METHODS

A structured telephone questionnaire was carried out on all major emergency departments in England that receive children. Major departments were defined as those receiving more than 18 000 children a year and a list was provided by the Royal College of Paediatrics and Child Health.

A single interviewer (WK) interviewed the nurse in charge of the department during daytime working hours on weekdays. If the nurse in charge was unavailable, arrangements were made to telephone on a subsequent occasion. Questions were based on the standards listed in box 1. The questionnaire was piloted on staff from four emergency departments in our local region and the questions clarified accordingly.

RESULTS

The RCPCH provided a list of 37 major emergency departments that receive paediatric emergency cases. All responded to the telephone survey.

Assessment and management of the child suspected to be at risk

Altogether 36 of 37 (97%) departments had set guidelines for child protection. It was the policy that children suspected to be at risk must be seen by a senior emergency department doctor in 14 of 37 (38%), by a consultant paediatrician in 2 of 37 (5%), a paediatric registrar in 19 of 37 (51%), and a paediatric senior house officer in 2 of 37 (5%) cases. Varying criteria were routinely used to select patients to be seen by a senior emergency department doctor or paediatrician, and included deliberate self harm, fractures in children less than 2 years old, domestic violence, and repeat attendances.

Structure and staffing

The trust’s named doctor with expertise in child protection was known to the respondent in 27 of 37 (73%) and the named nurse was known in 30 of 37 (81%); their contact details were known in 29 (78%) and 27 (73%) cases respectively. Specialist paediatric advice was available to the emergency department at all times in all cases; the most senior specialist available at all times was a paediatric emergency medicine consultant in 7 of 37 (19%) cases, consultant paediatrician in 15 of 37 (41%), paediatric registrar in 11 of 37 (30%), and paediatric senior house officer in 4 of 37 (11%). Eleven of 37 (30%) of departments had a trained paediatric emergency medicine consultant. Nineteen of 37 (51%) had a registered children’s nurse on duty at all times.

Specific training “to be alert to the potential indicators of abuse in children” was provided for senior house officer and senior grade doctors in 28 of 37 (76%) (none in 1 of 37 (3%), and don’t know in 7 of 37 (19%)) and for nursing staff in 32 of 37 (86%) (none in 3 of 37 (8%), and don’t know in 1 of 37 (3%)).

The emergency department is on the same hospital site as the paediatric inpatient facilities in 33 of 37 (89%), being on a separate site in 4 of 37 (11%).

Administration and information

The child protection register (CPR) is checked routinely in all children in 8 of 37 (22%), and only in those suspected to be at risk in 29 (78%). The individual responsible for checking the CPR varied considerably between departments, and included the receptionist, triage nurse, other nurse, emergency department senior house officer, senior emergency department doctor, paediatric registrar, emergency department health visitor, child protection officer, and consultant’s secretary. In 12 cases (32%) the emergency department did not have direct access to the CPR.

The previous number of emergency department attendances for each patient was routinely recorded in 33 of 37 (89%) and patient records from previous attendances routinely made available in 20 of 37 (54%) of cases. The records from any siblings who have attended were never routinely obtained, but were made available in cases of concern in 24 of 37 (65%).

GPs were routinely notified of a child’s attendance in 31 of 37 (84%) cases, and only in cases of concern in six (16%). Health visitors were routinely notified in 22 (59%) and only in cases of concern in 15 (41%) of cases. In 15 departments (41%) the health visitor routinely checked all paediatric attendances. The school nurse was routinely contacted in 3 of 37 (8%) and only in cases of concern in 23 (62%) of cases. In 9 of 37 (24%) the school nurse was never notified of paediatric attendances.

DISCUSSION

The survey revealed that almost all major paediatric receiving emergency departments have child protection guidelines and access to senior staff with paediatric training. Patients’ previous attendances are often highlighted. Most attendances are communicated to the general practitioner and health visitor, and even more when there are particular concerns about a child.

Not all of the Department of Health’s 1999 recommendations are observed. The designated named doctor and nurse with expertise in child protection are not known to all staff. There is considerable variability between departments regarding access to local CPRs and inconsistency regarding which member of staff is permitted to check them and for what indication. School nurses are infrequently notified of attendances despite the recommendations. There is room for improvement in providing adequately trained staff: 30% of departments have a consultant in paediatric emergency medicine at this time, and only 51% have a registered children’s nurse on duty at all times. A significant proportion of medical and nursing staff receive no formal training in identifying potential indicators of child abuse.

Limitations of study

A telephone survey was conducted as this is more effective than a postal questionnaire in achieving complete participation.^5,6 In addition it more reliably tests the interviewee’s knowledge of local procedures and personnel. There are however possible areas for the introduction of bias. Interviewer bias occurs when the views of the interviewer affect the way the questions are asked.⁷ This was minimised by the use of a structured, predominantly closed questionnaire with a scripted introduction and questions. All interviews were conducted by the same person, and the responders were allowed to complete their answers to open questions without interruption. The questionnaire was piloted on a sample from a relevant population (local emergency departments). Another potential source of bias is social desirability bias,⁷ where the responder chooses particular answers to a question to present a positive image.

We elected to interview the nurse in charge as they would be available in the department, and as a permanent member of staff would be most familiar with the procedures and staffing within that particular hospital. It is also the member of staff most likely to be making or suggesting decisions of administration, communication and referral to particular members of the medical team, and would probably be involved in overseeing the triage of children at risk. Asking a member of the nursing staff about the training of medical staff is however a potential weakness.

The relevance of the results may be limited. We surveyed large departments only, which may represent best practice and more optimal staffing.⁸ However, it is these departments to which the RCPCH guidelines in particular apply. It is possible that the RCPCH’s list of major departments is incomplete, and other sources, such as the BAEM directory, could have been consulted. The shortcomings in achieving recommended targets may nevertheless be pertinent, and the lessons applicable, to smaller emergency departments.

Child protection is an emotive area of health care where inadequacies in management may result in disastrous and upsetting consequences, but for which there is a limited evidence base to support diagnostic tests or therapeutic interventions. Clarification of protocols with staff training and regular audit can improve awareness of those children in whom there are features that might cause concern.⁹ It is therefore vital that formally agreed guidelines are adhered to, and that senior and appropriately trained medical and nursing staff are involved. The recommendations of the Department of Health and the RCPCH provide such guidelines, and departments should endeavour to achieve these standards where resources permit. This audit demonstrates the degree to which this has been achieved, and highlights areas where further improvements can be made.