[Propensity score analysis in observational research: application to a study of prophylaxis against venous thromboembolism]. / L'analyse par score de propension Exemple d'application à une étude observationnelle sur la prophylaxie de la maladie thromboembolique veineuse.

INTRODUCTION: In observational studies, a significant difference in the outcomes between treated and untreated patients may be observed in absence of treatment effect and caused by differences in baseline characteristics. EXEGESIS: Propensity score analysis is a post hoc adjustment method which consists in deriving the conditional probability of receiving the treatment for a patient given his measured baseline characteristics (i.e., the propensity score). Matching each treated patient to an untreated one who has the nearest propensity score tends to balance baseline characteristics between the two groups and reduce the risk for overt bias. Then, the outcomes can be compared between matched treated and untreated patients. CONCLUSION: Propensity score analysis is relevant for clinical conditions and treatments for which randomized controlled trials are unlikely to be conducted. However, propensity analysis cannot adjust for unmeasured characteristics and sensitivity analysis should be performed to assess how sensitive the conclusions are to potential confounding factors.

Investigation of the ability of the Pneumonia Severity Index to accurately predict clinically relevant outcomes: a European study.

In order to confirm the validity of the Pneumonia Severity Index (PSI) for patients in Europe, data from adults with pneumonia who were enrolled in two prospective multicentre studies, conducted in France (Pneumocom-1, n = 925) and Spain (Pneumocom-2, n = 853), were compared with data from the original North American study (Pneumonia PORT, n = 2287). The primary outcome was 28-day mortality; secondary outcomes were subsequent hospitalisation for outpatients, and intensive care unit admission and length of stay for inpatients. All outcomes within individual risk classes, and mortality rates in low-risk (PSI I-III) and higher-risk patients, were compared across the three cohorts. Overall mortality rates were 4.7% in Pneumonia PORT, 6.3% in Pneumocom-2 and 10.6% in Pneumocom-1 (p <0.01), ranging from 0.4% to 1.6% (p 0.06) for low-risk patients and from 13.0% to 19.1% (p 0.24) for high-risk patients. Despite significant differences in baseline patient characteristics, none of the study outcomes differed within the low-risk classes. The sensitivity and negative predictive value of low-risk classification for mortality exceeded 93% and 98%, respectively. Thus, in two independent European cohorts, the PSI predicted patient outcomes accurately and reliably, particularly for low-risk patients. These findings confirm the validity of the PSI when applied to patients from Europe.

Temporal trends in outcomes of older patients with pneumonia.

BACKGROUND: It is unclear how outcomes of care for patients hospitalized for pneumonia have changed as patterns of health care delivery have changed during the 1990s. This study was performed to determine trends in outcomes of care for older patients hospitalized for pneumonia. METHODS: This retrospective analysis was based on Medicare claims and included most patients with pneumonia who were older than 65 years and admitted to acute care hospitals in Connecticut between October 1, 1991, and September 30, 1997 (fiscal years 1992-1997). We assessed the trends in hospital costs, discharge destination, hospital mortality rates, mortality rates within 30 days of discharge, and 30-day readmission rates for pneumonia. Multivariate logistic regression analyses were used to adjust for differences in patient characteristics. RESULTS: The mean (+/- SD) length of stay declined from 11.9 + 11.4 days to 7.7 + 7.2 days between 1992 and 1997. During this period, adjusted in-hospital mortality rates declined (P =.02), while the adjusted risk of discharge to a nursing facility increased (P<.001) and the adjusted risk of hospital readmission for pneumonia within 30 days of discharge increased (P =.05). The adjusted risk of death 30 days after discharge increased, although the difference was not statistically significant (P =.09). CONCLUSIONS: Between 1992 and 1997, the adjusted risks of mortality after discharge, placement in a nursing facility, and hospital readmission for pneumonia increased among older patients hospitalized for pneumonia, in association with a decline in mean hospital length of stay. These findings raise the question of whether the declining hospital length of stay has negatively affected patient outcomes. Arch Intern Med. 2000;160:3385-3391.

Associations between initial antimicrobial therapy and medical outcomes for hospitalized elderly patients with pneumonia.

BACKGROUND: Although medical practice guidelines exist, there have been no large-scale studies assessing the relationship between initial antimicrobial therapy and medical outcomes for patients hospitalized with pneumonia. OBJECTIVE: To determine the associations between initial antimicrobial therapy and 30-day mortality for these patients. METHODS: Hospital records for 12945 Medicare inpatients (> or = 65 years of age) with pneumonia were reviewed. Associations between initial antimicrobial regimens and 30-day mortality were assessed with Cox proportional hazards models, adjusting for baseline differences in patient characteristics, illness severity, and processes of care. Comparisons were made with patients treated with a non-pseudomonal third-generation cephalosporin alone (the reference group). RESULTS: Initial treatment with a second-generation cephalosporin plus macrolide (hazard ratio [HR], 0.71; 95% confidence interval [CI], 0.52-0.96), a non-pseudomonal third-generation cephalosporin plus macrolide (HR, 0.74; 95% CI, 0.60-0.92), or a fluoroquinolone alone (HR, 0.64; 95% CI, 0.43-0.94) was independently associated with lower 30-day mortality. Adjusted mortality among patients initially treated with these 3 regimens became significantly lower than that in the reference group beginning 2, 3, and 7 days, respectively, after hospital admission. Use of a beta-lactam/beta-lactamase inhibitor plus macrolide (HR, 1.77; 95% CI, 1.28-2.46) and an aminoglycoside plus another agent (HR, 1.21; 95% CI, 1.02-1.43) were associated with an increased 30-day mortality. CONCLUSIONS: In this study of primarily community-dwelling elderly patients hospitalized with pneumonia, 3 initial empiric antimicrobial regimens were independently associated with a lower 30-day mortality. The more widespread use of these antimicrobial regimens is likely to improve the medical outcomes for elderly patients with pneumonia.

Efficacy of pneumococcal vaccination in adults. A meta-analysis of randomized controlled trials.

BACKGROUND: Because of the prevalence of pneumococcal pneumonia, the substantial morbidity and mortality associated with many pneumococcal infections, and an increase in the incidence of antibiotic resistance among pneumococcal isolates, considerable efforts for disease prevention have been made using a polyvalent polysaccharide pneumococcal vaccine. Despite numerous clinical trials of the vaccine, its efficacy in the prevention of pneumococcal infections and other clinically relevant medical outcomes in adults remains uncertain. METHODS: To assess quantitatively the efficacy of pneumococcal vaccination, a MEDLINE literature search, manual reviews of article bibliographies, and communications with pneumococcal vaccine investigators were used to identify randomized controlled trials of the pneumococcal vaccine. Independent review of 594 articles revealed nine randomized trials with 12 vaccine and control study groups that evaluated clinically relevant outcomes in adults. To estimate a summary effect size for all outcomes, Mantel-Haenszel odds ratios (ORs) and Dersimonian and Laird rate differences (RDs) and their associated 95% confidence intervals (CIs) were computed. RESULTS: Summary ORs demonstrated a statistically significant protective effect of the vaccine for four pneumococcal infection-related outcomes: definitive pneumococcal pneumonia (OR = 0.34; 95% CI = 0.24 to 0.48), definitive pneumococcal pneumonia for vaccine-containing pneumococcal antigen types only (vaccine types only) (OR = 0.17; 95% CI = 0.09 to 0.33), presumptive pneumococcal pneumonia (OR = 0.47; 95% CI = 0.35 to 0.63), and presumptive pneumococcal pneumonia (vaccine types only) (OR = 0.39; 95% CI = 0.26 to 0.59). The summary RDs, which account for heterogeneity among studies, confirmed a statistically significant protective effect for two of these same outcomes: definitive pneumococcal pneumonia (RD = 4/1000; 95% CI = 0/1000 to 7/1000) and definitive pneumococcal pneumonia (vaccine types only) (RD = 8/1000; 95% CI = 1/1000 to 16/1000). Summary ORs and RDs failed to demonstrate a protective effect for pneumonia (all causes), bronchitis, and mortality (all causes) or mortality due to pneumonia or pneumococcal infection. Subgroup analyses showed that for all four pneumococcal infection-related outcomes, vaccine efficacy differed for high- and low-risk subjects, demonstrating efficacy for low-risk subjects and lack of efficacy for high-risk subjects. CONCLUSIONS: Pneumococcal vaccination appears efficacious in reducing bacteremic pneumococcal pneumonia in low-risk adults. However, evidence from randomized controlled trials fails to demonstrate vaccine efficacy for pneumococcal infection-related or other medical outcomes in the heterogeneous group of subjects currently labeled as high risk.

Influence of age on symptoms at presentation in patients with community-acquired pneumonia.

BACKGROUND: Advanced age has become a well-recognized risk factor for death in patients with pneumonia. It may also be associated with reduced symptom reporting, raising the possibility that diagnosis and treatment may be delayed in older patients. OBJECTIVE: To evaluate the association between age and the presenting symptoms in patients with community-acquired pneumonia. METHODS: This study was conducted at inpatient and outpatient facilities at 3 university hospitals, 1 community hospital, and 1 staff-model health maintenance organization. Patients included adults (age > or = 18 years) with clinical and radiographic evidence of pneumonia, who were able to complete a baseline interview. The presence of 5 respiratory symptoms and 13 nonrespiratory symptoms were recorded during a baseline patient interview. A summary symptom score was computed as the total number of symptoms at presentation. RESULTS: The 1812 eligible study patients were categorized into 4 age groups: 18 through 44 years (43%), 45 through 64 years (25%), 65 through 74 years (17%), and 75 years or older (15%). For 17 of the 18 symptoms, there were significant decreases in reported prevalence with increasing age (P < .01). In a linear regression analysis, controlling for patient demographics, comorbidity, and severity of illness at presentation, older age remained associated with lower symptom scores (P < .001). CONCLUSIONS: Respiratory and nonrespiratory symptoms are less commonly reported by older patients with pneumonia, even after controlling for the increased comorbidity and illness severity in these older patients. Recognition of this phenomenon by clinicians and patients is essential given the increased mortality in elderly patients with pneumonia.

Do pulmonary radiographic findings at presentation predict mortality in patients with community-acquired pneumonia?

BACKGROUND: Previous studies have reported conflicting results on whether pulmonary radiographic findings predict mortality for patients with community-acquired pneumonia (CAP). OBJECTIVE: To determine whether pulmonary radiographic findings at presentation are independently associated with 30-day mortality in patients with suspected CAP. METHODS: This study was conducted as part of the Pneumonia Patient Outcomes Research Team multicenter, prospective cohort study of ambulatory and hospitalized patients with clinical and radiographic evidence of CAP. For each patient with CAP, a structured data form was completed by a panel of radiologists to evaluate the radiographic pattern of infiltrate, number of lobes involved, presence of pleural effusion, and 8 other radiographic characteristics. Cox proportional hazards models were used to evaluate the independent association between radiographic findings and 30-day mortality, while controlling for patient demographic, clinical, and laboratory characteristics with a known association with this outcome. RESULTS: Of 2287 patients enrolled in the Pneumonia Patient Outcomes Research Team cohort study, 1906 patients (83.3%) had a pulmonary radiographic infiltrate confirmed by the radiology panel. Overall, 30-day mortality in this cohort was 4.9%. Univariate regression analyses demonstrated the following radiographic characteristics to be significantly associated with 30-day mortality: (1) bilateral pleural effusions (risk ratio [RR], 7.0; 95% confidence interval [CI], 3.9-12.6); (2) a pleural effusion of moderate or greater size (RR, 3.4; 95% CI, 1.4-8.4); (3) 2 or more lobes involved with infiltrate (RR, 2.5; 95% CI, 1.6-3.8); (4) bilateral infiltrate (RR, 2.8; 95% CI, 1.9-4.2); (5) bronchopneumonia (RR, 1.6; 95% CI, 1.0-2.7); and (6) air bronchograms (RR, 0.5; 95% CI, 0.2-0.9). Multivariate analysis of radiographic features and other clinical characteristics showed the presence of bilateral pleural effusions (RR, 2.8; 95% CI, 1.4-5.8) was independently associated with mortality. CONCLUSIONS: In patients with CAP, the presence of bilateral pleural effusions is an independent predictor of short-term mortality. This finding, which is available at presentation, can help guide physicians' assessment of prognosis in CAP.

Preferences for home vs hospital care among low-risk patients with community-acquired pneumonia.

OBJECTIVE: To measure preferences for initial outpatient vs hospital care among low-risk patients who were being actively treated for community-acquired pneumonia (CAP). METHODS: Study patients included 159 patients with CAP, 57 (36%) initially hospitalized, who were identified as being at low risk for early mortality using a validated prediction model. Subjects were enrolled from university and community health care facilities located in Boston, Mass, Halifax, Nova Scotia, and Pittsburgh, Pa, participating in the Pneumonia Patient Outcome Research Team prospective cohort study of CAP. Three utility assessment techniques (category scaling, standard gamble, and willingness to pay) were used to measure the strength of patient preferences for the site of care for low-risk CAP. At the time of initial therapy or during the early recuperative period, patient preferences were assessed across a spectrum of potential clinical outcomes using 7 standardized pneumonia clinical vignettes. RESULTS: Responses to the 7 pneumonia scenarios indicated that most patients consistently preferred outpatient-based therapy. This pattern was observed regardless of whether patients had actually been treated initially at home or in a hospital. Patients (74%) who stated that they generally preferred home care for low-risk CAP were willing to pay a mean of 24% of 1 month's household income to be assured of this preference. Preference for home care, as measured by the category scaling and the willingness to pay, persisted after adjustment for sociodemographic and baseline health status covariates. Sixty nine percent of interviewed patients said that their physician alone determined whether they would be treated in the hospital or at home. Only 11% recalled being asked if they had a preference for either site of care. CONCLUSIONS: Most patients, even those treated initially in a hospital, who were at low risk for mortality from CAP prefer outpatient treatment. However, most physicians appear not to involve patients in the site-of-care decision. More explicit discussion of patient preferences for the location of care would likely yield more highly valued care by patients as well as less costly treatment for CAP.

The hospital discharge decision for patients with community-acquired pneumonia. Results from the Pneumonia Patient Outcomes Research Team cohort study.

BACKGROUND: The hospital discharge decision directly influences the length of stay in patients with community-acquired pneumonia, yet no information exists on how physicians make this decision. OBJECTIVES: To identify the factors physicians considered the factors responsible for extending length of hospital stay in clinically stable patients, and the outpatient medical services that would allow earlier hospital discharge for patients with community-acquired pneumonia. METHODS: Physicians responsible for the hospital discharge decision of patients with community-acquired pneumonia were asked to identify the factors responsible for extending stay in patients hospitalized beyond stability, and the medical services that could have allowed earlier hospital discharge to occur. RESULTS: For the 418 eligible patients with community-acquired pneumonia identified during the study, 332 questionnaires (79%) were completed by 168 physicians. Physicians believed 71 patients (22%) were discharged from the hospital 1 day or more (median, 2.5 days) after reaching clinical stability. The most common factors rated as being "very important" in delaying discharge were diagnostic evaluation or treatment of comorbid illness (56%), completion of a "standard course" of antimicrobials (15%), and delays with arrangements for long-term care (14%). Among the 302 patients with available information on both length of hospital stay and stability at discharge, median length of stay was 7.0 days for the 29 low-risk patients hospitalized beyond reaching clinical stability and 5.0 days for the remaining 128 low-risk patients (P < .005); median length of stay was 12.5 days for the 42 medium- and high-risk patients hospitalized beyond reaching clinical stability and 8.0 days in the remaining 113 medium- and high-risk patients (P < .001). Frequently cited medical services that "probably" or "definitely" would have allowed earlier discharge to occur included availability of home intravenous antimicrobial infusion (26%) and home visits by nurses (20%). CONCLUSIONS: Physicians believed that diagnostic evaluation or treatment of comorbid illness, completion of a standard course of antimicrobial therapy, and delays with arrangements for long-term care delayed hospital discharge in clinically stable patients. Addressing the efficiency of these aspects of inpatient medical care, as well as providing home treatment programs, could decrease the length of hospital stay in patients with community-acquired pneumonia.

The hospital admission decision for patients with community-acquired pneumonia. Results from the pneumonia Patient Outcomes Research Team cohort study.

BACKGROUND: The hospital admission decision directly influences the magnitude of resource use in patients with community-acquired pneumonia, yet little information exists on how medical practitioners make this decision. OBJECTIVES: To determine which factors medical practitioners consider in making the hospital admission decision and which health care services they believe would allow ambulatory treatment of low-risk hospitalized patients with community-acquired pneumonia. METHODS: Medical practitioners responsible for the hospital admission decision for low-risk patients with community-acquired pneumonia were asked to describe patient characteristics at initial examination that influenced the hospitalization decision, and to identify the health care services that would have allowed initial outpatient treatment of hospitalized patients. RESULTS: A total of 292 medical practitioners completed questionnaires for 472 (76%) of the 624 low-risk patients eligible for this study. Although all patients had a predicted probability of death of less than 4%, practitioners estimated that 5% of outpatients and 41% of inpatients had an expected 30-day risk of death of more than 5%. Univariate analyses identified 3 practitioner-rated factors that were nearly universally associated with hospitalization: hypoxemia (odds ratio, 173.3; 95% confidence interval, 23.8-1265.0), inability to maintain oral intake (odds ratio, 53.3; 95% confidence interval, 12.8-222.5), and lack of patient home care support (odds ratio, 54.4; 95% confidence interval, 7.3-402.6). In patients without these 3 factors, logistic regression analysis demonstrated that practitioner-estimated risk of death of more than 5% had a strong independent association with hospitalization (odds ratio, 18.4; 95% confidence interval, 6.1-55.7). Practitioners identified home intravenous antibiotic therapy and home nursing observation as services that would have allowed outpatient treatment of more than half (68% and 59%, respectively) of the patients initially hospitalized for treatment. CONCLUSIONS: Practitioners' survey responses suggest that the availability of outpatient intravenous antimicrobial therapy and home nursing care would allow outpatient care for a large proportion of low-risk patients who are hospitalized for community-acquired pneumonia. These data also suggest that methods to improve practitioners' identification of low-risk patients with community-acquired pneumonia could decrease the hospitalization of such patients. Future studies are required to help physicians identify which low-risk patients could safely be treated in the outpatient setting on the basis of clinical information readily available at presentation.

Processes and outcomes of care for patients with community-acquired pneumonia: results from the Pneumonia Patient Outcomes Research Team (PORT) cohort study.

BACKGROUND: Although understanding the processes of care and medical outcomes for patients with community-acquired pneumonia is instrumental to improving the quality and cost-effectiveness of care for this illness, limited information is available on how physicians manage patients with this illness or on medical outcomes other than short-term mortality. OBJECTIVES: To describe the processes of care and to assess a broad range of medical outcomes for ambulatory and hospitalized patients with community-acquired pneumonia. METHODS: This prospective, observational study was conducted at 4 hospitals and 1 health maintenance organization in Pittsburgh, Pa, Boston, Mass, and Halifax, Nova Scotia. Data were collected via patient interviews and reviews of medical records for 944 outpatients and 1343 inpatients with clinical and radiographic evidence of community-acquired pneumonia. Processes of care and medical outcomes were assessed 30 days after presentation. RESULTS: Only 29.7% of outpatients had 1 or more microbiologic tests performed, and only 5.7% had an assigned microbiologic cause. Although 95.7% of inpatients had 1 or more microbiologic tests performed, a cause was established in only 29.6%. Six outpatients (0.6%) died, and 3 of these deaths were pneumonia related. Of surviving outpatients, 8.0% had 1 or more medical complications. At 30 days, 88.9% (nonemployed) to 95.6% (employed) of the surviving outpatients had returned to usual activities, yet 76.0% of outpatients had 1 or more persisting pneumonia-related symptoms. Overall, 107 inpatients (8.0%) died, and 81 of these deaths were pneumonia related. Most surviving inpatients (69.0%) had 1 or more medical complications. At 30 days, 57.3% (non-employed) to 82.0% (employed) of surviving inpatients had returned to usual activities, and 86.1% had 1 or more persisting pneumonia-related symptoms. CONCLUSIONS: In this study, conducted primarily at hospital sites with affiliated medical education training programs, virtually all outpatients and most inpatients had pneumonia of unknown cause. Although outpatients had an excellent prognosis, pneumonia-related symptoms often persisted at 30 days. Inpatients had substantial mortality, morbidity, and pneumonia-related symptoms at 30 days.

Hospitalization decision in patients with community-acquired pneumonia: a prospective cohort study.

PURPOSE: To identify a low-risk subset of patients with community-acquired pneumonia that could safely be treated in the ambulatory setting; and to assess how clinicians make the hospitalization decision. PATIENTS AND METHODS: We performed a prospective, observational study of 280 ambulatory and hospitalized adults with clinical and radiographic evidence of pneumonia. Patients were followed to assess all potential morbid complications and 6-week mortality. Physicians responsible for managing these patients were surveyed to assess the reasons for treating in a hospital or ambulatory setting and the therapies that dictate hospitalization. RESULTS: Sixty-one percent (170 of 280) of patients did not have an indication for admission at presentation using modified Appropriateness Evaluation Protocol criteria (a severe vital sign abnormality, alteration in mental status, suppurative complication, arterial hypoxemia, severe laboratory abnormality, or an acute coexistent medical problem requiring admission independent of the pneumonia). Among these 170 patients, 38% had a complicated course defined as death within 6 weeks, development of a new suppurative or medical complication due to pneumonia, intensive care unit admission, persistent fever or use of intravenous fluids or oxygen beyond 3 days, hospitalization lasting more than 3 days, or subsequent hospitalization in patients initially treated in the ambulatory setting. Five predisposing factors for a complicated course were identified in logistic regression models. The odds ratio for age more than 65 years was 2.7; for comorbid illness, 3.2; for temperature more than 38.3 degrees C (101 degrees F), 4.1; for immunosuppression, 12.0; and for a high-risk etiology, 23.3. The risk of a complicated course increased linearly with the number of risk factors, from 12% with none to 100% with four or more factors (p less than 0.001). Physicians most often relied on the general clinical appearance of the patient when making the triage decision, and most commonly cited intravenous antibiotics and chest physical therapy as treatments requiring hospitalization. CONCLUSIONS: If validated, our findings could improve physicians' assessment of prognosis, and may identify a low-risk subset of patients with community-acquired pneumonia who could safely be managed in the ambulatory setting.

Fever of nosocomial origin: etiology, risk factors, and outcomes.

OBJECTIVES: To identify the etiologies and risk factors associated with fever of nosocomial origin and compare the outcomes of patients with and without nosocomial fever. DESIGN: A prospective study with a case-control component. SETTING: The three medical services of a university hospital. PATIENTS AND METHODS: We evaluated 100 patients with fever of nosocomial origin. Etiologies were assigned by the consensus of three independent reviewers using Centers for Disease Control guidelines. Predictors of bacterial etiology were identified using logistic regression methods. Controls matched for age (+/- 5 years), sex, and comorbidity were identified for 65 of the nosocomial fever cases. Cases and controls were compared with conditional logistic regression to identify risk factors for fever of nosocomial origin. Cases and controls were followed at 1 month to compare length of stay, mortality, and hospital readmission rates. MEASUREMENTS AND MAIN RESULTS: An etiology was assigned in 81 cases: 51 had bacterial infection, 5 had nonbacterial infection, and 25 had a noninfectious etiology. Multivariate predictors of bacterial infection were the presence of diabetes mellitus (odds ratio [OR] = 5.74; 95% confidence interval of the odds ratio [CI] = 1.61 to 20.40), length of stay to fever onset after admission greater than 10 days (OR = 3.98, CI = 1.04 to 15.22), maximum temperature greater than 38.7 degrees C (OR = 3.37, CI = 1.28 to 8.88), and white blood cell count greater than 10 x 10(9)/L (OR = 4.64, CI = 1.68 to 12.77). Comparison of cases and controls indicated that patients with nosocomial fever had more invasive procedures in the 72 hours preceding enrollment in the study than controls (OR = 1.46, CI = 1.03 to 2.08). Although cases were hospitalized almost twice as long as controls (21.5 versus 12.5 days; p < 0.0001), neither 30-day mortality nor hospital readmission rates differed between cases and controls. CONCLUSION: Nosocomial fever often does not represent nosocomial infection and may constitute a wide variety of other conditions. Since bacterial etiology of fever is a major concern, clinical and laboratory findings can be used to predict patients at high risk for an infectious process. A reduction in the number of invasive procedures may decrease the incidence of nosocomial fever. New-onset nosocomial fever leads to an increase in length of hospital stay but is not a significant predictor of mortality.

Validation of a pneumonia prognostic index using the MedisGroups Comparative Hospital Database.

PURPOSE: Our purpose was to validate a previously developed pneumonia-specific prognostic index in a large, multicenter population. PATIENTS AND METHODS: We developed a pneumonia-specific prognostic index in a prospective, multicenter study of 346 patients with clinical and radiographic evidence of pneumonia admitted to 3 Pittsburgh hospitals (the derivation cohort), and validated the index in 14,199 patients with a principal ICD-9-CM diagnosis of pneumonia admitted to 78 hospitals in the 1989 MedisGroups Comparative Hospital Database (the validation cohort). The prognostic index classified patients into five ordered risk classes based on six predictors of mortality: age greater than 65 years, pleuritic chest pain, a vital sign abnormality, altered mental status, neoplastic disease, and high-risk pneumonia etiology. Each patient in the validation cohort was assigned to a risk class by obtaining values for the index's six predictors in the MedisGroups population. The performance of the prognostic index in the derivation and validation cohorts was assessed by comparing hospital mortality rates within each of the index's five prognostic risk classes. RESULTS: The hospital mortality rate was 13.0% in the derivation cohort, and 11.1% in the validation cohort (p = 0.26). The agreement in the risk class-specific mortality rates was striking with the exception of class V: in class I, mortality was 0% in the derivation cohort versus 1% in the validation cohort; in class II, 0% versus 1.1%; class III, 10.9% versus 8.6%; class IV, 21.8% versus 26.2%; and class V, 73.7% versus 37.7%. There were no statistically significant differences in mortality rates within the first four risk classes, which represented the vast majority of patients in the derivation (94%) as well as the validation (98%) cohorts. CONCLUSIONS: These data support the generalizability of a pneumonia-specific prognostic index. This index, which performs exceptionally well in classifying low-risk patients, may help physicians identify patients with community-acquired pneumonia who could safely be managed in the ambulatory setting, or if hospitalized, the patients that could be treated with abbreviated inpatient care.

Resting electrocardiograms as baseline tests: impact on the management of elderly patients.

PURPOSE: We evaluated management decisions of house staff in a university hospital emergency department (ED) to determine the impact of baseline electrocardiogram (ECG) results on the care of elderly patients. SUBJECTS AND METHODS: The study employed a prospective before-after research design. The subjects included internal medicine house staff assigned to the ED and the 47 elderly patients treated by them during 58 ED visits. The measures of impact on patient care included how often and in what ways comparison of baseline and ED ECG results changed the house staff's diagnostic impressions and treatment plans. To determine changes, we asked house staff to indicate their impressions and management plans before and after comparing baseline and ED ECG results. In addition, a panel of faculty members used a modification of the nominal group process to determine treatment changes and to estimate benefit to patients from the changes. RESULTS: House staff indicated that baseline ECG results led them to change diagnostic impressions during 13 (22%) and treatment plans during eight (14%) of the 58 visits. Six of the patients benefited from the treatment changes. The faculty judges indicated that baseline ECG findings would have altered their diagnostic impressions in eight (14%) and treatment decisions in three (5%) of the 58 cases, one of which would have benefited from the change. All patients for whom there were changes of impression or plan had prior histories of cardiovascular disease. CONCLUSION: We conclude that baseline ECGs provide ED house staff with diagnostic information that often leads to changes in management of elderly patients with prior histories of cardiovascular disease. Future studies should evaluate the efficacy of baseline ECGs in patients without such histories.

Relation between length of hospital stay and costs of care for patients with community-acquired pneumonia.

PURPOSE: Patients with pneumonia often remain hospitalized after becoming clinically stable, without demonstrated benefits on outcome. The purposes of this study were to assess the relation between length of hospital stay and daily medical care costs and to estimate the potential cost savings associated with a reduced length of stay for patients with pneumonia. SUBJECTS AND METHODS: As part of a prospective study of adults hospitalized with community-acquired pneumonia at a community hospital and two university teaching hospitals, daily medical care costs were estimated by multiplying individual charges by department-specific cost-to-charge ratios obtained from each hospital's Medicare cost reports. RESULTS: The median total cost of hospitalization for all 982 inpatients was $5, 942, with a median daily cost of $836, including $491 (59%) for room and $345 (41%) for non-room costs. Average daily non-room costs were 282% greater on the first hospital day, 59% greater on the second day, and 19% greater on the third day than the average daily cost throughout the hospitalization (all P <0.05), and were 14% to 72% lower on the last 3 days of hospitalization. Average daily room costs remained relatively constant throughout the hospital stay, with the exception of the day of discharge. A projected mean savings of $680 was associated with a 1-day reduction in length of stay. CONCLUSIONS: Despite institutional differences in total costs, patterns of daily resource use throughout hospitalization were similar at all institutions. A 1-day reduction in length of stay might yield substantial cost-savings.

Erythromycin ototoxicity: prospective assessment with serum concentrations and audiograms in a study of patients with pneumonia.

BACKGROUND AND METHODS: The incidence and risk factors for erythromycin-induced ototoxicity are unknown. We conducted a prospective, nested case-control study of assessment of auditory function in patients receiving erythromycin versus other antibiotics (control group) for community-acquired pneumonia. Sequential audiograms were performed during antibiotic therapy for both cases and controls by an audiologist unaware of the identity of the therapy administered. Erythromycin serum concentrations were obtained for all patients receiving erythromycin. RESULTS: Symptomatic ototoxicity (tinnitus or hearing loss) confirmed by audiograms was documented in five of 30 patients receiving erythromycin and none of 15 receiving other antibiotics. Ototoxicity was significantly related to high peak concentration and high AUC 0-infinity as a function of decreased total systemic clearance. Ototoxicity occurred only in those patients who received 4 g/day versus 2 g/day or no erythromycin (p = 0.05). Ototoxicity resolved in all patients within 6 to 14 days after discontinuation of therapy. CONCLUSIONS: Erythromycin ototoxicity is dose- and serum concentration-dependent. Patients receiving erythromycin, especially at a total daily dose of 4 g, should be monitored regularly for subjective evidence of sensorineural hearing dysfunction. Ototoxicity is reversible if the diagnosis is made early in the course.

Ambulatory patients with community-acquired pneumonia: the frequency of atypical agents and clinical course.

OBJECTIVES: To determine the etiology of community-acquired pneumonia in patients treated in an ambulatory setting, using serological methods, and to compare presenting symptoms, radiographic manifestations, and clinical outcomes of patients with pneumonia of "atypical" and undetermined etiology. PATIENTS AND METHODS: This prospective cohort study was conducted in emergency room and outpatient facilities of Victoria General Hospital, Halifax, Nova Scotia, and in offices of participating family doctors based in Halifax. One hundred forty-nine adults with acute onset of one or more symptoms or signs suggestive of pneumonia and radiographic evidence of pneumonia who provided informed consent were enrolled. Patients known to be HIV positive or who had been discharged from a hospital within the previous 10 days were ineligible for enrollment. Demographic features and clinical data were collected by direct patient interview and chart review by trained research nurses. Outcome measures included quantitative evaluation of pneumonia-specific symptoms, and responses to the Short Form 36 Health Survey at presentation and at 30 days after presentation. Information was also collected on each patient's health prior to pneumonia, as well as the time until each patient's self-reported return to work and to usual activities. The etiology of pneumonia was determined by testing acute and convalescent serum samples for antibodies to Legionella pneumophila serogroup 1, Mycoplasma pneumoniae, Chlamydia pneumoniae, Chlamydia psittaci, Coxiella burnetii, adenovirus, respiratory syncytia virus, influenza viruses A and B, and parainfluenza viruses 1, 2, 3. RESULTS: The study population consisted of 149 patients, 54 (36%) of whom were men, with a mean age (+/- SD) of 41 +/- 15 years. An etiological diagnosis was made in 74 (49.7%) patients using serological methods. Etiological agents included M pneumoniae 34 (22.8%); C pneumoniae 16 (10.7%); M pneumoniae and C pneumoniae 5 (3.4%); C burnetii 4 (2.7%); influenza A virus 4 (2.7%); and other agents 6% (7.4%). Three patients (2%) had a conventional bacterial etiology, and 72 patients (48.3%) had pneumonia of undetermined etiology. Patients with pneumonia of known (atypical) and undetermined etiology were similar in terms of age, gender, race, education, employment, and comorbidity. Despite a higher proportion of patients with pneumonia of known etiology reporting sweats, chills, and headache at presentation, the two groups were similar for symptom severity and bother. The patients with pneumonia of undetermined etiology were more likely to have multilobar pneumonia (P < 0.02). Both patients with atypical pneumonia and those with pneumonia of undetermined etiology suffered severe deterioration of physical functioning with a marked but incomplete recovery at 30 days. Those with atypical pneumonia had higher physical functioning and general mental health scores at 30 days. CONCLUSIONS: Nearly half the cases of ambulatory community-acquired pneumonia are due to "atypical" agents. It is not possible to reliably distinguish patients with atypical pneumonia from those with pneumonia of undetermined etiology by clinical features at baseline. The outcomes in terms of resolution of symptoms, functional status, return to work, and return to usual activities are essentially similar in the two groups.

Variation in length of hospital stay in patients with community-acquired pneumonia: are shorter stays associated with worse medical outcomes?

PURPOSE: To assess the variation in length of stay for patients hospitalized with community-acquired pneumonia and to determine whether patients who are treated in hospitals with shorter mean stays have worse medical outcomes. SUBJECTS AND METHODS: We prospectively studied a cohort of 1,188 adult patients with community-acquired pneumonia who had been admitted to one community and three university teaching hospitals. We compared patients' mean length of stay, mortality, hospital readmission, return to usual activities, return to work, and pneumonia-related symptoms among the four study hospitals. All outcomes were adjusted for baseline differences in severity of illness and comorbidity. RESULTS: Adjusted interhospital differences in mean length of stay ranged from 0.9 to 2.3 days (P <0.001). When the risk of each medical outcome was compared between patients admitted to the hospital with the shortest length of stay and those admitted to longer stay hospitals, there were no differences in mortality [relative risk (RR) = 0.7; 95% CI, 0.3 to 1.7], hospital readmission (RR = 0.8; 95% CI, 0.5 to 1.2), return to usual activities (RR = 1.1; 95% CI, 0.9 to 1.3), or return to work (RR = 1.2; 95% CI, 0.8 to 2.0) during the first 14 days after discharge, or in the mean number of pneumonia-related symptoms 30 days after admission (P = 0.54). CONCLUSIONS: We observed substantial interhospital variation in the lengths of stay for patients hospitalized with community-acquired pneumonia. The finding that medical outcomes were similar in patients admitted to the hospital with the shortest length of stay and those admitted to hospitals with longer mean lengths of stay suggests that hospitals with longer stays may be able to reduce the mean duration of hospitalization for this disease without adversely affecting patient outcomes.

Variations in antimicrobial use and cost in more than 2,000 patients with community-acquired pneumonia.

PURPOSE: To assess the patterns of antimicrobial use, costs of antimicrobial therapy, and medical outcomes by institution in patients with community-acquired pneumonia. PATIENTS AND METHODS: The route, dose, and frequency of administration of all antimicrobial agents prescribed within 30 days of presentation were recorded for 927 outpatients and 1328 inpatients enrolled in the Pneumonia Patient Outcomes Research Team (PORT) multicenter, prospective cohort study. Total antimicrobial costs were estimated by summing drug costs, using average wholesale price for oral agents and institutional acquisition prices for parenteral agents, plus the costs associated with preparation and administration of parenteral therapy. Thirty-day outcome measures were mortality, subsequent hospitalization for outpatients, and hospital readmission for inpatients. RESULTS: Significant variation (P <0.05) in prescribing practices occurred for 17 of the 23 antimicrobial agents used in outpatients across 5 treatment sites, and for 18 of the 20 parenteral agents used in inpatients across 4 treatment sites. The median duration of antimicrobial therapy for treatment site ranged from 11 to 13 days for outpatients (P=0.01), and from 13 to 15 days for inpatients (P=0.49). The overall median cost of antimicrobial therapy was $12.90 for outpatients, and ranged from $10.80 to $58.90 among treatment sites (P <0.0001). The overall median cost of antimicrobial therapy was $228.70 for inpatients, and ranged from $183.70 to $315.60 among sites (P <0.0001). Mortality and hospital readmission for inpatients were not significantly different across sites after adjusting for baseline differences in patient demographic characteristics, comorbidity, and illness severity. Although subsequent hospitalization for outpatients differed by site, the rate was lowest for the site with the lowest antimicrobial costs. CONCLUSION: Variations in antimicrobial prescribing practices by treatment site exist for outpatients and inpatients with community-acquired pneumonia. Although variation in antimicrobial prescribing practices across institutions results in significant differences in antimicrobial costs, patients treated at institutions with the lowest antimicrobial costs do not demonstrate worse medical outcomes.