Major trends in the microbial etiology of nosocomial infection.

To determine trends in the microbial etiology of nosocomial infections in the 1980s, surveillance data on the microbiology of documented nosocomial infection reported to the National Nosocomial Infections Surveillance System and from the University of Michigan Hospital were analyzed. Antimicrobial susceptibility data on selected pathogens from both sources were also reviewed. Overall, Escherichia coli decreased from 23% of infections in 1980 to 16% in 1986-1989, Klebsiella pneumoniae dropped from 7% to 5%, whereas coagulase negative staphylococci increased from 4% to 9% and Candida albicans increased from 2% to 5%. Staphylococcus aureus, Pseudomonas aeruginosa, Enterobacter species and enterococci had minor increases, but antimicrobial resistant strains for these pathogens as well as coagulase-negative staphylococci were seen more frequently. In contrast to the 1970s, major shifts in the etiology of nosocomial infection have occurred in the decade of the 1980s. Taken as a whole, the shifts are away from more easily treated pathogens toward more resistant pathogens with fewer options for therapy. These shifts underscore the continued need for prevention and control to accompany new developments in therapy.

Progress report on the evaluation of the efficacy of infection surveillance and control programs.

The main objective of the Study on the Efficacy of Nosocomial Infection Control (SENIC Project) is to determine whether infection surveillance and control programs have reduced the rates of nosocomial infection in United States hospitals. To study this question, we stratified all hospitals in the SENIC target population into 16 design strata defined by categories of a surveillance and a control index derived from hospitals' responses to a preliminary screening questionnaire, and estimated the nosocomial infection rates among 339,044 randomly selected patients admitted in 1970 and 1975 through 1976 to 338 hospitals selected randomly from the 16 design strata. Finding that the over-all infection rates, standardized for important confounding variables or covariates, in hospitals with higher intensity programs had increased less from 1970 to 1975-1976 than those of hospitals with low intensity programs would indicate the efficacy of these programs. Potentially important confounding variables and covariates being studied include individual patient risk factors, hospital characteristics and the completeness of hospitals' medical records. Since only the first has been explored sufficiently, no conclusions on efficacy can yet be drawn. The analytic techniques were illustrated with preliminary data on infection rates at the four individual sites of infection.

The joint associations of multiple risk factors with the occurrence of nosocomial infection.

To compare nosocomial infection rates estimated in different time periods or in different hospitals, it is necessary to control for differences in the distribution of factors that substantially influence a patient's susceptibility to infection. To evaluate the associations of multiple risk factors with the occurrence of infection at each of four major sites and to develop composite measures for use in controlling for differences in the distribution of risk among groups of patients, we used a multivariate categorical data analysis technique to study the infection experience of 169,518 patients admitted in 1970 to the 338 hospitals studied in the Study on the Efficacy of Nosocomial Infection Control (SENIC, Project). The relative importance of risk factors and their complex interactions varied by site. The factors found to be highly important for one or more sites were duration of urinary catheterization, the patients' intrinsic risk as reflected in their diagnoses and types of surgical procedures, duration of preoperative hospitalization, duration of operation, anatomic location of surgical procedure, previous infection and steroid or immunosuppressive therapy. Site-specific risk strata and estimates of each patient's probability of acquiring infection were developed from these data for use in future SENIC analyses.

Nosocomial infections in U.S. hospitals, 1975-1976: estimated frequency by selected characteristics of patients.

To obtain estimates of the frequency of nosocomial infections nationwide, those occurring at the four major sites--urinary tract, surgical wound, lower respiratory tract and bloodstream--were diagnosed in a stratified random sample of 169,526 adult, general medical and surgical patients selected from 338 hospitals representative of the "mainstream" of U.S. hospitals. We estimate that in the mid-1970s one or more infections developed in 5.23 percent (+/- 0.16) of the patients and that 6.62 (+/- 0.24) infections occurred among every 100 admissions. Risks were significantly related to age, sex, service, duration of total and of preoperative hospitalization, presence of previous nosocomial or community-acquired infection, types of underlying illnesses and operations, duration of surgery, and treatment with urinary catheters, continuous ventilatory support or immunosuppressive medications. Seventy-one percent of the nosocomial infections occurred in the 42 percent of patients undergoing surgery and 56 percent in the 38 percent financed by Medicare, Medicaid or other public health care plans.

The National Nosocomial Infections Surveillance System: plans for the 1990s and beyond.

The National Nosocomial Infections Surveillance (NNIS) System is an ongoing collaborative surveillance system among the Centers for Disease Control (CDC) and United States hospitals to obtain national data on nosocomial infections. This system provides comparative data for hospitals and can be used to identify changes in infection sites, risk factors, and pathogens, and develop efficient surveillance methods. Data are collected prospectively using four surveillance components: hospital-wide, intensive care unit, high-risk nursery, and surgical patient. The limitations of NNIS data include the variability in case-finding methods, infrequency or unavailability of culturing, and lack of consistent methods for post-discharge surveillance. Future plans include more routine feedback of data, studies on the validity of NNIS data, new components, a NNIS consultant group, and more rapid data exchange with NNIS hospitals. Increasing the number of NNIS hospitals and cooperating with other agencies to exchange data may allow NNIS data to be used better for generating benchmark nosocomial infection rates. The NNIS system will continue to evolve as it seeks to find more effective and efficient ways to measure the nosocomial infection experience and assess the influence of patient risk, changes in the delivery of hospital care, and changes in infection control practices on these measures.

Surgical wound infection rates by wound class, operative procedure, and patient risk index. National Nosocomial Infections Surveillance System.

To perform a valid comparison of rates among surgeons, among hospitals, or across time, surgical wound infection (SWI) rates must account for the variation in patients' underlying severity of illness and other important risk factors. From January 1987 through December 1990, 44 National Nosocomial Infections Surveillance System hospitals reported data collected under the detailed option of the surgical patient surveillance component protocol, which includes definitions of eligible patients, operations, and nosocomial infections. Pooled mean SWI rates (number of infections per 100 operations) within each of the categories of the traditional wound classification system were 2.1, 3.3, 6.4, and 7.1, respectively. A risk index was developed to predict a surgical patient's risk of acquiring an SWI. The risk index score, ranging from 0 to 3, is the number of risk factors present among the following: (1) a patient with an American Society of Anesthesiologists preoperative assessment score of 3, 4, or 5, (2) an operation classified as contaminated or dirty-infected, and (3) an operation lasting over T hours, where T depends upon the operative procedure being performed. The SWI rates for patients with scores of 0, 1, 2, and 3 were 1.5, 2.9, 6.8, and 13.0, respectively. The risk index is a significantly better predictor of SWI risk than the traditional wound classification system and performs well across a broad range of operative procedures.

Nosocomial infection rates in adult and pediatric intensive care units in the United States. National Nosocomial Infections Surveillance System.

To determine which intensive care unit (ICU) infection rate may be best for interhospital and intrahospital comparisons and to assess the influence of invasive devices and type of ICU on infection rates, we analyzed data from the National Nosocomial Infections Surveillance System. From October 1986 to December 1990, 79 hospitals reported 2,334 hospital-months of data from 196 hospital units. The median overall infection rate was 9.2 infections per 100 patients. However, this infection rate had a strong positive correlation with average length of ICU stay (r = 0.60, p less than 0.0001). When patient-days was used in the denominator, the median overall nosocomial infection rate was 23.7 infections per 1,000 patient-days. Although there was a marked reduction in the correlation with average length of stay, this rate had a strong positive correlation with device utilization (r = 0.59, p less than 0.0001). To attempt to control for average length of stay and device utilization, we examined device-associated nosocomial infection rates. Central line-associated bloodstream infection rates, catheter-associated urinary tract infection rates, and ventilator-associated pneumonia rates varied by ICU type. The distributions of device-associated infection rates were different between some ICU types and were not different between others (coronary and medical ICUs or medical-surgical and surgical ICUs). Comparison of device-associated infection rates and overall device utilization identified hospital units with outlier infection rates or device utilization. These data show that: (1) choice of denominator is critical when calculating ICU infection rates; (2) device-associated infection rates vary by ICU type; and (3) intrahospital and interhospital comparison of ICU infection rates may best be made by comparing ICU-type specific, device-associated infection rates.

Comparison of rates of nosocomial infections in neonatal intensive care units in the United States. National Nosocomial Infections Surveillance System.

To determine nosocomial infection (NI) rates among neonatal intensive care units (NICUs) that are useful for interhospital comparison, we analyzed data reported in 1986-1990 from 35 hospitals that have level III NICUs and used standard National Nosocomial Infections Surveillance protocols and NI site definitions. Overall rates of NI were calculated as the number of NI per 100 patients (overall NI patient rates) or the number of NI per 1,000 NICU patient-days (overall NI patient-day rates). A strong positive association was found between overall NI patient rates and the neonates' average length of stay, a marker for duration of exposure to important risk factors. No correlation was found between overall NI patient-day rates and average length of stay. However, a strong positive correlation between overall NI patient-day rates and a measure of device utilization (total device-days/total patient-days x 100) was found. Additionally, a positive correlation between overall NI patient rates and device utilization was found. Stratification among the three birthweight groups (less than 1,500 g, 1,500-2,500 g, greater than 2,500 g) did not eliminate the need to control for variations in these factors among NICUs. Device-associated, device-day infection rates, calculated as the number of umbilical or central line-associated blood-stream infections per 1,000 umbilical or central line-days and the number of ventilator-associated pneumonias per 1,000 ventilator days, were not correlated with a unit's site-specific device utilization. These data suggest that calculation of device-associated NI rates in NICUs using device-days as the denominator helps to control for the duration of exposure to the primary risk factor and will be more meaningful for purposes of interhospital comparison.

Nosocomial infections in elderly patients in the United States, 1986-1990. National Nosocomial Infections Surveillance System.

We analyzed 101,479 nosocomial infections in 75,398 adult patients (greater than 15 years) that were reported to the National Nosocomial Infections Surveillance (NNIS) system between 1986 and 1990 by 89 hospitals using the NNIS hospital-wide surveillance component. Overall, 54% of the infections occurred in elderly patients (greater than or equal to 65 years). In the elderly, 44% of the infections were urinary tract infections (UTIs), 18% were pneumonias, 11% were surgical wound infections (SWIs), 8% were bloodstream infections (BSIs), and the remainder were infections at other sites. When we compared the infections in elderly patients with those in younger adult patients, ages 15 to 64 years, a far greater percentage of the infections in elderly patients were UTIs, and there were more pneumonias than SWIs. Elderly and younger patients with ventilator-associated pneumonia were about 1.5 times more likely to develop a secondary BSI than those with pneumonia not associated with ventilator use. When the pathogens isolated from the infections were compared to those reported to the NNIS system in 1984, the percentage that were coagulase-negative staphylococci had increased in both elderly and younger patients. The patient died in 12% of all of the infections. Surveillance personnel reported that 54% of the infections in elderly infected patients who died were related to death compared with 59% in younger infected patients who died. When the infection was related to the patient's death, it was most often pneumonia or a BSI. The risk of an infection-related death was significantly higher when the infected patient developed a secondary BSI. Infection prevention efforts should target infections that occur frequently, are amenable to intervention, and have an adverse outcome.

Secular trends in nosocomial primary bloodstream infections in the United States, 1980-1989. National Nosocomial Infections Surveillance System.

More than 25,000 primary bloodstream infections (BSIs) were identified by 124 National Nosocomial Infections Surveillance System hospitals performing hospital-wide surveillance during the 10-year period 1980-1989. These hospitals reported 6,729 hospital-months of data, during which time approximately 9 million patients were discharged. BSI rates by hospital stratum (based on bed size and teaching affiliation) and pathogen groups were calculated. In 1989, the overall BSI rates for small (less than 200 beds) nonteaching, large nonteaching, small (less than 500 beds) teaching, and large teaching hospitals were 1.3, 2.5, 3.8, and 6.5 BSIs per 1,000 discharges, respectively. Over the period 1980-1989, significant increases (p less than 0.0001) were observed within each hospital stratum, in the overall BSI rate and the BSI rate due to each of the following pathogen groups: coagulase-negative staphylococci, Staphylococcus aureus, enterococci, and Candida species. In contrast, the BSI rate due to gram-negative bacilli remained stable over the decade, in all strata. Except for small nonteaching hospitals, the greatest increase in BSI rates was observed in coagulase-negative staphylococci (the percentage increase ranged between 424% and 754%), followed by Candida species (219-487%). In small nonteaching hospitals, the greatest increase was for S. aureus (283%), followed by enterococci (169%) and coagulase-negative staphylococci (161%). Our analysis documents the emergence over the last decade of coagulase-negative staphylococci as one of the most frequently occurring pathogens in BSI.

Risk of human immunodeficiency virus infection among emergency department workers.

PURPOSE: To estimate (1) the prevalence of human immunodeficiency virus (HIV) infection in emergency department (ED) patients, (2) the frequency of blood contact (BC) in ED workers (EDWs), (3) the efficacy of gloves in preventing BC, and (4) the risk of HIV infection in EDWs due to BC. PATIENTS AND METHODS: We conducted an 8-month study in three pairs of inner-city and suburban hospital EDs in high AIDS incidence areas in the United States. At each hospital, blood specimens from approximately 3,400 ED patients were tested for HIV antibody. Observers monitored BC and glove use by EDWs. RESULTS: HIV seroprevalence was 4.1 to 8.9 per 100 patient visits in the 3 inner-city EDs, 6.1 in 1 suburban ED, and 0.2 and 0.7 in the other 2 suburban EDs. The HIV infection status of 69% of the infected patients was unknown to ED staff. Seroprevalence rates were highest among patients aged 15 to 44 years, males, blacks and Hispanics, and patients with pneumonia. BC was observed in 379 (3.9%) of 9,793 procedures; 362 (95%) of the BCs were on skin, 11 (3%) were on mucous membranes, and 6 (2%) were percutaneous. Overall procedure-adjusted skin BC rates were 11.2 BCs per 100 procedures for ungloved workers and 1.3 for gloved EDWs (relative risk = 8.8; 95% confidence interval = 7.3 to 10.3). In the high HIV seroprevalence EDs studied, 1 in every 40 full-time ED physicians or nurses can expect an HIV-positive percutaneous BC annually; in the low HIV seroprevalence EDs studied, 1 in every 575. The annual occupational risk of HIV infection for an individual ED physician or nurse from performing procedures observed in this study is estimated as 0.008% to 0.026% (1 in 13,100 to 1 in 3,800) in a high HIV seroprevalence area and 0.0005% to 0.002% (1 in 187,000 to 1 in 55,000) in a low HIV seroprevalence area. CONCLUSIONS: In both inner-city and suburban EDs, patient HIV seroprevalence varies with patient demographics and clinical presentation; the infection status of most HIV-positive patients is unknown to ED staff. The risk to an EDW of occupationally acquiring HIV infection varies by ED location and the nature and frequency of BC; this risk can be reduced by adherence to universal precautions.

Benzyl alcohol toxicity: impact on mortality and intraventricular hemorrhage among very low birth weight infants.

Benzyl alcohol preservative in intravascular flush solutions has been reported to cause neurologic deterioration and death in very low birth weight infants. Following the widespread discontinuation of the use of such solutions in newborns, scattered reports of decreased mortality and decreased incidence of intraventricular hemorrhage among small premature infants appeared in the pediatric literature. To better assess the true impact of benzyl alcohol toxicity in this group of infants, we undertook a detailed review of the medical records of all babies less than 1,250 g birth weight admitted to our neonatal intensive care unit for 13 months before and 13 months after the use of solutions containing benzyl alcohol was stopped. Significant decreases were found in both mortality rate (from 80.7% to 45.7%) and incidence of grade III/IV intraventricular hemorrhage (from 46% to 19%) among infants less than 1,000 g birth weight who did not receive the preservative compared with those who did. No significant changes were found in several other prenatal factors that could have contributed to this improvement in survival. We conclude that benzyl alcohol toxicity contributed significantly to both mortality and the occurrence of major intraventricular hemorrhage among infants weighing less than 1,000 g at birth and that solutions containing benzyl alcohol should never again be used in the care of such infants.

Nosocomial infections among neonates in high-risk nurseries in the United States. National Nosocomial Infections Surveillance System.

BACKGROUND: Nosocomial infections result in considerable morbidity and mortality among neonates in high-risk nurseries (HRNs). PURPOSE: To examine the epidemiology of nosocomial infections among neonates in level III HRNs. METHODS: Data were collected from 99 hospitals with HRNs participating in the National Nosocomial Infections Surveillance system, which uses standard surveillance protocols and nosocomial infection site definitions. The data included information on maternal acquisition of and risk factors for infection, such as device exposure, birth weight category (< or = 1000, 1001 through 1500, 1501 through 2500, and > 2500 g), mortality, and the relationship of the nosocomial infection to death. RESULTS: From October 1986 through September 1994, these hospitals submitted data on 13 179 nosocomial infections. The bloodstream was the most frequent site of nosocomial infection in all birth weight groups. Nosocomial pneumonia was the second most common infection site, followed by the gastrointestinal and eye, ear, nose, and throat sites. The most common nosocomial pathogens among all neonates were coagulase-negative staphylococci, Staphylococcus aureus, enterococci, Enterobacter sp, and Escherichia coli. Group B streptococci were associated with 46% of bloodstream infections that were maternally acquired; coagulase-negative staphylococci were associated with 58% of bloodstream infections that were not maternally acquired, most of which (88%) were associated with umbilical or central intravenous catheters. CONCLUSIONS: Bloodstream infections, the most frequent nosocomial infections in all birth weight groups, should be a major focus of surveillance and prevention efforts in HRNs. For bloodstream infections, stratification of surveillance data by maternal acquisition will help focus prevention efforts for group B streptococci outside the HRN. Within the nursery, bloodstream infection surveillance should focus on umbilical or central intravenous catheter use, a major risk factor for infection.

Nosocomial infections in coronary care units in the United States. National Nosocomial Infections Surveillance System.

To describe the epidemiology of nosocomial infections in Coronary Care Units (CCUs) in the United States, we analyzed data collected between 1992 and 1997 using the standard protocols of the National Nosocomial Infections Surveillance (NNIS) Intensive Care Unit (ICU) surveillance component. Data on 227,451 patients with 6,698 nosocomial infections were analyzed. Urinary tract infections (35%), pneumonia (24%), and primary bloodstream infections (17%) were almost always associated with use of an invasive device (93% with a urinary catheter, 82% with a ventilator, 82% with a central line, respectively). The distribution of pathogens differed from that reported from other types of ICUs. Staphylococcus aureus (21%) was the most common species reported from pneumonia and Escherichia coli (27%) from urine. Only 10% of reported urine isolates were Candida albicans. S. aureus (24%) was the more common bloodstream isolate than enterococci (10%). The mean overall patient infection rate was 2.7 infections per 100 patients. Device-associated infection rates for bloodstream infections, pneumonia, and urinary tract infections did not correlate with length of stay, number of hospital beds, number of CCU beds, or the hospital teaching affiliation, and were the best rates for comparisons between units. Use of invasive devices was lower than in other types of ICUs. Overall patient infection rates were lower than in other types of ICUs, which is largely explained by lower rates of invasive device usage.

Resistance to imipenem among selected gram-negative bacilli in the United States.

OBJECTIVES: Identification of imipenem resistance among selected gram-negative bacilli, especially Pseudomonas aeruginosa and Enterobacter species. METHODS: We analyzed 1986-1990 National Nosocomial Infection Surveillance (NNIS) data from 3,316 P aeruginosa isolates and 1,825 Enterobacter species isolates for which susceptibility results to imipenem were reported. RESULTS: For P aeruginosa, 11.1% of the isolates were resistant to imipenem; 16.1% were either intermediate-susceptible or resistant to the drug. A logistic regression model found that resistance was more common among P aeruginosa isolated from the respiratory tract, patients in intensive care units, and in teaching hospitals. Additionally, resistance to imipenem increased by 25% in teaching hospitals from 1986-1988 to 1989-1990. For Enterobacter species, 1.3% of the isolates were resistant to imipenem; 2.3% were either intermediate-susceptible or resistant to the drug. However, imipenem resistance for these isolates did not differ between the two periods and was not more common in patients in an intensive care unit or infections at any specific site. CONCLUSIONS: The frequency of resistance to imipenem is greater among P aeruginosa than among Enterobacter species. Resistance to imipenem among the P aeruginosa isolates is more common from strains isolated from patients with nosocomial infections in an intensive care unit, from the respiratory tract, and from teaching hospitals. Resistance appears to be increasing among nosocomial P aeruginosa isolated in teaching hospitals.

Ciprofloxacin resistance among nosocomial Pseudomonas aeruginosa and Staphylococcus aureus in the United States.

OBJECTIVE: We attempted to determine if an increase in resistance to ciprofloxacin occurred among nosocomial pathogens, especially Pseudomonas aeruginosa and Staphylococcus aureus. METHODS: We examined 1989-1992 ciprofloxacin susceptibility results from 8,517 P aeruginosa and 9,021 S aureus isolates associated with nosocomial infections reported to the National Nosocomial Infections Surveillance System. RESULTS: For S aureus, 27.1% of isolates were resistant to ciprofloxacin; of methicillin-resistant S aureus isolates, 80% also were resistant to ciprofloxacin. A logistic regression model found that ciprofloxacin resistance was more common among S aureus isolated from the urinary and respiratory tracts than from other sites of isolation, and among isolates that were methicillin resistant. After controlling for these factors, the model showed a 123% increase in the odds of ciprofloxacin resistance from 1989-1990 to 1991-1992. For P aeruginosa, 4.7% of the isolates were resistant to ciprofloxacin. Resistance varied by site of infection and rose most dramatically for respiratory tract isolates from 2.0% in 1989-1990 to 5.3% in 1991-1992. CONCLUSION: Resistance to ciprofloxacin is more frequent among nosocomial S aureus than among P aeruginosa and is increasing rapidly among S aureus isolates and from selected sites among P aeruginosa isolates.

Comparing nosocomial infection rates among surgical intensive-care units: the importance of separating cardiothoracic and general surgery intensive-care units.

Among surgical intensive-care units (ICUs), we assessed differences in risk-adjusted nosocomial infection rates between cardiothoracic (CT) and general surgery ICUs, using National Nosocomial Infection Surveillance data from 1987 to 1995. Device-associated rates and average length of stay were significantly lower in CT ICUs. Comparisons of risk-adjusted nosocomial infection rates among CT ICUs should be made separately from rates from general surgery ICUs.

Nosocomial infections in combined medical-surgical intensive care units in the United States.

OBJECTIVE: To describe the epidemiology of nosocomial infections in combined medical-surgical (MS) intensive care units (ICUs) participating in the National Nosocomial Infection Surveillance (NNIS) System. DESIGN: Analysis of surveillance data on 498,998 patients with 1,554,070 patient-days, collected between 1992 and 1998 from 205 MS ICUs following the NNIS Intensive Care Unit protocol, representing 152 participating NNIS hospitals in the United States. RESULTS: Infections at three major sites represented 68% of all reported infections (nosocomial pneumonia, 31%; urinary tract infections (UTIs), 23%; and primary bloodstream infections (BSIs), 14%: 83% of episodes of nosocomial pneumonia were associated with mechanical ventilation, 97% of UTIs occurred in catheterized patients, and 87% of primary BSIs in patients with a central line. In patients with primary BSIs, coagulase-negative staphylococci (39%) were the most common pathogens reported; Staphylococcus aureus (12%) was as frequently reported as enterococci (11%). Coagulase-negative staphylococcal BSIs were increasingly reported over the 6 years, but no increase was seen in candidemia or enterococcal bacteremia. In patients with pneumonia, S. aureus (17%) was the most frequently reported isolate. Of reported isolates, 59% were gram-negative bacilli. In patients with UTIs, Escherichia coli (19%) was the most frequently reported isolate. Of reported isolates, 31% were fungi. In patients with surgical-site infections, Enterococcus (17%) was the single most frequently reported pathogen. Device-associated nosocomial infection rates for BSIs, pneumonia, and UTIs did not correlate with length of ICU stay, hospital bed size, number of beds in the ICU, or season. Combined MS ICUs in major teaching hospitals had higher device-associated infection rates compared to all other hospitals with combined medical-surgical units. CONCLUSIONS: Nosocomial infections in MS ICUs at the most frequent infection sites (bloodstream, urinary, and respiratory tract) almost always were associated with use of an invasive device. Device-associated infection rates were the best available comparative rates between combined MS ICUs, but the distribution of device-associated rates should be stratified by a hospital's major teaching affiliation status.

Accuracy of reporting nosocomial infections in intensive-care-unit patients to the National Nosocomial Infections Surveillance System: a pilot study.

OBJECTIVE: To assess the accuracy of nosocomial infections data reported on patients in the intensive-care unit by nine hospitals participating in the National Nosocomial Infections Surveillance (NNIS) System. DESIGN: A pilot study was done in two phases to review the charts of selected intensive-care-unit patients who had nosocomial infections reported to the NNIS System. The charts of selected high- and low-risk patients in the same cohort who had no infections reported to the NNIS System also were included. In phase I, trained data collectors reviewed a sample of charts for nosocomial infections. Retrospectively detected infections that matched with previously reported infections were deemed to be true infections. In phase II, two Centers for Disease Control and Prevention (CDC) epidemiologists reexamined a sample of charts for which a discrepancy existed. Each sampled infection either was confirmed or disallowed by the epidemiologists. Confirmed infections also were deemed to be true infections. True infections from both phases were used to estimate the accuracy of reported NNIS data by calculating the predictive value positive, sensitivity, and specificity at each major infection site and the "other sites." RESULTS: The data collectors examined a total of 1,136 patients' charts in phase I. Among these charts were 611 infections that the study hospitals had reported to the CDC. The data collectors retrospectively matched 474 (78%) of the prospectively identified infections, but also detected 790 infections that were not reported prospectively. Phase II focused on the discrepant infections: the 137 infections that were identified prospectively and reported but not detected retrospectively, and the 790 infections that were detected retrospectively but not reported previously. The CDC epidemiologists examined a sample of 113 of the discrepant reported infections and 369 of the discrepant detected infections, and estimated that 37% of all discrepant reported infections and 43% of all discrepant detected infections were true infections. The predictive value positive for reported bloodstream infections, pneumonia, surgical-site infection, urinary tract infection, and other sites was 87%, 89%, 72%, 92%, and 80%, respectively; the sensitivity was 85%, 68%, 67%, 59%, and 30%, respectively; and the specificity was 98.3%, 97.8%, 97.7%, 98.7%, and 98.6%, respectively. CONCLUSIONS: When the NNIS hospitals in the study reported a nosocomial infection, the infection most likely was a true infection, and they infrequently reported conditions that were not infections. The hospitals also identified and reported most of the nosocomial infections that occurred in the patients they monitored, but accuracy varied by infection site. Primary bloodstream infection was the most accurately identified and reported site. Measures that will be taken to improve the quality of the infection data reported to the NNIS System include reviewing the criteria for definitions of infections and other data fields, enhancing communication between the CDC and NNIS hospitals, and improving the training of surveillance personnel in NNIS hospitals.

Methicillin-resistant Staphylococcus aureus in U.S. hospitals, 1975-1991.

OBJECTIVES: Analyze changes that have occurred among U.S. hospitals over a 17-year period, 1975 through 1991, in the percentage of Staphylococcus aureus resistant to beta-lactam antibiotics and associated with nosocomial infections. DESIGN: Retrospective review. The percentage of methicillin-resistant S aureus (MRSA) was defined as the number of S aureus isolates resistant to either methicillin, oxacillin, or nafcillin divided by the total number of S aureus isolates for which methicillin, oxacillin, or nafcillin susceptibility test results were reported to the National Nosocomial Infections Surveillance (NNIS) System. SETTING: NNIS System hospitals. RESULTS: Of the 66,132 S aureus isolates that were tested for susceptibility to methicillin, oxacillin, or nafcillin during 1975 through 1991, 6,986 (11%) were resistant to methicillin, oxacillin, or nafcillin. The percentage MRSA among all hospitals rose from 2.4% in 1975 to 29% in 1991, but the rate of increase differed significantly among 3 bed-size categories: < 200 beds, 200 to 499 beds, and > or = 500 beds. In 1991, for hospitals with < 200 beds, 14.9% of S aureus isolates were MRSA; for hospitals with 200 to 499 beds, 20.3% were MRSA; and for hospitals with > or = 500 beds, 38.3% were MRSA. The percentage MRSA in each of the bed-size categories rose above 5% at different times: in 1983, for hospitals with > or = 500 beds; in 1985, for hospitals with 200 to 499 beds; and in 1987, for hospitals with < 200 beds. CONCLUSIONS: This study suggests that hospitals of all sizes are facing the problem of MRSA, the problem appears to be increasing regardless of hospital size, and control measures advocated for MRSA appear to require re-evaluation. Further study of MRSA in hospitals would benefit our understanding of this costly pathogen.