Editor’s Spotlight/Take 5: Does Hypothetical Centralization of Revision THA and TKA Exacerbate Existing Geographic or Demographic Disparities in Access to Care by Increased Patient Travel Distances or Times? A Large-database Study

In the study highlighted in this month’s Editor’s Spotlight/Take 5, Ramirez and colleagues [12] carried out a hypothetical evaluation of travel times for patients undergoing revision joint surgery who were brought from their local hospitals to busier, more-centralized facilities. In essence, they’re asking whether a hub-and-spoke model can be used on their patients.

The hub-and-spoke model comes originally from transportation. In 1955, Delta Air Lines began a system of flying passengers from outlying locales (the spokes) to a central airport (the hub), where they connected to other flights [3]. Other companies, like FedEx, followed suit, ultimately succeeding by taking every item of freight from each pickup point to a central clearinghouse, where the entire operation would be managed [5]. By any measure, it’s an efficient way to do business.

In healthcare, a hub-and-spoke system with centralized care can also have substantial benefits. There’s ample evidence that patients with complex indications or complications will benefit from specialized care. In orthopaedics, for example, it’s accepted that patients with bone or soft tissue tumors should receive their care in a tertiary care facility [4]. There are clear advantages in efficiency of care—less duplication across sites, easier availability of surgical instruments, and better access to other specialists—meaning that care may cost less due to economies of scale. And it allows critical access or local hospitals to focus on common problems (the “bread and butter”) and focus on serving their communities.

But there are inherent problems with this model. If one or more of the spokes becomes busier than normal, adjusting supply to meet demand is difficult. And if the hub is overwhelmed (bad weather in Atlanta for Delta, an outbreak of COVID-19 for a tertiary hospital), delays are inevitable. Missing a connecting flight can be annoying; losing access to care can be devastating.

And patients aren’t packages. Being away from family, friends, and home is hard enough for patients being seen at their local hospitals—it’s far worse if they are separated by hundreds of miles. Travel is expensive—not just for the index procedure, but for preoperative and postoperative appointments. If travel times are longer, that means more missed work days, a scramble to find care for children or elders, and finding a place to stay for family. As Schwartz et al. [15] note: “Taxpayers currently fund emergency medical services and transport for transplantable organs, but paying for similar transportation for elective surgical patients would represent a huge additional cost. High-volume centers … could contribute, but would be unlikely to do so voluntarily.”

Before embracing any change in policy, it’s important to think about its consequences. That’s why I was eager to speak with Benjamin F. Ricciardi MD, senior author of “Does Hypothetical Centralization of Revision THA and TKA Exacerbate Existing Geographic or Demographic Disparities in Access to Care by Increased Patient Travel Distances or Times? A Large-database Study,” [12] to get his thoughts on how our patients are impacted when they are taken to hospitals that perform high volumes of revision arthroplasty.

Take 5 Interview with Benjamin F. Ricciardi MD, senior author of “Does Hypothetical Centralization of Revision THA and TKA Exacerbate Existing Geographic or Demographic Disparities in Access to Care by Increased Patient Travel Distances or Times? A Large-database Study”

Paul A. Manner MD, FRCSC: You’ve looked at one state here. New York may be unique in some ways. It’s densely populated and has a plethora of high-volume orthopaedic hospitals. But that’s not going to be the case for other states. For example, Texas has more inhabitants, but they’re spread over a much bigger area, so the population density is less than a tenth that of New York, and it has fewer residency programs. What happens in New York might have to stay in New York. Can we apply your findings to other areas of the country, or other countries?

Benjamin F. Ricciardi MD: I agree with you that different states or countries may have different solutions to take advantage of the hospital and/or clinician expertise that comes with higher surgical volumes. We used New York State as a model because (1) the SPARCs database is an all-payer state administrative database that provides each patient a unique identifier so they can be tracked across admissions within the state and (2) New York has a wide range of hospital types (small, critical access hospitals [CAHs] to some of the largest tertiary joint replacement hospitals in the United States) and geographic environments (rural to urban). It would be interesting to see how our results differ across states with a variety of geographic and demographic profiles. In our prior work [13], we found that complication rates were similar across hospitals after revision total joint arthroplasty (TJA) in the top 75% by volume hospitals, which supports other work in primary TJA, suggesting that expertise at the hospital level translates into equivalent outcomes at relatively low surgical volumes. Allowing more hospitals to participate in revision arthroplasty by keeping volume thresholds as low as possible to reduce access to care issues in more rural states with a less dense hospital network would be beneficial. Studies utilizing other state databases or national databases that include robust state-level data would definitely help show how our analysis might vary across states.

Fig. 1.

Benjamin F. Ricciardi MD

At the national level, many European countries, Australia, and New Zealand have comprehensive national registries that would be amenable to this type of analysis. One study from the National Joint Registry of England, Wales, Northern Ireland, the Isle of Man, and Guernsey found that redistributing revision TKA from low-volume sites performing less than 20 per year to other regional sites would allow all centers to perform greater than 30 per year and would only increase the mean annual case volume per center by 11 cases [9]. This preliminary study suggests that at the national level in other countries, improving revision TJA care networks is being investigated and may be feasible without overburdening existing hospitals. The effect on patient travel and access to care would need to be further evaluated in these models.

Dr. Manner: You’ve cited several studies [7, 8, 18] (including your own recent excellent work [13]) showing an association between higher hospital and surgeon volume and lower rates of adverse outcomes after revision TJA. But these studies all show potentially important differences in patient characteristics between groups. In your own study, which compared the top 5% of hospitals by volume versus others, the top hospitals had a higher proportion of White patients, elective rather than urgent surgical procedures, patients admitted from home or office, and patients who were privately insured. Similarly, Jeschke’s group [7] found that low-volume hip revision hospitals were taking care of patients who were substantially sicker overall. How do we take this into account when parsing the numbers? More importantly, what does this mean for patient care?

Dr. Ricciardi: You make an important point regarding the centralization of revision TJA care. We (and others) have tried to control for the underlying patient demographics and admission characteristics within the statistical models, but there may be underlying unmeasured confounders that make these low-volume hospitals look “worse” when in reality, the acuity and challenge of their patient population is not being fully captured by the underlying dataset [7, 8, 13, 18]. More detail about the individual procedures, such as the complexity of the procedures performed, microbiological culture data in the setting of infection, and severity of the individual patient comorbidities would be needed to understand how complicated this patient population is relative to higher volume centers. My impression is that the highest volume revision hospitals don’t always take care of the sickest or most complicated orthopaedic conditions based on the studies you have cited above. Unfortunately, these may be the patients who would benefit most from the expertise of a high-volume center but don’t end up receiving care at these institutions for many reasons (insurance, lack of awareness, or travel barriers are just a few of them).

I see two potential solutions: (1) programs that facilitate or incentivize these hospitals to transfer patients with complex medical or surgical needs to centers with greater expertise or (2) improve the access of these low-volume hospitals to surgeons and medical physicians with greater expertise in managing these patients. The first solution is what we discussed in our study. The second solution could involve telemedicine consultations with physicians who have greater expertise than those in the lower volume hospitals to better inform decision-making regarding these more complex orthopaedic conditions. There is a precedent for this in complex elective surgery, such as Bernese periacetabular osteotomy (PAO); an excellent paper in CORR showed that low complication rates of PAO could be achieved by low-volume pelvic surgeons in a geographically remote setting with the use of a structured mentorship program involving distance mentorship [2]. This may be a model to consider in revision TJA. Alternatively, hospital networks may have programs where surgeons or physicians within the network travel to these lower volume hospitals at certain intervals of time to improve patient access to care while reducing travel burden.

Dr. Manner: Using the Centers of Excellence model for complex and relatively rare procedures seems like a reasonable approach—surgery for congenital heart disease comes to mind here. But knee revisions, unfortunately, aren’t that rare even if they are complex and resource-intensive. A recent study revisited Kurtz et al’s [10] highly cited projections of arthroplasty demand and estimated that those numbers were too high [16]. But we’re still looking at over 200,000 revisions per year by 2030. Would regional centers be able to handle that demand? How else might we manage these patients if regional centers can’t?

Dr. Ricciardi: I agree that revision hip and knee replacements are not as complex as something like congenital heart surgery, so the centralization models can likely be more inclusive. While the highest 5% of hospitals by volume in our previous study had the lowest readmission rates, it may not be practical (or necessary) to send all revisions to these hospitals [15]. For instance, we previously found that complication rates were not different after revision TJA in the top 75% of hospitals by volume, suggesting that revision expertise may be achieved in a broad range of hospitals at a relatively low volume [13]. Centralization of revision TJA care is already occurring naturally to some extent, and 90% of the revisions in New York State were performed in the top 50% by volume hospitals already [12]. In states with a similar structure to New York, this would likely not overburden existing centers given the broad range of facilities available among the higher volume centers.

Analyses in other states would be beneficial to confirm whether this has broad applicability. If regional centers can’t handle the demand, then providing access to the expertise of higher volume centers in a lower volume setting may achieve equivalent outcomes. For instance, discussing complex orthopaedic conditions with experienced surgeons and infectious disease physicians via technology such as telemedicine may provide the lower volume surgeon with enough support to improve their revision TJA outcomes. Like I mentioned above, distance-structured mentorship for PAO in a low-volume, rural practice resulted in low complication rates, and access to telemedicine has reduced the barriers for these programs.

Dr. Manner: Recently, CORR featured a study [11] evaluating how critical access (meaning rural) hospitals performed in treating patients with hip fractures compared to non-CAH hospitals. These patients were less likely to experience many serious complications, and the differences were even greater after controlling for patient demographics, hospital-level factors, and procedural characteristics. Can we extrapolate from this experience to revision arthroplasty?

Dr. Ricciardi: This is an excellent point and would be a great future study. No studies that I am aware of have specifically looked at CAHs and revision arthroplasty. In primary TJA, safety-net hospitals in recent years have been shown to perform worse in alternative-payment models; specifically, they had higher complication rates and they were less likely to meet targets set by the program [17]. Unfortunately, in primary TJA at least, safety-net hospitals tend to take care of more patients with complex orthopaedic conditions from both a medical and socioeconomic point of view, and studies probably don’t fully capture the complexities of these populations in safety-net hospitals or CAHs. This may also be the case for revision TJA given our findings (and those of others) [7, 8, 12, 13, 18], and future studies are warranted.

The hip fracture population may actually be the opposite to the arthroplasty population. The patients with the most medical complexity likely are transferred to regional centers given that CAHs may not have access to intensive care and cardiopulmonary management capabilities. By propensity matching the patient populations, the findings of Malik et al. [11] suggest that for a given underlying patient population, the CAHs seem to achieve comparable outcomes; however, they may have been taking care of a less sick population overall compared to regional centers. These are all interesting points and worth further investigation.

Dr. Manner: Several academic centers have been cited for running concurrent ORs, where attending surgeons are billing for simultaneous procedures while trainees perform much of the surgical work [1, 6, 14]. Leaving aside the legal or ethical issues, how would patients coming to high-volume centers know that their procedures will actually be performed by the big-name surgeons they came to see? Would it be safer to stick with the local surgeon, who might be better able to focus on the task at hand? Or is this outweighed by the increased resources and institutional experience of a larger regional center?

Dr. Ricciardi: I agree with you that outcomes of care don’t just vary at the hospital level, but at least in primary TJA, they vary at the surgeon level as well. This also is likely true for revision TJA. There is constantly a balancing act allowing surgeons to perform the surgical volumes to meet their clinical and reimbursement expectations while not discouraging them from taking on complex procedures as they gain expertise. Reimbursement for revision TJA has been a controversial subject for a long time, and given the complexity of these operations, current reimbursement discourages surgeons from performing complex revision TJA, at least compared to more straightforward primary TJA. It will be important to fix these reimbursement imbalances.

More specialized ancillary services and institutional experience with revision TJA may allow surgeons performing revision TJA in higher volume settings to address complex surgical or patient issues that may arise in these challenging cases relative to lower volume settings. For example, plastic surgery experience in the case of soft tissue deficiency or infectious disease physician experience for bone and joint infection may influence outcomes. Greater access to experienced surgeons at a regional center by their lower volume hospitals within their care network may also improve their surgical judgement despite their lower volumes. I do think good results are achievable in revision TJA in most cases at lower volume centers with an experienced physician, and innovative ways to provide access to more specialized services may expand the capabilities of lower volume revision centers even further.