Clinical and Experimental Gastroenterology (Oct 2018)
Agreement of triage decisions between gastroenterologists and nurses in a hospital endoscopy unit
Abstract
Stephen Inns,1,2 Jeffrey Wong,1 Dena McPhedran,1 Gladys De Guzman,1 Katherine Broome,1 Dalice Sim,2 Rosemarie Sandford1,† 1Hutt Valley DHB Endoscopy Unit, Hutt Valley DHB, Lower Hutt, Wellington, New Zealand; 2Department of Medicine, Otago University, Wellington School of Medicine, Wellington, New Zealand †Mrs Rosemarie Sandford passed away on December 1, 2015 Introduction: Efficient and accurate triage of endoscopy referrals is essential. Many of the decisions made are based on national and local triage criteria. Standardizing this approach for nurse use could maintain quality, address clinical risk and significantly improve resource utilization.Aims: This study aimed to compare gastroenterologist and nurse triage of unselected gastroenterology referrals in order to evaluate the proportion of referrals felt able to be triaged to endoscopy and the inter-rater agreement between a triage gastroenterologist and endoscopy nurses for clinical triage decisions regarding the urgency of gastroscopy and colonoscopy.Methods: The proportion of referrals triaged to endoscopy by a consultant gastroenterologist performing triage as a part of normal practice and two endoscopy nurses using a decision algorithm was measured. The inter-rater agreement for the triage category decision (urgency of referral) between the three triage clinicians was assessed. An adjudication panel provided a consensus decision triage category decision in cases where there was not complete agreement between the three triage clinicians.Results: Each clinician assessed 105 referrals. Nurse A was able to triage 54 (51%) referrals to a triage category and Nurse B 44 (42%) referrals. Cohen’s κ was run to determine if there was agreement between clinicians for the triage categories allocated. The agreement between the two nurses was substantial (k=0.645, P<0.0005). Between the gastroenterologist and each nurse, moderate agreement was seen (Nurse A, k=0.589, P<0.0005; Nurse B, k=0.437, P<0.0005). Moderate agreement was seen between the nurses and an adjudication panel (Nurse A, k=0.423, P<0.0005; Nurse B, k=0.464, P<0.0005). However, there was only slight agreement between the adjudication panel and the gastroenterologist (k=0.099, P=0.010).Conclusion: Nurse triage using a decision algorithm is feasible, and inter-rater agreement is substantial between nurses and moderate to substantial between the nurses and a gastroenterologist. An adjudication panel demonstrated moderate agreement with the nurses but only slight agreement with the triage gastroenterologist. This suggests that nurse triage using a decision algorithm can approximate decision making by an experienced gastroenterologist. Keywords: comparative study, endoscopy, nurses, referral and consultation, triage
