Management and Prevention of CommunityAcquired Pneumonia The Medicare National Pneumonia Project

1
Management and Prevention of Community-Acquired
PneumoniaThe Medicare National Pneumonia Project
Dale W. Bratzler, DO, MPH Principal Clinical
Coordinator Oklahoma Foundation for Medical
Quality
2
Community-Acquired PneumoniaEpidemiology

• Sixth leading cause of death
• 1 from infectious disease
• Up to 5.6 million cases per year
• gt10 million physician visits
• 1.1 million hospitalizations
• Mortality
• Outpatient - lt 1
• Admit (ward) - 10-14
• ICU - 30-40

Niederman MS, et al. Am J Respir Crit Care Med.
20011631730-1754. Bartlett JG, et al. Clin
Infect Dis. 200031347-382.
3
Performance Measure 1Antibiotic Timing

• Proportion of patients receiving first dose of
  antibiotics within4 hours

4
First Dose Timing and Outcomes
Using multivariate logistic regression the
model included the timing of antibiotic first
dose, PSI score, ICU admission, US census region,
race/ethnicity, other processes of care
(oxygenation assessment, performance of blood
cultures, and antibiotic selection). Patients
who were on antibiotics prior to admission are
excluded from this analysis. (Houck PM, Bratzler
DW, et al. Arch Intern Med. 2004.)
5
First Dose Timing and Outcomes
Using multivariate logistic regression the
model included the timing of antibiotic first
dose, PSI score, ICU admission, US census region,
race/ethnicity, other processes of care
(oxygenation assessment, performance of blood
cultures, and antibiotic selection).
Houck PM, Bratzler DW, et al. Arch Intern Med.
2004.
6
Performance Measure 2Initial Antibiotic
Selection

• Proportion of patients who receive an initial
  antibiotic regimen consistent with current
  guidelines

7
Etiology of Community-acquired Pneumonia
Ambulatory Patients Hospitalized
(Non-ICU)2 Severe (ICU)2 S. pneumoniae S.
pneumoniae S. pneumoniae M. pneumoniae M.
pneumoniae H. influenzae H. influenzae C.
pneumoniae Legionella spp. C. pneumoniae H.
influenzae Gram-negative bacilli Viruses Legio
nella spp. S. aureus
Aspiration ICU Intensive care unit 1 Based
on collective data from recent studies 2
Excluding Pneumocystis spp From File et al, Am
J Med Continuing Educat Series 1997
8
3rd Generation Cephalosporin
2nd or 3rd Generation Cephalosporin
macrolide Quinolone monotherapy
Gleason PP, et al. Arch Intern Med.
19991592562-2572.
9
Initial Antibiotic Selection and
Outcomes1998-1999

• These were adjusted for age, gender, neoplastic
  disease, cardiovascular disease, altered mental
  status, respiratory rategt30/min, systolic BPlt90
  mmHg, temperaturelt35 C or gt40 C, pulsegt125/min,
  blood pHlt7.35, BUNgt10.7 mmol/L, sodiumlt130 mEq/L,
  hematocritlt30, PO2lt60 mmHg, pleural effusion.
• Oral macrolides and quinolones were included in
  these groups as well as parenteral macrolides and
  quinolones.

Preliminary analysis
10
Initial Antibiotic Selection and
Outcomes2000-2001

• These were adjusted for age, gender, neoplastic
  disease, cardiovascular disease, altered mental
  status, respiratory rategt30/min, systolic BPlt90
  mmHg, temperaturelt35 C or gt40 C, pulsegt125/min,
  blood pHlt7.35, BUNgt10.7 mmol/L, sodiumlt130 mEq/L,
  hematocritlt30, PO2lt60 mmHg, pleural effusion.
• Oral macrolides and quinolones were included in
  these groups as well as parenteral macrolides and
  quinolones.

Preliminary analysis
11
Performance Measure 3Blood Cultures

• A. Proportion of patients who have a blood
  culture performed within 24 hours /- of hospital
  arrival
• B. Proportion of patients whose initial blood
  culture specimen is collected after hospital
  arrival and prior to initial antibiotic dose

Niederman MS, et al. Am J Respir Crit Care Med.
20011631730-1754. Bartlett JG, et al. Clin
Infect Dis. 200031347-382.
12
Performance Measure 3Blood Cultures

• Lower 30-day mortality was associated with blood
  culture collection within 24 hours of hospital
  arrival
• Odds Ratio 0.90 (95 CI 0.81-1.00)
• Could promote pathogen-directed therapy when
  positive

Meehan TP, et al. JAMA 19972782080-2084.
13
Blood Culture IsolatesNational Pneumonia
Project, 1998-1999, 2000-2001
Metersky ML, et al. Am J Resp Crit Care Med.
2004169342-347.
14
Independent Predictors of BacteremiaNational
Pneumonia Project, 1998-1999, 2000-2001
Metersky ML, et al. Am J Resp Crit Care Med.
2004169342-347.
15
Bacteremia Risk based on PredictorsNational
Pneumonia Project, 1998-1999, 2000-2001
Derivation cohort is based on 13 034 patients
hospitalized in 1998-1999 and the validation
cohort is based on 12 771 patients hospitalized
in 2000-2001.
Metersky ML, et al. Am J Resp Crit Care Med.
2004169342-347.
16
Predicting Bacteremia in CAP

• Implications of failure to diagnose bacteremia??
• 11 false negative rate using decision support
  tool
• Positive blood cultures do not result in
  adjustments of antibiotic therapy
• Pneumococcal bacteremia does not require longer
  treatment
• However, with current practice, 38 of bacteremic
  patients are missed
• Consequences of false positive blood culture
• Independently associated with 1-day excess length
  of stay
• More likely to receive vancomycin treatment

Metersky ML, et al. Am J Resp Crit Care Med.
2004169342-347.
17
Performance Measure 6Adult Smoking Cessation
Advice

• Proportion of patients who smoked during the
  previous year who receive smoking cessation
  advice or counseling prior to discharge

18
Performance Measure 6Adult Smoking Cessation
Advice

• Smoking cessation treatments ranging from brief
  clinician advice to specialist-delivered
  intensive programs are effective
• Advice from physician more likely to quit
• Ideal time to convince a patient that their
  symptoms may be related to smoking

Fiore MC, et al. Treating tobacco use and
dependence. Clinical Practice Guideline.
Rockville, MD U.S. Department of Health and
Human Services. Public Health Service. June
2000. Bratzler DW, et al. Smoking in the elderly
Its never too late to quit. J Okla State Med
Assoc. 200295185-191.
19
Performance Measure 7Oxygenation Assessment

• Proportion of patients whose arterial oxygenation
  is assessed by ABG measurement and/or pulse
  oximetry within 24 hours of hospital arrival

Niederman MS, et al. Am J Respir Crit Care Med.
20011631730-1754. Bartlett JG, et al. Clin
Infect Dis. 200031347-382.
20
Performance Measure 7Oxygenation Assessment

• Hypoxemia associated with worse outcomes from
  pneumonia
• Provides prognostic information that may assist
  in the decision to hospitalize

21
Performance Measure 4 5Inpatient Vaccination

• Proportion of patients, age 50 years and over,
  discharged during the months of October through
  February who are screened for and, if indicated,
  given influenza vaccine prior to discharge.
• Proportion of patients, age 65 years and over,
  who are screened for and, if indicated, given
  pneumococcal polysaccharide vaccine prior to
  discharge.

22
(No Transcript)
23
Deaths Due to Vaccine Preventable Diseases - US,
1989-1998
630k (90 in elderly)
11k (reported cases actual likely 5-10x more)
MMWR 2001 48 (RR-53) Thompson et al. JAMA 2003
289 179 Feikin DR, et al. Am J Public Health
2000 90 223-9. Data for influenza
pneumococcal diseases are estimates, data for
other diseases are reported cases.
24
Rationale for Hospital-based Influenza Vaccination
Proportion of the population with subsequent
hospitalization or death from pneumonia or
influenza who had been discharged from a
hospital. Fedson DS, Houck P, Bratzler D.
Editorial Infect Control Hosp Epidemiol.
200021692-699.
25
Missed OpportunitiesBackground
Fedson DS, et al. Infect Control Hosp Epidemiol.
200021692-699.
26
Missed OpportunitiesInfluenza vaccine
Medicare patients aged gt 65 years discharged
alive from the hospital between October 1 and
December 31, 1998. Prior based on Medicare
claims analysis or medical record abstraction for
1998 flu season. After discharge based on
Medicare claims analysis through January 31, 1999.
Bratzler DW, Houck PM, et al. Arch Intern Med.
20021622349-2356.
27
Missed OpportunitiesPneumococcal vaccine
Medicare patients aged gt 65 years discharged
alive from the hospital between July 1, 1998, and
March 31, 1999. Prior based on Medicare claims
analysis (Part B data back to 1991) or medical
record abstraction. After discharge based on
Medicare claims analysis for one month after
dismissal.
Bratzler DW, Houck PM, et al. Arch Intern Med.
20021622349-2356.
28
Success of Standing Orders as Part of a
Multifaceted Program
Nichol KL. Am J Med. 1998 105 385.
29
(No Transcript)
30
NRH Immunization Rate
31
Facility-Level Descriptive AnalysisProgram Type
for Vaccination
Bardenheier B, Shefer A, McKibben L, Roberts H,
Bratzler D, Miller J. Preliminary analyses,
CDC/CMS Immunization Standing Orders Project
32
Influenza Program Type and Coverage (N 249)
Usual (n53)
APO (n35)
PPAO (n58)
R/R (n57)
SOP (n44)
Program Type
Median
33
Pneumococcal Program Type and Coverage (N 249)
Usual (n118)
APO (n17)
R/R (n52)
PPAO (n36)
SOP (n24)
Program Type
Median
34
(No Transcript)
35
Vaccination Fee Schedule

• Influenza
• Acquisition fee - 10.10
• Administration fee - 6.70
• Pneumococcal
• Acquisition fee - 23.28
• Administration fee - 6.70

36
www.cms.hhs.gov/preventiveservices/2i.pdf
37
National Trends in Pneumonia Care - Medicare
38
National Pneumonia ProjectOngoing National
Surveillance
We did not exclude comfort care and chest X-ray
patients from any of the data sets, because
information on comfort care and chest X-ray was
not available in the MQIS and the 7th SoW
Surveillance data sets, respectively.
39
National Pneumonia ProjectOngoing National
Surveillance
40
Timing of Antibiotic First Dose
Bratzler DW, Houck PM. Academy for Health
Services Research and Policy. Washington, DC.
June 24, 2002.
41
Initial Antibiotic Selection
Selected combinations shown.
42
Improving Care for Pneumonia
43
Improving Pneumonia Care

• Study of 36 hospitals in Oklahoma
• Improvements in care seen in hospitals with
• standing orders
• clinical pathways

Chu LA, Bratzler DW, Lewis RJ, Murray CK, et al.
Arch Intern Med. 2003163326-332.
44
Example of a hospital pre-printed order form for
antibiotic selection.
45
Improving Care for PneumoniaImplementation of
Treatment Guidelines

• Intermountain Health Care
• Implemented treatment guideline that included
  admission decision support and recommendations
  for antibiotic timing and selection
• Reduced 30-day mortality
• Odds ratio 0.69 (95 CI 0.49-0.97)

Dean NC, et al. Decreased mortality after
implementation of a treatment guideline for
community-acquired pneumonia. Am J Med.
2001110451-457.
46
Standing Orders are Among the Most Effective
Strategies

• Non-physicians offer and administer vaccinations
  without direct doctor involvement at the time of
  the visit
• Established through approved policies protocols
• Locations clinics, hospitals nursing homes

47
Institutional VaccinationNew Medicare Regulation

• Federal Register, Vol. 67, No. 191 (October 2,
  2002)
• All orders for drugs and biologicals must be
  in writing and signed by the practitioner or
  practitioners responsible for the care of the
  patient as specified under 482.12(c) with the
  exception of influenza and pneumococcal
  polysaccharide vaccines, which may be
  administered per physician-approved facility
  policy after an assessment for contraindications.
  
• Includes similar provisions for nursing homes and
  home health agencies

48
National Benchmark PerformancePneumonia
State specific and national rates based on cases
discharged during the 4th quarter of 2003. Based
on the Achievable Benchmarks of Care
methodology for cases discharged from the
hospital during calendar year 20031,2 1Kiefe
CI, et al. Int J Qual Health Care.199810443-447.
2Weissman NW, et al. J Eval Clin Pract.
19995269-281.
49
www.medqic.org/pneumonia