuse [21]. This has important implications when applying practice to legal precedents, which are in place in Rwanda [10]. For example, while there is growing recognition that corporal punishment has negative consequences, many children around the world experience violent disciplinary practices at home [22]. These practices, in some countries, are viewed as an acceptable means of correcting a child's misbehavior but may be illegal in other states [21,22].

Study objectives: the objective of this study was to identify the attitudes and practices of paediatricians and paediatric residents regarding the reporting of child abuse in paediatricians and paediatric residents in Rwanda.

Methods     

Study design: a cross-sectional survey. Reporting of this study has been verified in accordance with the STROBE checklist (Strengthening the Reporting of Observational Studies in Epidemiology) [23,24].

Study setting and location: Rwanda is a landlocked country located in East Africa with a population of 12.2 million inhabitants. The gross domestic product (GDP) per capita more than tripled from $211 in 2001 to $787 in 2018 [25].

Physicians/subjects: all paediatricians, subspecialists, and paediatric residents in Paediatrics who signed the consent form were eligible for inclusion. There were no exclusion criteria. Participants were identified through the Rwanda Paediatric Association (RPA) and University of Rwanda (UR) for paediatricians and paediatric residents respectively. For recruitment, email addresses were gained from the RPA and UR. An email invitation was sent to all eligible physicians.

Sample size (power calculation): this was a descriptive study and therefore no sample size calculation was undertaken.

Questionnaire employed: the “Child Abuse and Neglect Questionnaire” was used [14,26]. The questionnaire was developed and validated in Australia to assess reporting practices in HCPs. It includes three sections: i. demographics (e.g, age, gender, years in practice, etc), ii. yes/no questions assessing recognition and reporting of all forms of suspected or definite cases of child abuse and the number of suspected or definite cases seen in the previous 12 months, iii. Likert questions (1-5) separated into sections on confidence in practice, decision making, beliefs and attitudes regarding child abuse. The Likert questions were all presented between 1 (very unlikely) and 5 (very likely), the exact wording of each Lickert options are provided with the result tables. In analyzing the Likerts items were either positive or reverse coded. A reverse coded question is one where the “correct answer” is presented negatively and for example scores 5 for “strongly disagree”

Data collection and management: the questionnaire was sent with the invitation email with a web-link (URL) to a Google Forms® version of the questionnaire. Two reminder invitations were sent to maximize response-rate. Questionnaire responses were stored in Microsoft Excel® and statistical analysis undertaken in SPSS® (statistical package for the social sciences) v24.0.

Statistical analysis: summary statistics were analyzed with means and proportions [27,28]. As the training and years of experience of residents is significantly different, the results for these groups are presented separately, with totals for the entire cohort.

Academic integrity: the study was undertaken as the MMed (MSc) thesis by the Principal Investigator (HBM). The study protocol was therefore reviewed and approved by the University of Rwanda academic team in February 2016.

Institutional review board (IRB): the research protocol was reviewed and approved by the University of Rwanda IRB (Ref: 052/CMHS IRB/2017).

Results     

Participants: a total of 102 practicing paediatricians (n=53) and paediatric residents (n=49) were eligible for enrolment. A total of 61 responses were received from these 102 invitations (response rate 60%) with residents being much more likely to respond than paediatric specialists (Table 1). The majority of physicians were male (69%) and working in the capital city, Kigali (79%).

Training and experience in reporting child abuse: only five physicians (8%) reported having received no training in child abuse at any point (Table 2). Physicians reported receiving lectures (69%) and training “in practice during residency” (62%). Only 31% of physicians reported that they felt that their education had adequately addressed child abuse reporting.

Reporting practices: twenty-nine physicians (48%) had never made a report or notified the authorities of child abuse (Table 2). Twenty-three physicians (38%) had at some point suspected child abuse and had decided to report the suspicion. Eighteen physicians (29.5%) were unaware that they had a legal responsibility to notify the authorities of child abuse (Table 3). Physicians were as likely to report to their Head of Department as they were to the police.

Confidence in identifying and reporting abuse: subjects felt most confident at identifying physical abuse (mean=4.02) (Table 4). Regarding confidence in reporting, physicians were less confident (mean=3.34) in reporting child abuse compared to other competencies, such as identifying abuse.

Response to suspicion of child abuse: several of the Likert questions in this section were “reverse coded” (italic text in Table 5). Physicians correctly responded that to “take no action” (mean=4.57) was not an appropriate response to suspicions of abuse. To “persuade the child to give details of the abuse” was the most likely action reported by physicians (mean=4.26).

Fear of reporting: all the fear of reporting questions were reverse coded Likert questions. Therefore, the biggest fears of reporting were that “Unpleasant events would follow” (mean= 3.10) (Table 6 and not wanting to get involved in legal proceedings (mean=3.10).

Beliefs and attitudes about abuse: “parents have the right to treat their children as they see fit” was a reversely coded item and was, therefore, the belief that physicians most recognized as being false (Table 7). Residents were statistically less likely (p=0.023) than paediatricians to recognize that “most parents are capable of hurting their children when they are under great pressure.”

Beliefs and attitudes affecting reporting: the factor most likely to influence a decision to report child abuse was “Being convinced there is clear evidence of abuse” (Table 7). Residents were more likely to be influenced by a feeling that the abuse was a single incident and unlikely to happen again (p=0.026).

Discussion     

The objective of this is study was to identify the attitudes and practices of paediatricians and paediatric residents regarding the reporting of child abuse in Rwanda.

Under-recognition of abuse: our results suggest an under-recognition of child abuse. Only ten (16%) physicians reported a case of child abuse in the previous 12 months. In Rwanda, residents provide the majority of face-to-face care in government tertiary facilities. Therefore, they are in an ideal position to report, and are legally obliged to do so. Nine physicians (15%) acknowledged seeing a suspected case of child abuse and not reporting it in the previous year. This would imply that a minimum of 42 (67.7%) physicians had not seen any child with suspected child abuse in the previous year. This would appear highly unlikely considering the high large Rwandan paediatric population and the known global prevalence of child abuse [7] and therefore suggests that the recognition of child abuse is a problem or that physicians were reluctant to report that they had intentionally not reported in the previous year. Reasons not to report are multifactorial and include lack of evidence that an injury seen was caused by abuse and the belief that they could personally manage it with the family more effectively in a different way [29].

Confidence in recognizing possible abuse: we found that the level of confidence in recognizing indicators of physical abuse, sexual abuse, and emotional abuse was variable. Many physicians reported a high level of confidence in appropriately identifying indicators of physical abuse (mean=4.02) (Table 4). There was a trend in the total confidence score towards residents feeling less confident then specialists. This is understandable as the physicians themselves acknowledged that a great deal of child abuse knowledge and training is gained from clinical practice and residents have fewer years of clinical experience.

Reporting practices: regarding reporting practices, 30% of physicians were unaware of their legal responsibility to report child abuse (Table 3) [13]. Despite having this legal responsibility to report concerns of child abuse in Rwanda it was interesting to note that 15% of physicians did not even believe that they had a professional or ethical responsibility to notify the authorities of a suspicion of child abuse. Even though the majority of physicians were aware of their legal responsibility to report suspected cases of child abuse, only 53% had actually ever reported a case of suspected child abuse (Table 2). This would suggest that there is a need to develop a culture of reporting and accountability to increase the likelihood that cases of suspected abuse are reported and that children are protected. Physicians were as likely to report to their Head of Department (79%) as they were to the police (78%) which reflects the hierarchical culture in Rwanda to inform a superior before informing anyone else (Table 3). Professional organisations such as the Ministry of Health (MoH), Rwandan Medical Dental Council (RMDC) and Rwanda Paediatric Association (RPA) have a potential to work collaboratively to raise awareness of the legal and professional responsibilities to report suspected child abuse. This should be facilitated with processes to support and protect doctors when they do report.

Response to suspected abuse: regarding physicians´ response to suspected child abuse; persuading the child to give details of the abuse was the highest scoring response (mean=4.26) (Table 5). In the reverse coded question regarding confidentiality, more residents (mean=2.15) than paediatricians (mean=3.27) reported that they would make promises to a child not to tell anyone if that was what the child desired.

Fear of reporting: the highest scoring fear of reporting for physicians was that they would “not want to get involved in legal proceedings” (Table 6). This is consistent with previous literature which has found that physicians are reluctant to report child abuse to avoid being involved in police or court proceedings which can be time-consuming [21]. This could be compounded in the resource-limited setting where the burden of disease is higher and where there are significant time pressures on a smaller workforce.

Beliefs and attitudes affecting reporting: we found that physicians who were convinced there is clear evidence of abuse would be most influenced to report. This has previously been reported [17]. This attitude scored higher then concerns for the child´s welfare or the personal satisfaction of making a report. Our results suggest that there remain many perceived myths about child abuse and its pervasiveness. In our study, residents were more likely to be influenced to not report if they held the belief that the incident of concern was an isolated one (p=0.026). This finding underscores the importance of education to dispel the widely held but false belief that abusive injuries are anomalous single occurrences. Child abuse is very commonly a chronic condition. Well recognized professional organizations, such as the American Academy of Pediatrics, publish practice protocols for health care providers to routinely screen for injuries when abuse is suspected because it is so frequent for additional traumas to be discovered.

Training in safeguarding: child abuse reporting by medical professionals requires both knowledge and experience [30]. Regarding training in safeguarding, only 8% of responding physicians reporting having received no training in child abuse reporting. However, only 31% reported that the training that they had received on reporting child abuse was adequate. This is an important factor as those physicians who consider their previous training and competencies to be inadequate may be less likely to report child abuse, placing children at risk. It has been shown that undertaking training of professionals working with children will significantly increase the number of cases that they report [31,32]

Study strengths: this is the first study of its kind to explore attitudes and practices of physicians towards reporting child abuse in Rwanda. The median age of the Rwandan population in 2012 was 18.1 years-of-age [33]. This figure is likely reflect in other resource-limited settings. Therefore, this is a key topic as large proportions of such populations are potentially vulnerable to child abuse. This study used a previously validated questionnaire, albeit for a resource-rich setting. The confidential nature of the questionnaire allowed physicians to reply free from any harm.

Validity of results: the questionnaire employed used only Likert style questions and therefore the responses were restricted to the scoring system. Likert style questions are prone to acquiescence bias. Future research could consider qualitative methods to explore more rich perspectives of this topic. The questionnaire was developed and validated in Australia, and therefore may not be fully appropriate to our setting, and no attempt was made to validate it in Rwanda. Regarding the external validity of the results, this study did not assess the attitudes and perceptions of nursing staff or general practitioners (medical officers) who also are in regular contact with vulnerable children. Medical officers are newly qualified doctors working in district hospitals and provide front-line care throughout the country. As paediatricians and residents exclusively treat children, these results are biased to the knowledge and experience they have. Our opinion is that it is likely that “non-paediatric” Rwandan physicians are more likely to under-recognize and under-report child abuse. These non-paediatric physicians make up the majority of the Rwandan clinical workforce and therefore training, and exposure needs to be upscaled across the board. It has also been found that nursing staff report suspected abuse less then doctors [34].

Further limitations: this study has a number of limitations; the use of an electronic form (Google Forms) may have biased the physicians to younger and more computer literate physicians. In Rwanda, we have few paediatricians for the size and proportion of the paediatric population. The data was therefore limited to the number of eligible physicians, and though the overall response rate was satisfactory, it was exacerbated by a low response rate amongst the qualified paediatricians. The lower response rate in qualified paediatricians, compared to residents, could be explained by the electronic format of our survey, extra responsibilities of paediatricians and finally, that the principal investigator was a resident and colleagues might have responded to be supportive.

Conclusion     

In conclusion, child abuse is a global epidemic. In Rwanda, child abuse is understood to be a problem, but not always recognized or reported in the individual child. We have identified many factors that are associated with under-recognition and under-reporting of child abuse. All healthcare professionals providing care to children in Rwanda will benefit from regular, mandatory training on child abuse and reporting of child abuse. Systems and cultures of accountability to ensure reporting should also be put in place.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Dr. Hippolyte Bwiza Muhire undertook the research as the memoire of his MMed in Paediatrics. Prof Dr. Lisine Tuyisenge, Dr Francois Budigu, and Dr Allison Ball were supervisors and therefore contributed to the conception and analysis of the research. Dr. Peter Cartledge analysed the data and undertook the interpretation with Dr. Bwiza. All authors approved the final manuscript.

Tables     

Table 1: baseline details of participants

Table 2: training and reporting in safeguarding

Table 3: reporting practices

Table 4: confidence - “In your opinion, how prepared do you feel to be able to carry out the following?”

Table 5: response to suspected child abuse -“In a situation where you suspect child abuse, indicate how you would respond in such a situation”

Table 6: fear of reporting

Table 7: beliefs and attitudes and child abuse and reporting suspected cases

References