A configural model of expert judgement as a preliminary epidemiological study of injury problems: An application to drowning

Abstract

22 23 24 25 26 27 28 2 29 A configural model of expert judgement as a preliminary epidemiological study of 30 injury problems: An application to drowning. 31 32 Abstract 33 Robust epidemiological studies identifying determinants of negative health outcomes require 34 significant research effort. Expert judgement is proposed as an efficient alternative or 35 preliminary research design for risk factor identification associated with unintentional injury.36 This proposition was tested in a multi-factorial balanced experimental design using specialist 37 judges (N=18), lifeguards and surfers, to assess the risk contribution to drowning for 38 swimming ability, surf bathing experience, and wave height. All factors provided unique 39 contributions to drowning risk (p<.001). An interaction (p=.02) indicated that occasional surf 40 bathers face a proportionally increased risk of drowning at increased wave heights relative to 41 experienced surf bathers. Although findings were limited by strict criteria, and no gold 42 standard comparison data were available, the study provides new evidence on causal risk 43 factors for a drowning scenario. Countermeasures based on these factors are proposed. 44 Further application of the method may assist in developing new interventions to reduce 45 unintentional injury. 3 47 Introduction 48 Analytic epidemiological studies test for the association of determinants with a negative 49 health outcome to support a theory of causality. Identified causal risk factors may then be 50 modified to improve health outcomes. Epidemiological research designs provide robust 51 evidence through observing candidate risk factors in the natural course of events. Means of 52 potential risk factor identification and specification include anecdotal evidence, case reports 53 or cross-sectional surveys.54 55 A significant challenge for observational epidemiological studies concerns the control of 56 confounding factors [1]. Elimination of competing explanations for study findings often 57 requires substantial study sizes. Accurate measurement of exposure to risk factors is complex 58 as is distinguishing factor causation from statistical association. Without attention to these 59 details, epidemiologic studies provide little substantial knowledge gain.60 61 Uncertainly about the derived health benefits leads to difficulty in justifying study costs and 62 may partially explain the lack of epidemiological studies for many significant injury 63 problems. To provide evidence on potential health gains, cost-effective preliminary studies 64 with acceptable internal validity and generalisability may guide later epidemiological 65 research. Such studies should replicate more rigorous epidemiological designs with respect to 66 risk factor identification and assessment. The study reported here tests and evaluates a 67 proposed method based on expert opinion for potential causal injury risk factor identification 68 and risk quantification as applied to unintentional drowning in an Australian surf bather 69 population, whe...