Pneumonia Nationwide in Veterans Health Administration (VHA) Medical Centers and Correlation with Immunizations: 1997-2009

Background: VHA is a large healthcare system in the U.S. which provided care to over 5.3 million persons with over 605,000 inpatient discharges in 2009.  In the mid 1990s, VHA Central Office adopted a paradigm for regional management wherein achievement of evidence-based clinical performance measures for the patient population under their jurisdiction contributes to 50% of the annual evaluation of administrators.  Two such measures present then (and currently) are receipt of pneumococcal and influenza vaccination. 

Objective: We undertook a review of VHA administrative data sources to determine if there has been change in pneumonia hospitalizations since instituting these measures.

Methods: All VHA inpatient hospitalizations between 1997 and 2009, nationwide, were reviewed for evidence of ICD-9-CM coded discharge diagnoses for lower respiratory infections/pneumonias (ICD-9-CM codes ranging from 480-489.99).  Immunization rates were obtained from VHA Office of Quality and Performance data for the same time period.  Statistical analysis included regression and Pearson correlation. 

Results: Overall pneumonia discharge rates per 100,000 have increased over time from 5432.2 to 6688.3 while death rates for those admitted with pneumonia have decreased from 16.0% to 8.4%.  The subset of pneumococcal pneumonia discharges have decreased from 405.9 to 195.7 while death rates have fluctuated between 11.6 and 4.6 during the time period, with a trend toward decrease over the last several years.  During this time, pneumococcal immunization rates have steadily increased from 60% to 95%.  These vaccination rates correlate inversely with the pneumococcal pneumonia rates (r=-0.81, p=0.0008); when correlation of immunization rates is compared to the subsequent year's pneumococcal pneumonia hospitalizations, there is still a strong inverse correlation (r=-0.87, p=0.0003).  Influenza pneumonia discharges and associated death rates varied during the time period with no discernible trends (26.2 to 105.8 per 100,000 discharges and 1.4% to 3.9%, respectively), though immunization rates have increased from 61% to 83%.

Conclusions: Though causation is not established, these data suggest that increasing pneumococcal vaccination coverage rates, which have occurred in conjunction with implementation of performance measures for vaccinations, are associated with decreasing pneumococcal pneumonia rates.  A similar association has not been noted for influenza pneumonia inpatient rates.