Infectious Disease Alert Updates
To gown and glove or screening surveillance ?
By Carol A. Kemper, MD, FACP
Harris AK, et al. Universal Glove and Gown use and acquisition of antibiotic-resistant bacteria in the ICU; A randomized trial. JAMA 2013; 310:1571-1580
A large-scale cluster randomized trial conducted in 20 U.S. intensive care units in 2012 sought to determine whether gowns and gloves for all patient contact significantly reduced acquisition of MRSA and VRE compared with usual precautions. Usual precautions are based on current CDC guidelines recommending contact precautions for patients with recognized infection or colonization with multi-drug resistant organisms. The ICU's were divided into intervention (universal gown/gloving) and usual precautions. The only exclusion criteria was prohibition against active surveillance for MDR.
A total of 92,241 screening swab cultures were collected from 26,180 patients throughout their ICU stay — and the ICU personnel were blinded to the results. The baseline rate of MRSA colonization during a 3 month period before study onset ranged from 3.02% to 16.37%, and the rate of VRE colonization ranged from 3.05% to 24.8%.
In the active intervention group (universal gowning/gloves), acquisition of MRSA and VRE decreased from 2135 events/1000 patient days to 1691 events/1000 patient days, respectively. A similar reduction was observed in the usual care group, with acquisition of MRSA decreasing from 1902 to 1629 events/1000 days. No statistically significant difference was observed between the two groups. Looking at individuals, there were fewer acquisitions of MRSA (-2.98/1000 patient days) within the intervention group compared with the usual care group (p = .042) but no difference was detected for VRE. Universal contact precautions decreased the number of hourly health care entries to the room by about 22% ( p= .02), but it did have a positive effect on hand washing compliance. There was no apparent reduction in health care associated infections such as catheter-associated UTIs, catheter-associated blood stream infections, and ventilator-associated pneumonia.
Many hospitals are moving in the direction of targeted surveillance of certain high-risk groups upon admission, with selective use of contact precautions as appropriate. For one thing, identification of MDR colonization upon admission may facilitate reduction in the rates of "hospital onset" infection a key goal of most infection control programs nowadays. But the success of this "intervention" likely rests on the frequency of pre-existing colonization with multi-drug resistant organisms. The question is: at what prevalence of colonization is it better to actively screen and isolate vs isolate certain high-risk admissions, vs isolate everyone. Obviously if the rate of MRSA colonization is only 3% in your ICU admissions, you're wasting a lot of effort to gown and glove everyone. But if the rate of colonization in your ICU is 25%, should you attempt to pre-screen and isolate as needed, or just throw everyone into an isolation room, and not waste money and effort on expensive PCR swabs? Someone should be able to model this.
Altered flu presentation in immunosuppression
Memoli MJ et al. CID 2014;58(2): 58:214.
Epidemiologic surveys have suggested that immunocompromised patients have greater morbidity and mortality with Influenza. In patients with leukemia/lymphoma, one study found that Influenza resulted in up to 80% incidence of pneumonia and 30% mortality. Similar observations were observed in bone marrow transplant patients. Another study of homologous stem-cell transplant (HSCT) patients undergoing chemotherapy with Influenza were at 30-40% risk for pneumonia, 16-17.5% required ICU care, and 7% died.
From 2008 to 2011, the NIH collected data on 86 patients seen at their facility with Influenza. Of these, 37% were severely immunocompromised (IC), including 59% who had undergone HSCT, 38% with hematologic malignancy; and the rest with solid tumors undergoing immunosupressive chemotherapy. One-fourth of IC patients had received seasonal Influenza vaccine.
The remaining 63% of patients with Influenza were not immunosuppressed, but 81% were overweight, 40% had a variety of other illness, such as diabetes, COPD, etc., and 22% smoked. About 60% of the non-IC patients had received seasonal Influenza vaccine with the previous year.
The non-IC patients presented with typical symptoms of Influenza with dry cough (96%), headache (89%), fever (83%), chills (81%), sweats (77%), body aches (68%); gastrointestinal complaints were not common; and 30% required hospitalization. In contrast, all of the IC patients required hospitalization, and their presentation was often less typical: cough (78%), fever (85%), headache (70%), chills (51%), muscle aches (41%), and GI symptoms were more common. IC patients were more likely to have CXR's performed, and of those who had CXRs performed, IC patients were more likely to have radiographic abnormalities than non-IC subjects (39% vs 23%), p=.015). They required longer stays of admission, more frequent admission to ICU, although only one IC patient died (3%).
IC patients also had significantly longer duration of viral shedding compared with non-IC subjects (19.0 vs 6.4 days) — many were still shedding virus despite resolution of their symptoms. Of the 80 Influenza A strains isolated, 6 (11%) were H1N1, 19.7% were H3N2, and 66.3% were pdm09H1N1 subtype. All 6 of the seasonal H1N1 strains were resistant to oseltamivir and possibly peramivir (carried the H2754 NAI resistance mutation), and all of the H3N2 and pdm09 subtypes were resistant to amantadine. Three of the 6 resistant H1N1 strains were observed at baseline, and 3 developed during treatment.
IC patients, most with hematologic malignancy or HCST, seen at the NIH were at high risk for more severe Influenza, more frequent pulmonary infiltrates (39%), and hospitalization (100%), with longer durations of hospitalization than their non-IC counterparts. Their presentations were often not typical, with a lower incidence of cough and inflammatory-mediated symptoms such as chills and body aches.
Before the Black Death, there was the first Plague
Wagner Dm, et al. Yersinia pestis and the Plague of Justinian 541-543 AD: a genomic analysis. Lancet 2014;http://dx.doi.org/10.1016/S1473-3099(13)70323-2.
Yersinia pestis was responsible for 3 devastating plagues. The first of these, referred to as the Justinian Plague, began in 541-543 AD, spreading through Asia, Africa, and into the Mediterranean Basin and Europe, and then continued to cycle through these countries every 8 to 12 years for the next 200 years. The death toll was a staggering 15-40%. Subsequent pandemic plagues occurred in the 14-17th c. (the Black Death) and again in the 19th and 20 centuries, even reaching San Francisco by boat in 1900-1904. It was the first time plague had reached the Continental United States. It has long been believed the first plague arose in Africa, travelling to Asia and China by the Silk Road, whereas subsequent plagues arose in China. Y. pestis generally arises from rodent reservoirs, and requires an existing healthy rodent reservoir to subsist (although is currently endemic in prairie dogs in the American Southwest).
These investigators sought to determine if ancestral strains of Y. pestis from the first pandemic were related to strains responsible for the two subsequent pandemics. DNA was extracted from the teeth of two individuals buried in a medieval cemetery in Bavaria Germany, estimated to be from 525-550 AD, some of which had multiple combined graves suspicious for pandemic victims. Both individuals were radiocarbon dated to have died in approximately 504 AD. Sequences from the core Y. pestis genome and three primary plasmids (pPCP1, pMT1, and pCDi) along with genes from 155 other Y. pestis strains were compared, creating a Y. pestis phylogeny.
They created a "maximum likelihood tree" of SNPs from all of the strains, using the genome from a strain of Y pseudotuberculosis as the root of the tree (since it is the presumed ancestral source of Y. pestis). Five major branches of the tree evolved, with the 3rd pandemic strain being located within the 5th and final branch. The two Justinian strains were isolated to their own novel branch, distant from the subsequent pandemic strains meaning this strain was unrelated and dead-ended. The two strains were "interleaved" between two extent groups of Y. pestis recognized for plague foci in Xinjiang, China, where they are still found in rodents and fleas. Thus this data suggests that the Justinian Plague arose in China, similar to other plagues, traveling across Asia to Europe along the Silk Road, but then died out for unclear reasons, either because the population was too sparse and failed to establish newer rodent reservoirs. The two later Plagues were then caused by later strains (distinctly different from the original Plague strain) that were able to establish new rodent reservoirs — spreading back and forth across Asia, Africa and Europe — and eventually reaching the United States.