Systematic review: impact of liver transplantation on employment.

Background: The majority of liver transplant recipients survive long term after the procedure. Aim: To assess if this positive outcome is associated with improved employment post-transplant. Methods: A systematic review of publications between 2001 and 2016 was performed. A standard procedure was used to search for suitable publications from two databases (PubMed and EMBASE). Duplicates were removed and abstracts screened by both authors for possible inclusion. Possible suitable publications were obtained and examined for the presence of pre- and post-employment information. Full articles that had this information were reviewed by standard methodology for assessment of bias. Results: A total of 162 individual abstracts were screened. Thirty-five full papers were reviewed and 13 papers included in the detailed review. Risk of bias was considered high due to low response rates, poor assessment of prognostic and confounding factors and varying definitions of employment. Heterogeneous data precluded meta-analysis. Eight studies focused on return to work as a primary outcome and five on quality of life with employment as a secondary outcome. Follow-up varied between 2 and 13 years. Rates of employment fell in all studies assessed. Employment rates ranged from 26 to 80% pre-transplant and 18 to 44% post-transplant. The proportion of those categorized as ill-health retired was 24% greater after orthotopic liver transplantation. Conclusions: Improved survival after liver transplantation was not reflected in a return to employment and retirement was common. Areas for further study include interventions to minimize physical deconditioning, depression associated with lower employment rates and type of work available after transplant.

A review of occupational disease surveillance systems in Modernet countries.

BACKGROUND: To improve occupational health public policies and to facilitate coordinated research within the European Union to reduce the incidence of occupational diseases (ODs), it is important to know what OD surveillance systems exist and how they compare. Monitoring trends in occupational diseases and tracing new and emerging risks in a network (Modernet) participants are well placed to provide this information as most either contribute data to and/or are involved in the management of OD systems. AIMS: To identify and describe OD surveillance systems in Modernet countries with the longer-term objective of identifying a core template to be used on a large scale. METHODS: A questionnaire sent to Modernet participants, seeking structured information about the OD surveillance system(s) in their country. RESULTS: Overall 14 countries (70%) provided information for 33 OD systems, among them 11 compensation-based (CB) systems. Six countries provided information for non-CB systems reporting for any type of OD. The other systems reported either only ODs from a prescribed list, or specific diagnoses or diagnostic groups, with reports to most schemes being physician-based. Data collected varied but all systems collected diagnosis, age, gender, date reported and occupation (and/or industry) and most collected information on exposure. CONCLUSIONS: This review provides information beneficial to both policy makers and researchers by identifying data sources useable to measure OD trends in European countries and opening the way to future work, both on trend comparisons within Europe and on the definition of a core template to extend OD surveillance on a larger scale.

Work-related ill-health: Republic of Ireland, Northern Ireland, Great Britain 2005-2012.

BACKGROUND: Data on work-related ill-health (WRIH) in the Republic of Ireland is inconsistent. AIMS: To compare the incidence of WRIH in the Republic of Ireland (ROI), Northern Ireland (NI) and Great Britain (GB) reported by clinical specialists in skin and respiratory medicine and by specialist occupational physicians (OPs). METHODS: Analysis of data reported to three surveillance schemes in The Health and Occupation Research (THOR) network in ROI and corresponding UK schemes. RESULTS: Contact dermatitis was the most frequently reported skin disease in the three areas. Asthma was the most frequently-reported respiratory disease in the ROI, while asbestos-related cases predominate in GB and NI. Mental health disorders, followed by musculoskeletal disorders were reported most frequently by OPs. Annual average incidence rates for skin disease were 2 per 100000 employed (95% confidence interval [CI] 1.9-2.8) in the ROI and 7 per 100000 for GB (95% CI 4.8-9.4). Unadjusted incidence rates for respiratory disease were 1 (95% CI 0.3-1) and 8 (95% CI 6.1-10.7) per 100000 in the ROI and GB, respectively; adjusted for reporter non-response, these figures increased to 15 (95% CI 11.3-19.6) and 32 (95% CI 28.4-35.6) per 100000 respectively. CONCLUSIONS: This is the first paper to include THOR data on WRIH from the ROI, NI and GB. Consistent and dedicated data collection in the ROI via the THOR schemes is viable and important in the light of a deficit of occupational ill-health data. Sustained efforts to improve participation are underway.

Sodium channel blockers for cystic fibrosis.

BACKGROUND: People with cystic fibrosis (CF) have increased transport of the salt, sodium across their airway lining. Over-absorption of sodium results in the dehydration of the liquid that lines the airway surface and is a primary defect in people with CF. OBJECTIVES: To determine whether the topical administration of drugs that block sodium transport improves the respiratory condition of people with CF. SEARCH STRATEGY: We searched the Cochrane Cystic Fibrosis and Genetic Disorders Group Trials Register comprising references identified from comprehensive electronic database searches, handsearching relevant journals and abstract books of conference proceedings. We contacted principal investigators known to work in the field, previous authors and pharmaceutical companies who manufacture ion transport agents for unpublished or follow-up data. Most recent search of the Group's register: March 2006 SELECTION CRITERIA: Published or unpublished randomised controlled trials (RCTs) or quasi-randomised controlled trials of sodium channel blockers compared to placebo or another sodium channel blocker or the same sodium channel blocker at a different dosing regimen. DATA COLLECTION AND ANALYSIS: Two authors independently extracted data. Meta-analysis was limited due to differing study designs. MAIN RESULTS: Four RCTs, with a total of 205 participants, examining the topical administration of the short-acting sodium channel blocker, amiloride, compared to placebo were identified as eligible for inclusion in the review. For three studies, interventions for six months were completed and it was possible to calculate relative change in respiratory function (FVC). There was a significant difference found in relative change in FVC in favour of placebo (GIV analysis of weighted mean difference for FVC; 1.51% (95% confidence interval -2.77 to -0.25). There were no significant differences identified in other clinically relevant outcomes. AUTHORS' CONCLUSIONS: We found no evidence that the topical administration of a short-acting sodium channel blocker improves respiratory condition in people with cystic fibrosis and some limited evidence of deterioration in lung function.

Bronchiolitis obliterans organizing pneumonia in a patient with non-Hodgkin's lymphoma following R-CHOP and pegylated filgrastim.

Bronchiolitis obliterans organizing pneumonia (BOOP) presents with fever, dyspnoea, pleuritic chest pain and hypoxia. The diagnosis can be made from radiological appearances on chest radiograph and CT scan correlated with histological findings following biopsy. We present a 52-year-old gentleman undergoing treatment for high grade non-Hodgkin's lymphoma who developed respiratory symptoms during chemotherapy. BOOP was diagnosed and he responded well to oral prednisolone. The cause of BOOP is often not certain. However, in this case we suspect pegylated filgrastim or rituximab as possible agents.

A double-blind, placebo controlled, dose ranging study to evaluate the safety and biological efficacy of the lipid-DNA complex GR213487B in the nasal epithelium of adult patients with cystic fibrosis.

UNLABELLED: GTAB1001: A Double-Blind, Placebo Controlled, Dose Ranging Study to Evaluate the Safety and Biological Efficacy of the Lipid-DNA Complex GR213487B in the Nasal Epithelium of Adult Patients with Cystic Fibrosis. OBJECTIVES: To evaluate the effectiveness of various dosages of the lipid-DNA complex GR213487B (0.4375mg and either 4.0mg or 0.0625mg) for producing CFTR gene transfer and correcting the chloride ion transport defect in the nasal epithelium of patients with cystic fibrosis. To assess the safety and tolerability of the lipid-DNA complex GR213487B when applied to the nasal epithelium of patients with cystic fibrosis. DESIGN: Single-center, double-blind, placebo controlled, dose ranging study. DURATION: Pre-treatment evaluations will be performed during two outpatient study visits (ie. between Day -7 to -3 and at Day -2). Patients will be admitted to the Clinical Research Unit (CRU) at the University of North Carolina at Chapel Hill on Day -1 for additional pre-treatment evaluations performed the day prior to administration of double-blind treatment (ie. gene transfer) on Treatment Day 0. Patients will remain in the CRU for 7 days (Day -1 to Day 6) and will be discharged on Day 6. Patients will subsequently be followed on an outpatient basis but will return for another assessment between Days 9-11, and may also return to the CRU for two optional study visits on Days 14 and 21. All patients will return to the CRU on an out-patient basis for follow-up evaluations on Day 28 +/- 3. SETTING: Patients will receive in-patient treatment in the CRU at the University of North Carolina at Chapel Hill and will remain in the CRU for 7 days. PATIENTS: A target enrollment of 12 evaluable patients is planned. STUDY TREATMENTS: Patients who meet all entry criteria will complete pre-treatment assessments, which will take place between Day -7 to Day -1, and will serve as a baseline for specific evaluations and to ensure clinical stability. Patients will return on Day -1 for admission to the CRU the day prior to gene transfer. Each nostril of the patients will be randomly assigned in a double blind manner to receive either GR213487B liquid nasal spray or the lipid alone (ie. control administered as liposome), by topical application directed at the inferior turbinate. The first four patients will receive an initial dosage of GR213487B containing 0.4375 mg of DNA. The decision to proceed to administer a higher dose (ie. 4.0mg DNA) or a lower dose (ie. 0.0625mg DNA) in the subsequent eight patients will be determined by the Principal Investigator in association with an FDA officer serving as an independent Clinical Ombudsman, according to the study plan (see Section 5.5 and Appendix 3-Dosing Flow Chart). MEASUREMENTS: Efficacy Evaluations The primary variables to determine the efficacy of transgene expression will be: * Evidence of vector derived CFTR (cystic fibrosis transmembrane conductance regulator) mRNA, as measured by reverse transcriptase polymerase chain reaction (RT-PCR) in nasal epithelial cells obtained from nasal scrapes on Day 3 and, nasal biopsies on Day 5, if sufficient tissue is available. * Correction of chloride ion transport across the nasal epithelium as measured by the transepithelial electrical potential difference (TEPD). The baseline TEPD will initially be measured, and again subsequently following perfusion of: --zero chloride perfusion containing amiloride (to induce chloride secretion) --zero chloride perfusion containing amiloride and isoproterenol (to increase cAMP-mediated chloride secretion) Secondary measures to determine the efficacy of gene transfer will be: * Evidence of delivery of plasmid DNA in the nasal lavage (Day 1-5, Day 9-11 and Day 28) * Evidence of vector derived CFTR mRNA from nasal scrapes performed after the nasal biopsy (ie. Day 9-11 and/or Day 28) * Percentage of cells from nasal biopsies expressing vector derived CFTR mRNA as measured by in situ hybridization * Evidence of vector derived CFTR

Prospective study of amikacin versus netilmicin in the treatment of severe infection in hospitalized patients.

PURPOSE: Four previous studies comparing netilmicin and amikacin have yielded inconclusive results concerning efficacy and rates of nephrotoxicity and ototoxicity. For this reason, we conducted a prospective, randomized, controlled trial of the two drugs in the treatment of hospitalized patients with severe infection. PATIENTS AND METHODS: A total of 202 patients were enrolled in the study; 100 received netilmicin and 102 received amikacin. Concomitant antimicrobials were restricted to metronidazole and benzylpenicillin. Peak and trough aminoglycoside levels were assayed within the first 36 hours and at least every 72 hours thereafter. A full blood cell count, serum electrolytes, creatinine, bilirubin, and liver enzymes were measured before therapy, weekly thereafter, and within 48 hours after the discontinuation of therapy. Nephrotoxicity and ototoxicity were assessed in patients. A standard agar dilution procedure was used to determine minimal inhibitory concentrations. RESULTS: No significant pretreatment differences were found between the two groups. Patients in the amikacin group responded significantly better to treatment than did patients in the netilmicin group (90% versus 79%; p less than 0.05). A notable finding was the markedly inferior response rate of Pseudomonas aeruginosa infections to netilmicin as compared with amikacin (13 of 24 with a favorable response compared with 25 of 26). No significant difference in ototoxicity was found, whereas nephrotoxicity appeared to be significantly less with amikacin (4% versus 12%, p less than 0.05). Although amikacin seemed less nephrotoxic than netilmicin, this may have been related to the significantly greater number of patients with initial renal dysfunction who received netilmicin. CONCLUSIONS: Amikacin appears to be significantly more efficacious than netilmicin for the treatment of P. aeruginosa infections, especially those in non-urinary tract sites. There is no apparent difference between the two drugs in terms of ototoxicity.

Sisomicin, netilmicin and dibekacin. A review of their antibacterial activity and therapeutic use.

Sisomicin is a naturally occurring aminoglycoside antibiotic produced by Micromonospora inyoensis, while dibekacin and netilmicin are both semisynthetic aminoglycosides. Dibekacin is 3',4'-dideoxykanamycin B and netilmicin is 1-N-ethyl sisomicin. In both cases, these modifications render the agents insusceptible to some of the enzymes found in resistant strains of bacteria which inactivate the parent compounds. Antibacterial activity: All 3 drugs show bactericidal activity against a wide range of Gram-negative bacteria (including E. coli, Enterobacter, Klebsiella and Proteus spp. and Ps. aeruginosa) and also against staphylococci; however, in common with other amino-glycosides, streptococci are usually resistant (except when beta-lactam antibiotics are used in combination) and anaerobic organisms are not sensitive. Sisomicin is closely related structurally to gentamicin Cla, but in vitro studies have shown it to have superior activity to gentamicin against Ps. aeruginosa, closely paralleling the activity of tobramycin, while still possessing the high activity of gentamicin against Serratia and other Gram-negative rods. However, sisomicin is inactivated by virtually all bacterial enzymes which inactivate gentamicin and tobramycin. Nevertheless, it retains useful activity against a number of gentamicin-resistant strains of Ps. aeruginosa which are resistant by non-enzymatic (possibly permeability barrier) mechanisms; in this respect it closely resembles tobramycin. Dibekacin closely resembles tobramycin structurally and in vitro it seems to have a very similar antibacterial spectrum, including activity against some strains of Ps. aeruginosa resistant to gentamicin. Netilmicin has a generally broader antibacterial spectrum than gentamicin, tobramycin, sisomicin or debekacin and is resistant to inactivation by phosphorylating and adenylylating enzymes; however, it is inactivated by all acetylases, apart from acetylase 3-I. Its spectrum is therefore not as wide as that of amikacin against 'gentamicin-resistant' strains. Nonetheless, it is intrinsically more active than amikacin, weight-for-weight, against sensitive strains, apart possibly from Ps. aeruginosa. In fact, its activity against species of the Enterobacteriaceae and staphylococci sensitive to gentamicin is of the same order as the latter and possibly better for Klebsiella-Enterobacter species. All 3 agents show marked antibacterial synergy with a variety of beta-lactam antibiotics against a range of bacteria. Pharmacokinetically, sisomicin, netilmicin and dibekacin all behave like gentamicin. All 3 drugs are excreted in the urine unchanged and have beta-phase elimination half-lives of around 2 to

Discordant organ laterality in monozygotic twins with primary ciliary dyskinesia.

Primary ciliary dyskinesia (PCD) is a genetic disease characterized by abnormal ciliary structure and function, impaired mucociliary clearance, and chronic middle ear, sinus, and lung disease. PCD is associated with situs inversus in approximately 50% of the patients. One proposed explanation for this relationship is that normal ciliary function plays a role in normal organ orientation, whereas organ orientation in PCD is a random event because of dysfunctional cilia in early embryonic development. Another hypothesis for the association between PCD and situs inversus is that mutated genes in PCD not only cause defective cilia, but are also linked to the control of organ laterality, such that abnormalities in this molecular pathway result in random left-right asymmetry. We report on a set of monozygotic twin women with PCD. In both patients, deficiency of the inner dynein arms was noted on ciliary ultrastructural analysis, associated with a clinical syndrome of bronchiectasis, chronic sinusitis, and middle ear disease. One of the twins has situs solitus, the other has situs inversus totalis. DNA analysis confirmed that the twins are monozygotic. This is consistent with the hypothesis that situs inversus occurring in patients with primary ciliary dyskinesia is a random but "complete" event in the fetal development of patients with PCD.

Airway deposition and clearance and systemic pharmacokinetics of amiloride following aerosolization with an ultrasonic nebulizer to normal airways.

STUDY OBJECTIVES: Airway epithelial ion transport is an important component of the airway defense mechanism, and new therapies that target ion transport are being developed. Amiloride is an example of such a new drug, exerting a dose-dependent action to inhibit Na+ transport. Amiloride may be useful in cystic fibrosis, blocking the characteristic airway epithelial Na+ hyperabsorption that occurs in the disease. To evaluate airway and systemic delivery of amiloride via an ultrasonic nebulizer (Omron NE-UO7), we measured the airway surface concentrations of amiloride in normal volunteers via a novel approach, together with the systemic pharmacokinetics of amiloride. DESIGN: Direct measurement of airway surface liquid, plasma, and urine amiloride concentrations following ultrasonic nebulization. PARTICIPANTS/INTERVENTIONS: Seven normal subjects were studied in the General Clinical Research Center of the University of North Carolina. Following inhalation with amiloride (1 mg/mL, 4.5 mL) for approximately 12 min, a bronchoscopy was performed. Amiloride deposition and clearance from airway surfaces over 1 h were evaluated by transbronchoscopic sampling using preweighed filter papers. Pulmonary and systemic absorption was assessed by measuring drug concentrations in blood and urine. RESULTS: The mean volume aerosolized was 3.5+/-0.3 mL during 12 min of aerosolization time; the mean initial concentration of amiloride on airway surfaces after nebulization was 1.6 x 10(-4) mol/L, with an elimination half life of approximately 23 min. Peak plasma concentrations of amiloride (30 min, 3.36+/-0.70 ng/mL) suggest early absorption across lung surfaces, rather than via the GI route. Mean urinary excretion of amiloride over 72 h was 0.63+/-0.07 mg, with 87% excreted in the first 24 h. CONCLUSIONS: The ultrasonic nebulizer rapidly delivers amiloride to normal conducting airways as assessed by the transbronchoscopic sampling technique. Early blood concentrations of amiloride probably reflect initial absorption across lung surfaces and are a useful index of the efficiency of the machine.

A rapid assay method for cephalosporins.

A simple, rapid assay for cephalosporins is described. The method is based on the inhibition by cephalosporins of the fermentation of glucose or inositol by a strain of Providence resistant to aminoglycoside antibiotics. The method gives answers which are as accurate as those obtained by standard agar diffusion techniques within four hours, and utilizes skills and resources readily available in most routine bacteriology departments. Results are not affected by gentamicin or kanamycin concurrently administered to the patient. This method will be of value in helping to monitor cephalosporin therapy in patients with serious sepsis, especially those with impaired renal function, and may help in elucidating and preventing the problem of nephrotoxicity associated with cephalosporin administration.

Klebsiella capsular type versus site of isolation.

More than 1750 clinical isolates of klebsiella were collected over a period of six years from two different hospitals and capsular typed by the fluorescent antibody technique. A correlation was made between type and site of isolation. Many types were found to be associated more frequently with one site, which suggested a predilection of some capsular types for certain sites of infection. The site may also be a factor contributing to the virulence of a particular type. A greater antibiotic resistance was often noted in types isolated from their predominant sites; however, antibiograms were not consistent for a type from a given site. The combination of site specificity, resistance, and another 'virulence factor' may all be involved in the determination of a pathogenic strain.

Controlling infection in a district general hospital.

The way in which hospital-acquired infection has been brought under control over a three-year period in a district general hospital is described. The main success has been achieved in reducing sepsis caused by Staph. aureus, especially methicillin-resistant strains, and Pseudomonas aeruginosa. These reductions were achieved in spite of inadequate ward isolation and operating theatre facilities, and before there was any marked change in patterns of prescribing antibiotics. Our experiences indicate the significant improvements that can be made in controlling nosocomial sepsis even without structural or other major alterations in a hospital, providing that the problem is fully appreciated and the infection control team are concerned enough to act vigorously in influencing their clinical colleagues.

The effect of sepsis rates of closing and cleaning hospital wards.

Sepsis rates were studied in five hospital wards before and after closure for cleaning. Each ward was closed because of an outbreak of Staph. aureus infection caused by a cloxacillin-resistant strain. The study shows that sepsis rates, especially sepsis caused by hospital strains of Staph. aureus, were greatly reduced in the three-month period following re-opening of the ward, provided that patients infected with such organisms were not readmitted to or allowed to remain in the ward.

Pneumonia caused by coliforms and Pseudomonas aeruginosa.

The diagnosis and treatment of 20 hospital patients seen in the past year with proven pneumonia caused by coliforms and Pseudomonas aeruginosa are discussed. Predisposing factors and methods for improving laboratory and clinical diagnosis are analysed, the main problem being to discriminate between genuine pneumonia caused by these organisms and mere contamination of sputum samples resulting from colonization of the upper respiratory tract following broad-spectrum chemotherapy. Overall initial chemotherapy with gentamicin cured 75% (15 out of 20) of the patients in spite of unfavourable underlying pathology. Where gentamicin was given in adequate dosage, which in practice meant that dose which produced peak serum concentrations of 8 mug/ml or more, the cure rate was 91% (11 out of 12). In those patients achieving (measured) peak serum concentrations of less than 8 mug/ml the cure rate was only 33% (4 out of 12). These figures include four patients who failed to respond to doses of gentamicin producing peak concentrations of 5-0-6-0 mug/ml in each case. These patients responded promptly to higher doses (or accumulation), producing peak serum concentrations of 8 mug/ml or more and were then cured within three to five days. Toxicity from gentamicin was not observed in any patient. These results indicate that it is necessary to monitor gentamicin therapy by laboratory assay to ensure adequate dosage and that peak serum concentrations of 8 mug/ml or more are significantly correlated with successful treatment of pneumonia caused by coliforms and Ps. aeruginosa.

A new serotyping method for Klebsiella species: development of the technique.

A new serotyping method for Klebsiella species using indirect immunofluorescence is described. Nonspecific fluorescence has been minimized by carrying out the capsular antigen-antibody reaction at pH 9.0. Commercial antisera have been tested with the 72 antigenic types of Klebsiella, and appropriate dilutions of each pool and specific antisera have been proposed for use in routine typing. Dilutions were chosen to allow strong fluorescence with each type and its specific antiserum and minimal fluorescence with cross reacting antisera. Where the pool antisera gave a weak reaction for one or more of the component types, it is recommended that the specific antisera for these types be added to the pool dilution. The few remaining cross reactions, with the pool and specific antisera in test dilution, are listed in a table. The unique cross reacting patterns of particular types have been found to be useful in identification. Typing Klebsiella by the fluorescent antibody technique is easy to perform and interpret; the results are reproducible, and it is less expensive than the existing capsular swelling method as it is more sensitive and requires less concentrated antisera. This new method of typing should facilitate detailed epidemiological studies of the mode of transmission of Klebsiella species in hospitals and thus allow more effective infection control measures to be instituted.

A new serotyping method for Klebsiella species: evaluation of the technique.

A new indirect fluorescent typing method for Klebsiella species is compared with an established method, capsular swelling. The fluorescent antibody (FA) technique was tested with standards and unknowns, and the results were checked by capsular swelling. Several unknowns were sent away for confirmation of typing, by capsular swelling. The FA method was also tried by a technician in the routine department for blind identification of standards. Fluorescence typing gives close correlation with the established capsular swelling technique but has greater sensitivity; allows more econimical use of expensive antisera; possesses greater objectivity as it requires less operator skill in the reading of results; resolves most of the cross reactions observed with capsular swelling; and has a higher per cent success rate in identification.

Epidemiological study of klebsiella infection in the special care baby unit of a London hospital.

Of the babies admitted to the Special Care Baby Unit of the Royal Free Hospital over 20 months, 10.2% were infected or colonised by klebsiella. The fluorescent antibody technique was used to identify epidemics caused by three strains: capsular type 8 K. aerogenes, type 68 K. oxytoca, or type 13 K. aerogenes, each of which was predominant at a different time, exhibited a difference in virulence, and showed a predilection for different sites of infection. Intestinal colonisation was frequently followed by the presence of sepsis in other sites by the same capsular type. Antibiotic administration led to a higher incidence of klebsiella infection, while the widespread use of compounds containing hexachlorophane could have contributed to skin colonisation and infection by klebsiella. An environmental survey indicated that 1% Hycolin failed to disinfect the incubators, that the babies were the reservoirs of the organisms, and that transmission was due to inadequate hand-washing of nurses and mothers. The mothers were found to have been uninformed of hygienic techniques. They were observed in various practices which could have contributed to the spread of the organism, including contaminating communal areas and handling babies other than their own. It has been recommended that the mothers of premature infants be instructed in the hygienic measures required in dealing with this susceptible population and that the nursing and medical staff be more strict in their own observance of these procedures.

Carriage of Candida species and C albicans biotypes in patients undergoing chemotherapy or bone marrow transplantation for haematological disease.

Six hundred and seventy four yeast isolates obtained from routine microbiological screening of 153 patients with haematological disease were identified and Candida albicans isolates biotyped over nine months to determine longitudinal and cross sectional patterns of yeast colonisation. A yeast microflora persisted in many patients despite the routine prophylactic use of oral antifungal agents. Analysis of the yeast species isolated on a cross sectional basis showed that C albicans accounted for 65% of yeasts isolated from the oral cavity but only 45% of the faecal yeast flora. Longitudinal changes in yeast flora occurred significantly more often in faecal samples than in oral samples and significantly less often in sites colonised with C albicans than in sites colonised with other species. No associations were found between the yeasts isolated and the nature of antifungal prophylaxis used, or the extent of a patient's stay in hospital.