Criterion validity of a computer-assisted instrument of self-triage (ca-ISET) compared to the validity of regular triage in an ophthalmic emergency department

OBJECTIVES The computer-assisted version of a self-triage tool (ca-ISET) for an ophthalmic emergency department (ED) was developed to increase the validity of the triage procedure when trained ED staff is absent. METHODS We tested whether sensitivity, specificity, Negative Predictive Value (NPV) and Positive Predictive Value (PPV) of the ca-ISET deviated from regular triage. Patients ≥18 years visiting the ED of the Rotterdam Eye Hospital in the Netherlands were invited to participate in this prospective study. This ED focuses on eye-related problems. Patient recruitment was carried out during working hours. The ca-ISET is a touch operated software application and the algorithm of the triage is based in the Manchester triage system. For all participants three triage scores were determined by (1) the participant using the ca-ISET; (2) triage by a regular, trained triage assistant and (3) triage by one physician who was specially trained in ophthalmic triage. The diagnosis of the physician was chosen as the reference standard to define criterion validity. The order of triage administration was alternated per patient. Only cases with triage scores from the two triage systems and the reference standard were included. The outcome variables, four triage colours, were transformed into a binary score: high urgent and low urgent. The difference between the ca-ISET and regular triage in terms of sensitivity, specificity, NPV and PPV was tested by Z-scores. RESULTS Of 247 eligible patients, data was elicited from 189 patients (average age 54 years, range 18-89). The sensitivity of the ca-ISET (0.89, CI: 0.75-0.96) did not differ from the sensitivity of the regular triage (0.69, CI: 0.53-0.82, Z=1.74, p=0.08). The ca-ISET was less specific (0.78, CI: 0.71-0.84) than the regular triage (0.92, CI=0.86-0.95, Z=3.04, p=0.00). We found no significant difference between the ca-ISET and regular triage for PPV (Z=0.19, p=0.85) and NPV (Z=0.03, p=0.98). CONCLUSIONS The sensitivity, PPV and NPV of the ca-ISET does not differ from the sensitivity of the regular triage, while the ca-ISET retained a reasonable level of specificity. Therefore the ca-ISET can be recommended as a tool for ophthalmic emergency departments, and could be used when trained ED staff is absent.