Proton pump inhibitors increase significantly the risk of Clostridium difficile infection in a low-endemicity, non-outbreak hospital setting.

BACKGROUND: Proton pump inhibitors (PPI) have been linked to higher risk of Clostridium difficile infection (CDI). The relevance of this association in hospitals with low disease activity, where an outbreak strain is nondominant, has been assessed in relatively few studies. AIM: To assess the association of PPI and CDI in a setting of low disease activity. METHODS: A retrospective cohort study was conducted at two hospitals. Patients admitted for > or = 7 days receiving antibiotics were included. Demographics, exposure to PPI, antibiotics and other drugs in relation to diagnosis of CDI were assessed by univariate and multivariate analyses. RESULTS: Of 14 719 patients, 149 (1%) first episode CDI were documented; PPI co-exposure increased CDI [1.44 cases/100 patients vs. 0.74 cases/100 non-exposed (OR: 1.96, 95% CI: 1.42-2.72)]. By logistic regression, PPI days (adjusted OR: 1.01 per day, 95% CI: 1.00-1.02), histamine-2 blockers, antidepressants, antibiotic days, exposure to medications, age, admission service and length of admission were significant predictors. CONCLUSIONS: A statistically significant increase in CDI was observed in antibiotic recipients who received PPI, but the absolute risk increase is modest. In settings of with low rates of CDI, the benefit of PPI therapy outweighs the risk of developing CDI. These data support programmes to decrease inappropriate use of PPI in hospitalized patients.

Bloodstream infection-associated sepsis and septic shock in critically ill adults: a population-based study.

BACKGROUND: Few studies have investigated the epidemiology of sepsis and septic shock in a large population and none have been from Canada. The objective of this study was to define the epidemiology of bloodstream infection (BSI)-associated sepsis and septic shock among all critically ill patients in a large, fully integrated health region in Canada. PATIENTS AND METHODS: All critically ill adults admitted to multidisciplinary intensive care units (ICU) in the Calgary Health Region during May 1, 1999 to April 30, 2000 were included. Clinical, microbiologic and outcome information was obtained from regional databases. RESULTS: We surveyed 1981 patients having at least one ICU admission. Systemic inflammatory response syndrome (SIRS) was diagnosed in 92%, BSI-associated sepsis (BSI with SIRS) in 6% and BSI-associated septic shock (BSI with SIRS and hypotension) in 3%; respective hospital mortality rates were 36%, 40% and 49%. The most common BSI etiologies were Staphylococcus aureus, Escherichia coli and Streptococcus species; only one isolate (1%) was highly antibiotic resistant. Independent risk factors for death among patients with SIRS included age (>or= 65), hypothermia (< 35 degrees C), and higher APACHE II and TISS scores. A surgical diagnosis was associated with decreased mortality risk. Neither a positive blood culture nor hypotension at presentation independently predicted death. CONCLUSION: Knowledge of the epidemiology of these syndromes is important for assessing the burden of disease and providing background information for investigating new therapies.

Randomized, double-blind, multicenter trial comparing clinafloxacin with imipenem as empirical monotherapy for febrile granulocytopenic patients.

In a double-blind, multicenter trial, 541 febrile granulocytopenic patients were randomized to receive either intravenous (iv) clinafloxacin (200 mg every 12 h) or i.v. imipenem (500 mg every 6 h) as empirical monotherapy. More baseline pathogens were susceptible to clinafloxacin (259 [99%] of 262 organisms) than to imipenem (253 [95%] of 265; P=.03). Initial favorable clinical response rates for clinafloxacin (88 [32%] of 272 patients) and imipenem (89 [33%] of 269) were similar. After addition of other antimicrobial agents, overall response rates were 259 (95%) of 272 for clinafloxacin and 251 (93%) of 269 for imipenem. During the study, only 13 clinafloxacin (5%) and 18 imipenem (7%) recipients died. Both drugs were generally well tolerated. Drug-related skin rash occurred more often with clinafloxacin (11% vs. 6%; P=.07), whereas nausea (2% vs. 5%; P=.16), Clostridium-difficile-associated diarrhea (3% vs. 8%; P=.02), and seizures (0% vs. 2%; P=.06) occurred more often with imipenem. These results suggest that clinafloxacin and imipenem have similar efficacy as empirical monotherapy in febrile granulocytopenic patients.

Comparison of higher-dose intradermal hepatitis B vaccination to standard intramuscular vaccination of healthcare workers.

OBJECTIVE: To compare the immunogenicity of hepatitis B vaccine administered via intradermal (ID) versus intramuscular (IM) route. METHODS: Subjects chose either to specify the route of immunization or to undergo random allocation to vaccination by the ID (0.15 mL) or the IM (1.0 mL) route. Yeast-derived recombinant hepatitis B vaccine was given at 0, 30, and 180 days. Hepatitis B surface antibody (HBsAb) and hepatitis B core antibody (HBcAb) were measured by microparticle enzyme immunoassay. RESULTS: 763 subjects were enrolled. Baseline screening identified 65 subjects (8%) who were positive for HBsAb or HBcAb. Vaccination was completed by 590 (85%) of 698 enrollees (370 ID, 220 IM). Seroconversion rates (geometric mean titers [GMT]>0 IU/mL HBsAb) for those vaccinated ID were 99% and 96% for screening at 9 months and 1 year post-vaccination, respectively; subjects vaccinated intramuscularly had similar rates of 95% and 96%. Seropositivity rates (GMT > or = 10 IU/mL HBsAb) showed a similar pattern, with 95%, 92%, and 73% at 9 months and 1 and 2 years, respectively, for those vaccinated ID, and 94%, 93%, and 81% for those having IM vaccination. GMT for HBsAb was significantly higher for individuals vaccinated IM than for those vaccinated ID (P<.0001). The GMT ratio for the IM and ID routes decreased over time, being 9.3 at 9 months, 7.8 at 1 year, and 5.9 at 2 years. An unanticipated side effect of intradermal vaccination was skin discoloration at injection sites, which persisted for at least 2 years postvaccination. Two thirds (112/166) of respondents reported that they would have selected the ID route despite the discoloration. CONCLUSIONS: Higher-dose ID vaccination (3 vs 1 microg per injection) uses one sixth of the dose required for standard IM vaccination. It is a cost-effective way to vaccinate populations against hepatitis B virus, but the long-term efficacy of the ID route must still be investigated.

Comparison of ceftiofur sodium and oxytetracycline for treatment of acute interdigital phlegmon (foot rot) in feedlot cattle.

OBJECTIVE: To determine whether ceftiofur sodium would be useful for treatment of acute interdigital phlegmon (foot rot) in cattle. DESIGN: Randomized controlled trial. ANIMALS: 308 cross-bred yearling steers with clinical signs of acute interdigital phlegmon (i.e., lameness with interdigital swelling, interdigital lesions, or both). PROCEDURE: Steers were randomly assigned to 1 of 3 treatment groups: ceftiofur at a dosage of 0.1 mg/kg (0.045 mg/lb) of body weight, IM, every 24 hours; ceftiofur at a dosage of 1.0 mg/kg (0.45 mg/lb), IM, every 24 hours, and oxytetracycline at a dosage of 6.6 mg/kg (3 mg/lb), IM, every 24 hours. All animals were treated for 3 days; treatment was considered successful if animals were no longer lame on day 4. Biopsy specimens were collected prior to treatment from 5 animals in each group and submitted for anaerobic bacterial culture and histologic examination. RESULTS: Success rates for the high-dosage ceftiofur (94/129; 73%) and oxytetracycline (87/128; 68%) groups were significantly higher than that for the low-dosage ceftiofur group (5/50; 10%), but there were no significant differences between the high-dosage ceftiofur and oxytetracycline groups. Anaerobic bacteria most frequently isolated from biopsy specimens were Porphyromonas levii and Provetella intermedia. CLINICAL IMPLICATIONS: Use of ceftiofur at a dosage of 1.0 mg/kg for treatment of cattle with acute interdigital phlegmon was as effective as use of oxytetracycline at a dosage of 6.6 mg/kg. However, ceftiofur has a negligible withdrawal time and, therefore, may be a better choice for treatment of near-market weight animals.

Quinolone-based antibacterial chemoprophylaxis in neutropenic patients: effect of augmented gram-positive activity on infectious morbidity. National Cancer Institute of Canada Clinical Trials Group.

OBJECTIVE: To determine whether augmented quinolone-based antibacterial prophylaxis in neutropenic patients with cancer reduces infections caused by gram-positive cocci and preserves the protective effect against aerobic gram-negative bacilli. DESIGN: Open, randomized, controlled, multicenter clinical trial. SETTING: Centers participating in the National Cancer Institute of Canada Clinical Trials Group. PATIENTS: 111 eligible and evaluable patients hospitalized for severe neutropenia (neutrophil count < 0.5 x 10(9)/L lasting at least 14 days) who were receiving cytotoxic therapy for acute leukemia or bone marrow autografting. INTERVENTION: One of three oral antibacterial prophylactic regimens (norfloxacin, 400 mg every 12 hours; ofloxacin, 400 mg every 12 hours; or ofloxacin, 400 mg, plus rifampin, 300 mg every 12 hours) beginning with cytotoxic therapy. MEASUREMENTS: Incidence and cause of suspected or proven infection. RESULTS: Microbiologically documented overall infection rates for norfloxacin, ofloxacin, and ofloxacin plus rifampin were 47%, 24%, and 9%, respectively (P < 0.001). Corresponding rates were 24%, 13%, and 3%, respectively for staphylococcal bacteremia (P = 0.03) and, 21%, 3%, and 3%, respectively for streptococcal bacteremia (P < 0.01). The pattern of bacteremia suggested that rifampin played a role in suppressing staphylococcal infection. Both ofloxacin alone and ofloxacin plus rifampin had a clinically significant antistreptococcal effect. Aerobic gram-negative rods were cleared from rectal surveillance cultures in all patients after a median of 5.5 days and caused infection in only one patient (0.9%). The reductions in the number of microbiologically documented infections among ofloxacin recipients and ofloxacin plus rifampin recipients were offset by concomitant increases in the number of unexplained fevers (24% of norfloxacin recipients, 53% of ofloxacin recipients, and 49% of ofloxacin plus rifampin recipients; P = 0.02). No statistically significant difference was found among the treatment arms with respect to the overall incidence of febrile neutropenic episodes as defined for this trial (79% for the norfloxacin group, 82% for the ofloxacin group, and 77% for the ofloxacin plus rifampin group). CONCLUSIONS: Quinolone-based antibacterial chemoprophylaxis protected patients from aerobic gram-negative bacillary infections. Augmentation of the gram-positive activity reduced the incidence of gram-positive infections but did not influence the overall incidence of febrile neutropenic episodes.

Should infection control practitioners do follow-up of staff exposures to patient blood and body fluids?

BACKGROUND: The purpose of this study was to determine the efficiency of a joint infection control/occupational health program for the follow-up of accidental blood or bloody body fluid exposures in health care workers. METHODS: A comprehensive staff follow-up program for all blood exposures with known patient sources was initiated in 1989, consisting of patient follow-up by the Infection Control Department (risk assessment for hepatitis B virus [HBV] and [HIV] infection and obtaining of consent for HIV testing) and staff follow-up by the Occupational Health Department. In 1992 a mailed survey was conducted to examine exposure follow-up policies and responsibilities in large teaching hospitals across Canada. RESULTS: A total of 924 blood exposures with known patient sources were reported between January 1989 and December 1993. HIV and HBV screening was obtained for 67.9% and 87.6% of patients assessed as at low risk and 82.3% and 92.2% of those assessed as at high risk for infection, respectively. Two previously unknown HIV-seropositive patients were identified, one of whom had been classified as at low risk (one of 530 [0.19%] patients at low risk who underwent screening). Primary reasons for screening being missed were patient discharge (46.3%) or communication problems (18.0%). The requirement for informed written consent before HIV screening accounted for the difference in completed HIV and HBV screens. Results of the hospital survey indicated that 40.8% of Canadian hospitals follow up all patients who are involved in blood exposures; however, most hospitals still rely on the physician to obtain consent (87.6%). CONCLUSIONS: Use of ICPs to screen patients involved in staff blood exposures during regular hours may be the most efficient method of follow-up, particularly if supplemented by a backup team of health professionals on nights and weekends. Although screening all patients for HBV/HIV may detect patients with undisclosed high-risk behaviors, institutions must decide whether the practice is cost-effective in areas of low prevalence.

Impact of free verotoxin testing on epidemiology of diarrhea caused by verotoxin-producing Escherichia coli.

During a 10-week period in the summer of 1990, an epidemiologic investigation of the prevalence of verotoxin (VT)-producing Escherichia coli infection was conducted in Calgary, Alberta, Canada. Consecutive stool specimens (n = 3,577) were cultured for E. coli O157:H7, and fecal filtrates were tested for free VTs (FVTs). E. coli O157:H7 was recovered from 22 specimens (0.6%), but VT was detected in 74 specimens (2.1%). Sixty-nine stool specimens positive for FVTs or E. coli O157:H7 were probed for VT genes by colony blot hybridization; 22 of 38 VT gene probe-positive isolates were non-O157:H7 E. coli organisms. Fourteen of 22 strains could not be induced to produce VT in vitro, despite the presence of FVTs in the stool sample, positivity on colony blot hybridization, positive PCR probes with the primers described by Pollard et al. (D. R. Pollard, W. M. Johnson, H. Lior, S. D. Tyler, and K. R. Rozee, J. Clin. Microbiol. 28:540-545, 1990) or Gannon et al. (V. P. Gannon, R. K. King, J. Y. Kim, and E. J. Golsteyn-Thomas, Appl. Environ. Microbiol. 58:3809-3815, 1992) (but not those described by Karch and Meyer [H. Karch and T. Meyer, J. Clin. Microbiol. 27:2751-2757, 1989]), and positive Southern blot analysis of isolates in 10 of 14 strains. The patient survey questionnaire showed that E. coli O157:H7 infection was associated with bloody diarrhea of short duration, whereas infection with other serotypes or persistence of FVT only was associated with longer-duration nonbloody diarrheal illness. We conclude that (i) detection of FVT in stools enhances the diagnosis of VT infection threefold over cultures for E. coli O157:H7, (ii) cultures for E.coli O157:H7 detect the majority of organisms of that serotype, (iii) the spectrum of disease produced by organisms of non-O157:H7 serotypes may include less severe but more protracted illness, and (iv) differences in the in vivo and in vitro expression of toxin and results of genetic probe studies highlight the need to examine control mechanisms of toxin production.

Molecular epidemiology of Xanthomonas maltophilia colonization and infection in the hospital environment.

Between April 1992 and December 1993, 80 Xanthomonas maltophilia isolates were collected from 63 patients in three acute-care hospitals in Calgary, Alberta, Canada. On the basis of Centers for Disease Control and Prevention definitions, 48 patients had nosocomial and 15 had community-acquired X. maltophilia. Thirty-eight of the patients were colonized and 25 were infected. Sixty-four percent of patients who acquired X. maltophilia in the intensive care unit (ICU) became infected, whereas 32% of patients in a non-ICU setting became infected. ICU patients tended to be hospitalized for a shorter period of time than non-ICU patients before the onset of X. maltophilia infection. Regardless of being colonized or infected, all patients had debilitating conditions, with respiratory disease being the most common underlying illness (35%). Forty-two patients (88%) with hospital-acquired X. maltophilia received prior antibiotic therapy which included gentamicin, tobramycin, ceftazidime, piperacillin, and imipenem. Agar dilution MICs showed that patient isolates were resistant to these antimicrobial agents that patients had received. Pulsed-field gel electrophoresis of SpeI-digested genomic DNA revealed that six epidemiologically linked patient isolates from the ICU of one acute-care hospital had identical DNA profiles. In contrast, isolates from patients from the other two hospitals had unique genotype profiles (n = 57) regardless of the presence or absence of an epidemiologic association. In these patients there was genetic evidence against the acquisition of a resident hospital clone. These results indicate that pulsed-field gel electrophoresis can resolve genotypically distinct strains of X. maltophilia and, consequently, is a useful tool for evaluating nosocomial infections caused by X. maltophilia.

Direct detection of verotoxin-producing Escherichia coli in stool samples by PCR.

A method for the rapid detection of verotoxin-producing Escherichia coli in stool samples by PCR was evaluated. Verotoxin-1 and verotoxin-2 genes in DNA extracted directly from stool samples were amplified with oligonucleotide primers. Stools spiked with control organisms, E. coli C600 (H19B) (verotoxin-1) or E. coli C600 (933W) (verotoxin-2), demonstrated that verotoxin-1-containing organisms could be detected at 10(2) CFU per 0.1 g of stool and verotoxin-2-containing organisms could be detected at 10(7) CFU per 0.1 g of stool. Testing of stool samples from patients with diarrhea showed a high concordance between PCR positivity and the presence of verotoxin-producing E. coli, determined by isolation of serotype O157:H7 on sorbitol-MacConkey medium (34 of 35 stool samples) or by colony blots with gene probes (19 of 21 stool samples). Conversely, only 1 of 20 (5.0%) stool samples that were O157:H7 culture negative and colony blot negative and that contained free verotoxin only was positive by PCR. As well, only 4 of 145 (2.8%) stool samples that were negative for serotype O157:H7 or free verotoxin were PCR positive. PCR of DNA extracted directly from stool samples provides a rapid method for the detection of stool samples containing verotoxin-producing E. coli compared with colony blot testing.

Treatment of complicated urinary tract infections with lomefloxacin compared with that with trimethoprim-sulfamethoxazole.

The efficacy of lomefloxacin given at 400 mg once daily for 14 days compared with that of trimethoprim-sulfamethoxazole at 160 and 800 mg, respectively, given twice daily for 14 days in the treatment of symptomatic complicated urinary tract infections was studied in a prospective, randomized, single-blind multicenter study. A total of 133 subjects presenting with signs and symptoms of urinary tract infection and an underlying abnormality consistent with complicated urinary tract infection were enrolled in the study. Bacteriologic cure was significantly better in 68 subjects randomized to lomefloxacin than in 65 subjects randomized to trimethoprim-sulfamethoxazole at short-term follow-up (88 versus 52%; 95% confidence intervals [CIs] 77 and 94% and 39 and 65%, respectively) this difference was no longer significant at long-term follow-up (64 versus 47%; CIs, 52 and 75% and 32 and 57%, respectively). Clinical outcomes were similar for both therapeutic regimens at short- and long-term follow-ups. The organisms that infected the subjects pretherapy were more frequently resistant to trimethoprim-sulfamethoxazole, and drug therapy was discontinued more frequently in subjects treated with trimethoprim-sulfamethoxazole because of adverse antimicrobial effects. In secondary analyses, outcomes did not differ with age or underlying genitourinary abnormality.(ABSTRACT TRUNCATED AT 250 WORDS)

Ciprofloxacin: an oral quinolone for the treatment of infections with gram-negative pathogens. Committee on Antimicrobial Agents. Canadian Infectious Disease Society.

OBJECTIVE: To optimize the use of ciprofloxacin, a synthetic fluoroquinolone, in community and hospital practice, and to minimize overprescribing by providing an overview of the literature and recommendations for appropriate use. OPTIONS: First-line treatment for infections in which ciprofloxacin is shown to be effective, treatment with oral ciprofloxacin rather than parenteral therapy with another drug, and step-down therapy with oral ciprofloxacin after initial parenteral treatment. OUTCOMES: In order of importance: efficacy, side effects and cost. EVIDENCE: A MEDLINE search of articles concerning ciprofloxacin, including in-vitro and pharmacokinetic studies using recognized standard methods with appropriate controls and published in recognized peer-reviewed journals, and randomized, controlled, double-blind clinical trials. VALUES: The Committee on Antimicrobial Agents of the Canadian Infectious Disease Society (CIDS) and a recognized expert (T.J.L.) recommended use of ciprofloxacin to treat infections against which it has proved effective both in vitro and in randomized controlled trials. They took into account its value as an oral replacement for other drugs given parenterally and development of resistance. BENEFITS, HARMS AND COSTS: With more appropriate use of ciprofloxacin there will be less development of resistant pathogens. For certain infections patients who would otherwise require admission to hospital could be treated at home and patients initially admitted for parenteral therapy could be discharged sooner on oral therapy with ciprofloxacin. RECOMMENDATIONS: Ciprofloxacin may be considered as first-line treatment for a number of infections in which gram-negative pathogens are proven or strongly suspected, including complicated urinary tract infections, bacterial prostatitis, bacterial diarrhea, selected bone and joint infections, malignant otitis externa, bronchopulmonary infections in patients with cystic fibrosis and selected pneumonia cases. VALIDATION: The paper was prepared, reviewed and revised by the Committee on Antimicrobial Agents of the CIDS. It was then reviewed and revised further by the Council of the CIDS. SPONSOR: The CIDS is solely responsible for developing, funding and endorsing these guidelines.

Evaluation of enzyme immunoassay for Candida cytoplasmic antigens in neutropenic cancer patients.

A Candida albicans cytoplasmic antigen enzyme immunoassay (CACP antigen EIA) was developed with antibodies raised against antigens prepared from yeast cells grown under standardized growth conditions. The C. albicans components reactive in the EIA were shown to be predominantly proteins with associated carbohydrates. Denaturing gel electrophoresis revealed the presence of five major CACP proteins with molecular weights between 36,000 and 44,000. The clinical usefulness of the CACP EIA was evaluated by retrospective blinded measurement of 89 serum samples from 31 granulocytopenic patient episodes. Twice-weekly surveillance cultures, sequential serum samples (approximately once per week or with change of the clinical course), and standard diagnostic criteria of fungal infection were used to categorize patients. The sensitivity and specificity of the CACP assay on the basis of serum samples were 82 and 100%, respectively (67 and 100% on the basis of patient episodes). The positive and negative predictive values were 100 and 97% for serum (100 and 93% for patient episodes). By comparison, the CANDTEC assay had low sensitivity (33%) and poor positive predictive values (50%). The CACP EIA may be a useful test suitable for further evaluations as a method for the diagnosis of invasive Candida infection in neutropenic cancer patients.

Prolonged and multipatient use of prefilled disposable oxygen humidifier bottles: safety and cost.

OBJECTIVES: Multipatient use and prolonged use of prefilled disposable oxygen humidifier bottles (Aquapak 301, Respiratory Care, Inc., Arlington Heights, IL) were evaluated by performing microbiologic monitoring and a cost analysis on bottles used for varying numbers of patients and lengths of time. METHODS: Humidifiers were hung for a maximum of one month. Monitoring was conducted in 6 different nursing areas. Quantitative cultures were done for aerobes and Legionella. Reusable humidifier bottles also were monitored. RESULTS: Cultures were obtained from 1,311 disposable and 60 reusable humidifiers. No significant bacterial contamination was detected in the prefilled disposable oxygen humidifier units. Ten percent of the reusable bottles were contaminated by organisms associated with skin flora. CONCLUSIONS: Multipatient use and increased duration of use of disposable humidifiers result in cost savings without increasing patient risk. Restricted multipatient use of prefilled disposable oxygen humidifier bottles for a period of one month is a safe and cost-efficient practice.

Reduction in clinical response to empiric antimicrobial therapy of febrile granulocytopenic patients receiving TMP/SMX infection prophylaxis.

In the course of a multicentre clinical trial evaluating two antibacterial regimens for the empiric treatment of suspected infection in febrile neutropenic cancer patients, a suboptimal response was noted among recipients of antibacterial prophylaxis with trimethoprim/sulphamethoxazole (TMP/SMX). Multivariate analysis identified TMP/SMX prophylaxis as a predictor of poor outcome independent of other variables such as classification of infection, marrow recovery, neutrophil count at first fever, indwelling central venous catheter use, and underlying disease. This effect appeared to be restricted to recipients of tobramycin plus ticarcillin (TT). TMP/SMX suppresses potentially pathogenic aerobic Gram-negative bacilli and allows colonization and subsequent infection by Gram-positive microorganisms against which TT-like regimens have limited activity. Recognition of this phenomenon may permit a more appropriate selection of antibacterial agents for the therapy of suspected infection in the neutropenic patient.

Effect of interventions on prescribing of antimicrobials for prophylaxis in obstetric and gynecologic surgery.

The effect of interventions on the conformity of physicians with guidelines for the appropriate use of antimicrobial prophylaxis in obstetric and gynecologic surgery is reported. Guidelines on the appropriate use of antimicrobial prophylaxis in common obstetric and gynecologic surgical procedures were developed in late 1986 by the antibiotic subcommittee at a 1100-bed tertiary-care teaching facility. The guidelines were not adopted immediately by the department of obstetrics and gynecology (OB-GYN). An audit of the medical records of women who had received antimicrobial therapy for abdominal and vaginal hysterectomies and emergency cesarean sections during January through March 1987 showed that cefoxitin was used in 68% of the cases instead of the less expensive and equally efficacious cefazolin as recommended in the guidelines. The projected annual cost of this nonconformity was $26,500. After the subcommittee informed the physicians about the guidelines and the audit results, the OB-GYN department adopted the guidelines. A second audit performed one year later showed that cefazolin was used in the recommended manner in 93% of cases; projected annual cost savings were $25,000. Both audits showed that prophylactic treatment was inappropriately prolonged in 6% of cases. Substantial cost savings were realized by minimizing inappropriate antimicrobial drug use through efforts to promote rational and cost-effective therapy.

Handwashing practices in an intensive care unit: the effects of an educational program and its relationship to infection rates.

Handwashing is the single most important procedure in the prevention of nosocomial infections and yet it remains the most violated of all infection control procedures. With a sequential intervention study in an intensive care unit we have demonstrated that poor handwashing practices are associated with a high nosocomial infection rate, whereas good handwashing practices are associated with a low nosocomial infection rate. An educational and enforcement program designed to improve handwashing procedures can significantly reduce endemic nosocomial infection rates.

Emerging role of quinolones in the prevention of gram-negative bacteremia in neutropenic cancer patients and in the treatment of enteric infections.

The pharmacokinetic and antibacterial characteristics of the fluorinated carboxyquinolones make them attractive candidates for the treatment and prevention of infections arising from the alimentary canal. Norfloxacin and ciprofloxacin have been shown to suppress and eliminate the pool of potentially pathogenic aerobic gram-negative rods colonizing the alimentary canal of neutropenic patients with acute leukemia, thereby reducing infection-related morbidity and mortality due to gram-negative sepsis. Although norfloxacin appears to have limited efficacy in the prevention of gram-positive infections in neutropenic patients, other quinolones with improved in vitro activity against such organisms are being evaluated. Co-trimoxazole is considered a standard of therapy for many bacterial infectious diarrheal illnesses. However drug resistance, and the absence of coverage of Campylobacter jejuni are important deficiencies. Early trials suggest that the quinolones are safe and effective treatment of a wide variety of bacterial diarrheal illnesses. Despite these encouraging results, further controlled trials will be necessary to clarify how the quinolones should best be used.

Comparison of norfloxacin with cotrimoxazole for infection prophylaxis in acute leukemia. The trade-off for reduced gram-negative sepsis.

A total of 63 neutropenic patients receiving cytotoxic therapy for acute leukemia were randomly allocated to receive norfloxacin (400 mg every 12 hours) or cotrimoxazole (160/800 mg every 12 hours) to prevent bacterial infection. Compliance was more than 95 percent and no adverse effects attributable to the study drugs were observed. The overall incidence of febrile illness (67 percent) was similar between the groups; however, no gram-negative bacillary infections were observed in 31 norfloxacin recipients compared with four of 32 cotrimoxazole recipients. Furthermore, nine norfloxacin recipients had 17 gram-positive bacteremias compared with two in two cotrimoxazole recipients (p = 0.0034). Norfloxacin was more effective than cotrimoxazole for preventing acquisition of aerobic gram-negative bacilli in surveillance cultures. Neither study drug allocation nor the presence of an indwelling central venous catheter influenced outcome among the 42 patients who subsequently received empiric systemic antibiotics for suspected infection. Although gram-positive infection remains an unsolved problem, norfloxacin appears to be a safe, effective, well-tolerated alternative to cotrimoxazole for preventing gram-negative infection in neutropenic patients with acute leukemia.

Changes in endogenous microflora among febrile granulocytopenic patients receiving empiric antibiotic therapy: implications for fungal superinfection.

The ecologic effect of empiric systemic antibiotic therapy on the endogenous microflora was evaluated in 83 febrile granulocytopenic patients with cancer who were randomly allocated to receive moxalactam plus ticarcillin (45 patients) or tobramycin plus ticarcillin (38 patients) for suspected infection. Serial surveillance cultures of the nasal passages, oropharynx and feces performed twice a week showed that patients who received the former regimen had higher elimination rates and significantly lower acquisition rates (p = 0.027) for aerobic gram-negative bacilli than did patients who received the latter regimen. However, therapy with moxalactam plus ticarcillin also resulted in significantly higher acquisition rates for yeasts (p = 0.004). This was associated with a significantly higher fungal superinfection rate among these patients than among those who received tobramycin plus ticarcillin (40% v. 16%) (p less than 0.05). Moxalactam plus ticarcillin therapy created a greater microbial ecologic vacuum by the elimination of intestinal anaerobes, which, in turn, permitted fungal colonization and an increased risk of superinfection. Our results support the recommendation that an antipseudomonal penicillin plus an aminoglycoside be selected as empiric therapy for suspected infection in febrile granulocytopenic patients with cancer. Such a regimen would spare the anaerobic intestinal microflora, thereby reducing the risk of fungal colonization and infection.