Federation University Australia

Transcription

1 Trial of a new incident reporting system for the outdoor sector (UPLOADS): initial data and lessons learnt Natassia Goode 1, Paul M. Salmon 1, Michael G. Lenné 2, Caroline F. Finch 3 1 University of the Sunshine Coast Accident Research 2 Accident Research Centre, Monash University Introduction 3 Australian Centre for Research into Injury in Sport and its Prevention (ACRSIP) Federation University Australia The goal of the UPLOADS (Understanding and Preventing Led Outdoor Accidents Data System) project is to develop a standardised, national approach to incident reporting for the outdoor sector in Australia, and a corresponding national incident dataset. The project is a collaboration between the authors and outdoor education and recreation associations, outdoor activity providers and government departments. The system is primarily aimed at organisations which facilitate supervised or led outdoor activities (e.g. organisations operating under the banners of outdoor education, school camps, adventure tourism, outdoor recreation and outdoor therapy). As Fig. 1 shows, the UPLOADS project consists of a number of studies that contribute towards the development of a reliable and valid incident reporting, storage and analysis tools. This paper focuses on the 6 month trial of the prototype system. Organisations involved in the trial were asked to collect incident data using the UPLOADS software tool for six months. The software tool was stored on a computer within their organisation, and was not accessible by the research team. It included a function to automatically remove names from the data, so it can be sent to the research team for analysis. Organisations were asked to contribute incident data and participation days per activity to the project on a monthly basis. This paper summarises: 1) The level of participation of organisations involved in the trial over the six months period; 2) The data collected in relation to bushwalking incidents, in order to demonstrate how a national dataset might provide a picture of the risks associated with particular activities; and 3) The lessons that were learnt from the trial regarding the needs of the sector.

2 Paper accepted in the 18th National Outdoor Education Conference, April 2014, Adelaide. Fig. 1: Overview of the studies within the UPLOADS project, which contribute towards the development of the incident reporting system. The bold box highlights where the study reported in this paper sits within the project. Organisations involved in the trial Fifteen organisations participated in the trial. Five organisations operated in New South Wales, 4 in Queensland, 4 in Victoria, 1 in South Australia, 1 in Tasmania and 1 in Western Australia. Five organisations were commercial enterprises, 5 were not- for- profits, 2 were schools, 2 were registered training organisations and 1 was from the public sector. On average, organisations operated from 3 locations (SD = 2.07, range 1 to 7). 11 organisations were members of outdoor education/recreation industry bodies. Contribution of data For each month of the trial, participants either: contributed data; responded that no activities or incidents had occurred during the month; or did not respond to the request for data. Fig. 2 shows the number of organisations that contributed participation data in each month of the trial. Five organisations contributed participation data or responded that no activities had been conducted every month. Four organisations did not respond to any requests for data.

3 Fig. 2: Number of organisations contributing participation data in each month of the trial Fig. 3: Number of organisations contributing incident data in each month of the trial Incident data During trial, 184 incidents were reported, including 157 incidents associated with adverse outcomes and 25 near misses (2 missing classification). The reporter was present at the scene of 176 incidents (6 not present, 2 missing classification). Incidents by activity type Fig. 4 shows the number of incidents per activity type. The activity type most frequently associated with incidents was walking/running outdoors (71 incidents). The primary activity in this category was bushwalking, accounting for 70 incidents.

4 Fig. 4 Number of incidents by activity type Analysis of bushwalking incident data The following analysis is provided to demonstrate how data might provide a picture of the risks associated with particular activities. Incident severity Adverse outcomes were rated in terms of actual impact, while near misses were rated in terms of potential impact. On average, adverse outcomes were rated as 2.57 in actual severity (SD = 1.27). On average, near misses were rated as 2.67 in potential severity (SD = 1.53). According to the severity scale this is a short term impact on individual/s that doesn t have large effect on participation. People involved in the activity prior to the incident The reports ask for details regarding the people involved in the activity prior to the incident. 69 activities involved participants, 68 involved instructors and 24 involved other supervisory staff. On average, 18 participants were involved in the activity prior to the incident (SD = 18), 2 instructors (SD = 1) and 2 other supervisory staff (SD = 2). Adverse outcomes Of the incidents reported, there were 59 injuries, 8 illnesses, and 1 case of equipment damage. No reported incidents involved social or psychological impacts, missing or overdue people, or environmental damage. Injury details Fig. 5 shows the type of injuries sustained according to body location. In addition to those shown on the diagram, 4 injuries were to multiple body regions (classified as 2 superficial injuries and 2 other and unspecified effects of external causes), 1 was to unspecified parts of truck, limb or body regions (classified as 1 superficial injury) and 1 was missing classification of body location.

5 Of the 59 people injured, 50 were female and 7 male (2 unclassified). Average age was years (SD = 5.73, range 8 to 41, n = 23). 22 were classified as participants (37 unclassified). 7 people had some prior experience bushwalking, 49 were classified as unknown prior experience and 3 were unclassified. Fig.5 Type of injuries sustained by body location (n = 59) Illness details Seven different illness types were reported across 8 incidents. Of the 8 people reported ill, 5 were female and 3 were male. Average age was years (SD = 1.79, range 15-19, n = 5). In terms of their role, 6 were classified as participants (2 unclassified). In terms of prior experience bushwalking, 2 had some prior experience, 5 were classified as unknown prior experience and 1 was unclassified. Causal factors and relationships

6 Organisations were asked to code the causal factors and relationships evident in each report. 71 different causal factors were identified across 59 reports. Fig. 6 shows the causal factors, and their frequencies, identified across the reports in the context of Rasmussen s (1997) Risk Management Framework. 23 relationships were identified across 15 reports. Fig. 7 shows the relationships between the factors identified, again presented in the context of Rasmussen s (1997) Framework. Fig.6: Frequency of causal factors identified across the reports, presented in the context of Rasmussen s (1997) Risk Management Frameworks (n = 59) Fig.7: Relationships between causal factors identified across the reports, presented in the context of Rasmussen s (1997) Risk Management Framework (n = 15) Lessons learnt from the trial The key lessons learnt from the trial were that:

7 1) Some organisations do not run enough activities or have enough incidents to require a tool that analyses trends. Organisations should be able to select features that suit their needs. 2) The structure of activity programs differs considerably across organisations. The tool should focus on the collection of incident and participation data, rather than trying to provide a holistic system for tracking participants and activities. 3) Some led outdoor activity providers have a high staff turnover. The software tool should be intuitive and require minimal training to enter data. 4) Some led outdoor activity providers have minimal IT infrastructure. The software tool needs to run across multiple platforms, and take into account that organisations may have old software platforms. The need to install software updates is a significant barrier to participation. Based on these findings, the UPLOADS Software Tool has been redeveloped. In addition, UPLOADS Lite has been developed for organisations that do not require complex data analysis tools. An online survey tool allows organisations to contribute anonymous incident reports and participation data, and save the data they enter for their own records. Reference List Goode, N., Finch, C.F., Cassell, E., Lenne, M.G. & Salmon, P.M. In Press. What would you like? Identifying the required characteristics of an industry- wide incident reporting and learning system for the led outdoor activity sector. Australian Journal of Outdoor Education. Goode, N., Salmon, P.M., Lenne, M. & Finch, C.F. (2014) A test of a systems theory- based incident coding taxonomy for risk managers. AHFE 2014, Krakow, Poland July 19th 23rd. Rasmussen, J Risk management in a dynamic society: A modelling problem. Safety Science, 27(2/3): Salmon, P.M., Goode, N., Lenne, M.G. Cassell, E. & Finch, C Injury causation in the great outdoors: a systems analysis of led outdoor activity injury incidents. Safety Science, 63: Salmon, P., Williamson, A., Lenne, M.G. Mitsopoulos- Rubens, E. & Rudin- Brown, C.M. (2009). The role of Human Factors in led outdoor activity incidents: Literature review and exploratory analysis, Monash University Accident Research Centre. Salmon, P., Williamson, A., Lenne, M.G. Mitsopoulos- Rubens, E. & Rudin- Brown, C.M. (2010). Systems- based accident analysis in the led outdoor activity domain: Application and evaluation of a risk management framework. Ergonomics, 53(8): Salmon, P.M., Cornelissen, M. & Trotter, M. (2012). Systems- based accident analysis methods: A comparison of Accimap, HFACS, and STAMP. Safety Science, 50(4): igure Captions