The surgical robot comes with a hefty price tag. It is currently on the market for around $2 million, on par with the most expensive capital items that any OR would ever purchase. And this isn’t the only cost to consider. Also important are the annual maintenance contract for the robot and the need to train a team to use it, which requires extensive work hours that are not reimbursed by customers or insurance companies. During the start-up phase, the robotics team is inexperienced so the hospital encumbers additional costs on robotic cases due to longer OR times and higher risk of complications. This period, called the “learning curve”, can add to per case costs until the steep phase of the learning curve is completed at around 100-200 cases of experience. Added all together, it is a reasonable estimate that the total start-up cost of a robotics program exceeds $3 million.
The decision to take on “large ticket” items like a robot is based in part on an analysis of return on investment (ROI). In the case of robotics, the question is whether the added value from using the robot justifies its $3 million up-front costs. In theory, “added value” means that quality has been improved at equal or lower costs. In the business world, a new program’s ability to produce profits is a reasonable proxy for added value. In healthcare, the profit motive has been more opaque and almost any increase in costs have been tolerated as long as there is a measureable improvement in patient outcome. Without prior clinical experience, hospital administrators interested in ROI find it difficult to discern the added value of a robot. The financial value soon becomes clear if new patients interested in shortening their recovery time seek out less invasive options from lucrative services like cardiothoracic surgery at the hospital that otherwise wouldn’t have come. Declines in cardiac surgical volume at most programs that don’t offer robotics leads to empty hospital beds that were staffed to be occupied. This adds significant avoidable costs to a hospital (http://www.ncbi.nlm.nih.gov/pubmed/3759473?dopt=Abstract), a problem which is circumvented by program growth. There is also value when less invasive surgery allows earlier discharge from the hospital. This frees up beds for other patients that would not otherwise be available, thus improving the efficiency of hospital operations. There is value to the hospital’s marketing and HR departments when the new procedure creates excitement that aides in the retention and recruitment of patients, staff and physicians. These benefits depend on the institution; the case for robotics may be less compelling at prestigious programs with a full OR schedule and routine practice of early hospital discharge.
Yet healthcare dollars are not unlimited. Critics argue that US hospitals rushed to purchase surgical robots without understanding ROI or at least articulating the value proposition. The robot is often cited as a symbol of waste in our healthcare system, a testament to the success of critics. But it is important to remember hospitals are in business to please their customers. Like a hotel, the hospital invests in new buildings, comfortable amenities, gourmet food, etc. that may or may not help accomplish this goal. Offering innovative procedures that improve patient outcomes is another effective way to meet the needs of customers but can be associated with even more expensive investments. For example, a novel procedure called transcatheter valve implantation (TAVI) reduces the invasiveness of valve replacement and leads to a dramatic improvement in clinical outcomes. Medicare regulations require this procedure to be performed in a hybrid OR suite, a facility that doesn’t exist at most hospitals and costs millions of dollars to build. The hybrid OR is analogous to fancy furniture and a grand piano in a hospital lobby. Hospitals acquire these amenities and account for the fact that their value of is depreciated with every use when they calculate their overhead costs. Concern about high overhead has been largely absent from the debate about controlling healthcare costs because these costs are not reimbursed by insurance or patients. Since the public doesn’t have to foot the bill, the decision to run a business with high overhead is left to the discretion of hospital boards. For a variety of reasons, the robot seems to be the sole exception to this laissez faire approach to capital purchases. Most economic studies in the literature allocate the purchase costs of the robot in their estimates of per case costs for surgery that utilizes the robot, adding $2000-$8000 per case and making robotic surgery look financially unattractive.
Perhaps the logic why its costs are accounted for differently is that a robot is a more intricate asset that depreciates faster than the equipment used for traditional surgery. Intuitively, it makes sense that any profits from this program would be balanced against the costs from the quick depreciation of the robot in order to have a complete picture of cost-effectiveness. However, this analytic approach confuses up-front, one time fixed costs, often referred to as “sunk costs”, with the variable costs generated on a per case basis. Once the robot and its maintenance contract are purchased, the team is trained and the learning curve is completed, the $3 million entry fee is never paid again. A financial analysis that includes start-up costs is answering a question about ROI, not cost effectiveness. There is no question that if the start-up costs for robotics are excessive, then ROI and other metrics used to evaluate investments will be poor and hospitals won’t buy robots. Clinicians who aren’t responsible for decisions to purchase capital (e.g. heart surgeon, department chairman or division chief) usually aren’t focused on ROI. Financial concerns from clinicians usually question whether the robotic or traditional open approach is more cost effective after the robot has already been purchased. It is human nature to let excessive sunk costs distort downstream but unrelated financial decisions, a fallacy that goes by the well-known phrase “throwing good money after bad”. The argument against using one of the 2500 robots already sitting in US hospitals based on concerns that start-up costs for robotic program were too high is a demonstration of the sunk cost fallacy.
The information needed to define cost effectiveness is an accurate assessment of costs measured on a per case basis. Remarkably, hospitals are rarely able to determine this critical information, largely because they are not required to measure costs this way. Heart surgery is reimbursed by insurance at a flat fee that varies little between cases. Information on aggregate costs – all costs encumbered in the heart surgery division over a given timeframe divided by all the cases performed– are sufficient to negotiate for a fee that retains a profit for the hospital. The aggregate approach means that the hospital can’t determine when one case has higher costs than any other. The hospital knows what it is being reimbursed. If all procedures are performed the same way, an occasional high cost outlier is compensated for by the other cases with less costs that make a profit. The problem arises when changes are made in how the procedures are performed. Aggregate accounting means that profitability cannot be determined right away, but has to wait until the next quarterly or annual cost report is available. Therefore, a change in practice that increases cost is revealed too late because “behavior that can’t be measured, can’t be improved”.
To understand how this system evolved, consider a comparison of the heart surgeon and the plumber. Heart surgeons are sophisticated plumbers who clear pipes in the heart using different amounts of labor and supplies according to the needs of the customer. However, a plumber is paid a different amount by each customer on the basis of the varying costs of parts and labor for each service job. On the other hand, cardiac surgery is reimbursed by insurance companies, not the customer. Neither the hospital nor insurance company wants the hassle of having to account for “parts and labor” on each case. So they agree to simplify the billing process by negotiating a flat fee for heart surgery.
Hospitals do keep track of charges on a per case basis using a process known as the “chargemaster”, but this ends up being a cruel hoax that only adds to the confusion. It is a poorly kept secret that the list of charges from the chargemaster is grossly inflated (e.g. $200 for a dose of Tylenol). It also provides no real information on actual costs of care because prices that are charged vary widely between hospitals for no obvious reason. In addition, only charges that are potentially reimbursable are included in the chargemaster; all other costs that can’t be put on a patient’s bill are excluded. Virtually all the research literature analyzing the costs of robotic surgery have defined costs from the chargemaster data based on accepted cost to charge ratios. As a cardiac surgeon that has used the robot for 8 years, I have never had a hospital administrator suggest costs of my cases that use the robot are too high based on chargemaster data. The people that manage the chargemaster realize these data are just too unreliable. Imagine how the plumber’s customer would feel if presented with a bill for service that was incomplete in terms of what was done while costs of those things that were listed were grossly exaggerated.
The way around the cost accounting problems is a classic economic tool called “opportunity costs”. This analysis focuses on the incremental costs and benefits of one approach relative to the other. Traditional, open chest bypass surgery itself is a well-established example of an expensive therapy applied to the right patient (e.g. a patient with blockage of their left main coronary artery) yielding a powerful reduction in overall long-term healthcare costs relative to a cheaper but less effective alternative (e.g. non-surgical, medical management). Opportunity costs analyses like this have been a cornerstone of the evidence base in the fields of both medicine and business. In addition to quantifying all the value added by robotics, an opportunity cost analysis can reveal the problems of not having detailed data about costs on a per case basis. Behavior that is not measured can’t be improved.
Surgeons are scientists that appreciate solid data. The persistent lack of this type of precision in cost data has been a key source of tensions and stalemates between administrators and cardiac surgeon over cost-cutting initiatives. Without the framework of opportunity costs, new technologies like robotics are susceptible to biases of the sunk cost fallacy and the criticism that any given piece of expensive equipment is “a bridge too far”. If we want to maximize value in healthcare, a good place to start would be changing this backward financial thinking.