BALTIMORE — Influenza vaccination is associated with lower rates of hospitalization and outpatient visits for flulike illness, but only after statistical adjustment for the fact that people who get vaccinated are less healthy than are those who don't, Dr. Roger P. Baxter said at a vaccine research conference that was sponsored by the National Foundation for Infectious Diseases.
Raw data from Kaiser Permanente's medical records of approximately 500,000 adults aged 65 and older during each of four consecutive influenza seasons showed that hospitalization and outpatient visit rates were actually higher for those who received the influenza vaccine than for those who didn't, primarily because the population is sicker.
“People who get the flu vaccine are less healthy and utilize the medical system much more than those who don't get the vaccine. … It's a real confounder in all these studies,” said Dr. Baxter, associate director of the Vaccine Study Center at Kaiser Permanente, Oakland, Calif.
Another study confounder is the fact that the vaccine works better in some years than in others, a phenomenon believed to be caused at least in part by how good a “match” there is between the vaccine and the season's circulating strains.
Unadjusted, the rates of hospitalization per 1,000 person-years during the influenza season ranged from 169/1,000 vaccinated vs. 159/1,000 unvaccinated in 2003–2004 (risk ratio 1.06 in favor of no vaccination), to 182/1,000 vs. 150/1,000 in 2002–2003 (1.22 in favor of no vaccination).
Even after adjustment for age, gender, and three underlying diagnoses (diabetes, coronary artery disease, and heart failure), all of the values for the vaccine's effectiveness were negative, giving the counterintuitive impression that the vaccine actually causes disease. The values were −8.4% for preventing all hospital stays, −25.5% for all pneumonia and influenza outpatient visits, −21.4% for other respiratory outpatient visits, and −23.3% for all respiratory visits, all significant values.
But a second analysis that focused on health care utilization outside of the flu season suggests that “people who get the vaccine are high utilizers” year-round, Dr. Baxter said.
He and his associates compared the data during each year's influenza season with the off-season control period of June 1-Aug. 31. This level of analysis revealed that the difference in utilization between vaccinated and unvaccinated patients was even greater during the off-season than during influenza season, with a risk ratio of about 1.25 for each of the four seasons.
Dr. Baxter and his associates are further analyzing the Kaiser data to examine vaccine effectiveness in younger, healthier people, as well as adding mortality to the outcomes and looking more closely at the relationship between “good” vs. “bad” vaccine matches in the face of more or less virulent circulating influenza strains. Conducting such studies is “not easy,” he noted.
The Kaiser Permanente Vaccine Study Center receives research grants from the influenza vaccine manufacturers Sanofi-Pasteur, GlaxoSmithKline, Novartis, and MedImmune.