To take into account the proportion of patients lost to follow-up when calculating surgical-site infection (SSI) rates.

A multicenter SSI monitoring network in Basse-Normandie, France, using the definitions for SSI of the National Nosocomial Infections Surveillance System of the Centers for Disease Control and Prevention.

Between January 1, 1998, and December 31, 1999, 3,705 patients were operated on in 25 units of 10 institutions.

Of the patients, 41.2% (range, 5.1% to 95.5%) were seen 30 days or more after their operation. The global SSI attack rate was 2.19% (95% confidence interval, 1.72% to 2.66%). With the use of the Kaplan–Meier method, the incidence rate was 3.11% (95% confidence interval, 3.06% to 3.16%). The difference between the attack rate and the Kaplan–Meier incidence rate for each unit varied according to the percentage of patients seen on or after day 30 postoperatively and the number of SSIs diagnosed in patients seen on or after day 30.

Practice guidelines are needed for the international monitoring for postdischarge SSIs and the calculation of SSI rates. The proportion of patients seen 30 days after their operation is a major quality criterion for SSI monitoring and should be routinely given in monitoring reports, oral communications, and publications to compare results obtained by different teams.

Following an outbreak of methicillin-resistant Staphylococcus aureus (MRSA) infection in our acute rehabilitation unit in 1987, all patients except in-house transfers (because of their low prevalence of MRSA colonization) underwent MRSA screening cultures on admission.

To better characterize the current profile of patients with positive MRSA screening cultures at the time of admission to our acute rehabilitation unit, and to determine the relative yield of nares, perianal, and wound screening cultures in this population.

Prospective chart review with ongoing active surveillance for infections associated with the acute rehabilitation unit.

The rate of MRSA isolation from one or more body sites increased significantly from 5% (1987–1988) to 12% (1999–2000) (P = .0009) for newly admitted patients and from 0% to 7% (P < .0001) for in-house transfers. A negative nares culture was highly predictive (98%) of a negative perianal culture. Prior history of MRSA infection or colonization and transfer from outside sources were independently associated with positive MRSA screening cultures.

The rate of MRSA isolation from screening cultures of newly admitted patients, including in-house transfers, has increased significantly during the past decade in our acute rehabilitation unit. When paired with nares cultures, perianal cultures were of limited value in this patient population.

To reduce the number of nosocomial infections (NIs) in surgical patients by a quality management approach.

Prospective, controlled study in 8 medium-sized hospitals during a 26-month period.

Four study hospitals and 4 control hospitals.

In two 10-month intervention periods, 4 external physicians introduced quality circles and ongoing surveillance in the 4 study hospitals. There were three 8-week observation periods in all 8 hospitals with the same physicians before, during, and after the intervention periods.

During the first observation period, almost identical overall incidence densities were found for the study hospitals and the control hospitals. During the course of the study, the overall incidence density decreased significantly in the study hospitals (risk ratio [RR], 0.74; 95% confidence interval [CI95], 0.59 to 0.94) and nonsignificantly in the control hospitals (RR, 0.90; CI95 0.70 to 1.16). With the use of a Cox regression model to evaluate the impact of the intervention periods while taking into account the distribution of risk factors for NI in both groups, a significant risk reduction (RR, 0.75; CI95, 0.58 to 0.97) was observed after the first intervention period when comparing study and control hospitals. At the end of the study (ie, after the second intervention period), the difference between the study hospitals and the control hospitals was not significant (RR, 0.78; CI95, 0.60 to 1.01). This was due to no further improvement at the end of the study in the study hospitals and a decrease in the control hospitals.

This study demonstrates that NI rates can be significantly reduced by appropriate intervention methods in hospitals that are interested in quality management activities. However, continuous intense efforts are necessary to maintain these improvements.

To describe the cumulative incidence of and risk factors for tuberculosis (TB) infection among medical students.

In 1999, a cohort study of medical students with negative results (induration < 10 mm) on tuberculin skin test (TST) was performed. Students who had undergone two-step testing in 1998 were retested.

University and teaching hospital and referral center for TB and acquired immunodeficiency syndrome, and the Health Sciences Building of the Medical School of the Federal University of Rio de Janeiro, Brazil.

A sample of 618 consecutive medical students with negative TST results who had been tested 12 months before were approached. Information about sociodemographic characteristics, bacille Calmette-Guérin vaccination history, and potential exposures to TB was obtained using a standardized questionnaire. Four hundred fourteen (67%) students completed the study. Students were at two different levels of their training programs (juniors = no contact with patients; seniors = intensive contact with patients).

Of 414 participants, 16 (3.9%; 95% confidence interval, 1.06% to 12.1%) had converted to a positive reaction after 1 year. In a multivariate logistic regression analysis, higher level of clinical training was confirmed to be an independent factor associated with TST conversion (odds ratio, 4.77; 95% confidence interval, 1.01 to 22.46; P = .048).

Senior medical students are at increased risk of Mycobacterium tuberculosis infection in this setting. Therefore, a program of routine tuberculin skin testing and specific TB infection control guidelines are needed for this population (Infect Control Hosp Epidemiol 2002;23:591-594).

To describe the investigation and management of an outbreak due to multiresistant Acinetobacter baumannii and to determine risk factors for acquisition of the organism.

A 14-bed regional burn unit in a Canadian tertiary-care teaching hospital.

Case-control study with multivariate analysis of potential risk factors using logistic regression analysis. Surveillance cultures were obtained from the hospital environment, from noninfected patients, and from healthcare providers.

A total of 31 (13%) of 247 patients with acute burn injuries acquired multiresistant A. baumannii between December 1998 and March 2000; 18 (58%) of the patients were infected. The organism was recovered from the hospital environment and the hands of healthcare providers. Significant risk factors for acquisition of multiresistant A. baumannii were receipt of blood products (odds ratio [OR], 10.8; 95% confidence interval [CI95], 3.4 to 34.4; P < .001); procedures performed in the hydrotherapy room (OR, 4.1; CI95, 1.3 to 13.1; P = .02); and increased duration of mechanical ventilation (OR, 1.1 per day; CI95, 1.0 to 1.1; P=. 02).

Improved compliance with hand hygiene, strict patient isolation, meticulous environmental cleaning, and temporary closure of the unit to new admissions.

Acquisition of multiresistant A. baumannii was likely multifactorial, related to environmental contamination and contact with transiently colonized healthcare providers. Control measures addressing these potential sources of multiresistant A. baumannii were successful in terminating the outbreak. Ongoing surveillance and continued attention to hand hygiene and adequate environmental cleaning are essential to prevent recurrent outbreaks due to antibiotic-resistant bacteria in burn units.

To test the evidence that the risk of infection related to central venous catheters (CVCs) is decreased by anti-infective coating or cuffing.

Systematic review of randomized, controlled trials comparing anti-infective with inactive (control) CVCs.

Average insertion times were taken as a measurement of the length of insertion. Dichotomous data were combined using a fixed effect model and expressed as odds ratio (OR) with 95% confidence interval (CI95).

Two trials on antibiotic coating (343 CVCs) had an average insertion time of 6 days; the risk of BSI decreased from 5.1% with control to 0% with anti-infective catheters. There were no trials with longer average insertion times. In three trials on silver collagen cuffs (422 CVCs), the average insertion time ranged from 5 to 8.2 days (median, 7 days); the risk of BSI was 5.6% with control and 3.2% with anti-infective catheters. In another trial on silver collagen cuffs (101 CVCs), the average insertion time was 38 days; the risk of BSI was 3.7% with control and 4.3% with anti-infective catheters. In five trials on chlorhexidine-silver sulfadiazine coating (1,269 CVCs), the average insertion time ranged from 5.2 to 7.5 days (median, 6 days); the risk of BSI decreased from 4.1% with control to 1.9% with anti-infective catheters. In five additional trials on chlorhexidine–silver sulfadiazine coating (1,544 CVCs), the average insertion time ranged from 7.8 to 20 days (median, 12 days); the risk of BSI was 4.5% with control and 4.2% with anti-infective catheters.

Antibiotic and chlorhexidine–silver sulfadiazine coatings are anti-infective for short (approximately 1 week) insertion times. For longer insertion times, there are no data on antibiotic coating, and there is evidence of lack of effect for chlorhexidine-silver sulfadiazine coating. For silver-impregnated collagen cuffs, there is evidence of lack of effect for both short- and long-term insertion.

In December 1999, an outbreak of diarrhea was reported in a general hospital neonatal medium care unit (NMCU) caused by a novel strain of rotavirus with genotype P[6], G9. An investigation was conducted to determine risk factors for illness among neonates.

Rotavirus diagnosis was by latex agglutination and typing by reverse transcriptase polymerase chain reaction. A case-control study was performed using data collected from medical records on exposures in a 3-day period before illness (cases) or a random 3-day period (controls). Environmental swabs were tested for rotavirus. Antenatal blood samples from mothers and blood samples provided by hospital staff were analyzed for rotavirus antibodies.

Fifty-six cases of rotaviral illness were confirmed by latex agglutination. Forty-seven of these were among 118 neonates exposed to the NMCU (attack rate, 40%). There was a 4-week period with no clinical cases in the course of the outbreak. Increased frequency (s≥ 15 times in 3 days) of ungloved nasogastric feeding was a significant risk factor (adjusted odds ratio, 8.79), controlling for birth weight and gestational age. Environmental sampling showed persistence of the virus on ward surfaces despite cleaning. None of 24 NMCU staff members had high levels of antibodies against P[6], G9. Three (8%) of 38 mothers had high antibody levels; 2 had infants who became ill. The outbreak ended with a 7-day ward closure, disinfection, and introduction of gloved nasogastric feeding.

Case-control studies can be successful in identifying risk factors for nosocomial outbreaks of diarrhea. High levels of rotavirus antibodies in mothers may not protect infants. The environment may be the most important reservoir of rotavirus during outbreaks.

To determine the frequency of colonization by Enterobacter aerogenes in patients in the intensive care unit (ICU) for more than 48 hours and to evaluate the risk factors for infection in patients colonized by this bacteria.

An 8-month prospective study.

A 12-bed medical–surgical ICU in a 450-bed, university-affiliated, tertiary-care hospital in Belgium.

Pulsed-field gel electrophoresis was used to determine the genotypes of E. aerogenes isolates.

We observed two major clones of E. aerogenes in the ICU. Interestingly, 87.5% of infected patients had the same genomic profile for colonization and infection. Risk factors for infection in this particular population included younger age, prolonged hospital stay, mechanical ventilation, and bronchoscopy.

Colonization is a major prerequisite for infection. The identification of risk factors for infection in colonized patients can optimize the quality of treatment in the ICU.

To assess the ability of hospital air handling systems to filter Aspergillus, other fungi, and particles following the implosion of an adjacent building; to measure the quantity and persistence of airborne fungi and particles at varying distances during a building implosion; and to determine whether manipulating air systems based on the movement of the dust cloud would be an effective strategy for managing the impact of the implosion.

Air sampling study.

A 976-bed teaching hospital in Baltimore, Maryland.

Single-stage impactors and particle counters were placed at outdoor sites 100, 200, and 400 m from the implosion and in five locations in the hospital: two oncology floors, the human immunodeficiency virus unit, the cardiac surgical intensive care unit, and the ophthalmology unit. Air handling systems would operate normally unless the cloud approached the hospital.

Wind carried the bulk of the cloud away from the hospital. Aspergillus counts rose more than tenfold at outdoor locations up to 200 m from the implosion, but did not increase at 400 m. Total fungal counts rose more than sixfold at 100 and 200 m and twofold at 400 m. Similar to Aspergillus, particle counts rose several-fold following the implosion at 100 and 200 m, but did not rise at 400 m. No increases in any fungi or particles were measured at indoor locations.

Reacting to the movement of the cloud was effective, because normal operation of the hospital air handling systems was able to accommodate the modest increase in Aspergillus, other fungi, and particles generated by the implosion. Aspergillus measurements were paralleled by particle counts.

To describe the patterns of nosocomial infections in patients with traumatic injuries and to compare the associations between injury severity, derived from various severity scoring systems, and subsequent nosocomial infections.

Prospective observational study.

A 750-bed university hospital serving as a medical school and referral center for the southern part of Thailand.

All trauma patients admitted to the hospital for more than 3 days during 1996 to 1999 were eligible for this study.

The severity of injuries was measured in terms of injury severity score (ISS), revised trauma score (RTS), new injury severity score (NISS), and trauma injury severity score (TRISS). Infections acquired during hospitalization were categorized using Centers for Disease Control and Prevention criteria. The association between severity of injury and nosocomial infection was examined with Poisson regression models.

There were 222 nosocomial infections identified among 146 patients, yielding an infection rate of 0.8 infections per 100 patient-days. Surgical-site infection was the most common site-specific infection, accounting for 31.1% of all infections. The incidence of intravenous catheter–related bloodstream infection was 1.6 infections per 100 catheter-days. The bladder catheter–related urinary tract infection rate was 2.8 infections per 100 catheter-days. The rate of ventilator-associated pneumonia was 3.2 infections per 100 ventilator-days. The incidence of infection correlated well with injury severity. The infection incidence rate ratios for one severity category increment of ISS, NISS, RTS, and TRISS were 1.65 (95% confidence interval [CI95, 1.42 to 1.92), 1.79 (CI95, 1.55 to 2.05), 1.64 (CI95, 1.43 to 1.88), and 1.32 (CI95, 1.14 to 1.52), respectively.

Surgical-site infection was the most common site-specific nosocomial infection. The NISS might be the most appropriate severity scoring system for adjustment of infection rates in trauma patients.

To determine the motivating and behavioral factors responsible for improving compliance with hand washing among physicians.

Five unobtrusive, observational studies recording hand washing after direct patient contact, with study results reported to physicians.

A 450-bed hospital in a health maintenance organization with an 18-bed medical-surgical intensive care unit (ICU) and a 12-bed cardiac care unit.

An infectious disease physician met individually with participants to report study results and obtain a commitment to hand washing guidelines. Follow-up interviews were conducted to evaluate behavioral factors and educational programs. Hand washing study results were presented to all staff physicians by live and videotaped inservice presentations and electronic mail (e-mail) newsletters. The importance of influencing factors and the educational effectiveness of the hand washing program were evaluated.

Five observational hand washing studies were conducted in the ICU between April 1999 and September 2000. Rates of physician compliance with hand washing were 19%, 85%, 76%, 74%, and 68%, respectively. There were 71 initial encounters and 55 follow-up interviews with the same physicians. Physician interviews revealed that 73% remembered the initial encounter, 70% remembered the hand washing inservice presentations, and 18% remembered the e-mail newsletters. Personal commitment and meeting with an infectious disease physician had the most influence on hand washing behavior. Direct inservice presentations (either live or videotaped) had more influence than did e-mail information. Rates of ventilator-associated pneumonia did not significantly change before and during the study periods. A decrease in the rate of central-line–related bloodstream infections from 3.2 to 1.4 per 1,000 central-line days was found, but could not be solely attributed to improved physician compliance with hand washing.

Physician compliance with hand washing can improve. Personal encounters, direct meetings with an infectious disease physician, and videotaped presentations had the greatest impact on physician compliance with hand washing at our medical center, compared with newsletters sent via e-mail. Local data on compliance with hand washing and physician involvement are factors to be considered for physician hand washing compliance programs in other medical centers.

To describe a nosocomial outbreak of Salmonella serotype Saintpaul gastroenteritis and to explore risk factors for infection.

Case-control study.

A 208-bed, university-affiliated children's hospital.

Patients hospitalized at Children's Hospital and Regional Medical Center, Seattle, Washington, during February 2001 who had stool specimens obtained for culture at least 24 hours after admission. Case-patients (n = 11) were patients with an indistinguishable strain of Salmonella Saintpaul cultured from their stool. Control-patients (n = 41) were patients hospitalized for problems other than gastroenteritis whose stool cultures were negative for Salmonella.

Risk factors were evaluated using the chisquare test or Fisher's exact test. Continuous variables were compared using the Mann–Whitney U test. A multivariable analysis was performed using logistic regression. The predictor of interest was the receipt of enteral feeding formula mixed by the hospital.

Case-patients were more likely than control-patients to have received formula mixed by the hospital (OR, 4.2; 95% confidence interval, 1.04 to 17.16). Other variables evaluated were not significant predictors of Salmonella Saintpaul infection.

Formula mixed by the hospital appears to have been the source of this Salmonella outbreak. Strict sanitation measures must be ensured in formula preparation and delivery, and bacterial pathogens should be included in the differential diagnosis for nosocomial gastroenteritis.

To investigate and describe an outbreak of Serratia marcescens in a neonatal intensive care unit (NICU) and to report the interventions leading to cessation of the outbreak.

A 2,168-bed, tertiary-care, university teaching hospital in Vienna, Austria, with an 8-bed NICU.

We conducted a case–control study to identify risk factors for colonization and infection with S. marcescens. A case-patient was defined as any neonate in the NICU with a positive culture for S. marcescens between October 1, 2000, and February 28, 2001. Polymerase chain reaction was applied to type isolates.

During unannounced observations, the NICU was examined and existing policies were reviewed. Staff were reinstructed in hand antisepsis and gloving policies. Admissions were halted on December 27. During previously planned technical maintenance of the ward, the NICU was closed for 10 days and thorough aldehyde-based disinfection of the NICU was performed.

Ten neonates met the case definition: 6 with infections (among them 3 with cerebral abscesses) and 4 with asymptomatic colonization. Previous antibiotic treatment of the mothers with cefuroxime was the single significant risk factor for colonization or infection (P = .028; odds ratio, 17; 95% confidence interval, 1.3 to 489.5).

S. marcescens can cause rapidly spreading outbreaks associated with fatal infections in NICUs. With aggressive infection control measures, such outbreaks can be stopped at an early stage. Affected neonates themselves may well be the source of cross-infection to other patients on the ward. Antibiotic treatment of mothers should be reevaluated to avoid unnecessary exposure to antibiotics with the potential of overgrowth of resistant organisms.

This study presents the results of a 5-year surveillance program involving the prospective follow-up of health-care workers (HCWs) in the Veneto region of Italy exposed to blood-borne viruses.

All HCWs who reported an occupational exposure to blood-borne infection joined the surveillance program. Both HCWs and patients were tested for viral markers (hepatitis B surface antigen [HBsAg], antibody to hepatitis B surface antigen [anti-HBs], antibody to hepatitis B core antigen [anti-HBc], antibody to hepatitis C virus [anti-HCV], HCV RNA, and antibody to human immunodeficiency virus [HIV]) and had these markers plus transaminases assayed at 3, 6, and 12 months and then yearly thereafter. Moreover, a program of hepatitis B virus (HBV) prophylaxis was offered to those whose anti-HBs levels were less than 10 IU/mL.

Two hundred forty-five HCWs (156 women and 89 men) with a mean age of 37 (± 10) years who reported occupational exposure during the 5-year period.

At the time of exposure, 1HCW was positive for HBsAg (0.4%) and 2 were positive for HCV RNA (0.8%). Among the patients involved, 28 (11.4%) were positive for HBsAg, 68 (27.8%) were positive for HCV RNA, 6 (2.4%) were positive for HIV, and 147 (60.0%) were negative for all viral markers (4 patients were positive for both HCV and HIV). During the follow-up period after exposure (mean, 2.7 [± 1.6] years), there was no increase in transaminases or seroconversions to any of the viral markers.

Our accurate postexposure follow-up revealed a lack of transmission of HBV, HCV, and HIV.