Generalizable Deep Learning-based Sleep Staging Approach for Ambulatory Textile Electrode Headband Recordings.

Reliable, automated, and user-friendly solutions for the identification of sleep stages in home environment are needed in various clinical and scientific research settings. Previously we have shown that signals recorded with an easily applicable textile electrode headband (FocusBand, T 2 Green Pty Ltd) contain characteristics similar to the standard electrooculography (EOG, E1-M2). We hypothesize that the electroencephalographic (EEG) signals recorded using the textile electrode headband are similar enough with standard EOG in order to develop an automatic neural network-based sleep staging method that generalizes from diagnostic polysomnographic (PSG) data to ambulatory sleep recordings of textile electrode-based forehead EEG. Standard EOG signals together with manually annotated sleep stages from clinical PSG dataset (n = 876) were used to train, validate, and test a fully convolutional neural network (CNN). Furthermore, ambulatory sleep recordings including a standard set of gel-based electrodes and the textile electrode headband were conducted for 10 healthy volunteers at their homes to test the generalizability of the model. In the test set (n = 88) of the clinical dataset, the model's accuracy for 5-stage sleep stage classification was 80% (κ = 0.73) using only the single-channel EOG. The model generalized well for the headband-data, reaching 82% (κ = 0.75) overall sleep staging accuracy. In comparison, accuracy of the model was 87% (κ = 0.82) in home recordings using the standard EOG. In conclusion, the CNN model shows potential on automatic sleep staging of healthy individuals using a reusable electrode headband in a home environment.

Reduced plasma PCSK9 response in patients with bacteraemia is associated with mortality.

BACKGROUND: The proprotein convertase subtilisin/kexin type 9 (PCSK9) enzyme controls blood cholesterol levels by downregulating the expression of the low-density lipoprotein receptor (LDLR). Pathogenic lipids (e.g. lipopolysaccharide) are removed from the circulation by an LDLR/PCSK9-dependent mechanism; thus, it has been suggested that PCSK9 inhibitors may be beneficial in the treatment of infections. We measured plasma PCSK9 levels in patients with culture-positive bacteraemia and explored pathogen-dependent and infection site-dependent effects as well as correlations between patient characteristics and outcome. METHODS: Proprotein convertase subtilisin/kexin type 9 in the plasma was measured with an enzyme-linked immunosorbent assay from 481 patients with blood culture-positive infection on days 0 to 4 after admission to the emergency department. Patient outcome and clinical and laboratory data were gathered retrospectively from patient records. RESULTS: The plasma PCSK9 level was elevated equally in patients with Gram-positive or Gram-negative bacterial infections; particularly high levels were seen in patients with a lower respiratory tract infection and Streptococcus pneumoniae bacteraemia. PCSK9 levels showed a significant positive correlation with C-reactive protein (CRP) level. Bacteraemia patients with liver disease or a history of alcohol abuse had significantly lower levels of plasma PCSK9. Reduced PCSK9 plasma responses in patients were significantly associated with mortality at days 7, 28 and 90. CONCLUSION: Proprotein convertase subtilisin/kexin type 9 is upregulated in blood culture-positive infections. Plasma PCSK9 resembles acute-phase proteins; its expression is induced during an infection, reduced in liver disease and correlates positively with CRP level. We have shown that PCSK9 levels are lower in patients with a fatal prognosis.

The effect of preoperative oral antibiotic use on the risk of periprosthetic joint infection after primary knee or hip replacement: a retrospective study with a 1-year follow-up.

OBJECTIVES: Antibiotics are used for various reasons before elective joint replacement surgery. The aim of this study was to investigate patients' use of oral antibiotics before joint replacement surgery and how this affects the risk for periprosthetic joint infection (PJI). METHODS: Patients having a primary hip or knee replacement in a tertiary care hospital between September 2002 and December 2013 were identified (n = 23 171). Information on oral antibiotic courses purchased 90 days preoperatively and patients' chronic diseases was gathered. Patients with a PJI in a 1-year follow-up period were identified. The association between antibiotic use and PJI was examined using a multivariable logistic regression model and propensity score matching. RESULTS: One hundred and fifty-eight (0.68%) cases of PJI were identified. In total, 4106 (18%) joint replacement operations were preceded by at least one course of antibiotics. The incidence of PJI for patients with preoperative use of oral antibiotics was 0.29% (12/4106), whereas for patients without antibiotic use it was 0.77% (146/19 065). A preoperative antibiotic course was associated with a reduced risk for subsequent PJI in the multivariable model (OR 0.40, 95% CI 0.22-0.73). Similar results were found in the propensity score matched material (OR 0.34, 95% CI 0.18-0.65). CONCLUSIONS: The use of oral antibiotics before elective joint replacement surgery is common and has a potential effect on the subsequent risk for PJI. Nevertheless, indiscriminate use of antibiotics before elective joint replacement surgery cannot be recommended, even though treatment of active infections remains an important way to prevent surgical site infections.

Plasma cell-free DNA and qSOFA score predict 7-day mortality in 481 emergency department bacteraemia patients.

BACKGROUND: A few studies have shown that both quick Sequential Organ Failure Assessment (qSOFA) score and cell-free DNA (cfDNA) have potential use as a prognostic marker in patients with infection. We studied these two markers alone and in combination to identify those emergency department (ED) patients with the highest risk of death. METHODS: Plasma cfDNA level was studied on days 0 to 4 after admittance to the ED from 481 culture-positive bloodstream infection cases. The qSOFA score was evaluated retrospectively according to Sepsis-3 definitions. The primary outcome was death by day 7. RESULTS: CfDNA on day 0 was significantly higher in nonsurvivors than in survivors (2.02 µg mL-1 vs. 1.35 µg mL-1 , P < 0.001). CfDNA level was high (>1.69 µg mL-1 ) in 134 (28%) of 481 cases, and the qSOFA score was ≥2 in 128 (28%) of 458 cases. High cfDNA and qSOFA score ≥2 had 70% and 77% sensitivity and 76% and 76% specificity in predicting death by day 7, respectively. High cfDNA alone had odds ratio (OR) of 7.7 (95% CI 3.9-15.3) and qSOFA score ≥2 OR of 11.6 (5.5-24.3), but their combination had OR of 20.3 (10.0-41.4) in predicting death by day 7 when compared with those with low cfDNA and qSOFA score <2. Amongst the five cases with the highest cfDNA levels, there were three patients with severe disseminated intravascular coagulation. CONCLUSION: CfDNA and qSOFA score can be used independently to identify those bacteraemia patients at high risk of death, and combining these two markers gives additional advantage.

The impact of preoperative bacteriuria on the risk of periprosthetic joint infection after primary knee or hip replacement: a retrospective study with a 1-year follow up.

OBJECTIVES: Patients who undergo elective joint replacement are traditionally screened and treated for preoperative bacteriuria to prevent periprosthetic joint infection (PJI). More recently, this practice has been questioned. The purpose of this study was to determine whether preoperative bacteriuria is associated with an increased risk of PJI. METHODS: Patients who had undergone a primary hip or knee replacement in a tertiary care hospital between September 2002 and December 2013 were identified from the hospital database (23 171 joint replacements, 10 200 hips, and 12 971 knees). The results of urine cultures taken within 90 days before the operation were obtained. Patients with subsequent PJI or superficial wound infection in a 1-year follow-up period were identified based on prospective infection surveillance. The association between bacteriuria and PJI was examined using a multivariable logistic regression model that included information on the operated joint, age, gender and the patients' chronic diseases. RESULTS: The incidence of PJI was 0.68% (n = 158). Preoperative bacteriuria was not associated with an increased risk of PJI either in the univariate (0.51% versus 0.71%, OR 0.72, 95% CI 0.34-1.54) or in the multivariable (OR 0.82, 95% CI 0.38-1.77) analysis. There were no cases where PJI was caused by a pathogen identified in the preoperative urine culture. Results were similar for superficial infections. CONCLUSIONS: There was no association between preoperative bacteriuria and postoperative surgical site infection. Based on these results, it seems that the preoperative screening and treatment of asymptomatic bacteriuria is not required.

The Plasma Level of Proprotein Convertase FURIN in Patients with Suspected Infection in the Emergency Room: A Prospective Cohort Study.

There is an increasing need for novel biomarkers that enable better diagnostic and prognostic stratification of patients with suspected infection. A proprotein convertase enzyme FURIN is upregulated upon immune cell activation, and it promotes infectivity by cleaving and activating pathogens. In this study, we determined FURIN levels in plasma using ELISA from 537 patients that were admitted to emergency room with suspected infection. Patients were sorted to high- and low-level FURIN groups with a cut-off level of 370 pg/ml. The study cohort included five diagnostic groups: Group 1, no systemic inflammatory response syndrome (SIRS, n = 59 patients); Group 2, bacterial infection without SIRS (n = 67); Group 3, SIRS, but no bacterial infection (n = 308); Group 4, sepsis without organ failure (n = 308); and Group 5, severe sepsis (n = 49). Statistically significant associations were not found between the plasma level of FURIN and the prevalence of sepsis (P = 0.957), diagnostic group of a patient (P = 0.737) or the bacteria in blood culture (P = 0.499). Additionally, the concentration of FURIN did not predict the severity or case fatality of the infectious disease. However, statistically significant associations were found between high plasma level of FURIN and diagnosed rheumatic disease (P < 0.001) as well as with the prevalence of non-smokers (P = 0.034). Thus, albeit the plasma level of FURIN does not predict the severity of infectious disease, it may be of use in the diagnostics of autoimmune diseases.

Evidence of streptococcal origin of acute non-necrotising cellulitis: a serological study.

Bacteriological diagnosis is rarely achieved in acute cellulitis. Beta-haemolytic streptococci and Staphylococcus aureus are considered the main pathogens. The role of the latter is, however, unclear in cases of non-suppurative cellulitis. We conducted a serological study to investigate the bacterial aetiology of acute non-necrotising cellulitis. Anti-streptolysin O (ASO), anti-deoxyribonuclease B (ADN) and anti-staphylolysin (ASTA) titres were measured from acute and convalescent phase sera of 77 patients hospitalised because of acute bacterial non-necrotising cellulitis and from the serum samples of 89 control subjects matched for age and sex. Antibiotic treatment decisions were also reviewed. Streptococcal serology was positive in 53 (69%) of the 77 cases. Furthermore, ten cases without serological evidence of streptococcal infection were successfully treated with penicillin. Positive ASO and ADN titres were detected in ten (11%) and three (3%) of the 89 controls, respectively, and ASTA was elevated in three patients and 11 controls. Our findings suggest that acute non-necrotising cellulitis without pus formation is mostly of streptococcal origin and that penicillin can be used as the first-line therapy for most patients.

Healthcare workers as vectors of infectious diseases.

Nosocomial infections cause considerable morbidity and mortality. Healthcare workers (HCWs) may serve as vectors of many infectious diseases, many of which are not often primarily considered as healthcare-associated. The probability of pathogen transmission to patients depends on several factors, such as the characteristics of a pathogen, HCW and patient. Pathogens with high transmission potential from HCWs to patients include norovirus, respiratory infections, measles and influenza. In contrast, human immunodeficiency virus (HIV) and viral hepatitis are unlikely to be transferred. The prevention of HCW-associated transmission of pathogens include systematic vaccinations towards preventable diseases, continuous education, hand hygiene surveillance, active feedback and adequate staff resources.

Urine soluble urokinase-type plasminogen activator receptor levels correlate with proteinuria in Puumala hantavirus infection.

OBJECTIVES: Urokinase-type plasminogen activator receptor (uPAR) is upregulated during inflammation and known to bind to ß3 -integrins, receptors used by pathogenic hantaviruses to enter endothelial cells. It has been proposed that soluble uPAR (suPAR) is a circulating factor that causes focal segmental glomerulosclerosis and proteinuria by activating ß3 -integrin in kidney podocytes. Proteinuria is also a characteristic feature of hantavirus infections. The aim of this study was to evaluate the relation between urine suPAR levels and disease severity in acute Puumala hantavirus (PUUV) infection. DESIGN: A single-centre, prospective cohort study. SUBJECTS AND METHODS: Urinary suPAR levels were measured twice during the acute phase and once during convalescence in 36 patients with serologically confirmed PUUV infection. Fractional excretion of suPAR (FE suPAR) and of albumin (FE alb) was calculated. RESULTS: The FE suPAR was significantly elevated during the acute phase of PUUV infection compared to the convalescent phase (median 3.2%, range 0.8-52.0%, vs. median 1.9%, range 1.0-5.8%, P = 0.005). Maximum FE suPAR was correlated markedly with maximum FE alb (r = 0.812, P < 0.001) and with several other variables that reflect disease severity. There was a positive correlation with the length of hospitalization (r = 0.455, P = 0.009) and maximum plasma creatinine level (r = 0.780, P < 0.001) and an inverse correlation with minimum urinary output (r = -0.411, P = 0.030). There was no correlation between FE suPAR and plasma suPAR (r = 0.180, P = 0.324). CONCLUSION: Urinary suPAR is markedly increased during acute PUUV infection and is correlated with proteinuria. High urine suPAR level may reflect local production of suPAR in the kidney during the acute infection.

Obesity and nosocomial infections.

BACKGROUND: The prevention of healthcare-associated infections (HCAIs) is a major goal in modern healthcare. Intrinsic, patient-related factors may contribute to the risk of HCAIs. AIM: To review the association between obesity and the risk and outcome of HCAIs. METHODS: A PubMed search of relevant studies on obesity and nosocomial infections and obesity and dosing of antimicrobials. Search terms were: 'obesity', 'infection', 'nosocomial infection', 'surgical site infection', 'critical care unit', 'bacteremia', 'urinary tract infection', 'health care associated infection'. FINDINGS: Obesity has been shown to be associated with an increased risk of HCAIs in several studies. The association is most clear in cardiac, vascular, orthopaedic and gastrointestinal surgery. Body mass index (BMI) data are frequently recorded in patients undergoing surgical and invasive procedures. The recording of BMI data is not systematic in the literature and in many studies median BMI of the control group or reference group (normal weight) also indicates overweight or obesity. Thus, clear BMI cut-offs for increased infection risk cannot be determined. Obesity is frequently associated with underdosing of antimicrobials in both prophylaxis and treatment of HCAIs. Studies indicate that obesity affects the pharmacokinetics of antimicrobial drugs. However, there are no dosing recommendations for antimicrobial use in obesity. CONCLUSIONS: Obesity increases the risk of nosocomial infections and is frequently associated with underdosing of antimicrobials in both prophylaxis and treatment of HCAIs. A challenge in future hospital hygiene prevention lies in our capacity to combat obesity epidemics.

Obesity and the risk and outcome of infection.

The interactions between obesity and infectious diseases have recently received increasing recognition as emerging data have indicated an association between obesity and poor outcome in pandemic H1N1 influenza infection. Obesity is an established risk factor for surgical-site infections, nosocomial infections, periodontitis and skin infections. Several studies indicate that acute pancreatitis is more severe in the obese. Data are controversial and limited as regards the association between obesity and the risk and outcome of community-acquired infections such as pneumonia, bacteremia and sepsis and obesity and the course of HIV infection. As the cause-effect relationship between obesity and infection remains obscure in many infectious diseases, further studies are warranted. The consequences of obesity may have substantial effects on the global burden of infectious diseases.

Soluble urokinase-type plasminogen activator receptor in patients with suspected infection in the emergency room: a prospective cohort study.

OBJECTIVES: The soluble form of urokinase-type plasminogen activator (suPAR) was evaluated as an early prognostic marker of sepsis in patients with suspected infection. DESIGN: A single-centre prospective cohort study. METHODS: The cohort comprised 539 patients in the emergency department with suspected infection: 59 without systemic inflammatory response syndrome (SIRS) and without bacterial infection (group 1), 68 with bacterial infection and without SIRS (group 2), 54 with SIRS and without bacterial infection (group 3), 309 with sepsis (SIRS and bacterial infection) and without organ failure (group 4) and 49 with severe sepsis (SIRS, bacterial infection and organ failure) (group 5). suPAR was measured on admission using a commercial solid-phase enzyme-linked immunosorbent assay. RESULTS: The median soluble form of the receptor (suPAR) concentrations in groups 1-5 were 4.7, 5.0, 4.4, 4.8 and 7.9 ng mL(-1) , respectively (P < 0.001). The levels were significantly higher in nonsurvivors compared with survivors (8.3 vs. 4.9 ng mL(-1) , P < 0.001) and in patients with severe sepsis (group 5) compared with those in the other groups (7.9 vs. 4.8 ng mL(-1) , P < 0.001). Area under the receiver operating characteristics curve (AUC(ROC) ) for the prediction of case fatality was 0.79 (95% confidence interval [CI]: 0.72-0.86, P < 0.0001) and 0.75 for severe sepsis (95% CI: 0.68-0.81, P < 0.0001). At a cut-off level of 6.4 ng mL(-1) , suPAR had 76% sensitivity and 69% specificity for fatal disease; at a cut-off level of 6.6 ng mL(-1) , the sensitivity and specificity for severe sepsis were 67% and 72%, respectively. In multivariate models, high suPAR remained an independent predictor of case fatality and severe sepsis after adjusting for potential confounders. CONCLUSIONS: A high suPAR level predicts case fatality and severe sepsis in patients with suspected infection.

Meticillin-resistant Staphylococcus aureus (MRSA) bacteraemia in Tampere University Hospital: a case-control study, Finland October 2002 to January 2010.

Meticillin-resistant Staphylococcus aureus (MRSA) has emerged as a frequent pathogen in blood cultures in Pirkanmaa Hospital District (HD), Finland. To study risk factors for MRSA bacteraemia and the adequacy of empirical antimicrobial treatment, we retrospectively reviewed the hospital records of 102 patients, 51 with MRSA, and 51 with meticillin-sensitive Staphylococcus aureus (MSSA) bacteraemias respectively, who had been admitted to Tampere University Hospital in Pirkanmaa HD, from October 2002 to January 2010. For each patient with MRSA bacteraemia, one consecutively detected unmatched patient with MSSA bacteraemia was chosen as control. Patients with MRSA bacteraemias were significantly older (median age: 73 years vs 59 years, p=0.001), were more likely to have been transferred directly from another healthcare facility or were already in the hospital at the onset of bacteraemia (39/51 vs 26/51, p=0.007) and had a higher McCabe class than patients with MSSA bacteraemia (p=0.005). Patients with MRSA bacteraemia more seldom received adequate empirical antimicrobial therapy when compared to those with MSSA bacteraemia (13/51 vs 43/51, p<0.001). Of previously known MRSA carriers 10 of 29 received adequate empirical antimicrobial therapy for their condition. The percentage of MRSA bacteraemias among all S. aureus bacteraemias in Pirkanmaa HD is high compared to corresponding figures for the whole of Finland.

Plasma level of soluble urokinase-type plasminogen activator receptor as a predictor of disease severity and case fatality in patients with bacteraemia: a prospective cohort study.

OBJECTIVES: Urokinase-type plasminogen activator receptor (uPAR) is expressed on a variety of different immune cells and vascular endothelial cells during inflammation. Previous studies indicate that a high plasma concentration of the soluble form of the receptor (suPAR) predicts poor outcome in infectious diseases. DESIGN: A prospective cohort study. SUBJECTS AND METHODS: Plasma suPAR levels were measured in 132 patients with bacteraemia caused by Staphylococcus aureus, Streptococcus pneumoniae, ß-haemolytic streptococcae or Escherichia coli using a commercial enzyme-linked immunosorbent assay (ELISA). Values were measured on days 1-4 after a positive blood culture, on days 13-18 and on recovery. RESULTS: The maximum suPAR values on days 1-4 were markedly higher in nonsurvivors compared to survivors (15.8 vs. 7.3 ng mL(-1) , P < 0.001) and the area under the receiver operating characteristic curve (AUC(ROC) ) in the prediction of case fatality was 0.84 (95% confidence interval (CI) 0.76-0.93, P < 0.001). At a cut-off level of 11.0 ng mL(-1) , the sensitivity and specificity of suPAR for fatal disease was 83% and 76%, respectively. A high level of suPAR (≥ 11 ng mL(-1) ) was associated with hypotension (mean arterial pressure < 70 mmHg) (odds ratio (OR) 6.5; 95% CI 2.9-14.6) and high sequential organ failure assessment score (≥ 4) (OR 9.3; 95% CI 4.0-21.9). A high suPAR level remained an independent risk factor for case fatality in a logistic regression model adjusted for potential confounders. CONCLUSION: Plasma suPAR level is a sensitive and specific independent prognostic biomarker in patients with bacteraemia.

Smoking and the outcome of infection.

Smoking has substantial local and systemic adverse effects on the immune system, respiratory tract and skin and soft tissues. Smokers are at increased risk of invasive pneumococcal disease, pneumonia, periodontitis, surgical infections, tuberculosis, influenza and meningococcal disease. The results of several studies indicate that smokers with periodontitis or tuberculosis suffer more severe disease. Data on the impact of smoking on sepsis and pneumonia are controversial and limited, and systematic data regarding the outcome of the majority of infections in smokers are scarce. Abundant data indicate that children exposed to environmental tobacco smoke (ETS) suffer from more severe infections. However, information regarding the effects of ETS on the outcome of infections in adults is limited. Various aspects of the relation between smoking and the outcome of bacterial infection (e.g. potential dose-dependent effects and the interactions between smoking and other environmental factors that may affect the course of infectious diseases) remain to be established.

Long-term follow-up of a randomized clinical trial of open versus laparoscopic appendicectomy.

BACKGROUND: The aim of this study was to assess long-term outcome following open versus laparoscopic appendicectomy. METHODS: A total of 105 patients with suspected acute appendicitis were randomized to LA (51) or OA (54) between 1997 and 1999 at one hospital. Perioperative factors and follow-up data from the outpatient clinic were recorded. Information about symptoms and overall satisfaction was obtained by telephone interview. In addition, appendicectomy data for 2008 were analysed retrospectively for comparison in a contemporary setting. RESULTS: Data from 52 patients who had OA and 47 who had LA were analysed. OA was performed mostly by trainees, but LA was more likely to be undertaken by a consultant. The open procedure was quicker than the laparoscopic operation in the trial period (median 38 versus 65 min respectively; P < 0.001), but the difference was only 10 min in 2008. The OA group returned to work later than the LA group (median 13 versus 8 days; P = 0.013) and had more complications (22 versus 6; P = 0.014). Only one patient (OA) had a reoperation, owing to abdominal adhesions. Among 76 patients available for telephone interview, satisfaction scores were marginally higher for LA than OA. CONCLUSION: LA has some advantages compared with an open approach. REGISTRATION NUMBER: NCT00908804 (http://www.clinicaltrials.gov).

Gene-environment interaction between MBL2 genotype and smoking, and the risk of gram-positive bacteraemia.

Mannose-binding lectin (MBL) insufficiency caused by point mutations in the MBL2 gene has been associated with increased susceptibility to bacteraemic infections. We here investigated the effect of MBL2 polymorphisms on the susceptibility and clinical course of bacteraemia. The study cohort comprised 145 patients with bacteraemia and 400 controls. In the case of patients with bacteraemia, laboratory findings and clinical data were registered on admission and during six consecutive days. MBL2 structural polymorphisms at codons 52 (CGT-->TGT; designated D or O), 54 (GGC-->GAC; B or O) and 57 (GGA-->GAA; C or O) in exon 1 of the MBL2 gene and promoter region polymorphisms at position -221 (G-->C, designated Y or X alleles) were determined. No difference in MBL2 genotype frequencies between the bacteraemic patients and controls was detected, and MBL2 genotype had no independent effect on mortality, nor disease severity. However, smoking proved a significant risk factor for Gram-positive (Staphylococcus aureus, Streptococcus pneumoniae or beta-haemolytic streptococci) bacteraemia in patients carrying the variant O allele (53% current smokers in Gram-positive bacteraemia patients compared with only 21% in controls, odds ratios 4.2, 95% confidence intervals 2.0-9.0; P < 0.001), while it did not have an effect in those homozygous for the A allele. The same effect was not detected in Escherichia coli bacteraemia. In conclusion, MBL2 genotypes representing MBL insufficiency were not associated with the overall risk of bacteraemia or disease severity, but smoking in carriers of the structural variant O allele may have a deleterious effect increasing the risk of Gram-positive bacteraemia.

Characterization of a distinct group of slowly growing mycobacteria by biochemical tests and lipid analyses.

A group of slowly growing mycobacterial strains (n = 14) isolated from respiratory tract specimens was collected from 1971 to 1990 on the basis of growth characteristics and uncommon biochemical and glycolipid profiles. Growth at 25 to 45 degrees C, a negative Tween 80 hydrolysis test, a strong positive reaction in a 14-day arylsulfatase test, and susceptibility to ethambutol in combination with resistance to cycloserine were important for the initial separation. The strains had a distinctive glycolipid pattern which was unlike those of other mycobacterial species. Analyses of cellular fatty acids by gas-liquid chromatography and mycolic acids by thin-layer chromatography further characterized this homogeneous group of mycobacteria. The presence of 2-eicosanol (2-OH-20:0alc) and hexacosanoic acid (26:0) combined with the lack of 2-docosanol (2-OH-22:0alc) differentiated this group from other slowly growing mycobacteria.

Injection sclerotherapy for esophageal varices.

Thirty-five consecutive patients with bleeding esophageal varices were treated by repeated endoscopic injection sclerotherapy. During each session the varices were injected with 14 +/- 4.2 ml (mean +/- SD) of 5% ethanolamine oleate submucosally or intravariceally. The varices were obliterated in 31 (89%) patients. On average 3.3 +/- 2.4 sclerotherapy sessions were required for eradication of the varices. Mild fever was noticed almost in every patient after sclerotherapy. Mediastinitis was a complication in one (2.8%) patient. Esophageal stricture ensued in two (5.7%) patients which did not require treatment. The cumulative survival rates at 1, 2, 3, 4 and 5 years were 83%; 65%; 52%; 52% and 47% respectively. The corresponding 95% confidence intervals were (0.7, 0.96); (0.48, 0.8); (0.34, 0.7); (0.3, 0.74) and (0.22, 0.7). Sclerotherapy is an effective and safe method to treat bleeding esophageal varices.

Piperacillin compared with cefuroxime plus metronidazole in diffuse peritonitis.

Eighty-five patients were randomly allocated to receive either piperacillin (n = 38) or cefuroxime plus metronidazole (n = 45) after surgical treatment of diffuse peritonitis; 78 were evaluable. A mean of 1.5 (piperacillin group) and 1.7 (cefuroxime/metronidazole group) pathogens/patient were identified. Twenty-seven patients (71%) were successfully treated in the piperacillin group compared with 29 (64%) in the cefuroxime/metronidazole group. These data suggest that piperacillin was neither better nor worse than cefuroxime/metronidazole in diffuse, secondary peritonitis.