An observational study examining the inter-rater reliability of parent compared to clinician clinical decision rule scoring in children with minor head injury
Introduction:
Paediatric traumatic brain injury is estimated to affect more than 3 million children globally every year, with between 80-90% of cases being classified as mild. The majority of children assessed in Emergency Departments (EDs) are discharged without investigation or treatment and this presents an opportunity for attendance avoidance strategies. Our overall aim therefore, is to derive and validate a decision aid that could be used by parents in the community to safely advise them whether to seek emergency assessment.
Systematic review:
We conducted a systematic review of the existing literature using variations of the free-text terms 'parent', 'head injury' and 'clinical decision rule' along with appropriate subject headings for each term. MEDLINE and EMBASE were searched but neither identified any studies with existing clinical decision rules (CDRs) created or adapted for parental use. The search did identify several validated CDRs for clinician use and of these, the most adaptable for use by non-medically trained individuals was the Paediatric Emergency Care Applied Research Network (PECARN) rule.
Aim of this study:
We will assess the feasibility of parental use of a decision aid by evaluating the inter-rater reliability between parents and clinicians when assessing children over 2 years of age using the PECARN criteria.
Plan of investigation:
(a) Design and setting
This prospective, observational study will recruit children over 2 years of age attending the Royal Manchester Children’s Hospital with a minor head injury. Recruitment is expected to commence from April 2018, with results available for dissemination from August 2018. Parents will be asked to provide written informed consent for the study and children over the age of 10 will be encouraged to assent. The study will be approved by the Research Ethics Committee.
(b) Sampling
Children will be recruited 7 days a week, at any time of day over a 2 month period. Only children who are alert on arrival and that do not require immediate resuscitation will be included. Children who are unable to verbally communicate or are not accompanied by a person with parental responsibility will be excluded. A sample of 45 participants will provide sufficient statistical power to identify a kappa of 0.8 with lower bound of the 95% confidence interval no lower than 0.6.
(c) Data collection
Parents will be given a questionnaire to complete whilst awaiting assessment. The questionnaire contains the PECARN criteria in plain English language, the wording of which has been refined through consultation with patients. The child will be assessed and managed as normal. Afterwards, the clinician will also complete a questionnaire containing the PECARN criteria, blinded to the parental assessment.
(d) Analysis
We will calculate the inter-observer reliability for each individual variable and for the overall rule outcome by calculation of kappa scores and 95% confidence intervals. In accordance with existing literature, a kappa score of >0.6 will be deemed to represent good inter-observer reliability and will be considered sufficient to demonstrate the feasibility of further evaluation through larger studies.
