Cumulative and temporal associations between antimicrobial prescribing and community-associated Clostridium difficile infection : population-based case control study using administrative data

Kavanagh, Kimberley and Pan, Jiafeng and Marwick, Charis and Davey, Peter and Wiuff, Camilla and Bryson, Scott and Robertson, Chris and Bennie, Marion (2016) Cumulative and temporal associations between antimicrobial prescribing and community-associated Clostridium difficile infection : population-based case control study using administrative data. Journal of Antimicrobial Chemotherapy. ISSN 0305-7453

[img]
Preview
Text (Kavanagh-etal-JAC2016-antimicrobial-prescribing-and-community-associated-Clostridium-difficile)
Kavanagh_etal_JAC2016_antimicrobial_prescribing_and_community_associated_Clostridium_difficile.pdf
Accepted Author Manuscript

Download (763kB)| Preview

    Abstract

    Background. Community-associated (CA) Clostridium difficile infection (CDI) is a major public health problem. This study estimates the magnitude of the association between temporal and cumulative prescription of antimicrobials in primary care and CA-CDI. CA-CDI is defined as cases without prior hospitalisation in the previous 12 weeks who were either tested outside of hospital or tested within 2 days of admission to hospital. Methods. Three National patient level datasets –covering CDI cases, community prescriptions and hospitalisations were linked by the NHS Scotland unique patient identifier, the community health index, CHI. All validated cases of CDI from August 2010 to July 2013 were extracted and up to six population-based controls were matched to each case from the CHI register for Scotland. Statistical analysis used conditional logistic regression. Results. 1446 unique cases of CA-CDI were linked with 7964 age, sex and location matched controls. Cumulative exposure to any antimicrobial in the previous 6 months has a monotonic dose-response association with CA-CDI. Individuals with excess of 28 defined daily doses (DDD) to any antimicrobial (19.9% of cases) had an odds ratio (OR)=4.4 (95% CI:3.4-5.6) compared to those unexposed. Individuals exposed to 29+ DDD of high risk antimicrobials (cephalosporins, clindamycin co-amoxiclav, or fluoroquinolones) had an OR=17.9 (95% CI:7.6-42.2). Elevated CA-CDI risk following high risk antimicrobial exposure was greatest in the first month (OR=12.5 (8.9-17.4)) but was still present 4-6 months later (OR=2.6 (1.7-3.9)). Cases exposed to 29+DDD had prescription patterns more consistent with repeated therapeutic courses, using different antimicrobials, than long term prophylactic use. Conclusions. This analysis demonstrated temporal and dose-response associations between CA-CDI risk and antimicrobials with an impact of exposure to high risk antimicrobials remaining 4-6 months later.