| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Medicine (Dr. Kanwar and Brar), Division of Infectious Diseases (Drs. Khatib and Fakih), St. John Hospital and Medical Center, Detroit, MI.
Correspondence to: Mohamad G. Fakih, MD, MPH, Division of Infectious Diseases, Department of Medicine, St. John Hospital and Medical Center, 19251 Mack Ave, Suite 340, Grosse Pointe Woods, MI 48236; e-mail: mohamad.fakih{at}stjohn.org
Abstract
Background: The 2003 Infectious Diseases Society of America guidelines for community-acquired pneumonia (CAP) recommend the initiation of antibiotic therapy within 4 h of hospitalization. This quality indicator has been linked to the incentive compensation of third-party payers to hospitals. We evaluated the impact of this recommendation on the diagnosis of CAP and the utilization of antibiotics.
Methods: All patients with a hospital admission diagnosis of CAP before publication of the guidelines (January to June 2003) and after publication of the guidelines (January-June 2005) were included. We collected data on clinical signs and symptoms on presentation, chest radiograph findings, blood cultures prior to therapy with antibiotics, time to antibiotic administration, pneumonia severity index (PSI) score, confusion, urea, respiratory rate, BP, and age 
65 years (CURB-65), and mortality.
Results: A total of 518 patients were included in the study. More patients in 2005 had a hospital admission diagnosis of CAP without radiographic abnormalities compared to 2003 (2005, 91 patients [28.5%]; 2003, 41 patients [20.6%]; p = 0.04), and more patients received antibiotics within 4 h of triage (2005, 210 patients [65.8%]; 2003, 107 patients [53.8%]; p = 0.007). Blood cultures prior to antibiotic administration increased (2005, 220 patients [69.6%]; 2003, 93 patients [46.7%]; p < 0.001). However, the final diagnosis of CAP dropped to 58.9% in 2005 from 75.9% in 2003 (p < 0.001). The mean (± SD) antibiotic utilization per patient increased to 1.66 ± 0.54 in 2005 compared to 1.39 ± 0.58 in 2003 (p < 0.001). There were no significant differences in PSI or CURB-65 scores, or mortality.
Conclusions: Linking antibiotic administration within 4 h of hospital admission (as a quality indicator) to financial compensation may result in an inaccurate diagnosis of CAP, inappropriate utilization of antibiotics, and thus less than optimal care.
Key Words: antibiotics • pneumonia • practice
Community-acquired pneumonia (CAP) is one of the leading causes for adult hospitalizations in the United States, accounting for 1.3 million hospital discharges each year, with significant morbidity and cost.1 The Infectious Diseases Society of America (IDSA) published an update for practice guidelines for management of CAP in November 2003,2 in which it recommended the administration of antibiotics within 4 h of registration in the emergency department (ED). This recommendation was based on an association between timely antimicrobial therapy and reduced mortality, decreased length of stay (LOS), and improved outcomes in patients presenting with CAP.34
There is controversy on the use of this and other performance measures as the basis of public reporting and pay for performance programs, as encouraged by the Centers for Medicare and Medicaid Services.56789 Linking the administration of antibiotics within 4 h of ED arrival to financial compensation has placed increased pressure on hospitals and ED physicians to deliver the antibiotics promptly before the diagnosis of CAP is established. Our hospital participated in the Blue Cross-Blue Shield of Michigan incentive program to improve practice, including antibiotic prescribing patterns, by way of added reimbursement for meeting targets for select performance measures. The goal of our study is to determine whether there is an association between the increase in the numbers of patients who are misdiagnosed with CAP and the implementation of guidelines to ensure antibiotic delivery within 4 h of triage.
Materials and Methods
Study Design
The study was performed in a 608-bed teaching hospital with 112,000 annual ED visits. We performed a retrospective study of all patients who were admitted to the ED with a diagnosis of CAP over two 6-month periods prior to the publication of IDSA guidelines2 (January to June 2003) and 1 year after publication (January to June 2005). We obtained an institutional review board approval prior to starting the study.
Patients
All patients 21 years of age who were admitted to the ED with a primary or secondary diagnosis of pneumonia (International Classification of Diseases, ninth edition, Clinical Modification, codes 480.0–483.8, 485–486, or 487.0) were evaluated. Patients presenting from home or a nursing home were included in the study. Patients who had no documentation of antibiotic administration in the ED or at triage time were excluded from the study. Other exclusions involved transfers from other hospitals and acute care facilities, patients who died on the day of ED admission, patients who left the hospital against medical advice, or patients who received comfort care before a confirmatory diagnosis of CAP was made.
Data Collection and Analysis
Data were obtained regarding the timing and selection of antibiotics, the collection of blood cultures prior to antibiotic therapy, transfer to ICUs, hospital LOS, in-hospital mortality, and mean antibiotic utilization per patient. The time to the first antibiotic dose was calculated as the time difference between the recorded time on admission to triage in the ED and the recorded time of the administration of the first dose of antibiotics by the nursing staff. A pneumonia severity index (PSI) score was calculated for each patient at the time of ED admission.10 We did not include arterial pH of < 7.35 in our PSI calculation as blood gas analyses were not performed in the majority of patients. The confusion, urea, respiratory rate, BP, the age 65 years (CURB-65) score for the assessment of disease severity, based on British Thoracic Society11 guidelines was also calculated. Chest radiograph (CXR) findings were considered to be consistent with pneumonia if the radiology report contained any of the following terms: pneumonia; consolidation; infiltrate; opacity; or pneumonitis. Those patients reported as being "unable to exclude infiltrate or atelectasis" were considered to be nonconfirmatory for pneumonia. A final (hospital discharge) diagnosis of pneumonia was noted if the patient received a diagnosis of pneumonia during their hospital stay (as documented in the progress notes) or on hospital discharge (per the hospital discharge summary) by the attending physician, infectious disease specialist, or a pulmonologist. Alternative definitions of pneumonia were evaluated as follows: (1) definition A, CXR showing an infiltrate or consolidation, and one or more among shortness of breath, cough, sputum production, and a temperature of > 37.8°C; and (2) definition B, an infiltrate seen on a CXR and two or more of the above symptoms and signs.
Statistical Analysis
A two-tailed Student t test was used for continuous variables and a two-sided Fisher exact test was used for categoric variables. A data analysis was performed using a statistical software package (SPSS, version 12.0; SPSS Inc; Chicago, IL). A p value of < 0.05 for a two-sided test was considered to be statistically significant.
Results
Patient Characteristics
A total of 734 patients (2003, 291 patients; 2005, 443 patients) were initially identified with an ED admitting diagnosis of pneumonia. Of these, 518 patients (2003, 199 patients; 2005, 319 patients) were included in our study. The reasons for study exclusion were age < 21 years (n = 118), incomplete data (n = 53), transfer from another acute care hospital (n = 27), or patients died, left the hospital against medical advice, or were given palliative care before a definite diagnosis was made (n = 18). The patient characteristics of the two study periods, including PSI and CURB-65 scores on ED admission, are shown in Table 1
. The characteristics of the two populations were similar except that a smaller percentage of patients were admitted from nursing homes in 2005 (2005, 13%; 2003, 20%; p = 0.05).
|
|
Correlation With Final Diagnosis
The proportion of patients with a final diagnosis of CAP dropped from 151 patients (75.9%) in 2003 to 188 patients (58.9%) in 2005 (p < 0.001). When evaluating only those patients with a final diagnosis of CAP, 84 patients (55.6%) in 2003 and 123 patients (65.4%) in 2005 had received antibiotics within 4 h of triage (p = 0.07). In addition, the mean time to antibiotic administration was 251 ± 165 min in 2003 compared to 250 ± 204 min in 2005 (p = 0.95). On the other hand, for those patients with a final diagnosis different from CAP, 23 (47.9%) received antibiotics within 4 h in 2003 compared to 87 (66.4%) in 2005 (p = 0.04). In 2003, most of the patients with a non-CAP final diagnosis (n = 48) were treated for noninfectious cardiac and pulmonary diagnoses; these final diagnoses increased in number (n = 131) in 2005. On the other hand, a final diagnosis related to infection other than CAP was less common (2003, 12 patients [6%]; 2005, 16 patients [5%]).
Comparing ED Admitting Diagnosis to Our Definitions
Based on our definition A (ie, definite CXR findings and one or more clinical findings), only 89 of those patients admitted to the ED with CAP (44.7%) in 2003 had the diagnosis of pneumonia, and this number fell to 115 patients (36%) in 2005 (p = 0.06). According to our more stringent definition B (ie, definite CXR findings and two or more clinical findings), 65 patients (32.7%) in 2003 and 86 patients (27.0%) in 2005 had a diagnosis of pneumonia (p = 0.17).
Compliance With Other Quality Indicators and Outcomes
Obtaining blood culture results prior to antibiotic administration improved in 2005 when compared to 2003 (2005, 220 blood cultures [69.6%]; 2003, 93 blood cultures [46.7%]; p < 0.001). All patients were fully compliant with oxygenation assessment on admission. The use of CAP standing orders did not significantly change over the two study periods (2003, 152 patients [ 76.4%]; 2005, 250 patients [79.1%]). There were no significant differences between years in terms of the number of transfers to the ICU, the average hospital LOS, or the in-hospital mortality rate between the two groups (Table 2).
Discussion
The association between early antibiotic administration for CAP and improved outcomes has been reported in two large retrospective studies. Meehan et al3 found that the administration of antibiotics within 8 h of arrival at the hospital for Medicare patients was associated a lower 30-day mortality rate (odds ratio, 0.85; 95% confidence interval, 0.75 to 0.96). As a result, the Medicare National Pneumonia Project, from 1999 through 2002, promoted the administration of antimicrobial agents within 8 h of arrival at the hospital. Houck et al4 reported similar findings in 2004 for 13,771 Medicare CAP patient hospitalizations from 1998 to 1999. Antibiotic administration within 4 h of hospital arrival was associated with a lower in-hospital mortality rate compared to the mortality rate in those who had received antibiotics more than 4 h after hospital arrival (6.8% vs 7.4%, respectively).4 The Medicare Prescription Drug, Improvement, and Modernization Act of 2003 linked the reporting of quality markers (including 4-h antibiotic delivery) to compensation from the Centers for Medicare and Medicaid Services starting in 2005.5 Although these measures promote the use of evidence-based medicine, concerns have been raised about the processes used to achieve compliance from hospitals and physicians. The major argument has been that doing so would necessitate antibiotic delivery to many patients in whom the diagnosis is still in doubt.
We found that compliance with the 4 h-antibiotic-administration rule led to an increase in the misdiagnosis of CAP, and subsequently to greater utilization of inappropriate antibiotics. We have seen almost a 60% increase in the hospital admitting diagnosis of CAP compared to a < 25% increase in the hospital discharge diagnosis of CAP. When we applied our definitions for CAP, the proportion of patients with pneumonia was even lower. It is important to note that a minority of the patients who did not have a final diagnosis of CAP had a diagnosis related to another infection. Thus, many patients received antibiotics inappropriately for a noninfectious process. In addition, resource utilization such as obtaining blood cultures for patients who were admitted to the hospital with CAP was also impacted.2 Moreover, specific standing orders for patients CAP that are used to improve quality of care will be of no benefit or may impact negatively care in those patients who do not have CAP. This underscores the importance of making the correct diagnosis rather than being forced to apply a quality indicator to the wrong patient.
Are we able to improve the process of evaluating for CAP and initiating therapy with antibiotics on a timely basis? Our study demonstrates that attempting to increase compliance with antibiotic administration within 4 h resulted in a marked increase in misdiagnosis and inappropriate antibiotic utilization. A reasonable approach is to establish a target that is more feasible. We suggest enlarging the antibiotic administration window to 6 h. This target may provide more time for physicians to provide a better evaluation of the patient.7 Moreover, the performance measures need to include, in addition to the assessment of those patients with the final diagnosis of CAP, the accuracy of the hospital admitting diagnosis. Evaluating the accuracy of the diagnosis on hospital admission and linking it to performance may provide an incentive to physicians to work on achieving the correct diagnosis in the ED. Very recently, the new IDSA/American Thoracic Society guidelines for CAP12 recommended using no specific time window for the delivery of the first antibiotic dose; the guidelines suggested administering the first dose in the ED.
Some limitations of our study must be acknowledged. Like all studies employing retrospective chart review, we were dependent on what was documented. Some factors may not have been documented in patient records and potentially may have affected the reviewer’s findings. However, this should not affect the comparison between the hospital admission and the final diagnoses, showing that many cases that were originally labeled as being CAP were labeled with a different diagnosis on hospital discharge.
Footnotes
Abbreviations: CAP = community-acquired pneumonia; CURB-65 = confusion, urea, respiratory rate, BP, and age 65 years; CXR = chest radiograph; ED = emergency department; IDSA = Infectious Diseases Society of America; LOS = length of stay; PSI = pneumonia severity index
This study received the Kass Award and was presented at the 44th Annual Meeting of the Infectious Diseases Society of America, October 2006.
The authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.
Received for publication January 17, 2007. Accepted for publication February 2, 2007.
References
This article has been cited by other articles:
|
|
 |
|
  P. R. Srivaths, V. F. Corrales-Medina, M. G. Fakih, R. Khatib, M. Kanwar, and N. Brar Timing of Antibiotics and Community-Acquired Pneumonia Chest, December 1, 2007; 132(6): 2060 - 2061. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
