How the selected use of telemedicine in intensive care units affects economic efficiency?

Background: Telemedicine in the Intensive Care Unit (Tele-ICU) is expected to reduce health disparities in ICU care through more efficient allocation of resources, e.g., intensivists.  However, the spread of Tele-ICU appears to have slowed, partly due to the high start-up costs and limited evidence in economic efficiency.  Our preliminary economic evaluation estimated Tele-ICU to be cost-effective in most cases and cost-saving some cases, compared to conventional ICU care (published in “Critical Care Medicine”). Further analyses are needed to evaluate the impact of a selected use of Tele-ICU for a subset of ICU patients who have the highest mortality rates and ICU care costs (called “highest-risk subpopulation” hereafter).  This impact accounts for larger reductions in mortality and ICU care costs, offset by increased per-patient fixed cost of Tele-ICU use.  

Objectives: The primary aim is to examine if a selected use of Tele-ICU will improve economic efficiency, either lowering the incremental cost effectiveness ratio (ICER under a cost-effectiveness analysis) or maximizing the cost-saving amount (under a cost-benefit analysis). The secondary aim is to conduct a break-even analysis to explore conditions that make Tele-ICU cost-saving. 

Methods: We developed standard decision models, making a set of assumptions on the impact of Tele-ICU on mortality and per-patient costs during hospitalization, derived from the literature.  The intervention was the selected use of Tele-ICU, which was assumed to affect per-patient ICU cost and hospital mortality among a hypothetical highest-risk subpopulation (10% to 90% of all ICU patients) defined by an ICU scoring system, i.e., Acute Physiology and Chronic Health Evaluation (APACHE).  Tele-ICU operation costs included equipment-installation (start-up) costs and operation cost.  Tele-ICU effectiveness was measured by cumulative quality adjusted life years (QALYs) for five years after ICU discharge, with a 3% annual discount rate.  Monte Carlo simulation was performed to address model uncertainties.  All costs were adjusted to 2014 US dollars using the consumer price index.

Results: As hypothesized, the primary analysis found a U-shaped association between the economic efficiency (e.g., measured by ICER) and the proportion of the highest-risk subpopulation that used Tele-ICU (ranged from 10% to 100%). The optimal point (Mean ICER =$21,500 per QALY) was achieved when Tele-ICU was applied to the 30% highest-risk patients of the total ICU patients. The ICER estimates for 100% (included as a reference group), 90% and 10% highest-risk subpopulation who used Tele-ICU were $50,300, $43,200 and $40,000 per QALY, respectively.  The conditions and the magnitude of the optimal efficiency depends on the distribution of mortality rates and ICU care costs among all ICU patients. 

The secondary break-even analysis found that Tele-ICU will be cost-saving if (a) Per-patient Tele-ICU equipment-installation (start-up) cost was less than $165, or (b) Per-patient per-hospital-stay ICU cost (excluding Tele-ICU related cost) was lowered by at least 8.9%.  Cost-saving was not achieved even if the Tele-ICU operation cost was zero.

Conclusions: The results suggest that the economic efficiency of Tele-ICU can be improved by a selected use for the highest-risk subpopulation.  These findings and estimated benchmarks will help promote the more efficient use of Tele-ICU.