Reliability of a patient survey assessing cost-related changes in health care use among high deductible health plan enrollees

Table 2 Comparison of delayed or foregone care between federal poverty limit < 300 and federal poverty limit >300 in the original and follow-up surveys

	Original (n = 387)†			Follow-up (n = 387)			Pvalues ‡
Type of Care	FPL < 300%	FPL ≥ 300%	Pvalue* Original Survey	FPL < 300%	FPL ≥ 300%	Pvalue* Follow-up Survey	FPL < 300 Between Original and Follow-up	FPL ≥ 300 Between Original and Follow-up
Any Delayed or Foregone Care
Pediatric care	24.0	13.2	0.014	24.0	10.8	0.002	1.000	0.439
Adult care	48.8	33.6	0.007	51.2	32.8	0.001	0.785	0.877
Family care	56.2	41.6	0.011	58.7	38.8	0.001	0.799	0.621
Types of delayed or foregone care
ED visit	61.8	52.9	0.322	69.0	45.4	0.004	0.463	0.269
Preventive care	27.9	18.3	0.191	25.4	18.6	0.384	0.869	0.869
Imaging test	27.9	23.1	0.589	31.0	25.8	0.569	0.639	0.886
Prescription medication	17.7	15.4	0.856	12.7	5.2	0.144	0.513	0.016
Specialist visit	19.1	12.5	0.334	18.3	13.4	0.514	1.000	1.000
Screening test	17.7	11.5	0.365	18.3	10.3	0.207	0.841	0.670
Laboratory test	14.7	8.7	0.323	19.7	10.3	0.134	0.414	0.819
Operation or procedure	19.1	5.8	0.013	18.3	11.3	0.293	1.000	0.225
Physical therapy	11.8	12.5	1.000	14.1	11.3	0.768	0.637	0.683
Outpatient visit for routine check-up	10.3	6.7	0.582	9.9	8.25	0.93	1.000	0.796

The p-values in columns 3 and 6 represent the observed type-1 error (α) for tests comparing income groups whereas the p-values in columns 7 and 8 are the observed α for tests comparing the original and follow-up surveys.
*P values are for chi-square tests between income groups within survey
† We restricted the original survey data to follow-up respondents
‡ P values are for McNemar tests within income groups and between surveys
Bolded items signify comparisons that are statistically significant at α = 0.05

ISSN: 1472-6963