The public release of hospital and physician mortality data in Pennsylvania. A case study.

OBJECTIVES: Using the public reports of the Pennsylvania Health Care Cost Containment Council on coronary artery bypass graft surgery for 1990 to 1992 as a case study, the authors assess the sensitivity of results to the choice of data and statistical methodology. METHODS: Using the Council's public-release data, surgical mortality and utilization were reanalyzed by standard linear models, empirical Bayes methods, Monte Carlo simulations, and hierarchical statistical models. RESULTS: Statistical power calculations demonstrate that the annual volume of bypass surgery for many hospitals and for most surgeons is too small for meaningful mortality comparisons. The number of hospitals and physicians designated as mortality "outliers" in the Council's reports results in part from a failure to adjust critical P values for multiple comparisons. Hierarchical statistical models implemented by mixed effects logistic regression, by contrast, can detect true differences in performance without producing false outliers. Mortality analyses are sensitive to the choice of comorbidities used for severity adjustment of a mortality model. Small-area analyses indicate large differences in the rates of bypass surgery across Pennsylvania, with lower population-based rates of surgery associated with higher population-based inpatient mortality. CONCLUSIONS: Analyses of mortality by operative procedure, rather than by patient diagnosis, should consider the potential for selection bias caused by the decision to elect surgery. The clinical and statistical issues of operative mortality are sufficiently complex to merit review by independent experts before public release of hospital and physician performance measures.

Efficacy of the Steri-Shield filtered exhaust helmet in limiting bacterial counts in the operating room during total joint arthroplasty.

Filtered exhaust helmets (space suits) have been reported as efficacious in decreasing wound infections following joint arthroplasty procedures; however, strict experimental control of the many variables related to the incidence of clinical infection has been lacking in these studies, making support of such conclusions difficult. Any potential benefit of filtered exhaust helmets in the control of infection rates can be logically assigned to the reduction of airborne bacterial counts within the operating room. A study was done using the Steri-Shield helmet (Bio-Medical Devices, Irvine, CA) to define its efficacy in limiting airborne bacterial contaminates during total joint arthroplasty. Forty-eight paired prosthetic hip and knee cases were evaluated using a block experimentation design. One case of each pair was done with all surgical and scrub personnel cloaked in Steri-Shield filtered exhaust helmets and the other with the same personnel in conventional paper hoods and masks. Air samples were obtained with an Anderson two-stage viable particle sampler (Anderson Samplers, Atlanta, GA) placed immediately adjacent to the wound. Quantitative microbial cultures were obtained. The filtered exhaust helmet evaluated in this study provided no increased protection against bacterial contamination in the area of the surgical field than conventional paper hoods and masks.

Comparison of three typing methods for clinical and environmental isolates of Aspergillus fumigatus.

To evaluate procedures used for epidemiologic analysis of outbreaks of aspergillosis, we analyzed a collection of 35 Aspergillus fumigatus isolates using three typing methods: isoenzyme analysis (IEA), random amplified polymorphic DNA (RAPD) analysis, and restriction endonuclease analysis (REA). Twenty-one isolates were from a single hospital, with four isolates coming from different patients. Three clinical isolates came from a different hospital, and 11 clinical or environmental isolates were derived from a culture collection. With IEA, the patterns of alkaline phosphatase, esterase, and catalase discriminated nine types. In contrast, 22 types were obtained with five different RAPD primers, and 21 types could be detected with three of these (R108, R151, and UBC90). Restriction endonuclease analysis of genomic DNA, digested with either XbaI, XhoI, or SalI, detected 3, 17, and 13 different REA types, respectively, and 22 types were identified by combining the data from the XhoI and SalI REAs. Twenty-eight types were obtainable with a combination of REA, IEA, and RAPD patterns. Overall, the results pointed to substantial genetic variation among the isolates. Though two isolates had markedly distinct genotypes, their morphologic features and exoantigens were consistent with their being A. fumigatus. The analysis will help in planning epidemiologic studies of aspergillosis.

Comparing hospital mortality in adult patients with pneumonia. A case study of statistical methods in a managed care program.

OBJECTIVE: To compare and contrast a managed care program's analysis of differences in hospital mortality with results obtained by accepted statistical methods. DESIGN: A re-analysis of computerized discharge data using the same method used by a managed care program, and using conventional methods of categorical data analysis. One thousand computer simulations of a method for comparing hospitals by severity-adjusted mortality were done to determine the probability of falsely identifying hospitals as high-mortality outliers. SETTING: 22 acute care hospitals in central Pennsylvania. PATIENTS: All adult patients with pneumonia (n = 4587; diagnosis-related groups 089-090) less than 65 years of age who were discharged from the 22 hospitals in 1989, 1990, and 1991, excluding patients with the acquired immunodeficiency syndrome and transplant recipients. MEASUREMENTS: In-hospital mortality adjusted for age and severity of illness using MedisGroups admission severity group score. RESULTS: The hospital that had the highest mortality for adult pneumonia according to the managed care program's analysis did not, according to an appropriate analysis, differ significantly from other area hospitals (likelihood ratio test, P = 0.23). Random variation in this sample of patients with a low average mortality rate (3.5%) showed a 60% chance that 1 or more of the 22 hospitals would be falsely identified as a "high-mortality outlier" when simplistic statistical methods were used. CONCLUSION: Organizations seeking to compare the quality of hospitals and physicians through outcome data need to recognize that simplistic methods applicable to large samples fail when applied to the outcomes of typical patients, such as those admitted for pneumonia. Although these comparisons are much in demand, careful attention must be paid to their statistical methods to ensure validity and fairness.

A predictive risk index for nosocomial pneumonia in the intensive care unit.

PURPOSE: To develop a scoring system for stratifying patients in intensive care units (ICUs) by risk of developing nosocomial pneumonia (NP) and to identify the time period associated with the highest risk. PATIENTS AND METHODS: Two hundred and three patients 18 years of age or older and residing in the ICU for 72 hours or more were followed until development of NP or death or for 48 hours after discharge from the ICU. After the identification of independent risk factors for NP, a scoring system was developed to arrive at a predictive risk index for NP. RESULTS: Twenty-six (12.8%) patients developed NP. The presence of a nasogastric (NG) tube [odds ratio (OR) = 6.48, 95% confidence intervals (CI) = 2.11 to 19.82], upper abdominal/thoracic surgery (OR = 4.34, 95% CI = 1.43 to 13.14), and bronchoscopy (OR = 2.95, 95% CI = 1.02 to 8.52), most commonly performed for respiratory toilet, were identified as independent risk factors on multivariate analysis. The risks associated with endotracheal intubation and altered consciousness, although not independently significant, were highest when these factors were present for 1 to 4 days after the 72 hours required for study entry (endotracheal intubation, OR = 2.2 to 2.5; altered consciousness, OR = 1.4 to 2.0). The risk then declined; ORs of less than 1 were observed at 7 days. The risk associated with the NG tube was highest during the first 6 days (OR = 6.0 to 19.5). Although a subsequent decrease in risk was observed, the OR was still greater than 2 at 7 days. To obtain a predictive risk index for NP, a scoring system was developed using a multivariate model. This system has a sensitivity of 85% and a specificity of 66% in predicting NP in this ICU population. CONCLUSION: ICU patients can be stratified into high- and low-risk groups for NP using a bedside scoring system. Endotracheal intubation, altered mental status, and NG tube are associated with the highest risk of NP during the first 1 to 6 days of their presence after 72 hours of stay in the ICU. After this time period, the risk associated with these factors decreases. Bronchoscopy may be an independent risk factor for NP that has not been previously recognized. This procedure, often done in the ICU for respiratory toilet, may be an avoidable risk in this group of patients.

Self-reported HIV and AIDS experience, practices, and knowledge of Pennsylvania family physicians.

BACKGROUND: Patients with disease caused by the human immunodeficiency virus (HIV), while still more commonly treated in urban settings, are being seen in nonurban areas in numbers rapidly outstripping the local availability of specialists with expertise in HIV or acquired immunodeficiency syndrome (AIDS). METHODS: A questionnaire designed to measure self-assessed experience, practices, and knowledge regarding basic aspects of HIV was mailed in 1989 to the 2177 members of the Pennsylvania Academy of Family Physicians. RESULTS: The response rate was 72 percent. Approximately 95 percent of physicians surveyed had been asked questions by patients about AIDS, 30 percent had a patient with a confirmed positive blood test, and 27 percent had a patient with symptomatic HIV disease in their practice. CONCLUSIONS: Although most family physicians indicated that they were comfortable in recognizing persons at risk, counseling, and using tests to diagnose HIV and AIDS, more than one-half reported practice patterns at variance with published guidelines. Respondents were most uncomfortable with their knowledge and skills regarding legal issues, state and community resources, and caring for patients with AIDS. Continuing medical education courses at local hospitals and written materials were the two methods of AIDS education most likely to be used by respondents.

Infectious disease manpower in the United States--1986. 1. Description of infectious disease physicians. Manpower and Training Committee, Infectious Diseases Society of America.

A survey designed to assess the number, type, and current practice patterns of all infectious disease (ID) physicians active in the United States in 1986 was carried out in early 1987. Of 4328 mailed questionnaires, 48.3% were returned. One-third of respondents were in private practice, one-third in academics, and the rest in industry or government. Women accounted for 12.4% of the total; they were younger and as a group spent a greater proportion of total effort in ID. Sixty-five percent of all respondents had greater than or equal to 2 years training in ID. Overall, private practitioners worked longer hours than academicians but spent slightly less effort devoted solely to ID. The proportion of total effort devoted to ID has increased among physicians newly entering practice. Seventy-five percent of all respondents held a teaching appointment. Older ID physicians worked less than 50 h/week and tended to have more administrative than patient care responsibilities. In 1986, there were the equivalent of 1792 full-time ID physicians in the United States or 1:134,000 population.

Infectious disease manpower in the United States--1986. 2. Changes in practice patterns over time and training needs. Manpower and Training Committee, Infectious Diseases Society of America.

Infectious disease-trained internal medicine physicians responding to a questionnaire survey (n = 1802) reported minor differences in time spent in patient care versus laboratory-based research whether they subsequently became practitioners or academicians. Both practitioners and academicians ranked hospital epidemiology first, followed by knowledge of hospital antibiotic policies in order of importance for new trainees to be taught. Internists with greater than 12 months of training in infectious diseases were divided into private practice versus academically based groups, and their distribution of time spent in various professional activities was analyzed by 5-year intervals for each cohort. These studies confirmed an increasing proportion of time spent in infectious disease-related patient care for new practitioners. Over time, patient care activities decreased and administrative activities increased in all groups. These data are important for estimating future manpower needs.

The occurrence and distribution of burnout among infectious diseases physicians.

The occurrence and distribution of the three dimensions of the burnout syndrome (emotional exhaustion, depersonalization, and lowered feelings of personal accomplishment) were studied among infectious diseases physicians. A written survey was mailed to the entire identified US population of infectious diseases physicians (n = 4328); a 46.3% response rate provided 1840 usable surveys. Statistical analyses of the data demonstrated that 43.5% of the physician sample reported high scores on emotional exhaustion, and 40.3% scored high on depersonalization. Personal accomplishment scores remained high, despite burnout levels, with 91.8% reporting high personal accomplishment. The highest percentage of burnout occurred among physicians in private practice settings (55%), followed by government settings (39%), and academia (37%). The high percentage of infectious diseases physicians experiencing burnout suggests the need for further research to establish trends, to determine if other types of physicians experience similar levels of burnout, to identify casual factors, and to develop avenues to reduce stress and facilitate coping.

Drug-resistant Nocardia asteroides infection in a patient with acquired immunodeficiency syndrome.

We have reported what we believe to be the first case of disseminated infection due to a multiply drug resistant strain of Nocardia asteroides in a patient with the acquired immunodeficiency syndrome and concomitant disseminated histoplasmosis. This strain of the organism fits a pattern of susceptibility that is rare among N asteroides isolates in general and has been called the type 5 pattern, described as a resistance to broad spectrum cephalosporins, ciprofloxacin, and all aminoglycosides except amikacin. The recognition of disease due to this group of organisms is especially important in patients with AIDS because sulfonamides, considered the drugs of choice for treatment of N asteroides infection, are associated with a high incidence of adverse effects in these patients.

Endophthalmitis caused by non-albicans species of Candida.

While endophthalmitis due to Candida albicans is commonly seen in clinical practice, non-albicans species of Candida (NAC) are rarely reported to cause this condition. The English-language literature from 1965 through 1989 was reviewed for cases of endophthalmitis due to NAC. Clinical details of six well-documented cases and one unpublished case are presented here. Large studies of nosocomial fungemia are also summarized to emphasize the growing contribution of NAC to the total number of infections in humans, especially immunocompromised patients. The infrequency of ocular involvement in NAC infection is supported by studies in experimental animals that demonstrate a lower incidence of NAC endophthalmitis than of C. albicans endophthalmitis in the presence of disseminated infection. Since endophthalmitis in the absence of exogenous infection suggests disseminated candidal infection and since most NAC strains are relatively resistant to amphotericin B, it is important to recognize and appropriately treat endophthalmitis due to NAC.

Control of epidemic methicillin-resistant Staphylococcus aureus.

We controlled the spread of epidemic methicillin-resistant Staphylococcus aureus (MRSA) infection in an 884-bed veterans' facility by cohorting known active MRSA carriers and MRSA-infected patients on one nursing unit. Simultaneously, all previously-institutionalized transfers into the veterans' facility were screened with swab cultures for MRSA at the time of admission. All MRSA patients were maintained on contact (gown and glove) or strict isolation and treated aggressively with topical and enteral antibiotics with the assistance of the infectious disease consultant. The monthly incidence of new MRSA patients dropped from a maximum of 16 per month to three or less per month within six months of instituting these infection control measures. There were no further MRSA bacteremias after the establishment of the MRSA cohort in a single unit. Aggressive cohort management of known MRSA patients and screening of previously-institutionalized patients on admission for MRSA controlled epidemic MRSA in this large institution.

Dermatoses among floral shop workers.

Concern about the increasing incidence of hand dermatitis in floral shop workers in the United States and its possible association to the plant Alstroemeria, a flower that has become popular since its introduction in 1981, prompted investigation of the prevalence and cause of hand dermatitis in a sample of floral workers. Fifty-seven floral workers were surveyed, and 15 (26%) reported hand dermatitis within the previous 12 months. Sixteen floral workers (eight with dermatitis) volunteered to be patch tested to the North American Contact Dermatitis Group Standard and Perfume Trays, a series of eight pesticides and 20 plant allergens. Of four of seven floral designers and arrangers who reported hand dermatitis, three reacted positively to patch tests to tuliposide A, the allergen in Alstroemeria. Patch test readings for all other plant extracts were negative. A positive reading for a test to one pesticide, difolatan (Captafol), was noted, the relevance of which is unknown.

Nosocomial sepsis related to intravascular access.

Many aspects of intravascular catheter care require further detailed study, including the optimal duration of cannulation by particular devices at each site, the influence of newer catheter materials on rates of bacterial colonization and infection, and the degree to which handling the catheter or its connections influences the rate of cannula-associated complication. Use of topical antibiotics and antimicrobial substances, their frequency of application, and similar issues also require further study. If the available knowledge of infection control practice could be consistently applied, it has been estimated that the rate of nosocomial bacteremia (including those secondary to infection at another site) could be reduced by up to 35%.

Acute pancreatitis caused by Legionella pneumophila.

A 57-year-old man had Legionella pneumonia and acute pancreatitis. The diagnosis of legionellosis was established by both culture and significant rise in antibody titer to Legionella pneumophila. The pancreatitis was diagnosed by elevated amylase and lipase values, peaking on the fourth hospital day, roentgenologic findings, and a clinical picture compatible with the disease. The patient was not an alcohol consumer and had had no previous pancreatic disease. We conclude that acute pancreatitis can be a manifestation of legionellosis.

Sterility and efficiency of two methods of cardiac output determination: closed loop and capped syringe methods.

A prospective randomized study was done to compare the fluid contamination rates of two methods used to determine cardiac output by thermodilution under in-use conditions. Fifty patients were randomly assigned to have cardiac output measured by the capped syringe or closed loop method, and were followed with quantitative cultures of fluid and catheter. The two treatment groups (n = 25 each) were comparable in terms of age, sex, intensive care unit location, underlying diagnosis, and mean number of cardiac output determinations performed in the first 24 hours after placement of the catheter line. No differences in the rates of fluid contamination were seen at any time between the two groups. The overall fluid contamination rate was 2/124 cultures (1.8% with 95% confidence interval +/- 4.1%) with one colony of Staphylococcus epidermidis from a closed loop sample and one colony of Pseudomonas maltophilia in a capped syringe sample. Although no difference in contamination rates was documented, the closed loop system required less nursing time and effort.