1. Healthcare –Associated Pneumonia

2.  93 female LTCF patient presents with dyspnea and fever.  EMS notified; brought to ER  Medical hx: Afib, dementia, spinal stenosis, R hip OA, LVEF 55%  Vitals: 137/97 P-98, 102F, RR 22, O2 initially 75%, then 93% on 3L  Exam:  Decreased breath sounds on the right  Labs: WBC 8.8, Cr = 0.7, Hgb = 14  CXR: congestive changes; right basilar patchy opacity with positive infiltrate  Diagnosed with pneumonia and admitted

3.  93 y.o. F from LTCF with pneumonia. Treatment?  A. Levofloxacin  B. Levofloxacin + vancomycin  C. Levofloxacin + Pip/Tazo + vancomycin  D. Ceftriaxone  E. Ceftriaxone + azithromycin

4.  P: Elderly LTCF patients with pneumonia  I: Guideline based antibiotics  C: Non-guideline based antibiotics  O: Mortality or clinical outcomes

5.  HAP: hospital acquired pneumonia  Arises 48hrs or more after admission  HCAP: health-care associated pneumonia**  Hospitalized within last 90 days; LTCF, IV therapy, chemotherapy, wound therapy, or attended a hospital or hemodialysis clinic in last 30 days  VAP: ventilator associated pneumonia  Arises more than 48-72 hrs after intubation

6.  HCAP included in spectrum of HAP and VAP  HCAP: need therapy for MDR pathogens  Acknowledges most evidence for VAP  Timing of pneumonia is important  “Early onset HAP/VAP”: within 4 days of hospitalization (likely sensitive bacteria)  “Late onset HAP/VAP”: at 5 days or later (MDR pathogens more likely)

7.  HCAP etiology?  “elderly residents of LTCFs have a spectrum of pathogens that more closely resemble late-onset HAP and VAP”  Study 1 ▪ Staph aureus (29%), enteric GNRs (15%), Strep pneumoniae (9%), Pseudomonas (4%)  Study 2 – “failed to respond to 72 hrs of abx” ▪ MRSA (33%), GNRs (24%), Pseudomonas (14%)

11. Goal: evaluate effectiveness of guideline-based therapy (GBT) compared with other antimicrobial regimens and to identify subgroups of patients with HCAP who received greatest benefit from GBT.

12.  Cohort study; 346 U.S. hospitals  Inclusion  Patient discharge between Jul 2007 – Jun 2010  Age ≥ 18 yrs with ▪ Primary ICD-9 dx of pneumonia, OR ▪ Secondary dx of pneumonia, paired with primary dx of respiratory failure, ARDS, respiratory arrest, sepsis, or influenza  HCAP ▪ If dx of ESRD/dialysis in first 2 hospital days, OR if admit from a SNF, OR if DC from hospital in past 90 days, OR taking immunosuppressant drugs

13.  Exclusion  Transfer patients (could not assess initial severity or outcomes)  Length of stay ≤1 day  Cystic fibrosis  Attending not expected to treat pneumonia  DRG inconsistent with pneumonia  Any pt who did not have a CXR and begin antimicrobials within 48hrs of admission

14.  Data elements  Age, sex, race, marital and insurance status, comorbidities, tests, medications and treatments, physician specialty, comorbidities (via a software program)  Hospitals: region, bed size, rural/urban, teaching status

15.  GBT  1 abx against MRSA and 1 abx against Pseudomonas  Main predictor variable  Non-GBT: all other abx regimens  Primary outcome: in-hospital mortality  Secondary outcomes  7 day mortality, initiation of mechanical ventilation or admission to ICU, readmission in 30 days, cost, length of stay, clostridium difficile infection (CDI)

16.  Categorical variables  Frequencies and proportions  Continuous variables  Medians with IQRs  Logistic regression model for treatment  Propensity scores for GBT vs. non-GBT  Adjusted analyses  Sensitivity analysis  Explore effect of hypothetical unmeasured confounders

17.  N=85,097 patients from 346 hospitals  31,949 (37.5%) received GBT  Of those not receiving GBT, 82% received standard therapy for CAP  GBT patients  Younger  More likely male  More chronic disease  More severe pneumonia

21. Sensitivity Analysis: “A single potential confounder would have to be present in 30% of the GBT patients (and none of the non-GBT patients) and have an OR of 3.0 in order to find a statistically significant benefit to GBT”

23.  Explanations  Selection bias: physicians refer GBT for sickest patients  GBT might harm some patients ▪ ADE’s, resistance, CDI, complications of IV or prolonged hospitalization

24.  Observational nature  Worked solely with claims; could miss important confounders  No microbiologic data  Modified the ATS/IDSA guidelines  Used 1 vs. 2 drugs for pseudomonas  (results were same regardless)

25.  To date, no RCTs of GBT for HCAP  Findings question the necessity of treating all HCAP patients with GBT  Better models needed to identify at-risk patients for MDR pathogens  2010: Only 40% of patients with HCAP receive GBT

26.  Ewig S et al. Curr Opin Infect Dis 2012;25:166- 175.  HCAP poorly predictive of MDR pathogens  Frequency of MDR pathogens far lower than supposed in the original guideline document  HCAP concept results in tremendous overtreatment without any evidence for improved outcomes

27.  IDSA is currently revising HAP, VAP, and HCAP guidelines  1 MRSA antibiotic and 2 Pseudomonas antibiotics seem to be “too much” given the available evidence for routine treatment of a LTCF patient hospitalized with pneumonia

28.  93 y.o. F from LTCF with pneumonia. Treatment?  A. Levofloxacin  B. Levofloxacin + vancomycin  C. Levofloxacin + Pip/Tazo + vancomycin  D. Ceftriaxone  E. Ceftriaxone + azithromycin