Findings of a hospital surveillance-based outcome evaluation study for Clostridium difficile-associated colitis I. Varkonyi 1 , E. Rakoczi 1 , O. Misak 1 , E. Komaromi 1 , L. Kardos 1 , Z. Lampe 1 and Z. Szilvassy 2 1) Kenezy Gyula Hospital and Outpatient Facility and 2) University of Debrecen Medical and Health Science Centre, Debrecen, Hungary Abstract We completed a prospective study of 164 patients involved in a Clostridium difficile surveillance programme, evaluating a range of variables such as disease severity, treatment regimen and known clinical risk factors, for their effect on case lethality. The aim of this study was to determine if there are any additional clinical variables worth considering for inclusion in the therapeutic decision-making process. Beyond common risk factors, secondary immunodeficiencies such as diabetes mellitus, malignancy, autoimmune disease, immunosuppressive therapy and chronic hepatitis were assessed. Overall case lethality was 23%. There was a suggestion that regular proton pump inhibitor use in past medical history might be associated with greater lethality. Immunosuppressive therapy within 1 month before the onset of diarrhoea was associated with a significant four-fold lethality increase. This last finding may have the potential to further improve therapeutic judgement if used as an explicit component of a revised scoring system. In relation to Clostridium difficile-associated colitis, immunosuppressive therapy as a red flag entity, as described here, has not been previously published. Keywords: Clostridium difficile, colitis, hospital surveillance, immunosuppression, prognosis Original Submission: 21 December 2013; Revised Submission: 24 March 2014; Accepted: 23 April 2014 Editor: D. Raoult Article published online: 28 April 2014 Clin Microbiol Infect Corresponding author: I. Varkonyi, Kenezy Gyula Korhaz, Infektologiai Intezet, Pf. 195, 4001 Debrecen, Hungary E-mail: i.varkonyi@clintrial-audit.hu Introduction Clostridium difficile infection (CDI) remains one of the most challenging nosocomial infections worldwide. From 1986 to 2009 in Hungary, a total of four reported CDI epidemics took place (2001, 2003, 2004, 2009). In 2001, Urb an et al. [1] reported that out of 65 C. difficile isolates, the most common (39%) toxigenic type was PCR ribotype 087; there was no occurrence of the binary toxin. In 2003, the same team confirmed the presence of the binary toxin in two cases out of 112 C. difficile isolates; both were community-acquired infec- tions [2]. An annual average of 43.8 CDI cases were reported in Hungary over the years 2004–08 [3]. During those 5 years, average morbidity was 0.4 per 100 000 population, with average lethality at 0.9%. In 2009 and 2010, average annual morbidity (2.1/100 000) and lethality (1.9%) figures rose to markedly higher levels. Terhes et al. [4] detected the first Hungarian case of C. difficile PCR ribotype 027 in a hospital setting in 2008. In 2010, nine hospital epidemics were registered, in three of which PCR ribotype 027 strains were confirmed [3]. Methodology guidelines in effect in Hungary as of 2011 classify CDI as a notifiable disease [3]. According to Hungary’s National Nosocomial Surveillance System data, the number of sporadic cases in 2012 was 4506, as reported by 84 hospitals; the corresponding lethality figure was 22.3%. In 2011, the number of epidemics rose to 20, and in 2012 it was 21. However, no accurate national data on the PCR ribotype composition of the pathogenic strains are available [5]. In a recent Hungarian publication of ribotyping findings in 2010 and 2011, various C. difficile strains isolated from patients with diarrhoea and obtained from nationwide laboratory sources were found to be of the 027 ribotype in 30.4% of 601, and 50.2% of 699 cases, respectively; these results demonstrate an increasing tendency for the presence of ribotype 027 in the country [6]. ª2014 The Authors Clinical Microbiology and Infection ª2014 European Society of Clinical Microbiology and Infectious Diseases ORIGINAL ARTICLE 10.1111/1469-0691.12652