P538 The reliability of patient performed fecal calprotectin testing

Crouwel, F.(1);Dijkhuis, L.E.J.L.(1);Duijvestein, M.(1);Buiter, H.J.C.(2);De Boer, N.K.(1);Hamer, H.M.(3);

(1)Amsterdam UMC- location VUmc, Department of Gastroenterology and Hepatology, Amsterdam, The Netherlands;(2)Amsterdam UMC- location VUmc, Department of Radiology and Nuclear Medicine, Amsterdam, The Netherlands;(3)Amsterdam UMC- location VUmc, Department of Clinical Chemistry, Amsterdam, The Netherlands


For measurement of fecal calprotectin (FC) a stool sample is sent to the laboratory, where calprotectin is extracted and determined. With the CALiaGold (CG) tube, especially designed for home sampling, this labor-intensive extraction procedure in the laboratory can be bypassed. Furthermore, the home-use of this buffer containing extraction device can potentially prevent FC degradation. We aimed to determine the reliability of patient performed FC extraction, to assess whether degradation during transport can be prevented and to determine its usability by patients.


In this prospective cross-sectional observational study, 4 CG tubes and 2 regular tubes were filled from the same bowel movement by patients with inflammatory bowel disease. Half of the tubes were directly frozen, others were sent to the laboratory by mail. Four more CG tubes were filled at the laboratory; 2 from the native sample sent by mail and 2 from the directly frozen native sample. The FC levels were measured by a particle enhanced turbidimetric immunoassay. The directly frozen tubes were used for the comparison between patient and analyst performed extractions, while the tubes sent by regular mail were used to determine FC stability during transport. The usability was assessed with a questionnaire.


Fifty-three patients were included. No significant difference was found in patient performed extractions compared to analyst performed extractions in samples with FC levels <200 µg/g. However, in samples with FC levels ≥200 µg/g patient performed extractions were significantly lower (p=0.014). When patients were divided in 3 groups (i.e. FC levels <50ug/g, 50-200 µg/g and >200 µg/g), the resulting Cohen’s kappa coefficient was 0.787 (95% CI: 0.646-0.928), reflecting a substantial agreement between patient and analyst performed extraction. A median FC increase of 28.6% was found in the native samples after one freeze-thaw cycle, potentially explaining the higher FC levels in the analyst performed extractions from the directly frozen native sample.

Patients sampling imprecision was higher (p<0.01, median CV 17%) compared to the analyst (median CV 8%). Higher FC levels were found in the CG tubes sent by mail compared to native samples, reflecting less FC degradation. The questionnaire revealed that 62% of patients preferred the CG tube and 8% the conventional tube.


Patient performed FC extraction is a realistic alternative sampling method, especially since the freeze-thaw process in the native samples may have led to higher values in the analyst performed extractions. Moreover, usage may prevent FC degradation during transport and patients prefer the use of the extraction device.