51. Clinical and Cost-effectiveness of a Procalcitonin Test as a Prompt Indicator of Prodomol Meningococcal Disease in Febrile Children: Cost-effectiveness Analysis
- Author
-
Frank Kee, Fiona Lynn, Michael D. Shields, Kathryn Anne Dunlop, Thomas Bourke, Jennifer M Bell, and A. Angus
- Subjects
medicine.medical_specialty ,Receiver operating characteristic ,Cost effectiveness ,business.industry ,Health Policy ,Public Health, Environmental and Occupational Health ,Cost-effectiveness analysis ,Emergency department ,Confidence interval ,Procalcitonin ,Test (assessment) ,Clinical pathway ,Emergency medicine ,medicine ,Intensive care medicine ,business - Abstract
Bell J.M.1, Shields M.1, Angus A.2, Dunlop K.3, Bourke T.1, Kee F.1, Lynn F.A.1 1Queen’s University Belfast, Belfast, UK, 2Belfast Health & Social Care Trust, Belfast, UK, 3South Eastern Health and Social Care Trust, Belfast, UK OBJECTIVES: To establish if the procalcitonin (PCT) test’s diagnostic performance is more clinically and cost-effective than C reactive protein (CRP) and White Cell Count (WCC) tests for suspicion of prodomal stage Meningococcal disease (MD) in children presenting at emergency department (ED) with a fever without source. METHODS: A decision analytic model was designed to reflect realistic clinical pathways for a child presenting with non-specific fever to ED. Test accuracy was evaluated using data from independent studies carried out in developed countries identified through a systematic literature search. Studies were combined to determine the optimal cut-off value for the PCT, CRP and WCC tests, each as an indicator of MD. Summary Receiver Operator Curve (SROC) analysis was used to determine the inter-study and overall diagnostic performance of each test from the areas under the curve (AUC), with 95% confidence intervals (CIs). Components of each clinical pathway were costed in UK sterling using the National Schedule of Reference Costs 2010-2011. Hospital stays were costed using the appropriate Health Resource Group code. RESULTS: Seven studies involving 881 children with non-specific fever provided data for inclusion. The PCT test was more accurate (sensitivity= 89%, 95%CI= 75-96; specificity= 71%, 95%CI= 37-91) for early MD compared to CRP (sensitivity= 84%, 95%CI= 74-90; specificity= 60%, 95%CI= 44-74) and WCC (sensitivity= 50%, 95%CI= 39-60; specificity= 68%, 95%CI= 54-79). PCT had the best PLR (3.0, 95%CI= 1.8-7.8) to be viable as a rule-in test for MD and a borderline NLR as a rule-out test, making it a better option to either CRP (2.1, 95%CI= 1.4-3.1) or WCC (1.5, 95%CI= 1.0-2.3). Outcomes from the decision analytic model indicated that the PCT test was the most cost-effective (£2547 per patient treated), followed by the combined CRP and WCC test (£3069 per patient treated). CONCLUSIONS: The improved sensitivity and specificity of the PCT test provides a more cost-effective test than the currently recommended CRP and WCC tests.
- Published
- 2013
- Full Text
- View/download PDF