P374 Home or hospital-based analysis of stool calprotectin: assessing two methods for monitoring inflammatory bowel disease

Abstract

Background: For 8 years we have been following patients with in-flammatory bowel disease by periodically measuring calprotectin levels in their sent-in stool samples with an enzyme-linked immunosor-bent assay (ELISA). Levels below 250 [xg/g confirmed disease remis-sion, levels above 500 [xg/g indicated a disease flare, and indecisive values between 250 and 500 [xg/g required retesting after 1 month. Physicians and patients found repeated testing of calprotectin helpful to guide therapy, but both wished to receive the results without delay. BUHLMANN Laboratories recently developed a lateral flow-based calprotectin test and a software application (IBDoc) that turns an ordinary smartphone camera into a reader for quantitative measurements at home. We compared this new method with the established ELISA method to see whether they agreed sufficiently for the new to replace the old, or to use the two interchangeably. Method(s): Eligible teenagers and adults, who had a smartphone validated for the IBDoc app, received an instruction manual to perform the calprotectin stool test at home. The residual of the stool specimen was sent to the hospital for ELISA measurement of calprotectin. Agreement between methods was assessed, as well as critical misclassifications of disease activity (leading to over-or undertreat-ment). Predefined acceptable limits of agreement were +/-100 [xg/g in the lower ranges of calprotectin and +/-200 [xg/g in the higher ranges. Result(s): 85 participants produced 152 paired calprotectin measurements. In the lower ranges (i.e., between 40 and 400 M-g/g) 99 of 117 pairs (85%) were within acceptable limits of agreement, and in the higher ranges (>400 M-g/g) 20 of 35 pairs (57%). Eighty percent of all paired measurements were concordant (). Critical mis-classification (disease remission with one method and disease flare with the other) was observed in 4% of pairs. Two critical misclassi-fications leading to undertreatment (low IBDoc, high ELISA) were in fact invalid measurements by a single patient who continued the analytical step before the indicated incubation period. Conclusion(s): We found acceptable agreement between IBDoc home test and hospital-based ELISA in the critical lower ranges of calprotectin and therefore the new method can be used to monitor patients [Table Presented] in remission. Results in the higher range need to be confirmed before therapy adjustment. Misclassification can probably be further reduced with a face-to-face training of the patients.