TABLE 5.

Sensitivity of Cost-Effectiveness Ratios to Key Assumptions

ParameterEstimated Cost-Effectiveness Ratio
Lower BoundUpper Bound
1. Base case*$19 800$94 800
2. Do not control for observed SES variables (Medicaid status, race, maternal marital status, and maternal age) in the logistic regression$19 000$91 200
3. LOS specified as ln (LOS)$17 300$83 200
4. LOS specified as l/LOS$24 200$116 300
5. Life expectancy of 66 y$20 600$98 900
6. Life expectancy of 86 y$19 200$92 000
7. Control for clinical variables (infection, severe respiratory problems, trauma, other severe outcomes, preterm delivery) in logistic regression$16 900$81 100
8. Assume increase in LOS is half what would occur if all newborns and mothers used up the full amount of time allotted to them under federal law$19 600$90 200
9. Include an indicator variable for method of delivery in the logistic regression$11 500$55 000
10. Include all deaths that occurred within 90 d of birth in the dependent variable$7100$34 000
11. Include all deaths that occurred within 1 y of birth in the dependent variable$6200$29 900
12. Discount rate of 0% annually$8000$38 300
13. Discount rate of 5% annually$29 600$142 200
  • * LOS untransformed; life expectancy of 76 y after time of infant death; 3% annual discount rate; ignore postneonatal deaths.