Skip to content

Enhanced diagnostic algorithm to include assessment of disease severity, length of hospital stay and treatment choice for Clostridium difficile disease

Student thesis: Doctoral Thesis

Clostridium difficile (C.difficile) is a bacterium that can infect the bowel causing diarrhoea and lead to pseudomembranous colitis if left untreated. The infection most commonly affects people who have recently been treated with antibiotics and can spread easily to other people. The increase in incidence and severity of C.difficile infection (CDI) is a cause for concern, with serious implications for patients and healthcare systems. Accurate laboratory and clinical diagnosis of CDI can be both problematic and challenging. Varying performance has been reported in the literature for Toxin Enzyme Immunoassays (EIA) (sensitivity ranging from 53.3% – 97.4%), the current method of choice for routine C.difficile testing in many hospitals. Alternative diagnostic assays are now available such as, glutamate dehydrogenase (GDH) and polymerase chain reaction (PCR). These are being evaluated by many hospital laboratories to improve diagnosis of CDI.
This research study aims to develop an accurate and rapid diagnostic testing algorithm for CDI. Once diagnosed, it is difficult to assess the clinical severity of C.difficile disease. A second aim of this study is to evaluate a bowel inflammation marker, lactoferrin (LF), to establish if elevated LF concentration can predict disease severity, patient length of stay (LOS) and guide appropriate antibiotic treatment.
Stool samples were tested prospectively using three laboratory tests for the diagnosis of CDI: an enzyme immunoassay (EIA) for the detection of C.difficile toxin, a lateral flow assay for the simultaneous detection of glutamate dehydrogenase (GDH) & toxin and an automated real-time polymerase chain reaction (PCR) assay for detection of the toxin B gene (tcdB). Both toxin detection methods lack sensitivity, with sensitivities of (51.2% & 65.2%), for the EIA method and lateral flow assay, respectively. The GDH component of the lateral flow assay and the PCR assay yielded high sensitivities (99.9% & 97.7%, respectively). All tests evaluated yielded low positive predictive values (PPV) (range, 49.4% to 68.3%), compared with the gold standard cytotoxigenic culture (CYTGC).The optimum single tests were GDH and PCR, both reporting the highest negative predictive values (NPV) (99.9% & 99% respectively). No one test is suitable as a stand-alone test, therefore, a three-stage testing algorithm was designed to improve diagnostic accuracy, using a combination of all three tests evaluated. This algorithm yielded a higher sensitivity than the EIA or lateral flow assay alone (98.2%, 51.2% and 65.2%, respectively) and yielded an NPV as high as the GDH and PCR assays (99.9%, 99.9% & 99.9% respectively).
To predict severity and length of hospital stay (LOS) for patients diagnosed with CDI, quantitative LF was tested and concentrations recorded at day 0, 3 and 10 of C.difficile treatment. LF concentrations were compared with the published clinical severity Zar score used as a reference standard.
By grouping lactoferrin concentrations on day 0, a treatment algorithm was designed to inform clinicians of appropriate antibiotic treatment based on disease severity. This information has been added to the new diagnostic algorithm for accurate and rapid diagnosis of CDI enabling clinicians to better clinically manage and improve outcomes for this group of patients.
Original languageEnglish
Awarding Institution
    Award dateSep 2019
    Relations Get citation (various referencing formats)

    ID: 20507921