P355 Diagnostic accuracy and usability of home calprotectin testing
A. Thomas*1, M. Clarke1, V. Cairns1, J. Goodhand1, T. Ahmad1, N. Kennedy1
1Royal Devon and Exeter Hospital, Gastroenterology, Exeter, UK
Faecal calprotectin (FC) is a reliable and non-invasive stool biomarker that is useful in monitoring inflammation in patients with IBD. Testing is routinely performed in a laboratory by enzyme-linked immunosorbent assay (ELISA) which is time consuming and costly. QuantonCal (QC) is an immunologic rapid test that can measure calprotectin via a lateral flow device to a smart phone. It is designed to be used by the patient in their home and can deliver a result within 15 min. We sought to compare the accuracy and usability of QC against ELISA.
We approached patients attending IBD outpatient clinic at the Royal Devon and Exeter Hospital, UK. Inclusion criteria were owning a smartphone and need for measuring calprotectin to assess disease activity/inflammation. Patients collected a stool sample to perform QC and submitted a sample for ELISA analysis to assess correlation. Patients were invited to complete a questionnaire.
Of 59 patients 32 had Crohn’s disease, 23 ulcerative colitis and 4 IBDU. Thirty-eight (64.4%) submitted at least one sample for analysis, 26 (44.1%) submitted two samples, and 24 (42.3%) completed a satisfaction survey. 3 QC samples were excluded for invalid readings.
The QC and lab calprotectin readings were significantly correlated; Spearman’s rank correlation coefficient = 0.737, p < 0.005, however the QC overestimated the lab calprotectin value by 74% (Figure 1). Median QC was 385 µg/g [IQR 31–1850], median lab 148 [IQR 46–529]).
Correlation betwen QuantonCal (home calprotectin) and laboratory calprotectin.
Using 27 lab calprotectin readings >250 µg/g as gold standard against QC readings, test performance was: area under the curve (AUC) = 0.870 (95% confidence interval (CI) = 0.779–0.961), sensitivity 90%, specificity 78%, positive predictive value 70%, and negative predictive value 94%. 24 patients completed the questionnaire. Acceptability was high: 15 (62.5%) thought QC was ‘very easy’, 7 (29.2%) ‘easy,’ no patients reported the application was ‘difficult’ or ‘very difficult’ to use. There was a preference towards QC compared with lab test: equal preference 9 (37.5%), slight preference 7(29.2%) and strong preference 4 (16.7%). Patients cited real-time results and feeling ‘in control’ of their disease as reasons for this.
The QC overestimated the lab calprotectin reading by 74%, with only moderate specificity and positive predictive value, rendering the diagnostic accuracy of QC poor. Inaccurate QC results could lead to false reassurance, delayed treatment, or inappropriate escalation of therapy. Despite patients reporting good usability, QC should not replace ELISA.
1. Fitzgerald D, Sugrue K, McCarthy J,