Rural Hospital: METHODS
The process of developing the CAP guidelines began in 1999 at LRGH. The Infection Control and Epidemiology Committee had established a Respiratory Continuous Quality Improvement (CQI) subcommittee to address the rising rates of hospital-acquired (nosocomial) pneumonia. This multidisciplinary group reviewed all cases of nosocomial pneumonia and recommended improvements that resulted in a reduction of these cases to only 25% of former levels.
With the publication of the IDSA guidelines for CAP in 1998, the Respiratory CQI subcommittee decided to apply this successful methodology to CAP. A one-in-four sample of all patients admitted in 1999 for CAP was reviewed to identify the parameters outlined in the IDSA guidelines. Baseline data were collected for antibiotic selection, use of services, LOS, and charges. From these data and from those of the IDSA and ATS guidelines, the treatment algorithm shown in Table 1 was developed.
Table 1 Treatment of Community-Acquired Pneumonia
| Adults 18 Years or Older | Class of Care | Antibiotic Use | Laboratory/Procedure Orders |
| Outpatient | Class I or II outpatient | <60 yrs of age: >60 yrs of age or Macrolide comorbid illness Doxycycline Avelox generic/Tequin® | None |
| Day 1 (SOS) | Class III
hospitalization SOS (24 hr) |
Avelox®/Tequin® if taking food; otherwise, IV Tequin®; first dose given in ED within two hours | CBC with differential Sputum C&S Oximetry BMP, chest x-ray |
| Day 1 | Class IV or V hospitalization (acute stay) | IV Tequin or IV ceftriaxone and Avelox®/Tequin generic or ceftriaxone (Rocephin®) and a macrolide; first dose given in ED or nursing unit within two hours | CBC with differential
Sputum C&S Oximetry or ABG CMP, chest x-ray, UA Blood cultures if S&S of sepsis |
| Day 2 | Case manager to assess for patient discharge and to have VNA see patient at home on first day after discharge for skilled nursing visit | Daily assessment to switch IV to PO antibiotic based on vital signs, clinical condition, oxygen saturation, x-ray studies, and culture results | Laboratory tests as needed |
| Days 3-5 | Patient discharged home;
VNA to visit one day after hospital discharge |
PO antibiotic or home IV (rare) | Home oxygen as needed
Saturation below 90% Home nebulizer therapy as needed |
| Follow-up care | Physician to see patient one week after discharge, chest x-ray 4-6 weeks; to follow until patient values return to baseline | PO antibiotic 7-14 days | |
| ABG = arterial blood gas; BMP = basic metabolic panel; CBC = complete blood count; CMP = complete metabolic panel; C&S = culture and sensitivity; ED = emergency department; IV = intravenous; PO = oral; S&S = signs and symptoms; SOS = short-order stay; UA = urinalysis; VNA = Visiting Nurse Association. | |||
Visits were made to the Departments of Internal Medicine and Family Practice to solicit input from admitting providers; adjustments were made according to the feedback received, and training was provided to all involved hospital staff and home care agencies. Implementation of the initiative began in June 2000. For this analysis, patients requiring a hospitalization for CAP during 1999 were included in the pre-implementation phase and were compared with patients requiring hospitalization for CAP after implementation of the guidelines, from December 1, 2000, to November 30, 2001.
Figure 1 Pharmacotherapy distribution by therapeutic class for patients with community-acquired pneumonia. Pre = pre-intervention; Post = post-intervention.
Patients were included in the study if they (1) were at least 18 years of age and (2) had a primary diagnosis of CAP Patients were excluded if they had the following:
- a serious infection (e.g., meningitis, endocarditis, continuing bacteremia) or neutropenia, nosocomial pneumonia
- a primary diagnosis other than CAP
- a compromised immune system
- the need for prescribed vasopressor agents
The Director of Infection Control initially confirmed the CAP diagnoses and the inclusion and exclusion criteria, and an outside research consulting firm later verified these data after reviewing patients’ charts.
canada viagra online
The clinical data used in the study were also obtained through this chart review and included the collection of patient demographics, antibiotic therapy, the route of administration, and the LOS. Patient-specific charges were obtained from the accounting records of LRGH. These charges were adjusted according to the percent increase of specific charge categories over the study period; thus, all charges represented 1999 dollars.
Quality of care and adherence to the guidelines were defined as an appropriate initial antibiotic (see Table 1) and the administration of the antibiotic within two hours of a written order.
Antibiotic selection was deemed appropriate on the basis of a calculated risk score, the IDSA-based Fine score methodol-ogy. This classification system is used to predict the risk of death within 30 days and stratifies the patients into five classes (I to V). Higher-risk classes are considered to have more severe cases of CAP.
The score was calculated by assessment of the patients’ age, the presence of coexisting disease, abnormal physical findings (respiratory rate or temperature), and abnormal laboratory values (pH, blood urea concentration, and sodium concentration). Because these data were collected retrospectively, missing data complicated the derivation of the Fine score. The missing data were assumed to be negligible, and patients were assigned to the most conservative (lowest-risk) class, resulting in the potential to underestimate the actual Fine score.
Descriptive statistics, means, variances, and frequencies were initially used to compare outcomes in the two measurement periods. Differences in demographics and the proportion of patients receiving an appropriate antibiotic were analyzed using chi-square, Fisher’s exact test, and logistic regression, when appropriate. Variance and multivariate regression analyses were used to determine differences in the use of resources and LOS. Charges were log-transformed and modeled as a function of the intervention and relevant covariates (e.g., age, sex, and comorbidities). A P value of .05 or less was determined to be statistically significant. SAS statistical software, version 8.0 (SAS Institute Inc., Cary NC), was used to perform all analyses.
