Historically, safety and efficacy have been the most important considerations in evaluating new therapies. With rapidly accelerating healthcare expenditures, increasing attention has been paid to the costs of new procedures. Payers consider cost/benefit ratios in addition to safety and efficacy when making coverage and payment decisions. Those decisions exert increasing control over how rapidly and widely new procedures are adopted. The article by van den Biggelaar on cost-effectiveness of deep anterior lamellar keratoplasty (DALK) versus penetrating keratoplasty (PK) in this issue of the Journal addresses this interplay between clinical and economic factors.
Zirm performed the first successful penetrating keratoplasty in 1905. PK eventually became the procedure of choice for corneal blindness and has been predominant for the past half-century. Nevertheless, PK has its problems. Suprachoroidal hemorrhage is rare but devastating. Suture-related problems are common and can lead to graft failure. Visual rehabilitation is delayed by ocular surface problems and variable refractive error. Long-term astigmatism and ametropia leave some patients unable to attain useful visual function without contact lenses or further surgical intervention. Globe rupture after minor trauma is a permanent risk. Graft failure rates of 14% to 36% at 5 to 10 years and regraft failure rates of 50% or greater can be seen, with immunologic rejection in 15% to 20% of cases contributing, along with nonimmunologic endothelial cell loss, to late graft attrition.
Over the past decade, lamellar keratoplasty has begun to supplant PK. Endothelial keratoplasty (EK), by eliminating many of the ocular surface and wound integrity problems associated with PK and by offering faster visual rehabilitation, has become the procedure of choice for patients with endothelial disease. Anterior lamellar keratoplasty (ALK) techniques have evolved to the point that they now challenge the supremacy of PK for patients with stromal disease.
Advantages of DALK include retention of recipient endothelium, eliminating endothelial rejection, less steroid exposure and attendant risks of glaucoma and cataract formation, less distortion of angle anatomy and synechia formation, and an expanded donor pool. DALK results in less long-term endothelial cell loss and higher long-term graft survival than PK. There may be a marginal increase in wound integrity with retention of Descemet membrane. Visual acuity after DALK rivals and may surpass that after PK.
Despite its advantages, DALK has been slower to catch on than EK. From 2005 to 2009 the number of ALKs done in the United States has remained flat, at 1% to 2% of all corneal grafts and 4% to 6% of grafts done for keratoconus, while EK has grown to 43% of all grafts and 70% of those done for endothelial failure (Eye Bank Association of America. 2009 Eye Banking Statistical Report, 2010). The United States also lags behind other countries in embracing DALK. Over 30% of keratoplasties at the Singapore National Eye Center are ALKs. Complications inhibiting acceptance of DALK, such as Descemet perforations (6%–57%) and interface opacities, have decreased with refinements in technique. There is still a steep learning curve, however, and DALK does not offer any advantage over PK with respect to astigmatism. DALK remains technically more demanding and requires 54% more surgeon time than PK. Yet in the United States, surgeons are paid less for DALK than for PK.
While recent trends in keratoplasty have been driven primarily by clinical results, economic factors are looming larger. Total healthcare spending in the United States is growing rapidly, now consuming an estimated 16% of gross domestic product. Over the past 5 years, Medicare expenditures have increased from 336 to 509 billion dollars. The Patient Protection and Affordable Care Act portends seismic changes in payment policy to reverse increases in spending. The number of keratoplasties performed in the United States (42 606 in 2009) may be too small to attract the attention of regulators, who are most interested in high-volume, high-cost procedures. However, there have been instances of payers bundling tissue fees with facility fees, and keratoplasty will be subject to the same downward pressures on reimbursement that are affecting all of medicine. In this environment, economic analyses have the potential to become critical drivers of healthcare funding decisions.
The cost-effectiveness comparison by van den Biggelaar in this month’s Journal shows an incremental cost-effectiveness ratio for DALK of $7245 per patient compared to PK. Although this study demonstrates the greater cost-effectiveness of DALK over one year of follow-up, it has several limitations. Direct comparisons between costs in the United States and the Netherlands, where the study was performed, are difficult. The authors assume that direct nonmedical and indirect medical costs are the same between the groups, but the DALK group has more inpatient days and follow-up visits than the PK group. A single year of follow-up cannot capture costs associated with late complications and graft failures, which are expected to be greater in the PK group, resulting in an underestimation of the economic advantage of DALK. Effectiveness in this study is defined by a 10-point increase in the National Eye Institute Visual Function Questionnaire (VFQ-25) and by endothelial cell loss limited to 20% at one year, a proxy for decreased risk of late graft failure. As the authors note, it is difficult to know what society is willing to pay for these improvements. Cost-effectiveness analyses like this are specific to ophthalmology and are difficult to compare between different medical specialties competing for limited healthcare dollars.
Cost-utility analysis is another type of healthcare economic analysis. It measures dollars expended in return for a universal value gained, and is popular because it allows comparisons between specialties. Utility analysis measures quality of life on a scale of 0.0 (death) to 1.0 (perfect health). Visual acuity correlates to utility and has been validated to be stable across disease states, gender, race, time of visual loss, and age. Validity of the utility values could be strengthened by expanding the data set to include more patients with corneal disease and to account for peripheral field loss with intact central acuity. Nevertheless, improvement in visual acuity can be directly translated into a corresponding improvement in utility. Multiplying the improvement in utility by the expected duration of the benefit in years yields the number of quality-adjusted life years (QALYs) gained.
Cost-utility analyses compare dollars spent per QALY gained. Interventions costing $100 000 or less per QALY are considered “cost-effective,” while those costing $50 000 or less per QALY are considered “very cost-effective,” regardless of specialty. While these numbers may seem arbitrary, they are critical. A new technique that results in a better outcome may be more expensive than established therapy. However, if that new technique falls within the “very cost-effective” range based on cost-utility analysis, it is considered an important and valid therapy worthy of reimbursement. Cost-utility analyses of PK ($11 557/QALY) and keratoprosthesis ($16 140/QALY) indicate that they are very cost-effective procedures.
Over the course of one year, DALK and PK produce similar mean visual acuities, generating similar utility values. Costs, however, differ. The largest short-term cost differences in van den Biggelaar’s report were for the operating room and surgeon: 41% and 54% greater, respectively, for DALK, although in the United States surgeons are paid less for DALK than for PK. Differences in long-term graft survival affect duration of benefit and costs for repeat procedures to manage complications. Late graft failures occur at a greater rate after PK, resulting in increasing utility and cost advantages for DALK over time. In recent studies, DALK and EK have both demonstrated greater cost utility than PK, with incremental cost-utility ratios of $3025/QALY for DALK (Mehta JB, et al. Cornea Society/EBAA Fall Educational Symposium 2010; Abstract 13) and $5291/QALY for EK. Comparisons like this will be of growing importance with the introduction of promising but costly procedures such as femtosecond laser–assisted keratoplasty. Expensive new treatments are sure to face increasing economic scrutiny.