Taking the leap to make bundled payments work

While the concept of value-based healthcare is gaining traction around the world, much remains to be done. Healthcare providers still have few incentives to change their practices to align with these expectations of higher quality at lower costs. To establish the right sets of incentives for value-based healthcare, both Japan and the United States have been gradually moving toward more bundled payments for medical care. The diagnosis-related group (DRG) hospital payment that allows hospitals to accept a predetermined lump sum for individual diagnostic categories was introduced in the United States in the early 1980s; Japan introduced a payment program called the diagnostic procedure combination (DPC) in 2003.

While these provider payment systems represent a step in the right direction, to succeed, “payment levels must be carefully calibrated to ensure providers’ financial viability while providing incentives to reduce costs and safeguards to ensure high quality,” according to industry experts in a recent blog.

To that end, we examined data from 514 hospitals in Japan that performed total knee replacement (TKR) surgeries with the goal of designing a bundled payment method that would meet criteria proposed by industry experts. To begin, we analyzed data aggregated at the hospital level from acute care facilities in the Global Health Consulting, Tokyo, Japan (GHC Japan) hospital discharge database. Then we identified 549 hospitals in the GHC Japan hospital discharge database that performed at least one TKR between April 1 and Dec. 31, 2012. All hospitals where data were collected for less than six months were excluded from our analysis. In addition, all cases of double-joint replacements were excluded, giving us a final sample size of 514 hospitals and 11,289 TKR cases.

Improving results, value

The United States and Japan could help their health systems achieve higher value by changing present payment systems for hospitalized patients. Japan needs to move from a DPC graduated per diem payment, which might incent long lengths of stay (LOS). In fact, a number of reports show that Japan has the longest LOS for hospitalized patients, including the longest LOS observed for total joint procedures internationally,1 and there is significant variability in practice patterns. For example, Japanese hospitals with low volume (fewer than 40 to 45 cases per year) have a 38% longer LOS for TKR and 39% longer LOS for total hip replacement (THR) than high-volume hospitals in the same sample.2 These LOS trends and practice variation can also lead to higher rates of patient complications, according to published reports.3

Likewise, the United States needs to move from the DRG payment system, a bundled payment for hospital services that lacks an incentive to coordinate care between various physicians and between physicians and hospitals.4

Next steps

Based on these results, we propose that as the first step in the bundled payment journey, the United States and Japan move to one bundled payment that includes all the costs of hospital and physician services, from the time of admission to a set date (e.g., 30, 60 or 90 days) postadmission.5

We believe this method, called an extended DRG (EDRG), is the best way to start bundled payments because the primary objective is to encourage better integration of services among different physicians providing care to patients, and an EDRG leads to better integration between hospitals and physicians. In Sweden, EDRG implementation has been shown to improve quality while reducing costs.

- DPC: diagnostic procedure combination
- DRG: diagnosis-related group
- EDRG: extended diagnosis-related group
- LOS: length of stay
- THR: total hip replacement
- TKR: total knee replacement

The package should also include incentives for medical centers to achieve cost efficiency and better patient outcomes. This reality-based approach to setting payment amounts is based on the way in which medical centers with good patient outcomes have achieved cost efficiencies.

Implementation

The DPC system in Japan does not capture all of the condition-specific costs incurred in hospitals or any costs incurred after discharge, and others have recommended a move toward a flat bundled payment for all hospital-related costs.6 Given those caveats, we will use Japanese DPC TKR cost data to illustrate how to:

• Set an EDRG payment to provide financial incentives for lower resource use and costs
• Build in financial incentives to improve patient outcomes

The EDRG basic payment amount should be set at the 75th percentile of the high-value quadrant of medical centers (hospitals that get better than average patient outcomes and have lower than average cost per case). Figure 1 shows the distribution of Japanese hospitals by patient outcomes and cost per case. The high-value medical center quadrant is outlined in this graph. Figure 2 shows that the 75th percentile of the high-value quadrant would put the Japanese TKR EDRG payment amount at $18,678 (1,886,628 J yen: 1 yen = $0.0084 U.S.)7 from the DPC data in this example.

Setting the EDRG payment amount at the 75th percentile of hospitals in the high-value quadrant encourages hospital efficiency and allows hospitals in the lower left quadrant (good efficiency/poor outcomes) to do well financially. We suggest that payers establish a quality withhold of 5% to provide an incentive for efficiency and better outcomes. If a medical center’s outcomes are not as good as the average hospital, it would receive 5% less than the target amount, or $17,744 (1,792,297 J yen) in our example. Thus, all hospitals in the two upper quadrants would receive a single bundled payment of $18,678 (1,886,628 J yen) while hospitals in the lower quadrants would receive $17,744 (1,792,297 J yen). The 5% withhold may be reversed if the hospital improves its outcomes in subsequent years.

Combining both of these points would establish financial incentives for better patient outcomes and lower costs. By establishing a true bundled payment with a quality withhold as outlined above, payers (the government in Japan’s case) would be setting this expectation: They are willing to set a payment amount that pays medical centers at the level going to hospitals that provide the highest value care. However, they are not willing to pay extra for care that is not efficient and does not provide good patient outcomes.

Evolving toward high-value healthcare

Moving to an EDRG payment scheme cannot happen overnight and should follow these steps:

• Announce that the new payment will be introduced in two years to give medical centers time to organize and evaluate results on patient outcomes and costs per patient.
• In the first year of implementation, use the new bundled payment approach for the five diagnostic categories with the highest total cost.
• In the first of the two-year lead-in, ask providers delivering care for each diagnostic category to develop patient outcomes that can be measured. For example, orthopedic surgeons and rehabilitation specialists would develop TKR outcome measures, cardiologists and cardiac surgeons would develop cardiac bypass outcomes, and so on.
• Compare medical centers with similar organizations for cost and quality. For example, compare teaching hospitals with teaching hospitals. The actual payment amounts to a medical center could also be adjusted to account for underlying costs of doing business in different geographic areas.
• Repeat this process with the next 10 most expensive diagnostic categories the next year.

While most healthcare professionals say they want higher value healthcare, most wish for a better result while continuing to do things as they have in the past. If we want high-value healthcare, we are more likely to get it if we actually start paying for it.

Notes

1. Waddell J, Johnson K, Hein W, Raabe J, FitzGerald G, Turibio F. “Orthopedic practice in total hip arthroplasty and total knee arthroplasty: Results from the Global Orthopaedic Registry (GLORY).” Am J Orthop. 2010; 39(9 suppl):5-13.
2. Mitsuyasu S, Hagihara A, Horiguchi H, Nobutomo K. “Relationship between total arthroplasty case volume and patient outcome in an acute care payment system in Japan.” J Arthroplasty. 2006; 21(5):656-663.
3. Dulworth S, Pyenson B. “Healthcare-associated infections and length of hospital stay in the Medicare population.” Am J Med Qual. 2004; 19(3):121-127.
4. Emanuel, E J. “Saving by the bundle.” The New York Times. Nov. 16, 2011.
5. Luft, H S. Total cure: The antidote to the health care crisis. Cambridge: Harvard University Press, 2008.
6. Anderson G, Ikegami N. How can Japan’s DPC inpatient hospital payment system be strengthened? Lessons from the U.S. Medicare prospective system. Center for Strategic & International Studies, Health and Global Policy Institute, October 2011.
7. Revenue data were converted from Japanese yen to U.S. dollars assuming a ratio of 1 yen to $0.0099 U.S., the exchange rate on Jan. 4, 2014.