Skip to main content
. 2023 Jan 3;22:1. doi: 10.1186/s12939-022-01813-2

Table 2.

Marginal differences of included characteristics on medical expenditure during the last year of life

Characteristics Dy/dx 95% CI P value
Rurality (Ref: Urban)
 Rural -229 -378 -80 0.003
Primary Caregivers (Ref: Informal)
 Formal 278 -443 1,000 0.45
Place of death (Ref: Home)
 Hospital 798 518 1,077  < .001
 Nursing home facilities -440 -1,055 176 0.161
Age at death (Ref: 80–89)
 90–99 -180 -274 -86  < .001
 ≥ 100 -424 -546 -302  < .001
Per capita household income annually (Ref: < 391)
 391–942 4 -124 132 0.95
 942–3,060 206 -26 438 0.08
 > 3,060 726 452 999  < .001
Main financial source before dying
Retirement wage (Ref: No) 511 120 902 0.01
Timely medical services (Ref: Yes)
 No -335 -446 -223  < .001
 Was not ill -317 -451 -183  < .001
Health Condition
Disabilities in ADLs (Ref: No) 313 178 448  < .001
Number of Comorbidities (Ref: 0)
 1 -4 -127 120 0.95
 2 177 25 330 0.02
 ≥ 3 224 15 433 0.04
Region (Ref: Eastern)
 Central -264 -425 -103 0.001
 Western -282 -436 -128  < .001
 Northeast -154 -419 111 0.26

Note: Dy/dx Average marginal effect, CI Confidence interval, ADLs activities of daily living. Only results of predictors of statistically significance were presented. Results of associations between gender, ethnicity, marital status, years of schooling, number of children even born, family support as main financial source, once having white-collar jobs before retirement, living arrangement during the last year of life, self-rated health status, health behaviors and end-of-life medical expenditures were not presented. All the currency was presented in the US dollar