Ribotype 078 Clostridium difficile infection incidence in Dutch hospitals is not associated with provincial pig farming: Results from a national sentinel surveillance, 2009-2015

• Published in: PloS one 2017-12, Vol.12 (12), p.e0189183-e0189183
• Main Authors: van Dorp, Sofie M, de Greeff, Sabine C, Harmanus, Céline, Sanders, Ingrid M J G, Dekkers, Olaf M, Knetsch, Cornelis W, Kampinga, Greetje A, Notermans, Daan W, Kuijper, Ed J
• Format: Article
• Language: English
• Subjects: Agricultural laborers; Bacteria; Biology and Life Sciences; Clostridium difficile; Clostridium infections; Clusters; Communities; Demographic aspects; Density; Disease control; Diseases; Farming; Farms; Health aspects; Health surveillance; Hogs; Hospitals; Incidence; Infections; Infectious diseases; Laboratories; Livestock; Medicine and Health Sciences; Microbiology; Mortality; Patients; People and places; Permutations; Physical Sciences; Poisson density functions; Public health; Risk factors; Spacetime; Statistical analysis; Studies; Surveillance; Typing
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0189183

It has been suggested that the high incidence of ribotype 078 Clostridium difficile infections (CDI) in the Netherlands is related to pig farming. We used data of hospitalised CDI patients (>2yrs of age) diagnosed between May 2009 and May 2015 in 26 hospitals participating in a national sentinel surveillance. We compared clinical and geographical characteristics of 078 CDI to other CDI. We investigated the association between 078 CDI incidence and four indicators of pig farming (piglet, pig, piglet farm and pig farm density) by mixed-effects Poisson regression. We used a space-time permutation model to search for community-onset 078 CDI clusters (using SaTScan). A total of 4,691 CDI were identified. Ribotype 078 was isolated in 493 of 3,756 patients (13.1%) including a typing result. These patients had slightly higher community-onset disease and a 35% increase of 30-day mortality compared to non-078 CDI patients. The pooled overall and 078 incidence rates were 2.82 (95% CI, 2.42-3.29) and 0.26 (95% CI, 0.21-0.31) CDI per 10,000 patients-days respectively. Hospital 078 CDI incidence was not associated with provincial pig (IRR, 0.98; 95% CI, 0.89-1.08), piglet (IRR, 0.95; 95% CI, 0.75-1.19), pig farm (IRR, 1.08; 95% CI, 0.84-1.39), or piglet farm density (IRR, 1.00; 95% CI, 0.56-1.79). No clusters of community-onset ribotype 078 CDI were found. Our results do not indicate that the ribotype 078 CDI incidence in hospitals is related to pig (farm) or piglet (farm) density. However, transmission beyond provincial borders or in non-hospitalised patients cannot be excluded.