Journal of NeuroEngineering and Rehabilitation

$5268 more per person per year compared to NMPK users. In the K1 and K2 population, MPK is associated with a reduction of $4237 per person per year in direct cost and $928 in indirect cost. The total cost associated with MPK is $2022 higher per person per year compared to NMPK. In the best scenario, the total cost of MPK is $5671 less, and in the worst scenario, $6074 more ex- pensive than NMPK. Combining economic and clinical benefits When the base case input values were used, for a 10-year time period, MPK resulted in an increase of 0.91 QALY per person and an increase of $10,604 in total cost per person, as illustrated by the orange-red dot in Fig. 4 . The corresponding base case ICER is $11,606 per QALY. The blue dots in Fig. 4 were generated from the probabilistic sensitivity analysis. In summary, MPK de- vices are more effective in all of the simulated scenarios, but also more costly in 83% of the simulated scenarios. The probabilistic sensitivity analysis results in ICERs ranging from -$25,355 to $36,357 per QALY. In the K1 and K2 population, MPK has an ICER of $13,568 per QALY. MPK may dominate NMPK as sug- gested in the probabilistic analysis, with an ICER of -$28,302 per QALY, meaning that it incurs lower total cost while leads to higher health status than does NMPK. The highest ICER is $41,498 per QALY in the probabilistic sensitivity analysis. Discussion This study is the first of its kind in the prosthetics litera- ture in the U.S. that integrates both clinical and eco- nomic data to assess the cost-effectiveness of advanced prosthetics for transfemoral amputees, specifically MPK. The results suggest that MPK is associated with substan- tial clinical benefits and cost-effectiveness compared to NMPK. It has been consistently demonstrated in the published literature that MPK leads to clinical improve- ments, such as improved walking speed, gait symmetry, and obstacle assessment. These clinical improvements, in turn, are associated with sizeable reductions in injuri- ous falls and incident osteoarthritis and as a result, lower mortality rates. The substantial clinical benefits of MPK can be largely attributed to reductions of falls with injuries and osteo- arthritis incidences. This is plausible from a clinical per- spective. For example, the computer software in MPK allows for the knee to dynamically adjust to uneven ter- rain, leading to improved stability and user confidence. The increased stability is thought to reduce cognitive burden and energy expenditure [ 41 ] . Increased stability, improved user confidence, and reduced cognitive burden and energy expenditures could all contribute to reduc- tions in the risk of falls [ 15 , 42 ] . The reduction in the knee moment associated with MPK likely eases the burden on the intact limb and therefore, reduces the probability of developing osteo- arthritis in the intact knee. The knee moment is a surrogate for the force an individual absorbs when striking the ground during walking. Because of MPK ’ s ability to adjust dynamically to uneven ground, MPK users are exposed to weaker forces when striking the ground, or smaller knee moment, than NMPK users. Due to the need for compensation, such forces are absorbed by the lower limb joints of the intact limb, and increase the burden on the healthy knee, hip and ankle, which is the expected mechanism through which osteoarthritis develops in the healthy limb [ 43 ] . When examining the life years gained, on average, an MPK user lives 0.09 year longer than an NMPK user over a 10-year time period; but when life years are ad- justed by quality, an MPK user gains 0.91 QALYs over an NMPK user. The difference between the number of life years gained and the number of QALYs gained is at- tributed to the substantial improvement in the quality of life, ranging from 10% to 37%. This finding is consistent with prior evidence that MPK is associated with im- proved mobility, safety, user confidence, activities of daily living, the ability of living independently, and satis- faction [ 44 – 48 ] , and thus substantially the better quality of life for amputees. While there are some cases where a medical innovation leads to net cost savings, a majority of med- ical innovations would result in a positive ICER, where a new technology leads to better health but costs more than conventional technologies. The commonly accepted thresholds of fiscal costs for value varies from $50,000 to $150,000 [ 49 – 51 ] . In other words, if a new technology Fig. 4 Incremental Cost and Effectiveness of MPK in Comparison to NMPK in K3/K4 Amputees, Note: MPK: microprocessor-controlled knees; NMPK: non-microprocessor controlled knees. QALY: quality adjusted life year. All costs are in 2016 U.S. dollars Chen et al. Journal of NeuroEngineering and Rehabilitation 2018, 15 (Suppl 1):62 Page 55 of 72