Hyperoxia is associated with adverse outcomes in the cardiac intensive care unit: insights from the Medical Information Mart for Intensive Care (MIMI-III) database

Background

Hyperoxia produces reactive oxygen species, apoptosis, and vasoconstriction, and is associated with adverse outcomes in patients with heart failure and cardiac arrest. Our aim was to evaluate the association between hyperoxia and mortality in patients (pts) receiving positive pressure ventilation (PPV) in the cardiac intensive care unit (CICU).

Methods

Patients admitted to our medical center CICU who received any PPV (invasive or non-invasive) from 2001 through 2012 were included. Hyperoxia was defined as time-weighted mean of PaO2 >120mmHg and non-hyperoxia as PaO2 ≤120mmHg during CICU admission. Primary outcome was in-hospital mortality. Multivariable logistic regression was used to assess the association between hyperoxia and in-hospital mortality adjusted for age, female sex, Oxford Acute Severity of Illness Score, creatinine, lactate, pH, PaO2/FiO2 ratio, PCO2, PEEP, and estimated time spent on PEEP.

Results

Among 1493 patients, hyperoxia (median PaO2 147mmHg) during the CICU admission was observed in 702 (47.0%) pts. In-hospital mortality was 29.7% in the non-hyperoxia group and 33.9% in the hyperoxia group ((log rank test, p=0.0282, see figure). Using multivariable logistic regression, hyperoxia was independently associated with in-hospital mortality (OR 1.507, 95% CI 1.311–2.001, p=0.00508). Post-hoc analysis with PaO2 as a continuous variable was consistent with the primary analysis (OR 1.053 per 10mmHg increase in PaO2, 95% CI 1.024–1.082, p=0.0002).

Conclusions

In a large CICU cohort, hyperoxia was associated with increased mortality. Trials of titration of supplemental oxygen across the full spectrum of critically ill cardiac patients are warranted.

Funding Acknowledgement

Type of funding source: None

Non-commercial poultry industries: Surveys of backyard and gamefowl breeder flocks in the United States

The National Animal Health Monitoring System (NAHMS) Poultry '04 study was conducted to better describe non-commercial United States poultry populations, in particular, backyard and gamefowl breeder flocks. To estimate the density of backyard flocks in close proximity to commercial operations, a sample of 350 commercial poultry operations in 18 top poultry producing states was selected from the National Agricultural Statistics Service (NASS) list of poultry operations. A 1 mile radius circle was drawn around each operation, and door-to-door canvassing was conducted within these circles to enumerate premises with all species of birds. Premises with backyard poultry flocks completed a questionnaire focusing on bird health, bird movement, and biosecurity practices. A similar questionnaire, provided in both English and Spanish, was mailed to all members of State affiliates of the United Gamefowl Breeders Association (UGBA) as well as to members of State associations not affiliated with UGBA. An average of 29.4 residences was found within a 1 mile radius of commercial operations, of which 1.9 residences per circle had backyard poultry flocks. Gamefowl breeder flocks were larger, used more health care and biosecurity practices, and moved birds more frequently compared to backyard flocks.

The concept of compartmentalisation

The rationale for establishing trade 'regions' and 'zones' is based on principles of epidemiological science and risk analysis that assess and manage animal disease risks so that the safety of trade can be ensured. However, the boundaries of geographical regions and zones may readily be breached through numerous epidemiological pathways. The concept of a 'compartment' extends the application of a 'risk boundary' beyond that of a geographical interface and considers all epidemiological factors that can contribute to the creation of an effective boundary. The fundamental requirement for application of either concept (regions/zones or compartments) is that the population considered for trade is maintained within management or geographical boundaries which allow clear epidemiological differentiation to be made between those animals and surrounding populations of higher risk. Seven factors are presented that an exporting country might use to guide the identification and documentation of a compartment. Additionally, the steps that would be undertaken to implement trade based on the compartmentalisation concept are discussed.

The Isolation of Exotic Newcastle Disease (END) Virus from Nonpoultry Avian Species Associated with the Epidemic of END in Chickens in Southern California: 2002–2003

During the first 11 months of the 2002-2003 exotic Newcastle disease (END) epidemic in chickens in southern California, a total of 27,688 cloacal and tracheal (oropharyngeal) swab pools and/or tissue pools from 86 different avian species other than chickens and turkeys were submitted for Newcastle disease virus (NDV) isolation and characterization. Fifty-seven specimens (0.23%), representing 12 species of birds and 13 unspecified species, from a total of 24,409 accessions or submissions were positive for NDV. The NDV isolate was characterized as ENDV by real-time reverse transcription-polymerase chain reaction (RT-PCR). Of the 11,486 premises with other avian species, 1599 also had chickens. There were 1900 positive chicken samples from 164 premises, and 56 positive other avian species from 51 premises. Twelve premises had both positive chickens and positive other avian species. All positive other avian species were located on premises either on or within a 1 km radius of known infected premises. In this epidemic, premises with positive other avian species were significantly more likely to have chickens, and were significantly more likely to have positive chickens (OR = 3.7, P < 0.0001).

Epidemiologic and surveillance studies on avian influenza in live-bird markets in New York and New Jersey, 2001

In 2001, all 109 retail live-bird markets (LBMs) in New York and New Jersey were surveyed for the presence of avian influenza virus (AIV) by a real time reverse transcriptase/polymer chain reaction assay (RRT/PCR) and results compared to virus isolation (VI) in embryonating chicken eggs. The RRT/PCR had a 91.9% sensitivity and 97.9% specificity in detecting presence of AIV at the market level. However, the sensitivity at the sample level is 65.87%. The RRT/PCR is a reliable method to identify AIV at the market level. In addition, a cross-sectional epidemiologic study of the LBMs showed that, during the past 12 months, markets that were open 7 days per week and those that also sold rabbits had the highest risk for being positive for AIV. Markets that were closed one or more days per week and those that performed daily cleaning and disinfecting had the lowest risk for being AIV positive.

Descriptive and surveillance studies of suppliers to New York and New Jersey retail live-bird markets

Low pathogenicity avian influenza virus (AIV) H7N2 has been isolated since 1994 from retail live-bird markets (LBMs) in the northeastern United States. This study examines the suppliers to the LBMs in New York and New Jersey. In 2001, 185 supplier premises in nine states were surveyed for the presence of AIV by virus isolation (VI) in embryonating chicken eggs. No H7 or H5 virus was isolated. In addition, 104 producer premises in two states were serologically negative for H7 and H5 AIV. Information on management practices was obtained via questionnaire for 191 premises in 12 states. The survey results suggest that current biosecurity practices at supplier premises could be improved, especially regarding movement of birds. The study supports the hypothesis that H7N2 AIV is primarily maintained within the LBMs and, if reintroduction from suppliers is occurring, it is likely reintroduced at a very low level or from suppliers not included in this study.

Factors associated with fecal shedding of verotoxin-producing Escherichia coli O157 on dairy farms

Fecal samples were collected from 4,361 dairy cows on 91 dairy operations between 26 February and 8 July 1996. Fecal samples were cultured for Escherichia coli O157, and positive isolates were probed for verotoxin-producing genes. A total of 52 (1.2%) fecal samples on 22 (24.2%) operations were positive for verotoxin-producing E. coli O157. Herds in which samples were collected on or after 1 May 1996 were significantly more likely to test positive than herds sampled before that date (odds ratio = 7.7). Herds maintained on farms on which alleyways were flushed with water to remove manure were 8.0 times more likely to have samples test positive for verotoxin-producing E. coli O157 than were herds maintained on farms cleaned by use of other methods of manure removal.

Prevalence of enterohemorrhagic Escherichia coli O157:H7 in a survey of dairy herds

The prevalence of Escherichia coli O157:H7 in dairy herds is poorly understood, even though young dairy animals have been reported to be a host. From February to May 1993, 662 fecal samples from 50 control herds in 14 states, and from June to August 1993, 303 fecal samples from 14 case herds in 11 states were collected for isolation of E. coli O157:H7. Case herds were those in which E. coli O157:H7 was isolated from preweaned calves in a previous U.S. Department of Agriculture study, whereas control herds from which E. coli O157:H7 had not been isolated previously were randomly selected from the same states as case herds. Among the control herds, E. coli O157:H7 was isolated from 6 of 399 calves (1.5%) that were between 24 h old and the age of weaning and from 13 of 263 calves (4.9%) that were between the ages of weaning and 4 months. Eleven of 50 control herds (22%) were positive. Among the case herds, E. coli O157:H7 was isolated from 5 of 171 calves (2.9%) that were between 24 h old and the age of weaning and from 7 of 132 calves (5.3%) that were between the ages of weaning and 4 months. Seven of 14 case herds (50%) were positive. Sixteen of 31 isolates were obtained by direct plating, with populations ranging from 10(3) to 10(5) CFU/g. Fifteen of 31 isolates were isolated by enrichment only. Nineteen of the isolates produced both verocytotoxin 1 (VT-1) and VT-2, whereas 12 produced VT-2 only.