Using U.S. Medicare records to evaluate the indirect health effects on spouses: a case study in Alzheimer’s disease patients

Table 4 AD spouses vs. matched controls: Medicare expenditures, 2001-2005 ¹

	AD spouses	Matched controls	P-value
Pooled annual cohorts (N = 6,291)
Observation time
Median Medicare FFS months/year	12	12
Mean Medicare FFS months/yr. (95% CI)	11.83 (11.82, 11.84)	11.85 (11.83, 11.86)	NS
Medicare expenditures
Median total expenditures/year	$2,197	$1,658
Mean total expenditures/yr. (95% CI)	$8,206 ($7,967, $8,445)	$6,640 ($6,414, $6,866)	<.001
Mean total PMPM	$694	$561
Incident cohorts (N = 2,987)
Multiple regression analysis ²
Impact on Medicare expenditures (95% CI), AD spouse vs. control	29% (18%, 42%)		<.001

¹Spouses and controls were matched by gender, age category, race/ethnicity and urban/rural county status. The cohorts were restricted to spouses who, with their partners, were Medicare fee-for-service beneficiaries for at least 6 months of the follow-up year. The profiles for each year’s AD spouse and control populations were merged with the profiles for the corresponding cohorts for the other years in the observation period. The annual profiles were calendar-year specific and not linked to an AD diagnosis date.
²Result of a log-transformed regression model of Medicare expenditures that adjusted for socioeconomic status, US region of residence, hospitalization during the index month, JFI level during the index month, select pre-index chronic conditions, and length of follow-up. This analysis was restricted to couples in which both the AD patient and spouse were community-dwelling during the observed index month. Follow-up does not include the index month (all patients were required to survive the index month in this analysis).
FFS = fee-for-service; SD = standard deviation; JFI = JEN Frailty Index; PMPM = per Medicare beneficiary per month; CI = confidence interval.

ISSN: 1472-6963