Table 4. Pooled crude and adjusted PIVE.

Complete Case and Imputed Data Analysis	Crude and Adjusted PIVE Estimates	Included Population	n	Percent PIVE	95% CI
Complete case analysis a	Crudeb	All	1,502	79.0	55.8–90.0
		<65 y	1,367	83.3	61.2–92.8
		15–64 y	912	76.6	44.7–90.1
		<15 y	455	100	58.2–100.0c
		No chronic disease	1,190	81.5	53.0–92.7
	Adjusted modeld	All	1502	66.0	23.9–84.8
		<65 y	1,367	71.3	29.1–88.4
		15–64 y	912	65.5	12.3–86.5
		<15 y	455	100	Not calculablee
		No chronic disease	1,190	70.2	19.4–89.0
Imputed data f	Crudeb	All	2,902	82.8	68.6–90.6
		<65 y	2,688	86.9	73.9–93.4
		15–64 y	1,463	80.6	57.2–91.2
		<15 y	1,218	94.2	75.6–98.6
		No chronic disease	2,354	84.6	67.7–92.7
	Adjusted modeld	All	2,902	71.9	45.5–85.5
		<65 y	2,688	78.4	54.3–89.8
		15–64 y	1,463	73.3	36.7–88.7
		<15y	1,218	84.8	31.0–96.6
		No chronic disease	2,354	72.9	39.7–87.8

^aExcluding individuals with missing values.

^bStudy site included in the model as fixed effect.

^cExact logistic regression estimates with zero cases vaccinated.

^dModel adjusted for 2009–2010 seasonal influenza vaccination, any influenza vaccination in previous two seasons, presence of at least one chronic disease, sex, at least one hospitalization for chronic disease in the previous 12 mo, current smoker, age group, practitioner visits in previous 12 mo (0, 1–4, and 5+ visits), month of symptom onset (note: in the 15–64 y stratum no adjustment for age group; in the “no chronic disease” stratum no adjustment for chronic disease or hospitalizations for chronic disease).

^eIf one of the cases would have been vaccinated, the estimated PIVE would be 85.2% (95% CI 30.0–98.3).

^fMissing data imputed using imputation using chained equations.