Community-acquired pneumonia: compliance with Centers for Medicare and Medicaid Services, national guidelines, and factors associated with outcome.

Background: This study was performed to evaluate the impact of adherence to national guidelines for management of community-acquired pneumonia (CAP) on patient outcomes.

Methods: Compliance with published national guidelines was assessed. Mortality rate and length of hospital stay were determined.

Results: Patients who were administered antibiotics within 4 hours of admission had a shorter stay. Those treated at least 8 hours after admission had the highest mortality. Good compliance seen with 1998 guidelines of the Infectious Diseases Society of America declined substantially when 2000 Infectious Diseases Society of America guidelines were evaluated. Pediatric compliance was difficult to evaluate. Documentation of vaccination screening and administration was poor.

Conclusion: Antibiotic therapy should be started within 4 hours in patients with CAP. Using the most recent CAP guidelines as a benchmark may lower compliance unless providers are reeducated. National consensus guidelines for pediatric patients should be developed. Hospitals should evaluate documentation of vaccine screening and administration and should implement programs to increase vaccination rates if needed.

**********

Pneumonia has been recognized as a common and potentially lethal condition for nearly two centuries. Community-acquired pneumonia (CAP) is a common, serious, and costly illness with substantial geographic variation in treatment patterns. (1) CAP is the major cause of death attributable to infectious diseases globally and in the United States; overall it is the sixth leading cause of death in the United States. (2,3) According to comprehensive studies of this infection in the preantibiotic era, mortality rates were approximately 1/1,000/yr. In addition, 80% of the cases were caused by Streptococcus pneumoniae, with mortality rates of approximately 20 to 40%. (4,5) Current estimates are 4 million cases annually in the United States, an attack rate of 12/1,000 adults/yr, and approximately 600,000 hospitalizations/yr at an annual cost of $23 billion. (6,7) Community-acquired pneumonia (CAP) affects 1% of the population each year and is responsible for an estimated 4.5 million visits to physician's offices, emergency departments, and outpatient clinics, as well as 1.1 million hospitalizations and 45,000 deaths. (2,8) Despite newer and more efficacious methods of prevention and treatment, the mortality rate of CAP in certain groups of patients (ie, those requiring hospitalization, older adults, and the immunocompromised) has not changed. (9) Between 1979 and 1994, the overall age-adjusted death rates attributable to pneumonia actually increased by 22% to 24.8/100,000. (2) By 1998, this number had decreased by almost 50% to 13.2/100,000. (10) Improved prevention and optimal treatment may have a significant impact on the morbidity, mortality, and cost associated with CAP. (11)

Despite advances in the identification of new microbial pathogens and the development of new antimicrobial agents, few diseases are so characterized by disputes about diagnostic evaluation and therapeutic decisions. There are, however, areas of agreement regarding appropriate therapy and management. To assist with treatment decisions and to promote proper care, clinical practice guidelines for adult CAP have been developed by several sources. Practice guidelines were published by the Infectious Diseases Society of America (IDSA) in 1998 and revised in 2000. (2.12) The Drug-resistant Streptococcus pneumoniae Therapeutic Working Group from the Centers for Disease Control and Prevention (CDC) has also published a report on the treatment of CAP given the emergence of resistant pneumococci. Although not designed for pediatric patients, it does provide recommendations for younger patients. (13) The American Thoracic Society (ATS) published an update to their 1993 recommendations in March 2001. (14) Specific U.S. pediatric consensus guidelines are lacking and there are no published U.S. guidelines for children. Although not specific to the U.S. population, the Canadian Medical Association Journal published a set of consensus guidelines developed by Canadian experts and based on studies conducted in developed countries. (15) The CAP guidelines have evolved over time. The 1998 IDSA guidelines considered sole [beta]-lactam/[beta]-lactamase inhibitor regimens or macrolide regimens appropriate. The 2000 IDSA guidelines were revised to require the combination. In addition, only extended [beta]-lactam regimens are considered appropriate. (2,12) Fluoroquinolones alone have been considered appropriate therapy, but the 2000 CDC guidelines recommend their use only in selected cases. (2,12-14) Tables 1 and 2 summarize the various guidelines.

Given the significance of CAP, there has been an increasing need to assess its impact on the health care system. To this end, the Centers for Medicare and Medicaid Services (CMS) of the Quality Improvement System for Managed Care named CAP a required national project in 2000. The Joint Commission on Accreditation of Healthcare Organizations also announced that CAP was an initial focus area for core performance measures. In addition to the clinical and economic concerns mentioned previously, the emphasis placed on management of CAP by these oversight agencies makes review of immunization practices (ie, prevention and treatment) of vital importance.

The medical center at which this study was performed provides care for active-duty and retired military members and their dependents. This population spans a broad age range from newborns to geriatric patients. The medical center does not perform organ or bone marrow transplantation and treats only a small population of human immunodeficiency viruspositive patients. Other disease states are seen at rates similar to those in the general population.

The purpose of this project was to evaluate the institution's prevention and treatment of CAP and to assess outcomes The ultimate goal was to prevent occurrence of CAP or to decrease the morbidity and mortality associated with CAP. By determining the extent to which practice complied with the CMS quality indicators and evaluating mortality and length of hospitalization, practices associated with successful outcomes could be identified and shared.

Methods

An inpatient International Classification of Diseases, Ninth Revision, diagnostic report was generated through the inpatient clinical records department for all patients admitted from September 1999 through August 2000 with International Classification of Diseases, Ninth Revision Codes 480 through 487. These records were pulled for all patients, regardless of age. Charts were reviewed to determine whether the admitting diagnosis was CAP. The confirmed charts were evaluated for compliance with five of the CMS quality indicators (Table 3). Patient demographics, comorbid conditions, disposition, and hospital length of stay were also documented. Descriptive statistics were compiled regarding compliance with the CMS quality indicators and national consensus guidelines. Canadian guidelines were used in areas not covered by U.S. guidelines (eg, patients younger than 8 years of age). Given the differences in health care economics between the two countries and concerns about pneumococcal resistance, the Canadian guidelines were liberalized to include appropriate third-generation cephalosporins. Regimens were considered to be compliant, insufficient, or too broad. Patients meeting special population criteria according to the guidelines were evaluated according to those criteria. Thus, regimens too broad for a ward patient may have been considered appropriate if administered to a patient in the intensive care unit (ICU). Regimens were considered too broad if an antibiotic beyond those required by the guidelines was included or the antibiotic selected had a spectrum beyond that recommended. Regimens were considered too narrow if they did not include coverage required by the guidelines. These compliance rates were compared with national or regional benchmarks when appropriate. Mortality and length of stay were studied in various subgroups to detect any confounding factors in the data. The nonparametric Kruskal-Wallis test was used to detect differences between time-to-treatment groups. When highly significant differences were detected among time-to-treatment groups, the groups were combined to allow further analysis.

Results

The records of 155 patients with confirmed diagnoses of CAP who were admitted between September 1999 and August 2000 were evaluated. Table 4 shows the number of patients evaluated, the sex of each patient, and the three age groups into which the patients were assigned: 18 years of age and younger, from 18 through 64 years of age, and 65 years of age and older. The majority of patients were admitted from home; three were admitted from a nursing home. More than half (54%) of the patients evaluated had concomitant disease states or risk factors. Chronic obstructive pulmonary disease (COPD) was the most common, followed by tobacco use, diabetes, congestive heart failure, chronic renal failure, and alcohol use. Forty-five patients (29%) had more than one concomitant disease state. Although the number of patients with tobacco use and COPD were the same, not all patients with COPD used tobacco products. Mortality and length of stay data are shown in Table 5. There was a significant difference (P < 0.05) in length of stay between the time-to-treatment groups (Fig. 1). Combined group analysis showed the median stay for those treated 4 or more hours after presentation (4 days) was significantly longer than those treated within 4 hours (3 days) (P < 0.01). This significance was maintained when only the patients older than 18 years of age were evaluated (P < 0.01). Approximately 62% of patients (n = 95) received the initial dose of antibiotic within 4 hours of admission, and 19% of patients (n = 31) received the initial antibiotic dose more than 8 hours after admission. Whether the patients had undergone oral antibiotic therapy before admission could not be determined. In addition, data regarding the length of time that symptoms had been present before admission, and the prevalence of antibiotic use just before admission, although important predictors of outcome, were not collected as part of this study.

A significant difference in length of stay was found between the pediatric and both the adult and geriatric groups (P < 0.01). The median age of the group treated within 2 hours was significantly less than that of the other three groups (P < 0.01). There was not a significant difference in median stay as a function of number of concomitant disease states and/or risk factors or as a function of any one disease state and/or risk factor. Those with the highest number of concomitant disease states and/or risk factors had the longest median hospital stay, but this finding did not reach statistical significance (P = 0.10). The median number of concomitant disease states was the same across time-to-first-dose intervals.

The overall mortality was 4.5% (7 of 155 patients). When stratified by age, mortality among the adults younger than 65 years old was slightly higher, at 5.0% (2 of 40). The highest mortality rate of 8.6% (5 of 58) was seen in the group at least 65 years of age, and the lowest mortality rate of 0% was seen in those younger than 18 years old. This difference, however, did not reach statistical significance (P = 0.08). Mortality also was not significantly different between the time-to-treatment groups (P = 0.98), although the group that received treatment more than 8 hours after presentation had the highest mortality. Those with positive blood cultures also had a high mortality rate of 17% (2 of 12), although this finding, too, was not statistically significant.

The emergency department (ED) was responsible for starting 23% of the antibiotic doses, but further inspection showed that patients seen in the ED received no additional benefit from rapid antibiotic administration. A total of 77% of ED patients received their first dose of antibiotics within 8 hours versus all other admission sources (ward, floor, direct admission), which showed a compliance of 81%. Empiric antibiotic selection is presented in Figures 2 and 3. Compliance with national guidelines (where available) is presented in Table 3. Overall compliance with 1998 IDSA or Canadian pediatric guidelines (patients <8) was 89%. Compliance with CDC (patients aged 8 and older), CMS, and ATS guidelines was 83%, 85%, and 84%, respectively. The most common regimen was levofloxacin alone, prescribed in 31% of patients (48 of 155). The second most common regimen was a third-generation cephalosporin plus a macrolide, which was prescribed in 21.9% (34 of 155) of patients. Twenty regimens (19%) did not meet the 1998 IDSA preferred guidelines. Of the 20 regimens, 11 lacked a third-generation cephalosporin, 6 lacked a macrolide, and 6 were too broad. Nine regimens met neither of the IDSA guideline criteria. All of these regimens were excessively broad. Compliance decreased when the 2000 IDSA guidelines were applied. Of the 24 regimens that did not meet the guidelines, 11 lacked a third-generation cephalosporin, 9 lacked a macrolide, and 7 were excessively broad. In both cases, some regimens failed to meet guideline criteria for more than one reason.

Analysis of the data in Figure 3 shows that empiric therapy was appropriate in 38 (79%) of the patients <8 years of age. In the absence of U.S. consensus guidelines, the Canadian Medical Association Guidelines were adapted. On the basis of the Canadian guidelines, reasons for being deemed inappropriate were excessively broad coverage and use of oral agents as sole initial therapy in a hospitalized child.

Antibiotics were given at appropriate doses in 96% of all pneumonia cases. Two patients without proven renal insufficiency were administered 250 mg intravenous (IV) levofloxacin.

Blood cultures were performed before antibiotic therapy in 71% (n = 112) of patients. Of these 112 patients, 12 (11%) had positive blood culture results. There was no significant difference in outcome (P = 0.09) or length of stay (P = 0.22) between those with and without positive blood cultures. Only two charts had documentation of pneumococcal vaccine screening during hospitalization. No influenza vaccine screening was documented. There was no documentation of pneumococcal or influenza vaccine administration in any of the patient records (Tables 6-8).

[FIGURE 1 OMITTED]

Discussion

In this study, patients treated for CAP within 4 hours of presentation had a significantly shorter length of stay than patients treated more than 4 hours after presentation. Most studies that have evaluated outcomes associated with time to antibiotic therapy have evaluated mortality. (12,16) The literature contains only abstract reports relating time to treatment for CAP and length of hospitalization. (17,18) Yih et al (17) evaluated the impact of therapy within 8 hours of presentation and found a 0.5-day shorter stay for those treated within 8 hours. Amberik and Cummings (18) implemented a program that reduced time to therapy from a mean of 8.3 hours to a mean of 3.1 hours, with a resulting decrease in length of stay of 0.6 day. These studies did not evaluate the impact of prompt initiation of treatment. The findings reported here and in the abstract publications may represent an additional benefit of early therapy and should be further evaluated. In the meantime, although obtaining diagnostic information is important, antibiotic therapy should not be withheld from acutely ill patients because of delays in obtaining appropriate specimens or the results of Gram's stains and cultures. (19) Age also correlated with length of stay. Pediatric patients had a significantly shorter stay than their adult counterparts. This may reflect a lower threshold for admission of the pediatric population and thus a less ill group. It could also reflect a group more acutely responsive to therapy. There was a difference in age between the time-to-treatment groups. As stated previously, the <2-hour group was significantly younger. This may explain the shorter stay in this group but would not explain the continued trend across the other time-to-treatment groups. Spearman rank correlation still showed the length of stay to be significantly different (P < 0.01) after removing the youngest age group (<18 years of age) from the analysis. The median ages in the other time-to-treatment groups were close. Whether the patients had received oral antibiotic therapy before admission could not be determined. In addition, how long symptoms had been present before admission was unknown. These two entities may have been predictive of length of stay and mortality, but could not be determined for the study. A number of other potential confounders including positive blood cultures, number of concomitant disease states/risk factors, and type of concomitant disease state/risk were also evaluated for impact on length of stay and mortality. None of the potential confounders correlated significantly. According to these findings, every attempt should be made to start antibiotic therapy within 4 hours of admission for treatment of CAP.

[FIGURE 2 OMITTED]

Several studies have shown that a greater than 8-hour delay from the time of admission to initiation of antibiotic therapy was associated with an increase in mortality. (12,16) Although this study did not show a statistically significant increase in mortality with prolonged time to therapy, those treated more than 8 hours after presentation did have the highest mortality. Twice as many patients would need to be studied to have enough power to establish a significant difference in mortality. A larger sample size may likely have produced results similar to those previously published. State statistics for CAP mortality in patients older than 65 years of age were 9.4% for Mississippi and 9.9% for Alabama. (20) These two states represent the geographic region for the study institution and serve as a benchmark for local mortality rates. The institution CAP data compared favorably, with a lower mortality rate of 8.6%. The mortality rates in the group aged 18 to 65 were reported as 4.3% in Mississippi and 4.1% in Alabama. The institution mortality data in this same age group was only 2.1%. (20) Although mortality rates were low, there is still documented benefit to shortening time to therapy to within 8 hours, and some studies would suggest 4 hours. (21,22)

These data suggest that every attempt should be made to start therapy within this 4-hour window and that all patients should have therapy started within 8 hours of presentation.

[FIGURE 3 OMITTED]

The majority of patients in this study received empiric antibiotic therapy consistent with IDSA, CDC, ATS, or Canadian pediatric guidelines for the treatment of CAP. (12-15) Compliance with guidelines in place at the time of the evaluation (1998 IDSA guidelines) was good. Reasons for non-compliance included use of regimens that were overly broad. This approach can lead to increased drug resistance and should be avoided. Broad regimens should be reserved for those with risk factors for resistant organisms. (23) In contrast, when regimens were evaluated against the 1998 preferred regimens or the newly published standards, noncompliance increased, often because coverage was not broad enough. The 2000 guidelines recommend the use of third-generation versus secondgeneration cephalosporins and the inclusion of macrolides with [beta]-lactam therapy. (13) Because these guidelines were published at the end of the evaluation period, the decrease in compliance could have been because of a lack of prescriber knowledge. Providers need to be educated so that selected therapy adequately covers the organisms of concern without increasing the risk of resistance. Nathwani et al (11) reported that the most obvious barrier to the implementation of the guidelines is lack of knowledge about their contents. The ATS 2001 guidelines were similar to the IDSA guidelines, but more formally incorporated modifying factors that guide therapy. For example, if no cardiopulmonary disease or disease-modifying factors existed, a hospitalized (non-ICU) patient with CAP could be adequately treated with IV azithromycin alone or an antipneumococcal fluoroquinolone alone. (14) The positive outcomes seen in this study with the most common regimens are consistent with those of other researchers. (24-26) Gordon et al (24) found that patients receiving combination therapy had significantly less mortality than those given [beta]-lactam monotherapy. Waterer et al (25) found that patients who received monotherapy for severe bacteremic pneumococcal pneumonia were five to six times more likely to die than patients who received dual drug or multidrug regimens. The reasons for improved outcomes with the addition of a macrolide to [beta]-lactam monotherapy are not known, but may be attributable to a number of mechanisms including the fact that atypical pathogens, including Legionella, Chlamydia, and Mycoplasma species account for up to 20% of CAP cases. (27) There are concerns about the continued efficacy of macrolides, because resistance is increasing. (28) Currently, the most prevalent mechanism of resistance in the United States is through the mefE gene or efflux mechanism. This mechanism of resistance may currently be overcome with the high concentrations achieved by commonly used macrolides. High-level resistance is conferred through the ermAM gene that produces erythromycin-ribosomal methylase. (28) Should high-level resistance increase or the minimum inhibitory concentrations associated with mefE-resistance increase, the role of 14-member macrolides (clarithromycin, azithromycin) may need further evaluation.

A fluoroquinolone with improved activity against S. pneumoniae is an attractive alternative for adults. Recently, the drug-resistant Streptococcus pneumoniae Therapeutic Working Group from the CDC declined to advocate the use of newer fluoroquinolones for first-line treatment because of their broad spectrum of activity. The group believed fluoroquinolone use may result in resistance among gram-negative organisms. (13,15) Although the fluoroquinolone regimen represents a simpler IV-to-oral conversion and costs half as much as the cephalosporin-macrolide combination, resistance patterns such as that seen with ciprofloxacin a decade ago could occur from excessive use of drugs such as levofloxacin. (29) Current guidelines, though, still point out that for hospitalized (non-ICU) patients, initial treatment should include a parenteral [beta]-lactam and a macrolide or an extended spectrum fluoroquinolone as monotherapy. (12-14)

Given the positive clinical outcomes seen with these regimens, both are appropriate therapy for CAP. There are, however, valid concerns regarding resistance development with both regimens. It may be rational to use these national guideline regimens in a fashion that avoids the placement of undue pressure on any single regimen.

The positive blood culture rate of 10.7% is consistent with the 11% positive rate found in the literature. (19) The 1998 and 2000 IDSA/CDC guidelines recommend blood cultures before initiation of antibiotics in hospitalized patients to assist in the establishment of an etiologic diagnosis. (12) Although therapy should not be delayed to achieve this goal, efforts should be made to obtain preantibiotic therapy cultures whenever possible. There was no difference in outcome between those with positive and negative cultures seen in this evaluation. There was a clinically important difference in mortality between those who had positive blood cultures and those who had negative blood cultures. The patients with positive blood cultures were five times more likely to die (odds ratio, 5.44). The power, however, was only 17% and was not sufficient to determine statistical significance.

It was difficult to determine whether screening for influenza and pneumococcal immunizations was performed, because documentation was found in only two charts. No documentation of influenza or pneumococcal vaccine administration was found in the chart review. As a result of this study, a process has been developed to document screening and vaccination. The process is currently undergoing implementation. Improved administration, documentation, and tracking procedures should be adopted for administration of influenza and pneumococcal vaccine, because proper vaccinations may prevent subsequent episodes of pneumonia. Of the macrolide-resistant isolates reported by Gay et al, (28) 78% were of serotypes included in the seven-valent conjugate pneumococcal vaccine.

Conclusions

Patients treated for CAP more than 8 hours after presentation represent a group at risk for increased mortality and possibly longer hospitalization. Given the relationship between time to therapy and outcome documented in this study, steps should be taken to have therapy for all patients started within 4 hours of presentation and to minimize the time from presentation to treatment. The length of stay and mortality for the study patients were below the reported averages both locally and nationally. Although overall therapy was prescribed within guidelines up to 92% of the time, this rate may be improved by education regarding the current national consensus guidelines and the implementation of institutional guidelines for therapy. Institutional guidelines could also improve appropriate antibiotic dosing. Evaluation of therapy in children younger than 8 years old was difficult given the lack of U.S. guidelines. Consensus guidelines should be developed for younger children with CAP. Finally, evidence that preventive immunization measures were occurring was difficult to document. These findings identify an important opportunity for improvement. Emphasis should be placed on providing a mechanism for screening patients and documenting influenza and pneumococcal immunization in the risk-stratified group of patients who require vaccination.

In the same way that the strength of mind surpasses that of the body, the sufferings of the mind are more severe than the pains of the body.

-Cicero (106-43 BC)

Table 1. Empiric therapy guideline summary for adults with
community-acquired pneumonia
                       2001 ATS guidelines    2000 CDC guidelines

Inpatient (non-ICU)  IV azithromycin         Cefuroxime,
                     alone: if allergic,     cefotaxime,
                       doxy plus a           ceftriaxone
                       [beta]-lactam or APF    or ampicillin-sulbactam
                       alone. If modifying     plus a macrolide. In
                       factors present: IV     selected cases.
                       [beta]-lactam           fluoroquinolone alone.
                       (ceftriaxone,
                       cefotaxime,
                       ampicillin-sulbactam
                       high-dose
                       ampicillin) plus
                       IV/oral
                       macrolide or doxy,
                       or IV APF
                       alone.

                     Modifying factors      Selected:
                     include
                       cardiopulmonary        First-line therapy
                       disease and            failed, allergy
                       being from a           to alternative
                       nursing home.          agents, documented
                                              highly resistant
                                              pneumococci.

Inpatient (ICU) (c)  Cefotaxime,            Cefotaxime or
                     ceftriaxone plus a     ceftriaxone plus
                       macrolide or a         either a macrolide
                       fluoroquinolone. If    or a fluoroquinolone.
                       at risk for
                       Pseudomonas: IV
                       Antipseudomonal
                       [beta]-lactam
                       plus IV
                       ciprofloxacin, or
                       IV aminoglycoside
                       plus either IV
                       azithromycin or IV
                       APF

                       2000 IDSA guidelines   1998 IDSA guidelines

Inpatient (non-ICU)  Preferred: Extended     Preferred: Cefotaxime,
                     Spectrum cephalosporin  ceftriaxone.
                       (cefotaxime or          or [beta]-lactamase
                       ceftriaxone) or         inhibitor (b)
                       [beta]-lactam/          with or without a
                       [beta]-lactamase        macrolide, or a
                       inhibitor               fluoroquinolone
                       (ampicillin-            alone Alternative:
                       sulbactam or            Cefuroxime
                       piperacillin-           with or without a
                       tazobactam) plus        macrolide, or
                       a macrolide, or APF     azithromycin alone
                       alone.

Inpatient (ICU) (c)  Preferred: Extended     Preferred: Erythromycin
                     spectrum cephalosporin    azithromycin, or
                       (cefotaxime or          fluoroquinolone
                       ceftriaxone) or         plus cefotaxime.
                       [beta]-lactam           ceftriaxone, or a
                       [beta]-lactamase        [beta]-lactam/[beta]
                       inhibitor               -lactamase inhibitor. (b)
                       (ampicillin-
                       sulbactam or
                       piperacillin-
                       tazobactam) plus
                       either a macrolide,
                       or APF

                           CMS guidelines

Inpatient (non-ICU)  Cefuroxime, ceftriaxone,
                       cefotaxime, cefepime, ampicillin-
                       sulbactam, piperacillin-
                       tazobactam, imipenem-cilistatin.
                       meropenem alone or with
                       erythromycin, clarithromycin. or
                       azithromycin, or ciprofloxacin
                       Ofloxacin, or an APF alone.

Inpatient (ICU) (c)  Cefuroxime, ceftriaxone.
                       cefotaxime, cefepime, ampicillin-
                       sulbactam, piperacillin-
                       tazobactam, imipenem-cilistatin.
                       meropenem plus erythromycin,
                       clarithromycin. or azithromycin,
                       or ciprofloxacin. Ofloxacin,
                       levofloxacin, or an APF.

* Select the least expensive antibiotic with the narrowest spectrum as
 the regimen of choice.

* If aspiration is suspected, as has been reported in admissions from
 nursing homes. provide anaerobic coverage.

(a) ICU. intensive care unit: ATS. American Thoracic Society: CDC,
 Centers for Disease Control and Prevention: IDSA, Infectious Diseases
 Society of America: CMS. Center for Medicare and Medicaid Services.

(b) [beta]-Lactam/[beta]-lactamase inhibitor not defined.

(c) If structural lung disease is present, provide pseudomonal coverage,
 particularly in the ICU setting.

(d) If documented [beta]-lactam allergy: quinolone (as defined above)
 plus clindamycin or vancomycin. APF, antipneumococcal fluoroquinolone
 (levofloxacin, gatifloxacin, sparfloxacin, moxifloxacin, gemifloxacin,
 trovafloxacin).

Table 2. Empiric therapy guideline summary for children with
community-acquired pneumonia (a)

                          2000 CDC guidelines            1997 CMAJ
                                                     guidelines (b)

Inpatient (non-ICU)  Cefuroxime, cefotaxime,        Age 1-3 mo
                     ceftriaxone, or ampicillin-    erythromycin or
                       sulbactam plus a macrolide.  clarithromycin or
                                                    cefuroxime.
                                                 .  Age 3 mo-5 yr:
                                                    ampicillin or
                                                    cefuroxime.
                                                    Age 5-18 yr:
                                                    erythromycin or
                                                    clarithromycin alone
                                                    or with
                                                      cefuroxime or
                                                      ampicillin.
                     Selected: First-line therapy
                     failed, allergy to
                     alternative
                        agents, proven highly
                        resistant pneumococci.

Inpatient (ICU) (a)  Cefotaxime or ceftriaxone      Age 1-3 mo
                     plus a macrolide.              erythromycin or
                                                    clarithromycin, or
                                                    cefuroxime
                                                      or cefotaxime plus
                                                      cloxacillin.
                                                    Age 3 mo-5 yr:
                                                    cefuroxime plus
                                                    erythromycin or
                                                      clarithromycin.
                                                    Age 5-18 yr:
                                                    erythromycin or
                                                    clarithromycin plus
                                                      cefuroxime.

* Select the least expensive antibiotic with the narrowest spectrum as
the regimen of choice.

* If aspiration is suspected, as has been reported in admissions from
nursing homes, provide anaerobic coverage.

(a) CDC, Centers for Disease Control and Prevention: CMAJ, Canadian
 Medical Association Journal; ICU, intensive care unit.

(b) Adapted to include third-generation cephalosporins (cefotaxime,
 cefiriaxone) as appropriate.

(c) If structural lung disease is present, provide pseudomonal coverage,
 particularly in the ICU setting.

Table 3. HCFA quality indicators

                                                    Study results
                                                        (%)

1.  Proportion of patients who received the           81
    initial antibiotic dose within 8 h of hospital
    arrival.

2.  Proportion of patients given an initial
    antibiotic consistent with the following
    consensus guidelines:

    IDSA 1998                                         92 (81) (b)

    IDSA 2000                                         78

    CDC                                               83 (38) (b)

    ATS 2001                                          84

    CMS                                               85

3.  Proportion of patients who had blood cultures     71
     collected before antibiotics administered.

4.  Proportion of inpatients with pneumonia            2
    screened for or administered influenza
    vaccination.

5.  Proportion of inpatients with pneumonia            2
    screened for or administered pneumococcal
    vaccination.

(a) HCFA, Health Care Financing Administration: IDSA. Infections
Diseases Society of America: CDC. Centers for Disease Control and
Prevention: ATS. American Thoracic Society: CMS. Centers for Medicare
and Medicaid Services.

(b) The number in parentheses reflects the "preferred" regimen, which
excluded fluoroquinolones.

Table 4. Patient demographics (a)

Category                                      Value

Total admissions evaluated                  155
   [less than or equal to]18 yr (M/F)       57 (43/14)
   >18 to <65 yr (M/F)                      40 (18/22)
   [greater than or equal to]65 yr (M/F)    58 (38/20)

Age in years, median (range)                M     R
  Overall                                   57 (0.1-83)
  [less than or equal to]18                 2  (0.1-16)
  >18 to <65                                56 (19-64)
  [greater than or equal to]65              73 (65-83)

Concomitant disease states/risk factor (b)
  Chronic obstructive pulmonary disease     39 patients
  Diabetes                                  25 patients
  Congestive heart failure                  14 patients
  Chronic renal failure                      9 patients
  Alcohol use                               14 patients
  Asthma (RAD)                              10 patients
  Tobacco use                               39 patients

(a) M, male: F, female: RAD, reactive airway disease.
(b) Forty-five patients had more than one disease state/risk factor.

Table 5. Length of hospital stay and mortality

                                   Median stay in  Mortality
                                    days (range)     (%)      No.

Overall                                3 (1-47)       4.5     154

Age group
  [less than or equal to]18 yr         3 (1-14)       0.0      57
  >18 to <65 yr                        4 (1-47)       5.0      40
  [greater than or equal to]65 yr      4 (1-38)       8.6      57

Time to treatment
  <2 hr                                3 (1-14)       3.6      56
  2-4 hr                               3 (2-12)       5.1      39
  4-8 hr                               4 (2-47)       3.5      29
  >8 hr                              4.5 (2-38)       6.7      30

Fig.1 Time until first dose of antibiotic versus median length of stay.
P< 0.01

Time to First Dose  Median LOS

        1             <2 hrs
        2             2-4 hrs
        3             4-8 hrs
        4             >8 hrs

Note: Table made from bar graph.

Fig.2 CAP antibiotic distribution in adult patients (the remaining 13%
not shown consisted of numerous other antibiotic combinations

Levofloxacin                   48%
Ceftriaxone Plus Azithromycin  30%
2nd Gen Ceph/Macrolide          4%
Quinolone Combination           5%

Note: Table made from pie graph.

Figure 3 Distribution of CAP antibiotic regimens in pediatric patients
 (age, <8 yr)

Cefuroxime                   38%
Ceftriaxone                  13%
Cefuroxime  and Azithomycin  25%
[beta]-Lactam                -
Double [beta] Lactam         -
[beta]-Lactam/Macrolide      -
Macrolide                    -

Note: Table made from pie graph.

Table 6. Initial drug therapy appropriateness: Patients >18 yr old (a)

Drug therapy                            No. of    Meets 1998 IDSA
                                        patients  guidelines/preferred

Azithromycin                                1            Yes/No
Cefotaxime + azithromycin                   2            Yes/Yes
Ceftriaxone                                 1            Yes/yes
Ceftriaxone + azithromycin                 30            Yes/yes
Ceftriaxone + gentamicin                    1             No/no
Cefuroxime + erythromycin                   2            Yes/no
Cefuroxime + azithromycin                   1            Yes/no
Gentamicin + azithromycin                   1             No/no
Levofloxacin + gentamicin                   2             No/no
Levofloxacin + piperacillin/tazobaetam      2          No/no (not ICU)
Levofloxacin + cefixime                     1             No/no
Levofloxacin                               48            Yes/yes
Imipenem + aztreonam                        1             No/no
Ticareillin/clavulanate                     1            Yes/yes
Ampicillin/sulbactam + azithromycin         1            Yes/yes
Piperacillin/tazobactam                     2            Yes/yes
Piperacillin/tazobactam + azithromycin      1            Yes/yes
Total                                      98             90/86
Percentage                                                92/88

Drug therapy                             Meets 2000 ID      Meets CMS
                                        Guidelines/          guidelines
                                        preferred

Azithromycin                                  No/No              Yes
Cefotaxime + azithromycin                    Yes/yes             Yes
Ceftriaxone                                   No/no              Yes
Ceftriaxone + azithromycin                   Yes/yes             Yes
Ceftriaxone + gentamicin                      No/no               No
Cefuroxime + erythromycin                     No/no              Yes
Cefuroxime + azithromycin                     No/no              Yes
Gentamicin + azithromycin                     No/no               No
Levofloxacin + gentamicin                     No/no               No
Levofloxacin + piperacillin/tazobaetam     No/no (not ICU)        No
Levofloxacin + cefixime                       No/no         No (not ICU)
Levofloxacin                                  Yes/yes            Yes
Imipenem + aztreonam                           No/no              No
Ticareillin/clavulanate                        No/no              No
Ampicillin/sulbactam + azithromycin           Yes/yes            Yes
Piperacillin/tazobactam                        No/no             Yes
Piperacillin/tazobactam + azithromycin        Yes/yes            Yes
Total                                          82/82              89
Percentage                                     84/84              91

Drug therapy                             Meets 2000 CDC
                                        guidelines/preferred

Azithromycin                                   No/no
Cefotaxime + azithromycin                     Yes/yes
Ceftriaxone                                    No/no
Ceftriaxone + azithromycin                    Yes/yes
Ceftriaxone + gentamicin                       No/no
Cefuroxime + erythromycin                     Yes/yes
Cefuroxime + azithromycin                     Yes/yes
Gentamicin + azithromycin                      No/no
Levofloxacin + gentamicin                      No/no
Levofloxacin + piperacillin/tazobaetam         No/no
Levofloxacin + cefixime                        No/no
Levofloxacin                                   Yes/no
Imipenem + aztreonam                           No/no
Ticareillin/clavulanate                        No/no
Ampicillin/sulbactam + azithromycin           Yes/yes
Piperacillin/tazobactam                        No/no
Piperacillin/tazobactam + azithromycin         No/no
Total                                          84/37
Percentage                                     86/38

(a) IDSA, Infections Diseases Society of America: CMS, Centers for
Medicare and Medicaid Services: CDC, Centers for Disease Control and
Prevention.

Table 7. Initial drug therapy appropriateness: Patients 8 to 18 yr old
(a)

Drug therapy
               No. of   Meets adapted Canadian   Meets 2000 CDC
              patients    pediatric guidelines  guidelines/preferred

Ceftriaxone       1                 No                  No/no

Ceftriaxone +     1                 Yes                Yes/yes
azithromycin

Cefuroxime        2                 No                  No/no

Cefuroxime +      4                 Yes                Yes/yes
azithromycin

Imipenem +        1                 No                  No/no
gentamicin

Total             9                  5                   5/5

Percentage                          56                  56/56

(a) CDC, Centers for Disease Control and Prevention.

Table 8. Initial drug therapy appropriateness: Patients <8 yr old (a)

Drug therapy                          No. of   Meets 1997
                                     patients     CMAJ
                                               guidelines (b)

Azithromycin                             1          1
Cefotaxime + ampicillin                  2          0
Cefotaxime + clindamycin                 1          0
Ceftriaxone                              6          6
Ceftriaxone + amoxicillin                1          0
Ceftriaxone + azithromycin               1          1
Cefuroxime                              18         18
Cefuroxime + amoxicillin                 1          0
Cefuroxime + Azithromycin               12         12
Cefuroxime + clindamycin                 1          0
Cefprozil                                1          0
Cefprozil + azithromycin                 1          0
Pediazole                                1          0
Piperacillin/tazobactam + aztreonam      1          0
Total                                   48         38
Percentage                                         79

(a) CMAJ, Canadian Medical Association Journal.

(b) Based on the Canadian Medical Association guidelines modified for
local sensitivity concerns.

Acknowledgments

We thank Major Patricia Blake, RN, for collecting data and reviewing medical records, and Walter Brehm, MS, of the 81st Medical Group Clinical Research Laboratory for reviewing the data and performing the appropriate statistical tests.

From the Department of Pharmacy Research, Keesler Medical Center, Keesler AFB, MS.

Presented in part at the Combined Forces Pharmacy Seminar. November 1, 2001, Orlando, FL.

Carinda Field's position is supported through an unrestricted grant from Schering Corp. Allan Stowers, in addition to serving in the U.S. Air Force Reserves. works for Pfizer, Inc., in the capacity of clinical education consultant.

Reprint requests to D. Randall Ziss, PharmD, BCPS, FASHP, 81MDSS/SGSP, 301 Fisher Street, 1A 132, Keesler AFB, MS 39534-2519. Email:[e-mail] 

Accepted July 23, 2002.

Copyright [c] 2003 by The Southern Medical Association 0038-4348/03/9610-0949

References

1. National Institute of Allergy and Infectious Diseases. Community-acquired pneumonia in adult and elderly populations. Clin Courier 1999;16(55):1-16 (CME).

2. Bartlett JG, Breiman RF, Mandell LA, File TM Jr. Community-acquired pneumonia in adults: Guidelines for management-The Infectious Diseases Society of America. Clin Infect Dis 1998;26:811-838.

3. Fedson DS. Pneumococcal vaccination in the prevention of community-acquired pneumonia: An optimistic view of cost-effectiveness. Semin Respir Infect 1993;8:285-293.

4. Heffron R. Pneumonia, with Special Reference to Pneumococcus Lobar Pneumonia. Cambridge, MA, Harvard University Press, 1979.

5. Bullows JGM. The reliability of sputum typing and its relation to serum therapy. JAMA 1935;105:1512-1518.

6. Garibaldi RA. Epidemiology of community-acquired respiratory tract infections in adults: Incidence, etiology, and impact. Am J Med 1985;78:32-37.

7. Marston BJ, Plouffe JF, Breiman RF, et al. Preliminary findings in a community-based pneumonia incidence study, in Barbaree JM, Breiman RF. Dufour AP (eds): Legionella: Current Status and Emerging Perspectives. Washington, DC, American Society for Microbiology, 1993, pp 36-37.

8. Niederman MS, McCombs JS, Unger AN, Kumar A. Popovian R. The cost of treating community-acquired pneumonia. Clin Ther 1998;20:820-837.

9. American Thoracic Society. Guidelines for the initial management of adults with community-acquired pneumonia: Diagnosis, assessment of severity, and initial antimicrobial therapy. Am Rev Respir Dis 1993;148:1418-1426.

10. Murphy SL. Deaths: Final data for 1998. Natl Vital Stat Rep 2000;48(11):1-105.

11. Nathwani D. Rubinstein E, Barlow G. Davey P. Do guidelines for community-acquired pneumonia improve the cost-effectiveness of hospital care? Clin Infect Dis 2001;32:728-741.

12. Bartlett JG, Dowell SF, Mandell LA, File TM Jr, Musher DM. Fine MJ. Practice guidelines for the management of community-acquired pneumonia in adults: Infectious Diseases Society of America. Clin Infect Dis 2000;31:347-382.

13. Heffelfinger JD, Dowell SF, Jorgensen JH, Klugman KP, Mabry LR, Musher DM, et al. Management of community-acquired pneumonia in the era of pneumococcal resistance: A report from the Drug-Resistant Streptococcus pneumoniae Therapeutic Working Group. Arch Intern Med 2000;160:1399-1408.

14. Niederman MS, Mandell LA, Anzueto A, Bass JB, Broughton WA, Campbell GD, et al. Guidelines for the management of adults with community-acquired pneumonia: Diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001;163:1730-1754.

15. Jadavji T. Law B, Lebel MH, Kennedy WA, Gold R, Wang EE. A practical guide for the diagnosis and treatment of pediatric pneumonia. CMAJ 1997;156:S703-S711.

16. Meehan TP, Fine MJ, Krumholz HM, Scinto JD, Galusha DH, Mockalis JT, et al. Quality of care, process, and outcomes in elderly patients with pneumonia. JAMA 1997;278:2080-2084.

17. Yih JY, Black SA, Beck CK. Retrospective evaluation of the management and outcomes in hospitalized patients with community acquired pneumonia applying an outcome prediction score. Abstracts of the ASHP Clinical Mid-year Meeting, December 3-7, 2001, New Orleans, LA.

18. Amberik JC, Cummings TL. Impacting patient outcomes in hospitalized patients with community acquired pneumonia. Abstracts of the ASHP Clinical Mid-year Meeting, December 3-7, 2001, New Orleans, LA.

19. Marston BJ, Plouffe JF, File TM Jr, Hackman BA, Salstrom SJ, Lipman HB, et al. Incidence of community-acquired pneumonia requiring hospitalization: Results of a population-based active surveillance study in Ohio-The Community-Based Pneumonia Incidence Study Group. Arch Intern Med 1997;157:1709-1718.

20. Atlas Scoring: A Technical White Paper. Westborough, MA, MediQual Systems, Feb 1996.

21. Kahn KL, Rogers WH, Rubenstein LV, Sherwood MJ, Reinisch EJ, Keeler EB. et al. Measuring quality of care with explicit process criteria before and after implementation of the DRG-based prospective payment system. JAMA 1990;264:1969-1973.

22. McGarvey RN, Harper JJ. Pneumonia mortality reduction and quality improvement in a community hospital. QRB Qual Rev Bull 1993;19:124-130.

23. Mandell LA. Antibiotic therapy for community-acquired pneumonia. Clin Chest Med 1999;20:589-598.

24. Gordon GS, Throop D, Berberian L, et al. Validation of the therapeutic recommendations of the American Thoracic Society guidelines for community-acquired pneumonia in hospitalized patients, in Abstracts of the CHEST World Congress, October 27-31, 1996, San Francisco, CA.

25. Waterer GW, Somes GW, Wunderink RG. Monotherapy may be sub-optimal for severe bacteremic pneumococcal pneumonia. Arch Intern Med 2001;161:1837-1842.

26. Gleason PP, Meehan TP, Fine JM, Galusha DH, Fine MJ. Associations between initial antimicrobial therapy and medical outcomes for hospitalized elderly patients with pneumonia. Arch Intern Med 1999;159:2562-2572.

27. Bartlett JG, Mundy LM. Community-acquired pneumonia. N Engl J Med 1995;333:1618-1624.

28. Gay K, Baughman W, Miller Y, Jackson D, Whitney CG, Schuchat A, et al. The emergence of Streptococcus pneumoniae resistant to macrolide antimicrobial agents: A 6-year population-based assessment. J Infect Dis 2000;182:1417-1424.

29. Chen DK, McGeer A, de Azavedo JC, Low DE. Decreased susceptibility of Streptococcus pneumoniae to fluoroquinolones in Canada: Canadian Bacterial Surveillance Network. N Engl J Med 1999;341:233-239.

RELATED ARTICLE: Key Points

* Compliance with national guidelines for the management of community-acquired pneumonia is possible only through education and reinforcement of the medical staff, particularly when conflicting consensus guidelines are published.

* Patients who were treated with antibiotics within 4 hours of admission had shorter hospital stays than did patients who were administered the first dose of antibiotics more than 4 hours after presentation.

* Patients who were not administered antibiotics until 8 hours after admission had the highest mortality rate of the patients studied.

* Screening and documentation of influenza and pneumococcal immunization for inpatients with community-acquired pneumonia must be completed to prevent avoidable readmissions and to reduce morbidity and mortality.

D. Randall Ziss, PHARMD, BCPS, FASHP, Allan Stowers, RPH, MS, and Carinda Feild, PHARMD