Paediatric trauma systems and their impact on the health outcomes of severely injured children: protocol for a mixed methods cohort study

NSW has the highest population of all states and territories in Australia, at June 2015, there were 1.42 million children under 15 years of age, encompassing an area of around 800,000 km² [18]. The transfer and treatment of the injured child in NSW is governed by the NSW Ministry of Health’s state wide trauma services plan [19]. There are three designated PTCs in NSW [20].

Pre- and inter-hospital trauma care is delivered and governed by the NSW Ambulance via land and aeromedical rotary and fixed retrieval. Transport decisions are governed by the pre-hospital trauma (T1) Protocol which provides a state-wide triage process for the prehospital identification of severely injured patients for the transport to definitive treatment at a major trauma centre within a 60 min time frame. Ambulances are to bypass local hospitals when they have a severely injured patient, in preference for a major trauma centre [13]. The protocol does not mandate transport of severely injured children to a PTC.

This mixed methods study will be conducted in four phases, using quantitative and qualitative data. The study process is outlined in Fig. 1.

There are three groups of participants: child, parent and staff.

All injured children (aged <16 years) requiring intensive care, or with an Injury Severity Score (ISS) ≥ 9 treated in NSW [22], or, who died following injury in NSW in the 2015–16 financial year, will be eligible for participation. In NSW only patients aged < 16 are accepted for treatment in a Paediatric Trauma Centre. Differences of care required among age groups will be reflected in the results, with specific ages grouped and compared. There is no one data source from which to identify severely injured paediatric trauma patients in NSW. Hence, injured participants will be identified through one of four mechanisms outlined in Table 1. A small proportion of children severely injured in rural NSW will likely be transported across borders for treatment due to their proximity to major cities in other states; we will identify these cases using the NSW Medical Retrieval Registry (Air Maestro, Avinet, Australia). Based on 2013 data from all these sources, we anticipate we will review 400 cases. A waiver of consent has been granted to retrospectively review medical records as: (i) obtaining consent from all relevant health consumers for their health information to be included in the study was considered impracticable, and (ii) an incomplete data set would substantially impair the research by introducing bias, and reducing the validity and generalizability of the research to the paediatric trauma population. There will be no contact with the injured children.

Staff involved in the patient phase of care where a potential problem was identified by the expert peer-review process will be invited to participate in a telephone semi-structured interview. The purpose of the interview will be to determine the range of human/clinical and systems based factors that may have contributed to the problem [23]. A general list of potential staff participants (that is, those rostered to the particular work area during the time period of interest) will be provided confidentially by the relevant hospital manager. Written informed consent will be obtained from clinical staff who agree to participate. Of the cases that may require peer-review, there may be up to 6 staff eligible to participate per case.

Six and 12 month HRQoL of children suffering major injury will be obtained by proxy by interviewing consenting parents. Parents of children with severe injury treated at one of the three PTCs will be eligible for participation. Discussion with the clinical team will include considerations such as the child’s condition and the timing for approaching parents, minimising the risk of any potential coercion. When an appropriate time is identified, the study site trauma coordinator(s) who are assisting with the study will inform parents/caregivers of the study and what their role as participants in the study would entail, including time involved to participate in follow-up telephone questionnaire interviews. Parents will be provided with a participant information package including the participant information sheet, consent form, and the baseline questionnaires (see Data Collection). It will be made clear that participation in the study is voluntary, and the decision not to participate can be made at any time, without reason, and this decision will not affect their child’s care. We anticipate there will be ~ 400 parent groups eligible to participate. A participation rate of approximately 53 % [24] to 86 % [25] is expected (n = 200-340).

There are five sources of data collection.

Eligible and consenting staff will participate in a telephone semi-structured interview to determine a range of systems factors such as patient factors, task factors, individual (staff) factors, team factors, work environment factors, organizational and management factors and institutional context factors. Discussions will be audio recorded and include the interviewee’s role in the case; it will be made clear that the purpose of the discussion is not to allocate blame, but is to understand what occurred and to identify any factors that may have contributed to the outcome [28].

Clinical outcomes of children suffering major injury will be obtained by proxy in a structured interview with the child’s parents. Health-related quality of life (HRQoL) data will be collected from recruited via proxy from the parent participants by the nurse surveyors using the Pediatric Quality of Life inventory (PedsQL 4.0) and the EuroQol five-dimensional EQ-5D-3 L™ face-to-face for the pre-injury (baseline) and in-hospital, and by telephone at 6 and 12 months post-injury. Each tool measures different aspects of functioning and well-being post-injury, is validated for use in paediatric injury outcomes research and can be reliably administered via parent proxy, and paper-pencil or telephone [30, 31]. A response rate of between 53 % [24] to 86 % [25] of parents approached is expected.

Data extracts from the NSW Trauma Registry, the NSW Medical Retrieval Registry, coronial information, and patient health care costs will be linked using unique variables, such as medical record number, first and last name, age, gender, date of admission and date of death. This study will use manual record linkage using the available identifiers. Where there is only a partial identifier available, other data variables will also be used in the linkage process. Higher weighting will be given to identifiers, such as first name, last name, date of birth, gender, MRN and hospital during the linkage process.

Each individual in the study will be given a ‘unique study number’ in each data extract, where relevant, and once linked, data will be de-identified. The identifying information (ie, first and last name, MRN and date of birth), along with the unique study number will be kept securely in a password protected file. Further information will only be added to the database should the case require peer-review and/or have the additional 6 and 12 month health outcome data added.

Based on the results of the data analyses and integration, a series of recommendations to improve pre-hospital transfer, in-hospital service delivery and between-hospital transfers for severely injured children will be developed. To obtain consensus on the suitability and the importance (i.e. prioritisation) of these recommendations, a modified-Delphi study will be conducted with the project’s Translation to Health Care Policy and Practice Group. This group is composed of representatives from each partner organisation, government, NSW ambulance, and consumer groups. The modified-Delphi will consist of two rounds of questionnaires; each questionnaire will be pilot tested on individuals not involved in the research for content ambiguities. In Round 1, the group will be invited to rate the suitability and importance of each recommendation (on a 5-point Likert scale), to suggest modifications to recommendations (where relevant), and to list the key factors that led them to rate the recommendations as they did. Determining when a panel of experts has reached consensus is not straightforward, so different levels of consensus will be specified (ie. high, moderate, low) [37]. The panel will be considered to have reached high consensus on a recommendation when the proportion of all of the panel’s ratings reaches ≥70 %, moderate consensus when the proportion is 50–69 %, and low consensus if the proportion is <50 %. In Round 2, the modified recommendations will be submitted for rating of suitability and importance [37]. If necessary, a third modified-Delphi round will be conducted.

This study protocol was approved by the NSW Population & Health Services Research Ethics Committee; reference number HREC/15/CIPHS/6 and the National Coronial Information Systems via the Department of Justice and Regulation Human Research Ethics Committee; reference number CF/15/18354. Site Specific Approvals were gained at each hospital site.