Health Insurance Is Associated With Preventive Care but Not Personal Health Behaviors

Anthony Jerant; Kevin Fiscella; Daniel J. Tancredi; Peter Franks

doi:10.3122/jabfm.2013.06.130054

Article Figures & Data

Tables

View popup

Table 1. Baseline Sociodemographics, Healthcare Utilization, Health Characteristics, Preventive Care, and Health Behaviors by Study Insurance Status Group

	No Insurance Change^* (n = 89,616; 94.2%)	Gained Insurance in Follow-up Year (n = 3212; 2.9%)	Lost Insurance in Follow-up Year (n = 3193; 2.9%)	P Value^{^†}
Mean age, years	46.2	36.5	34.8	<.001
Female sex	51.9	50.1	49.6	.03
Race/ethnicity				<.001
Hispanic (any race)	11.1	14.9	13
Non-Hispanic white	69.8	56.5	61.2
Non-Hispanic black	12.4	21.8	19.1
Non-Hispanic other	6.7	6.8	6.6
Education level				<.001
Less than high school	6.4	7.5	5.9
Some high school	12	17.8	19.6
High school graduate	32	37	37
Some college	23	22.3	24.1
College graduate	26.6	15.5	13.4
Household income (% of FPL)				<.001
<100	10.3	20	20
100–124	4	6.8	7.3
125–199	12.9	22.7	19.9
200–399	31.1	31.3	32.3
>400	41.7	19.2	20.6
Employed	72.5	81.3	83.9	<.001
U.S. Census region				<.001
Northeast	19.2	16.1	14.9
Midwest	22.3	20.5	19.9
South	35.9	39.6	39.6
West	22.6	23.9	25.5
Urban residence^{^‡}	82.9	83.3	79.8	.01
Insurance status				<.001
Private	73.5	—^{^§}	76.9
Public	14.6	—^{^§}	23.1
None	11.9	—^{^§}	—
Utilization and health characteristics
Mean healthcare expenditures, $	3961.2	1283.1	2327.0	<.001
Office visits (mean)	3.9	1.7	2.5	<.001
Prescription drugs (mean)	13.2	5.7	6.6	<.001
Report having a USOC	78.4	52.1	63.9	<.001
Health status
Physical (PCS-12) (mean)	49.5	51.1	51.2	<.001
Mental (MCS-12) (mean)	50.9	49.1	48.9	<.001
Preventive care
Influenza vaccine	31.2	13.3	13.0	<.001
CRC screening	39.8	17.7	28.6	<.001
Pap testing	81.1	77.0	82.7	<.001
Mammography	62.4	40.5	66.0	<.001
PSA testing	51.7	25.9	28.8	<.001
Health behaviors
Always using seatbelts	81.7	77.1	76.9	<.001
Not smoking	79.1	68.6	65.4	<.001
Not obese	74.4	74.0	74.5	.64

Data are percentages unless otherwise indicated. All percentages are population-weighted.
↵* Includes both continuously insured and continuously uninsured participants.
↵† χ² Test for categorical variables, linear regression for continuous variables.
↵‡ Defined as residence in a metropolitan statistical area.
↵§ Of patients, 69.7% gained private insurance and 30.3% gained public insurance in the follow-up year.
CRC, colorectal cancer screening; FPL, Federal Poverty Level; MCS-12, 12-item Short Form mental component summary score; Pap, Papanicolaou; PCS-12, 12-item Short Form physical component summary score; PSA, prostate-specific antigen; USOC, usual source of healthcare.

View popup

Table 2. Odds of and Marginal Changes in Preventive Care and Health Behaviors Associated With Changes in Health Insurance Status^*

	Patients (n)^{^†}	Base Unadjusted Models		Adjusted Models^{^‡}
	Patients (n)^{^†}	OR (95% CI)	Marginal Change^{^§} (95% CI)	Adjusted OR (95% CI)	Adjusted Marginal Change^{^§} (95% CI)
Preventive care
Influenza vaccine	28,686	1.28 (1.08–1.52)	6.1 (1.9–10.2)	1.27 (1.04–1.56)	3.6 (0.5–6.7)
CRC screening	12,636	1.51 (0.99–2.30)	9.9 (0.6–19.2)	1.48 (0.96–2.29)	2.5 (0.0–4.9)
Pap testing	11,436	1.72 (1.37–2.15)	12.7 (8.1–17.4)	1.56 (1.22–2.00)	8.5 (3.7–13.3)
Mammography	8,586	1.66 (1.21–2.29)	12.0 (5.3–18.8)	1.70 (1.21–2.38)	10.4 (3.6–17.1)
PSA testing	9,206	1.42 (1.01–2.00)	8.5 (0.6–16.4)	1.42 (0.98–2.05)	3.4 (−0.1 to 6.9)
Health behaviors
Always using seatbelts	22,112	0.91 (0.78–1.07)	−2.3 (−6.3 to 1.7)	0.96 (0.81–1.14)	−0.6 (−3.4 to 2.1)
Not smoking	9,550	1.04 (0.82–1.31)	0.9 (−4.9 to 6.8)	0.96 (0.75–1.24)	−0.5 (−3.6 to 2.7)
Not obese	17,992	1.00 (0.81–1.22)	−0.1 (−5.2 to 5.0)	0.94 (0.74–1.19)	−1.5 (−7.0 to 4.1)

↵* The reference group in all analyses is “no insurance.” Hence, the odds ratios (ORs) presented for each outcome pertain to an individual gaining insurance form baseline to follow-up, and the reciprocal of the ORs pertain to an individual losing insurance. The marginal changes presented pertain to an individual gaining insurance, whereas the negatives of the marginal changes pertain to an individual losing insurance.
↵† Sample size of people who reported a change in the status of the preventive care or health behavior from their baseline year to their follow-up year.
↵‡ Adjusted for income, employment, total healthcare expenditures, number of office visits to healthcare providers, number of drug prescriptions, availability of a usual source of care, and the 12-item Short Form physical and mental component summary scores, Census region, urbanicity, and Medical Expenditure Panel Survey calendar year.
↵§ Marginal change is the percentage of increase (if positive) or decrease (if negative) in the preventive care or health behavior associated with changing insurance status (ie, gaining or losing insurance).
CI, confidence interval; CRC, colorectal cancer; Pap, Papanicolaou; PSA, prostate specific antigen.