New and emerging etiologies for community-acquired pneumonia with implications for therapy. A prospective multicenter study of 359 cases

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.

A three-year study of severe community-acquired pneumonia with emphasis on outcome

Fifty-eight consecutive patients with severe community-acquired pneumonia were studied prospectively during a three-year period. The group included 44 men and 14 women (mean age: 45.0 +/- 15.7 years). The cause of pneumonia was diagnosed in 35 (60.3 percent) cases, and the most common pathogens were Streptococcus pneumoniae (37.1 percent), Legionella pneumophila (22.8 percent) and Gram-negative bacilli (11.4 percent). The fact that Mycobacterium tuberculosis was present in four (11.4 percent) patients and Pneumocystis carinii in three (8.5 percent) is worthy of note. The overall death rate was 22.4 percent. More than 50 percent of deaths occurred within the first five days and were caused by septic shock, hemoptysis (tuberculosis) or hypoxia. However, hypoxia remains the main fatal complication and all late-occurring deaths (> 5 days) observed were due to this cause. These data could be important in planning strategies and protocols to improve prognosis.

Antibiotic prescribing patterns in Italian hospital inpatients with pneumonia, chronic obstructive pulmonary disease, and urinary tract infections

OBJECTIVE

The primary objective of this study was to describe the antibiotic prescribing patterns in hospital inpatients with pneumonia (PN), exacerbations of chronic obstructive pulmonary disease (COPD), and urinary tract infections (UTIs). A second objective was to verify if some selected variables (i.e., risk factors, patient age, size of hospitals) were affecting the therapeutic choice.

DESIGN

Survey was performed on 1609 patients. The data were collected by physicians using a special form, covering a six-month period.

SETTING

Twenty-six medical wards and 8 geriatric wards in 24 acute-care hospitals in the Veneto Region in Northern Italy.

PATIENTS

A consecutive sample of PN, COPD, and UTI patients treated with antibiotics.

MAIN OUTCOME MEASURES

The following information was collected: patient and hospital demographics, risk factors, diagnoses, and antibiotic regimens.

RESULTS

Sixty-three antimicrobial agents used, with 2115 administrations, 1227 of which were single-drug therapy. The most frequently used drugs were third-generation cephalosporins (24.6 percent), fluoroquinolones (15.4 percent), aminopenicillins (15.0 percent), and ureidopenicillins (9.7 percent). There is great variability of therapeutic regimens in the various hospitals for the same disease. However, this variability is not explained by the different types of hospitals or by the patients' characteristics (e.g., age, risk factors).

CONCLUSIONS

Our results show that the use of broad-spectrum antibiotics probably is excessive. Moreover, the treatment seems to be based more on the opinion of the treating physician and the local habits rather than objective criteria. These factors may have negative repercussions not only in economic terms, but also in terms of alteration of the bacterial ecology.

Quinolones in the treatment of lower respiratory tract infections in adult patients

The potential role of quinolones is discussed on the basis of data obtained during the past 2 years from epidemiological studies, in vitro investigations, animal experiments and clinical trials. Although the newest compounds exhibit good activity against Streptococcus pneumoniae and intracellular pathogens in animal models, the role of quinolones as first line therapy in community-acquired pneumonia is still debatable and may be modified according to clinical presentation and the rate of resistance of pneumococci to beta-lactams and macrolides. Cost-utility and cost-benefit studies are required to delineate precisely the role of quinolones in the treatment of acute exacerbations of chronic bronchitis. In addition, promising results indicating a possible future for the clinical use of quinolones in the therapy of mycobacterial infections have been obtained.

Five cases of nosocomial and community-acquired Legionnaires' disease in São Paulo, Brazil

Legionella sp has been emerging over the last decade as an important cause of pneumonia both hospital and community-acquired. Following an outbreak in a Renal-Transplant Unit stocked serum was tested for antibodies against Legionella pneumophila serogroup 1, and 5 cases of Legionnaires' Disease were reviewed. Two of the cases were nosocomial and three cases were community-acquired. Clinical and laboratorial aspects were similar to those expected for other causes of pneumonia, however jaundice was encountered in two cases. This study suggests that the real incidence of pneumonia caused by Legionella sp is being underestimated and the authors emphasize the importance of considering Legionnaires' Disease when empirically treating community-acquired pneumonia.

Community-acquired pneumonia. The changing picture

Important changes in the initial management of community-acquired pneumonia have been prompted by the discovery of new respiratory pathogens, the changing susceptibility of traditional pathogens to antimicrobial agents, and the introduction of new antimicrobial agents. Although the clinical presentation may suggest a specific pathogen, findings overlap too much to reliably distinguish the specific cause of the pneumonia on a clinical basis. Useful laboratory studies include Gram's stain and culture of sputum, blood culture, serologic studies, and new tests such as the urinary antigen test for Legionella pneumophila. Empirical antimicrobial treatment must take into consideration that 20% to 30% of cases of community-acquired pneumonia are due to atypical pathogens that are not susceptible to beta-lactam agents.

Similarity of Chlamydia pneumoniae strains in the variable domain IV region of the major outer membrane protein gene

DNA was amplified by polymerase chain reaction from the gene encoding the major outer membrane protein (MOMP) of Chlamydia pneumoniae in order to examine the relatedness of strains isolated from diverse geographical regions. Primers for this reaction were chosen to span a 207-bp region comparable to that of the fourth variable segment of the MOMP gene of Chlamydia trachomatis. Among C. trachomatis, sequence heterogeneity is characteristic within variable sequence domain IV (VDIV) and correlates with serovar type. In contrast, sequence analysis of polymerase chain reaction products from 13 C. pneumoniae isolates indicated that all tested strains were identical in this segment of the MOMP gene. The predicted amino acid sequences from the C. pneumoniae VDIV gene products shared only 13.3 to 30% homology with published VDIV regions from serovars of C. trachomatis. Homology of these VDIV amino acid sequences with sequences from strains of C. psittaci ranged from 45.7 to 60%. The sequence conservation of the VDIV region of the MOMP gene indicates that C. pneumoniae strains may be more genetically homogeneous than C. trachomatis or Chlamydia psittaci strains. Future investigations of antigenic diversity among C. pneumoniae strains should be aimed at the evaluation of variation in other regions of the C. pneumoniae genome.

Pneumococcal bacteremia in Monroe County, New York

OBJECTIVES

Knowledge of the epidemiology of pneumococcal disease is critical for public health planning, evaluation of preventive strategies, and development of immunization recommendations.

METHODS

We studied the incidence and case-fatality rates of pneumococcal bacteremia as a proxy for pneumococcal disease in Monroe County, New York, from 1985 through 1989 by reviewing the laboratory and clinical care records of all cases occurring among residents.

RESULTS

There were 671 cases identified, for an overall yearly rate of 18.8 per 100,000. The rates were highest in the very young, in the very old, and in non-White populations. Age-specific rates were consistently higher in Blacks than in Whites. Predisposing medical conditions were present in 61% of cases. Case-fatality rates were 15% overall, 27% in those with predisposing medical conditions, and approximately 30% in Blacks older than 55 years and Whites older than 65 years.

CONCLUSIONS

This study documents the incidence of and mortality from pneumococcal bacteremia. It supports previous observations that Black populations have an increased risk of invasive pneumococcal infection and suggests that immunization should be considered for Blacks older than 55 years.

The epidemiology and significance of Chlamydia pneumoniae

Chlamydia pneumoniae has recently been found to be a new chlamydial species transmitted directly from man to man, obviously through the respiratory tract. It is the commonest chlamydia of mankind but fortunately the overwhelming majority of infections are mild, it has been estimated that 10% of all pneumonias are caused by this species. Since the most important chlamydial infections are chronic in nature, the question of possible chronic C. pneumoniae infections is of paramount importance. Apart from being associated with chronic inflammatory processes of the respiratory system, a quite unexpected connection to coronary heart disease has now been suggested.

Haemophilus influenzae type b bacteremia in adults with AIDS and at risk for AIDS

PURPOSE

Our objective was to determine the number of cases of Haemophilus influenzae type b bacteremia in patients with and at risk for acquired immunodeficiency syndrome (AIDS) from January 1983 to June 1991 at a municipal hospital in Bronx, New York.

PATIENTS AND METHODS

We reviewed blood culture records of adult patients admitted to North Central Bronx Hospital from January 1983 to June 1991 to identify cases of bacteremic H. influenzae type b disease. The hospital charts and admission chest radiographs of bacteremic patients were then reviewed.

RESULTS

Ten of 15 cases of adult H. influenzae type b bacteremia occurred in patients with AIDS or who were at risk for AIDS. Seven had AIDS at presentation. Nine were active or former intravenous drug users (IVDUs). All 10 cases were associated with a respiratory source, and five of the 10 patients also had H. influenzae type b isolated from sputum. All H. influenzae type b strains were negative for beta-lactamase.

CONCLUSIONS

Human immunodeficiency virus-positive IVDUs may be at increased risk for bacteremic H. influenzae type b infections. Empiric antibiotic regimens for community-acquired pneumonia in these patients should include appropriate antibiotics for the treatment of H. influenzae type b. In addition, these patients may be candidates for conjugate H. influenzae type b vaccine trials.

Efficacy and safety of temafloxacin versus those of amoxicillin in hospitalized adults with community-acquired pneumonia

Temafloxacin, a new fluoroquinolone, was compared with amoxicillin in the treatment of adult hospitalized patients with community-acquired pneumonia. In this double-blind, multicenter study, patients were randomly assigned to treatment with temafloxacin at 600 mg twice daily (n = 125) or amoxicillin at 500 mg three times daily (n = 121); the average duration of treatment was 10 days. Clinical recovery rates were similar for patients treated with temafloxacin and amoxicillin (89 and 85%), as were bacterial eradication rates (99 and 97%). This was also true for subgroups of patients with pneumococcal pneumonia (n = 100), nonpneumococcal pneumonia (n = 122), or atypical pneumonia (n = 12). Outcomes for temafloxacin- and amoxicillin-treated patients were also similar in terms of defervescence, improvement in leukocytosis, and radiographic evidence of infection. The frequency and severity of adverse events were similar in both groups, consisting primarily of digestive disorders and skin manifestations. We conclude that temafloxacin may be recommended as an alternative antibacterial drug for patients with suspected pneumococcal pneumonia who fail to respond to benzylpenicillin or amoxicillin when the incidence of multiresistant pneumococcal strains is low. In countries where the incidence of these strains is high, temafloxacin may also be recommended.

Microbial etiology of acute pneumonia in hospitalized patients

The purpose of this study was to determine the microbial etiology of pneumonia by using strict criteria among a group of hospitalized patients. Patients with acute community-acquired or hospital-acquired pneumonia were studied in a systematic and comprehensive manner for bacterial, viral, chlamydial, mycobacterial, and fungal pathogens. A total of 198 patients with 204 episodes of pneumonia were evaluated. Despite 100 percent follow-up of all surviving patients, a specific etiologic agent could be found in only 103 episodes. Among 154 episodes of community-acquired pneumonia, a diagnosis was made in 79; the most common pathogen was from the genus Legionella, followed by various Gram-negative enteric bacteria, Gram-positive cocci, influenza A virus, and Mycoplasma pneumoniae. The etiologic agent was found in 24 of the 50 patients with hospital-acquired pneumonia; no pathogen predominated. We conclude that even when elaborate diagnostic studies are done, including many invasive procedures, the etiology can be determined in only about half of the patients with acute pneumonia. The pathogens of pneumonia in this study are not markedly different between community-acquired and hospital-acquired infection.

A prospective study of community-acquired pneumonia in Hong Kong

A prospective study of community-acquired pneumonia in Hong Kong was carried out between January and December, 1988. Ninety adults (57 male) with a mean age of 57.3 years were admitted to the Prince of Wales Hospital with community-acquired pneumonia. The etiologic diagnosis of pneumonia was made in 37 cases (41 percent). Pneumococcal infection was diagnosed in 11 patients (12 percent). The same number of patients had pulmonary tuberculosis presenting as acute pneumonia. It could not be differentiated from other causes of pneumonia on clinical and radiologic grounds, although pleural effusion and upper lobe involvement were more common in patients with tuberculosis. Chlamydia species were identified in five patients (6 percent) and Mycoplasma pneumoniae was identified in three patients (3 percent). There was no case of Legionnaires' disease. The etiologic agent could not be identified in 59 percent of cases. The low incidence of etiologic diagnosis of community-acquired pneumonia was probably related to the widespread use of antibiotics in private practice. Tuberculosis is an important cause of community-acquired pneumonia in Hong Kong and this diagnosis should be considered in patients who fail to respond to first-line antibiotics.

Potable water as a cause of sporadic cases of community-acquired legionnaires' disease

BACKGROUND

The environmental sources of sporadic, community-acquired legionnaires' disease are largely unknown, and culturing of water sources after identification of a case is currently not recommended. We conducted a prospective study of sporadic cases of community-acquired legionnaires' disease to determine whether the environmental reservoirs could be identified.

METHODS

We cultured samples of potable water obtained from sources to which each of 20 patients with culture-confirmed, community-acquired legionnaires' disease had been exposed during the two weeks before the onset of symptoms. Monoclonal-antibody subtyping and restriction-endonuclease analysis were performed on the legionella isolates recovered from both the patients and the associated environmental cultures.

RESULTS

For 8 of the 20 patients, isolates of Legionella pneumophila with identical subtypes were identified in cultures from both the patient and the potable water to which the patient had been exposed. The environmental reservoirs linked to the infections were the water supplies of two private residences, two nursing homes, two hospital outpatient clinics, and an industrial plant.

CONCLUSIONS

Potable-water supplies that harbor L. pneumophila are an important source of community-acquired legionnaires' disease. Future studies should include attempts to identify the environmental sources of this infection.

Intervention to discontinue parenteral antimicrobial therapy in patients hospitalized with pulmonary infections: effect on shortening patient stay

OBJECTIVES

Current efforts to contain anti-microbial costs in hospitals are based on restricting drugs. We explored the effects of unsolicited case-specific recommendations to physicians to discontinue parenteral antimicrobial therapy in medically stable patients with pneumonia, in order to shorten hospital length of stay.

METHODS

A nurse-interventionist, working as an emissary of an appropriate committee in 3 nonteaching community hospitals, presented randomly assigned physicians with nonconfrontational suggestions to substitute comparable oral antimicrobials for parenteral antimicrobials. Blinded observers evaluated in-hospital and 30-day postdischarge courses of patients of physicians who had been contacted by the nurse (cases) and those who had not (controls).

RESULTS

Eighty-two patient episodes (47 physicians) met study criteria. There were 53 cases and 29 controls. In 42 of 53 (79%) case episodes, physicians discontinued parenteral antimicrobials; patients' mean length of stay was 2.4 days less than for 29 control episodes (estimated cost savings was $884/patient). In 11 (21%) episodes, case physicians continued parenteral therapy; patients' mean length of stay was 1.9 days longer than for controls (estimated cost excess was $704/patient). Education, training and practice characteristics were comparable in physician groups. Severity of illness indicators and postdischarge outcomes were comparable in patient groups.

CONCLUSIONS

The major cost-saving potential for shifting from parenteral to oral antimicrobial therapy is shortened length of stay. Timely information about alternative drug therapies, offered on a patient-specific basis, appears to modify the treating behavior of physicians. The program as currently conducted is cost-effective, with an estimated net savings of $50,000 per 100 interventions.

Pneumonia. Patient profiles, choice of empiric therapy, and the place of third-generation cephalosporins

Choosing appropriate antimicrobial therapy for patients with pneumonia requires knowledge of the etiologic agents seen in specific kinds of patients at specific times and places. For community-acquired pneumonia, there is an important difference in the agents seen in the normal and the compromised host. The normal host most often presents with viral, mycoplasmal, or pneumococcal pneumonia. The exact place of Chlamydia pneumoniae is still under study. A normal host who aspirates is at risk of anaerobic pneumonia. Normal hosts with influenza may acquire superinfection with Streptococcus pneumoniae, Haemophilus influenzae, or Staphylococcus aureus. Under specific epidemiologic conditions, community-acquired pneumonia may be due to Legionella species, Yersinia pestis, Francisella tularensis, Coxiella burnetii, Chlamydia psittaci, a mycotic agent, or tuberculosis. Patients with chronic bronchitis and emphysema are predisposed to H. influenzae, Moraxella catarrhalis, and S. pneumoniae infections. HIV-infected patients are likely to have Pneumocystis carinii pneumonia and pneumonia due to cytomegalovirus, S. pneumoniae, and H. influenzae. Patients with diabetes, nursing-home patients, hospitalized patients, immuno-compromised patients, and patients with recent antibiotic therapy are predisposed to pneumonia due to Gram-negative aerobic bacilli of enteric and environmental origin. Initial therapy should be directed at the likely organism or organisms based on hospital susceptibility surveillance. In the normal host with community-acquired pneumonia, the therapy will often be penicillin G or erythromycin. In the patient predisposed to Gram-negative pneumonia, a third-generation cephalosporin with or without an aminoglycoside is the usual choice.

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.

Current trends in bacterial respiratory pathogens

The relationship of virulence and antimicrobial susceptibility with morbidity due to bacterial respiratory pathogens is complex and evolving. Ultimately, decreasing the incidence of pneumonia due to bacterial pathogens will be dependent on successful preparation and distribution of effective vaccines. Until effective vaccines are widely available, control of a majority of respiratory infections will depend on promotion of rational therapeutic strategies. Though limited to a few specific serotypes and strains, changes in virulence of bacterial respiratory pathogens have been noted. Co-infections due to multiple respiratory pathogens may increase morbidity; however, the epidemiology of co-infections is not clear. Relationships between respiratory viruses and bacteria may exist that increase virulence of both agents, but information regarding these relationships awaits further investigation. Resistance of respiratory pathogens to the more commonly used antimicrobials, such as penicillin, erythromycin, chloramphenicol, and cotrimoxazole, is being documented globally with increasing frequency. The evolution of antimicrobic resistance, especially among strains of Streptococcus pneumoniae, the most common and deadly agent of lower respiratory tract infections, provides impetus to develop and promote effective pneumococcal vaccines and to search for new and effective antimicrobials.

Legionella pneumophila. A cause of severe community-acquired pneumonia

In a prospective study of community-acquired pneumonias, 30 patients were diagnosed with Legionnaires' disease in 15 months. Clinical, laboratory and radiologic features of these patients are reviewed and compared with those who have pneumococcal pneumonia. Alcoholism, history of smoking, previous antimicrobial therapy, gastrointestinal and neurologic manifestations, elevations of serum transaminases, alkaline phosphatase and creatinine levels were more frequent in pneumonia due to Legionella pneumophila than in pneumococcal pneumonia. The presence of respiratory failure and radiologic progression were common findings that suggested L pneumophila as the etiologic agent of a community-acquired pneumonia. Development of respiratory failure was associated with involvement of several lobes and isolation of L pneumophila in any specimen. In 21 of 30 patients with Legionnaires' disease, L pneumophila was isolated from respiratory specimens. Overall mortality was 10 percent, but it increased to 27 percent in patients not treated with erythromycin initially.

Pneumonia. Pearls for interpreting patients' radiographs

The chest radiograph can detect pneumonia, but laboratory evaluation is needed to determine the specific causative organism. However, before these results become available, a presumptive diagnosis can be made with the help of chest radiography. A lobar pattern on radiography is usually produced by pneumococcal and Klebsiella infections. A lobular (bronchopneumonia) pattern may be produced by Staphylococcus, gram-negative organisms, and anaerobes. An interstitial pattern results from viral, Mycoplasma, and Pneumocystis carinii infections. Mixed patterns may also occur. There is great variation in the presentation of each infection. However, pertinent clinical information, epidemiologic factors, and associated radiographic findings (eg, adenopathy, pleural effusion, cavitation) are helpful in further limiting diagnostic possibilities.

Ceftriaxone in treatment of serious infections. Pneumonia and lower respiratory tract infections

One gram once daily of ceftriaxone is as effective as 2 gm for the common causative organisms of community-acquired and nosocomial pneumonias. It should not be used alone against L. pneumophila, the TWAR pneumoniae, and possibly anaerobic pathogens. Preliminary study suggests that a regimen of 500 mg a day may be effective in selected patients with simple pneumonias caused by sensitive pathogens. Ceftriaxone is safe, well tolerated, and cost-effective, which means it can play a significant therapeutic role in nursing homes and other long-term care facilities as well as in outpatient therapy.

Quinolones in the treatment of lower respiratory tract infections caused by intracellular pathogens

Intracellular pathogens are inhibited to varying degrees, depending upon the strain of the organism and the quinolone tested. Quinolones achieve levels in the lower respiratory tract that equal or exceed serum concentrations, and they also achieve good intracellular concentrations. Experimental models of intracellular infection have demonstrated the efficacy of ciprofloxacin, difloxacin, fleroxacin, ofloxacin and pefloxacin. Animal models of experimental legionellosis have confirmed in vivo their efficacy in this field. Thus, quinolones appear to be a safe and efficacious alternative treatment in lower respiratory tract infection (LRTI) due to intracellular pathogens. Considering the in vitro and experimental studies, quinolones should play an important role in the treatment of LRTI caused by intracellular pathogens, and prospective controlled studies are strongly recommended.

Evaluation of housestaff physicians' preparation and interpretation of sputum Gram stains for community-acquired pneumonia

OBJECTIVE

To evaluate the preparation and interpretation of sputum Gram stains by housestaff physicians in the assessment of patients with community-acquired pneumonia.

DESIGN

A prospective, multicenter study.

SETTING

Two university-affiliated hospitals in Pittsburgh.

PATIENTS

Ninety-nine cases of clinically and radiographically established pneumonia occurring in 97 patients. Diagnostic test assessment: Housestaff and microbiology personnel prepared a Gram stain for each case of pneumonia. Housestaff assessed the presence and identity of a predominant microbial organism on the slides they prepared. Two senior staff microbiologists, blinded to patient and preparer, evaluated all slides for preparation, sputum purulence, and identification of the predominant organism. Two reference standards were used to assess the sensitivity, specificity, and predictive values of housestaff's Gram-stain interpretations: 1) senior staff microbiologists' determinations of the microbes present using the slides without benefit of culture results, and 2) the etiologic agent derived from results of sputum culture, blood culture, or serology.

MEASUREMENTS AND MAIN RESULTS

Housestaff physicians completed a Gram stain in 58% of the pneumonia episodes. Gram stains were not made in 42% of cases, primarily because patients were unable to produce sputum. Fifteen percent of housestaff's smears were judged inadequately prepared, compared with 3% for the laboratory personnel (p less than 0.01). Housestaff obtained purulent sputum samples significantly more often than did nursing personnel (58% versus 38%; p less than 0.01). Housestaff's Gram stains were 90% sensitive for detecting pneumococcus, with a 50% false-positive rate. The sensitivity of the Gram stain was less for identification of Haemophilus influenzae than for identification of Streptococcus pneumoniae. A single antimicrobial agent was chosen as initial therapy for 50% of the patients in whom housestaff identified a predominant organism, compared with 30% in whom a predominant organism was not identified (p less than or equal to 0.05).

CONCLUSIONS

Although housestaff obtained purulent sputum samples more frequently than did nursing personnel, they made systematic errors in the preparation and interpretation of Gram-stained slides. Housestaff physicians should receive formal training in the preparation and interpretation of Gram stains; the specific defects elucidated in this study warrant special attention.