Underdiagnosis of Clostridium difficile across Europe: the European, multicentre, prospective, biannual, point-prevalence study of Clostridium difficile infection in hospitalised patients with diarrhoea (EUCLID)

Summary

Background

Variations in testing for Clostridium difficile infection can hinder patients' care, increase the risk of transmission, and skew epidemiological data. We aimed to measure the underdiagnosis of C difficile infection across Europe.

Methods

We did a questionnaire-based study at 482 participating hospitals across 20 European countries. Hospitals were questioned about their methods and testing policy for C difficile infection during the periods September, 2011, to August, 2012, and September, 2012, to August, 2013. On one day in winter, 2012–13 (December, 2012, or January, 2013), and summer, 2013 (July or August), every hospital sent all diarrhoeal samples submitted to their microbiology laboratory to a national coordinating laboratory for standardised testing of C difficile infection. Our primary outcome measures were the rates of testing for and cases of C difficile infection per 10 000 patient bed-days. Results of local and national C difficile infection testing were compared with each other. If the result was positive at the national laboratory but negative at the local hospital, the result was classified as undiagnosed C difficile infection. We compared differences in proportions with the Mann-Whitney test, or McNemar's test if data were matched.

Findings

During the study period, participating hospitals reported a mean of 65·8 tests (country range 4·6–223·3) for C difficile infection per 10 000 patient-bed days and a mean of 7·0 cases (country range 0·7–28·7) of C difficile infection per 10 000 patient-bed days. Only two-fifths of hospitals reported using optimum methods for testing of C difficile infection (defined by European guidelines), although the number of participating hospitals using optimum methods increased during the study period, from 152 (32%) of 468 in 2011–12 to 205 (48%) of 428 in 2012–13. Across all 482 European hospitals on the two sampling days, 148 (23%) of 641 samples positive for C difficile infection (as determined by the national laboratory) were not diagnosed by participating hospitals because of an absence of clinical suspicion, equating to about 74 missed diagnoses per day.

Interpretation

A wide variety of testing strategies for C difficile infection are used across Europe. Absence of clinical suspicion and suboptimum laboratory diagnostic methods mean that an estimated 40 000 inpatients with C difficile infection are potentially undiagnosed every year in 482 European hospitals.

Funding

Astellas Pharmaceuticals Europe.

Introduction

Clostridium difficile is the major infective cause of nosocomial diarrhoea in developed countries. Rapid and accurate diagnosis is important to optimise patients' care and infection prevention.¹ An increase has been recorded in the measured incidence of C difficile infection in countries with surveillance programmes, and a striking shift in epidemiology has been noted over the past decade.2, 3, 4 In Europe, C difficile is the most commonly reported pathogen associated with hospital-associated gastrointestinal disease,⁵ whereas in the USA it was the most commonly reported health-care associated microorganism in a survey of 183 hospitals in 2010.⁶

In the most recent European epidemiological survey of C difficile infection (2008), the incidence in 97 hospitals across 29 countries varied widely (range 0·0–36·3 per 10 000 patient bed-days per hospital; weighted mean 4·1).⁷ The reasons for this large variation are unclear. Predisposing factors to C difficile infection, including increasing age, comorbidities, and use of broad-spectrum antibiotics, might be similar across Europe, although exposure to different C difficile strains probably varies.1, 8 Notably, a more than 40-fold variation in testing frequency for C difficile infection was recorded across European countries, and a correlation between testing rate and reported infection rate was identified.⁷ Suboptimum case ascertainment, either because of inadequate laboratory diagnosis or absence of clinical suspicion, means that the true burden of C difficile infection is unclear.9, 10, 11, 12 In a point-prevalence study done in Spain, 66% of patients with C difficile infection on one day were undiagnosed or misdiagnosed because of either a lack of clinical suspicion (47%) or insufficient laboratory testing (19%).⁹ Optimum laboratory diagnosis of C difficile infection depends on testing relevant patients at the correct time with suitable tests. Importantly, detection of C difficile toxin in patients' faecal samples, compared with identification of toxigenic C difficile (strains that either produce toxin in vitro or have toxin genes present), correlates with disease severity and mortality.12, 13 To improve the suboptimum sensitivity of commercially available toxin detection assays, two-stage laboratory diagnosis, including a sensitive C difficile screening test followed by a C difficile toxin assay, is recommended.1, 12 Nevertheless, a questionnaire-based study in 125 European laboratories in 2010 showed wide variation in use of C difficile infection diagnostic methods, with a quarter still using one assay.¹⁴ Notably, however, such data do not ascertain the true extent of missed C difficile infection diagnoses.

We aimed to investigate the underdiagnosis of C difficile infection in hospitals in 20 European countries by asking participating hospitals to forward inpatients' diarrhoeal faecal samples to a national coordinating laboratory for C difficile infection testing by the study reference method, irrespective of whether microbiology tests had been requested or done locally, on 2 days—one day in winter (December, 2012, or January, 2013) and one day in the summer (July, 2013, or August, 2013). Participating hospitals were also asked to complete a study questionnaire about C difficile infection testing.