Healthcare economics is the branch of economics focused on the efficient allocation of available resources to maximize health benefits to a society. Broadly speaking, the field encompasses the performance of healthcare systems in their entirety as determined by the efficiency, efficacy, value, and behavior in the production and consumption of healthcare within that system. The core definitions of markets, including the economic models of supply and demand, apply but are complicated by several additional factors that make healthcare unique.¹ First, the healthcare sector is confounded by the presence of third parties such as insurers, governments, and medical technology companies, each with their own, often competing, interests. Second, there exists an asymmetry of information such that consumers (patients) usually do not know what they need and cannot evaluate the value of the treatment they receive. Finally, providers are not paid directly by consumers but rather by third parties. As such, market rules are established by third parties rather than market prices. For these reasons among others, resources are frequently misallocated resulting in the scenario in which all parties act in their own self-interest to the detriment of the whole, while healthcare markets become increasingly inefficient.

Healthcare spending within the United States continues to rise without any sign of cessation. According to the most recent data from the Organisation for Economic Co-operation and Development (OECD), the United States spent 17.1% of gross domestic product (GDP) on healthcare in 2017 compared with 4.2% to 12.3% of GDP for all remaining member countries.² The reasons for the continued increase in spending are many. Rather than being cost saving, recent advances in medical technology, including new devices and implants, have increased spending. With increasing life expectancy, the percentage of older patients within the population continues to increase and place higher demand on the healthcare system. Nine percent of the US population was greater than 65 years old in 1960 versus 15% in 2015 and that number continues to increase.¹ Additionally, healthcare and life itself are inelastic goods, such that demand is relatively static even when prices change. As society becomes richer, the marginal value of years of life declines slower than other goods, thus people spend a higher fraction of income on healthcare. Finally, recent analysis has determined that the rise in healthcare expenditures is mostly due to increased expense of procedures rather than increased utilization of services.³

In an effort to decrease the cost of US healthcare, there has been a recent shift from a volume-based fee-for-service (FFS) system to a system focused on value-based care. In a traditional FFS system, healthcare providers charge for each service or procedure delivered and patients and/or payors pay for each healthcare service upon receipt. This leads to conflicting interests for all involved parties and increasing amounts of system inefficiency. In a value-based system, the focus is on patient-centered outcomes, which is measured by quality rather than volume of services rendered.

Economist Michael Porter defined value in healthcare as health outcomes achieved per dollar spent.⁴ Value is centered on the patient, which should then determine the rewards for all other parties (payors, providers, suppliers). Outcomes are multidimensional and vary based on the condition treated. Costs encompass total costs for the full cycle of care for treatment of the condition, not just the cost of individual services. For example, in evaluation of shoulder arthroplasty for treatment of glenohumeral osteoarthritis, possible outcome measures include but are not limited to 30- and 90-day readmissions, 30- and 90-day surgical site infections, length of stay, and patient-reported outcomes such as pain and functional scores. Costs include not only the prosthesis and operating room (OR) time but also the hospitalization, postoperative care including nursing facilities, rehabilitation, and management of any postoperative
complications. In this model, benefits of any individual service ultimately depend on the effectiveness of all other interventions throughout the care cycle.

In the past, value in healthcare has been challenging to both measure and deliver. Commonly utilized metrics did not adequately portray value. For example, measures for a single department are too narrow to be pertinent to the individual patient, whereas measures for an entire hospital are too broad. Shifting focus from billing to patient outcomes tailored for a specific disease process and intervention simultaneously improves healthcare efficiency and coordination of care among providers. Thus, the ability to measure, report, and compare outcomes is the most important step to improve value.⁵

In other industries, decreasing cost to improve value can be achieved by decreasing outcomes within a tolerable range. This is not ethically acceptable in healthcare as there exists a standard of care that must be met. Value cannot be created by decreasing outcomes to save money. Alternative strategies to decrease healthcare costs include limiting access to care, stratifying procedures, and rationing care, although these are not ethically sound provided there is enough healthcare to distribute. Increasing access to care especially at an earlier stage can also potentially decrease costs by way of preventative care.

In an effort to control federal healthcare spending, the US Center for Medicare & Medicaid Services Innovation (CMMI) was established by the 2010 Affordable Care Act. CMMI’s stated purpose was to develop and test “innovative payment and service delivery models to reduce expenditures … while preserving and enhancing the quality of care” for recipients of Medicare, Medicaid, and Children’s Health Insurance Program. Specifically, they have supervised the creation and trialing of several Alternative Payment Models (APMs), forms of healthcare payment reform that incorporate value into reimbursements. This national effort to improve value by decreasing costs, improving outcomes, and increasing access was the catalyst that set in motion the shift toward value-based care in the United States.

Perhaps the most relevant APM for shoulder arthroplasty is the bundled payment model. A bundled payment is a combined payment for the physician, hospital, and all other healthcare provider services for a single episode of care. An episode of care is defined as all services provided to a patient in the treatment of a defined condition or as part of a procedure for an agreed-upon period. The payment is calculated based on expected costs of all items and services furnished to a beneficiary during an episode of care. The purpose of a bundled payment is to create incentive for providers to coordinate and deliver care more efficiently. In this form of agreement, the providers assume risk that their costs exceed their payment, but they also have the potential to profit when costs are below the fixed payment, thus providing incentive to increase the value of their care not only with the intervention provided but also throughout the entire episode of care. Gainsharing, or an incentivized financial agreement between a hospital or group and an individual physician, allows this reward to be realized by the individual practitioner so long as appropriate ethical safeguards exist.⁶

Joint replacement has been a special area of interest for bundled payments given the relatively homogeneous nature of the procedures, increasing volume, and the rising cost burden for Medicare and Medicaid programs. Thus, the CMMI has made continued efforts to implement bundled payments for arthroplasty (FIGURE 53.1). The Bundled Payments for Care Improvement (BPCI) initiative was initiated in 2013 in an attempt to improve fragmented care for multiple medical conditions with minimal coordination across providers and healthcare settings. Specifically, it was a voluntary program in which providers could opt in to one of four payment and delivery models with target prices derived solely from
an individual institution’s historical pricing. Based on expanding participation and initial success with BPCI, the Comprehensive Care for Joint Replacement (CJR) was initiated in 2016. In contrast to BPCI, this was a mandatory program, designed for hospitals, focused solely on joint replacement, requiring the participation of nearly 800 hospitals in 67 urban areas. Target prices were also derived from an institution’s historical pricing, but were integrated with regional historical pricing with a progressive shift toward prices solely derived from regional prices. The initiative also implemented a pay-for-performance system incorporating the use of quality scores. After 2 years, the program was modified and now only mandates participation in 34 of the original 67 metropolitan statistical areas.

The aforementioned bundled care initiatives have mobilized hospitals and physicians and set changes into motion which are still evolving today.⁷ Studies have demonstrated appropriate response by hospitals to these changes, including altering post-acute care (PAC) settings, decreasing hospitalization times, and decreasing utilization of skilled nursing facilities (SNFs) and inpatient rehabilitation facilities (IRFs).⁸,⁹ They describe overall reductions in Medicare episode payments versus comparison groups in areas of both historically high and low payments, in both elective and fracture settings, all without affecting quality of care. However, closer scrutiny of participating institutions shows differences that are not initially apparent.

Analysis of both models has shown that BPCI hospitals tend to be larger and more teaching intensive (vs safety-net hospital status as seen in many CJR hospitals), with a larger annual Medicare volume and higher cost attributable to institutional PAC. Both groups had a similar risk exposure and baseline episode quality and cost.¹⁰ These differences have implications on the generalizability of results seen between both models and begin to broach the idea that mandatory participation may have detrimental effects on access to care. A study at a high-volume, single-specialty orthopedic hospital in 2019 compared the economic implications of the institution’s participation with BPCI with CJR and showed a substantially reduced margin for savings for CJR versus BPCI, with a projected annual savings decrease of 83.3%.¹¹ As CJR drives down price targets for all hospitals in a given area, proposed savings are threatened as methods to cut costs are rapidly exhausted. This may ultimately jeopardize access to care for patients at low-volume hospitals who may find joint replacement unaffordable to perform given the rapidly decreasing reimbursements. It remains to be seen how policymakers will interpret and utilize the results of these two initiatives in the future.

Several other APMs have been proposed and implemented, although their effect on shoulder arthroplasty is less profound than that of bundled payments. An accountable care organization is a network of doctors, hospitals, and other healthcare providers that share responsibility for coordinating care and meeting quality and cost metrics for a defined patient population. A patient-centered medical home facilitates coordination of care through a patient’s primary care physician. Private payors were slow to adopt, but now several commercial models also exist that share frameworks with the aforementioned APMs but with less rigidly defined parameters. All APMs are united by the common goal of increasing quality of patient care while decreasing per capita costs—increasing value—with reimbursements tied to quality metrics and incentives to coordinate care.

As hospitals make changes to adapt to and survive in new systems, there may be unanticipated adverse consequences along the way. One such consequence is the stifling of new innovation. Maximizing value by focusing on outcomes such as 30- or 90-day outcomes may falsely shift treatment toward short-term solutions. For example, when comparing open reduction and internal fixation (ORIF) to reverse total shoulder arthroplasty (RTSA) for treatment of proximal humerus fractures, RTSA may be thought to provide inferior value due to increased short-term cost but may be of greater value in the long term. This is further complicated by a perverse financial incentive in which insurance companies are unlikely to be insuring the same patients for several years in the future; thus, they are willing to pay for a less expensive, more “valuable” solution in the short-term despite worse long-term outcomes. The reason being that the potential long-term savings will not be realized by the current insurance company, as the patient is unlikely to be insured by that company when the savings are manifested. The capacity to explore and research innovation in the form of new prostheses and new technologies such as robotics may be suppressed with similar reasoning.

As previously mentioned, another area of concern is an increase in healthcare disparities specifically with respect to access to care. With attention fixated on outcome and quality measures, there is a decreased incentive to operate on patients with unfavorable social determinants of health, such as those with complicated medical comorbidities or lower socioeconomic status. When attempting to maximize quality metrics, improve outcomes, and limit complications, patient selection and risk stratification unfortunately become a large focus. This brings an ethical dilemma into the conversation as care may be delayed for those who need it the most.¹²

These implications on access to care certainly extend to the growing field of shoulder arthroplasty. Waldrop et al showed that Medicaid and Medicare patients younger than 65 years who underwent primary total shoulder
arthroplasty (TSA) demonstrated poorer preoperative function and postoperative patient-reported outcomes than similar patients with private insurance despite similar functional improvements from baseline.¹³ In models incorporating quality measures such as CJR, these findings may influence surgeon and hospital decisions to proceed with surgery. Sheth et al demonstrated that more disadvantaged patients (lower socioeconomic status determined by Area Deprivation Index) had a higher body mass index, higher preoperative opioid use, higher rates of diabetes, more preoperative pain, and lower preoperative functional scores when undergoing TSA for glenohumeral osteoarthritis.¹⁴ Certainly, providers and payors both need to be cognizant of the potential impact of socioeconomic factors and preexisting conditions when determining fair bundle payments and reimbursements going forward.

A center of excellence (COE) is a program within a healthcare institution that is assembled to supply an exceptionally high concentration of expertise and related resources centered on a specific area of medicine, delivering associated care in a comprehensive, interdisciplinary fashion to afford the best patient outcome possible.¹⁵ As an increasing proportion of care becomes reimbursed by APMs, there is an incentive for healthcare systems to integrate both horizontally (providers of the same service) and vertically (providers of different points of a supply chain) in an attempt to maximize efficiency and value, forming highly specialized COEs. Formation of these centers requires coordination and a tremendous amount of resources, but ideally, they maximize value by increasing quality of care while minimizing costs with mutual benefits for both providers and patients. In a study of hospital volume and outcomes specifically for RTSA, an inverse relationship was demonstrated between hospital volume and all-cause 90-day readmission rates, cost, and overall resource utilization.¹⁶

A similar discussion can be had regarding high-volume providers. Extensive research has shown a volume-value relationship in hip and knee arthroplasty with improved outcomes associated with increased surgeon volume.¹⁷ Similar studies have also been conducted for shoulder arthroplasty, although the interpretation of their results has been conflicting.¹⁸,¹⁹,²⁰ Meta-analyses of these studies have shown that low volume (definition ranging from less than one to five procedures per year) was associated with increased length of stay, OR time, in-hospital complications, and cost, although not all of these differences were statistically significant.²¹,²² In addition, there are very little data on readmission rates, and perhaps most important, no patient-reported or functional outcomes were collected. Thus, although not conclusive, the literature appears to point toward increased value with increased surgeon volume in shoulder arthroplasty. The formation of COEs and concentration of high-volume providers, however, are not free from unintended consequences. As they form, a high level of care becomes concentrated at one geographic focus. This may decrease access to care for those with already limited access by increasing travel distance, transportation costs, time off from work, etc. Patients may be diverted from hospitals that traditionally serve patients of a lower socioeconomic background. If combined with mandatory participation in bundled payment models based on regional pricing such as CJR, these safety-net hospitals may be priced out from certain procedures and low-volume surgeons may be excluded from performing procedures altogether, even further decreasing access to care.

The number of TSAs performed in the United States continues to increase.²³,²⁴,²⁵ FDA approval of RTSA in 2004 along with expanding indications for both TSA and RTSA and continued development of dedicated shoulder and elbow fellowship training all play a role in the continued growth of TSA.²⁶,²⁷ Zmistowski et al used Medicare databases and hospital referral regions (HRRs) to study trends and variability in TSA use from 2012 to 2014. They showed that although overall access to TSA is expanding, including access to surgeons with high-volume TSA caseload (defined as more than 20 Medicare TSA cases per year), there continues to be notable geographic variations with socioeconomic underpinnings. Specifically, HRRs with a lower proportion of white patients and a higher percentage of poor patients (defined by eligibility for Medicaid) reported decreased use of TSA despite the presence of a sufficient Medicare population to support an arthroplasty surgeon. In addition, they found that more than 44% of Medicare patients at their institution traveled outside of their local HRR for their procedure.²⁸ These findings echo the previously stated concern of limiting access to care with the concentration of high-volume providers in COEs. As the field of TSA continues to grow, the changing face of the healthcare system will undoubtedly have implications.

Currently, the episode of care for shoulder arthroplasty is Medicare Severity-Diagnosis Related Groups (DRG) code 483 (major joint replacement or reattachment of upper extremity with or without major complications or comorbidities). This does not account for the type of procedure (anatomic TSA vs RTSA) or the indication for procedure (glenohumeral osteoarthritis, rotator cuff arthropathy, fracture, etc.). In addition, the Current Procedural Terminology code is the same for both TSA and RTSA (23,472). This method of coding grossly generalizes all shoulder arthroplasty procedures and indications without adjusting for increased risks
that may accompany different patient populations. For example, a retrospective database study has shown that patients who underwent TSA for proximal humerus fracture were significantly more likely to have a longer hospital stay, 30-day surgical and medical complications, need for revision surgery within 30 days, postoperative transfusions, nonhome discharge, and 30-day readmissions than patients undergoing TSA for osteoarthritis.²⁹ Thus, there is a clear difference in resource utilization when TSA is performed for different indications. In this example, the fracture group comprises a different patient population than the arthritis group, with physiologic and functional differences as well as need for a more time-sensitive procedure with less time for optimization. This is analogous to what transpired with hip arthroplasty bundles. Initially, hip arthroplasty for arthritis was treated the same as hip arthroplasty for fractures. After several studies demonstrated that costs and outcomes for these two procedures are different, all fracture patients were assigned to a higher intensity DRG.³⁰,³¹,³² With the shift toward bundled payment models, similar changes will have to occur for shoulder arthroplasty to appropriately code procedures and assign their true expected value.

Despite this, initial trials of bundled payments with shoulder arthroplasty have shown early success. Odum et al compared 132 FFS patients and 333 BPCI patients who underwent TSA in their practice and found that BPCI patients had lower rates of SNF admissions, IRF admissions, home health aide (HHA) utilization, and readmissions, with a 4% decrease in expenditures overall after controlling for postacute events.³³ Walters et al compared bundled and unbundled groups receiving TSA specifically in the outpatient setting and found significant cost savings in the bundled group versus the unbundled group both in total surgical day charges as well as total 90-day global period charges. Implant pricing, negotiated as part of the bundle, was the primary driver of cost reduction.³⁴ Outcomes were not included as part of this particular study.