Cost-Effectiveness Analysis of Heart Rate Characteristics Monitoring to Improve Survival for Very Low Birth Weight Infants

William King MS1, Waldemar A. Carlo MD2, T. Michael O’Shea MD3, Robert Schelonka MD4

1Medical Predictive Science Corporation, Charlottesville, VA, wking@heroscore.com
2University of Alabama at Birmingham, Birmingham, AL
3University of North Carolina, Chapel Hill, NC
4Oregon Health Sciences University, Portland, OR

Background:

In a randomized controlled trial (RCT) of 3,003 VLBW patients, heart rate characteristics (HRC) display was associated with a 22% reduction in all-cause mortality. (NCT00307333)

Objective:

Use data on resource utilization and outcomes from the RCT to analyze the cost-effectiveness of HRC monitoring.

Methods:

  • Secondary analysis of cost-effectiveness from birth to NICU discharge, up to 120 days.
  • Estimated each patient’s cost from a third-party perspective in 2021 USD.
  • Applied specific per diem rates for ventilated or non-ventilated status as follows,
    ⋄ Day 1: $4,241 or $1,442
    ⋄ Days 2 to 28: $3,562 or $1,164
    ⋄ Days 29+: $3,577 or $1,164
  • Computed the incremental cost-effectiveness ratio and then used nonparametric boot-strapping with 5000 replications to evaluate uncertainty.
  • Additionally evaluated cost-effectiveness from a hospital perspective.

Results:

  • From a third-party perspective, the incremental cost-effectiveness ratio of HRC-monitoring was $34,720 per additional life saved.
  • 95% of boot-strap replications fell below $449,291 per additional life saved (Fig 1a and 1c).
  • In 70% of bootstrap replications, cost per survivor was lower in the HRC-display arm than the non-display arm.
  • From a hospital perspective, implementing HRC-monitoring would result in net profit of $1,338 per VLBW patient.

Table 1. Demographics, Mortality, and Resource Utilization.

Figure 1: Cost-effectiveness Acceptability Curve

HeRO Cost-effectiveness Acceptability Curve

Cost-effectiveness Acceptability Curve

Figure 1: a- the Cost-effectiveness Acceptability Curve is derived by varying the Willingness-to-pay threshold from zero to infinity as the fraction of bootstrapping replications on the cost-effectiveness plane that fall below or to the right is calculated (b-d). b- Willingness-to-pay set to the value that bisects the boot-strapped replications into 50% cost-effective and 50% not cost-effective: $36,986 (blue dotted line). c- Willingness-to-pay set to the value for which 95% of replications are cost-effective: $449,291 (green dashed line). d- Willingness-to-pay set to $10,000,000 (red dot-dashed line), yielding 98% of replications as cost-effective.

Conclusions:

  • HRC-monitoring appears cost-effective for increasing survival among VLBW infants.
  • Universal HRC-monitoring of NICU patients is likely to improve both patient survival and hospital profit.