Marc G. Romney

Risk Factors for Cryptococcus gattii Infection, British Columbia, Canada

To determine whether particular environmental, medical, or behavioral risk factors existed among Cryptcoccus gattii–infected persons compared with the general population, we conducted a sex-matched case−control study on a subset of case-patients in British Columbia (1999–2001). Exposures and underlying medical conditions among all case-patients (1999–2007) were also compared with results of provincial population–based surveys and studies. In case−control analyses, oral steroids (matched odds ratio [MOR] 8.11, 95% confidence interval [CI] 1.74–37.80), pneumonia (MOR 2.71, 95% CI 1.05–6.98), and other lung conditions (MOR 3.21, 95% CI 1.08–9.52) were associated with infection. In population comparisons, case-patients were more likely to be >50 years of age (p<0.001), current smokers (p<0.001), infected with HIV (p<0.001), or have a history of invasive cancer (p<0.001). Although C. gattii is commonly believed to infect persons with apparently healthy immune systems, several immunosuppressive and pulmonary conditions seem to be risk factors.

Large Community Outbreak of Streptococcus pneumoniae Serotype 5 Invasive Infection in an Impoverished, Urban Population

BACKGROUND Streptococcus pneumoniae is a common cause of sporadic invasive infections, but outbreaks of invasive pneumococcal disease are infrequent. In August 2006, a sudden increase in the number of patients presenting with invasive pneumococcal disease was noted at St. Pauls Hospital (Vancouver, Canada). Most patients with severe disease resided in an area referred to as the Downtown Eastside, a neighborhood known for its high rates of poverty and illicit drug use. METHODS Prospective, laboratory-based surveillance for invasive pneumococcal disease was initiated, including on-site serotyping of S. pneumoniae isolates. A vaccination campaign using 23-valent polysaccharide pneumococcal vaccine was launched in the Downtown Eastside. Multiple logistic regression was used to examine the association of sociodemographic variables and medical risk factors with S. pneumoniae serotype status. RESULTS A single S. pneumoniae serotype (serotype 5) was responsible for 78% of invasive pneumococcal disease cases (137 of 175 cases) during the outbreak period of August 2006-July 2007. The outbreak strain, although fully susceptible to penicillin, caused significant morbidity and placed considerable strain on the acute care system within the Vancouver Coastal Health region. Crack cocaine use was found to be the main independent risk factor associated with invasive pneumococcal disease due to S. pneumoniae serotype 5 (odds ratio, 12.4; 95% confidence interval, 2.22-69.5). CONCLUSIONS A targeted vaccination campaign using polysaccharide pneumococcal vaccine appeared to help control this outbreak. In urban centers with high rates of illicit drug use, vaccination strategies for preventing invasive pneumococcal disease may need to be refined to include individuals who use crack cocaine.

Community-associated methicillin-resistant Staphylococcus aureus is prevalent in wounds of community-based injection drug users

Injection drug users (IDUs) have an elevated risk for carriage of Staphylococcus aureus, including methicillin-resistant S. aureus (MRSA). Cutaneous injection-related infections are common in IDUs but detailed studies are few. Based on a subsample of 218 individuals from a community-recruited cohort of IDUs at a supervised injection facility, we investigated the microbiology and related antibiotic susceptibility profiles of isolates from 59 wounds. Twenty-seven percent of subjects had at least one wound and 25 (43%) were culture positive for S. aureus alone [14 MRSA and 11 (19%) methicillin-susceptible (MSSA) isolates]. Sixteen of 18 MRSA isolates were classified as community associated (CA) by the presence of genes encoding for PVL. MRSA and MSSA occurred in mixed infection with other organisms on three and six occasions, respectively. All CA-MRSA isolates were susceptible to tetracycline, vancomycin and linezolid but only 13% were susceptible to clindamycin compared to 63% of MSSA isolates. The frequency of CA-MRSA is a cause for concern in wound infection in the IDU setting.

Length of stay and mortality due to Clostridium difficile infection acquired in the intensive care unit

PURPOSE The purpose of this study was to determine the attributable intensive care unit (ICU) and hospital length of stay and mortality of ICU-acquired Clostridium difficile infection (CDI). MATERIALS AND METHODS In this retrospective cohort study of 3 tertiary and 3 community ICUs, we screened all patients admitted between April 2006 and December 2011 for ICU-acquired CDI. Using both complete and matched cohort designs and Cox proportional hazards analysis, we determined the association between CDI and ICU and hospital length of stay and mortality. Adjustment or matching variables were site, age, sex, severity of illness, and year of admission; any infection as an ICU admitting or acquired diagnosis before the diagnosis of CDI and diagnosis of CDI were time-dependent exposures. RESULTS Of 15314 patients admitted to the ICUs during the study period, 236 developed CDI in the ICU. In the complete cohort analysis, the hazard ratios (95% confidence interval) for CDI related to ICU and hospital discharge were 0.82 (0.72, 0.94) and 0.83 (0.73, 0.95), respectively (0.5 additional ICU days and 3.4 hospital days), and related to death in ICU and hospital, they were 1.00 (0.73, 1.38) and 1.19 (0.93, 1.52), respectively. In the matched analysis, the hazard ratios for CDI related to ICU and hospital discharge were 0.91 (0.81, 1.03) and 0.98 (0.85, 1.13), respectively, and related to death in ICU and hospital, they were 1.18 (0.85, 1.63) and 1.08 (0.82, 1.43), respectively. CONCLUSIONS C difficile infection acquired in ICU is associated with an increase in length of ICU and hospital stay but not with any difference in ICU or hospital mortality.

Carriage of methicillin-resistant Staphylococcus aureus by wild urban Norway rats (Rattus norvegicus).

Methicillin-resistant Staphylococcus aureus (MRSA) is an important cause of multi-drug-resistant infections in people, particularly indigent populations. MRSA can be transmitted between people and domestic animals, but the potential for transmission between people and commensal pests, particularly rodents, had not been investigated. The objective of this study was to identify the presence and characterize the ecology of MRSA in rats (Rattus spp.) from in an impoverished, inner-city neighborhood. Oropharyngeal swabs were collected from rats trapped in 33 city blocks and one location within the adjacent port. Bacterial culture was performed and MRSA isolates were characterized using a variety of methods, including whole-genome sequencing (WGS). The ecology of MRSA in rats was described using phylogenetic analysis, geospatial analysis, and generalized linear mixed models. MRSA was identified 22 of 637 (3.5%) rats tested, although prevalence varied from 0 – 50% among blocks. Isolates belonged to 4 clusters according to WGS, with the largest cluster (n = 10) containing isolates that were genetically indistinguishable from community-acquired USA300 MRSA strains isolated from people within the study area. MRSA strains demonstrated both geographic clustering and dispersion. The odds of an individual rat carrying MRSA increased with increased body fat (OR = 2.53, 95% CI = 1.33 – 4.82), and in the winter (OR = 5.29, 95% CI = 1.04 – 26.85) and spring (OR = 5.50, 95% CI = 1.10 – 27.58) compared to the fall. The results show that urban rats carried the same MRSA lineages occurring in local human and/or animal populations, supporting recent transmission from external sources. MRSA carriage was influenced by season, most likely as a result of temporal variation in rat behavior and rat-human interactions.

Economic analysis of vancomycin-resistant enterococci at a Canadian hospital: assessing attributable cost and length of stay

BACKGROUND Competing resource demands have resulted in the de-escalation of vancomycin-resistant enterococcus (VRE) control programmes in some Canadian healthcare centres. AIM To determine the attributable costs and length of stay (LOS) of VRE colonizations/infections in an acute care hospital in Canada. METHODS Surveillance and financial hospital-based databases were used to conduct analyses with cases and controls from fiscal year 2008-2009 (1 April 2008 to 31 March 2009) at an acute care hospital in downtown Vancouver, Canada. A statistical analysis of attributable costs and LOS was conducted using a generalized linear model. In a secondary analysis, differences in costs and LOS were examined for VRE infections versus colonizations. FINDINGS A total of 217 patients with VRE and a random sample of 1075 patients without VRE were examined. VRE has a positive and significant impact on patient hospitalization costs and LOS. Overall, the presence of VRE increased the estimated mean cost per patient by 61.9% (95% confidence interval: 42.3-84.3) in relative terms and

Incidence and Costs of Clostridium difficile Infections in Canada

17,949 (13,949-21,464) in absolute Canadian dollars. For LOS, the attributable number of days associated with a VRE case mean was 68.0% (41.9-98.9) higher in relative terms and 13.8 days (10.0-16.9) in absolute days. In the secondary analysis comparing VRE infection and colonization costs, no statistically significant difference was found. CONCLUSIONS Based on this analysis, the attributable cost and LOS of VRE are considerable. These factors should be considered before de-escalation of a hospital VRE control programme.

Mycobacterium heckeshornense peritonitis in a peritoneal dialysis patient: a case report and review of the literature

CDI has a substantial epidemiologic, and economic, burden; and the largest proportion of costs arise from prolonged hospitalization. Interventions reducing the severity of infection and/or relapses requiring rehospitalization are likely to have the largest absolute effect on direct medical cost.

A cluster of Bacillus cereus bacteremia cases among injection drug users

We report the first case of peritonitis attributed to Mycobacterium heckeshornense. This is a rare, non-tuberculous mycobacterium that has been reported as an aetiological agent in a growing number and widening spectrum of infections.

Shigella flexneri serotype 1 infections in men who have sex with men in Vancouver, Canada.

Bacillus cereus is a bacteria commonly found in nature that is most frequently implicated as the cause of gastrointestinal illness caused by ingestion of contaminated food; however, there are also reports of extraintestinal infections, although rare. This article describes three cases in which B cereus bacteremia occurred among injection drug users living in Vancouver, British Columbia, within a short period of time. Pulsed-field gel electrophoresis was performed to determine whether these three isolates were related.