Research Roundup: Improving Intelligence Forecasts, Vertically Integrated Health Care, and “Worrisome” Health Care Costsby Knowledge@Wharton -
How can intelligence agencies improve accountability and forecasting accuracy? Can hospitals become more efficient through vertical integration with home health agencies and nursing homes? Do taxpayers fully understand how the expansion of health care will be financed? Wharton professors Philip Tetlock and Barbara Mellers; Guy David and Evan Rawley; and Mark Pauly, respectively, examine these issues — and what they mean for business — in recent research articles.
Helping Intelligence Agencies — and Companies — Avoid the Blame Game
When business leaders fail to make accurate forecasts, profitability is at risk. When intelligence agencies miss the mark on their predictions, however, the results can be far worse. In a new analysis of behavior in the intelligence community, with implications for business managers, Wharton management professor Philip E. Tetlockand Wharton marketing professor Barbara A. Mellers present a framework to improve accountability and forecasting accuracy, particularly in a politically polarized climate.
In their article, “Intelligent Management of Intelligence Agencies: Beyond Accountability Ping-Pong,” published in the September 2011 edition of American Psychologist, the authors note that forecasts by intelligence organizations frequently are open to harsh criticism for either underreporting potential danger or overreacting to threats that never materialize. A clear recent example of underreporting would be the September 11, 2011, terrorist attacks on the United States, Tetlock says. At the other extreme, he points to reports — which later proved to be unfounded — that Iraq had developed weapons of mass destruction.
“The intelligence community is often whipsawed between these conflicting criticisms,” says Tetlock. “The question is: Is it possible in this kind of political environment to learn anything beyond avoiding the last mistake?” The authors propose three steps to end the “blame game” in intelligence predictions and improve accountability and intelligence forecasting.
First, the authors argue that intelligence agencies and constituents in government and throughout society need to come together and agree to put an end to bitter, often ideologically driven, assignment of blame. Tetlock suggests that “thoughtful moderates” with a long-term view of policy will need to drive this part of the process, especially during periods of deep division.
Next, intelligence agencies need to step up and agree to have their forecasting assessed on clear metrics. Tetlock says that meaningful forecasts could result from reports that put a hard number on predictions. For example, analysts could be required to put specific percentage odds on the likelihood that a coup, or uprising in a given country, would occur in a certain period of time. Agencies would amass large databases of predictions that could, over time, be reviewed to assess which were accurate and why.
Finally, in the authors’ view, intelligence groups and their overseers should acknowledge that ideology plays a part in forecasting. “If you want … the left and right to hold back their fire on unfair criticism, the best way to do that is to reassure people on the left and the right that their points of view are at least being used in the prediction process,” Tetlock notes.
The authors argue that the potential benefits of improved forecasting are so large that even a modest, incremental boost in accuracy could have major benefits. According to the article, “The intelligence community does not have to lower the probability of multibillion-dollar fiascos by much to recoup a multimillion-dollar investment.”
The same framework the authors outline could be useful in the business world, where deeply polarized camps can also arise. The divisions are not along liberal or conservative lines, Tetlock says, but there are often different views in the corporate suite about how a company should approach future strategy.
“There can be tension between those who prefer an expansionary strategy and people with a more cautious strategy,” Tetlock notes. With solid metrics, predictions could be assessed for relative accuracy. “When people think their accuracy is being monitored, they are a lot less extreme and a lot more reasonable,” Tetlock points out. “It depolarizes the discussion.”
Tetlock and Mellers acknowledge in the paper that their proposals may come across as “ridiculously naïve.” However, Tetlock notes that some large companies, including Hewlett Packard and Google, have developed sophisticated metrics to assess internal forecasting. “Market pressure in the private sector is driving companies toward becoming more explicit in measuring the accuracy of their expectations about the future,” Tetlock says. “I think the world is moving in this direction — haltingly.”
Vertical Integration in Health Care: When and Why It Helps Reduce Costs
To explain the results of his research into integration and task allocation, Wharton health care management professor Guy David refers back to a 2006 article he read in Modern Healthcare, a magazine covering health care business and policy news. The article’s authors argued that hospitals that are vertically integrated with post-acute outlets — i.e., home health agencies, nursing homes, rehab clinics and other facilities where patients go to recover after surgery or serious illness — end up mismanaging them.
The authors reached this conclusion, says David, because they observed that freestanding health agencies tend to make money while the home health agencies integrated with hospitals usually lose money. “The premise in the article was that if hospitals knew how to manage those entities, they would be profitable,” notes David. “I started wondering whether integrated home health agencies serve a different purpose than freestanding ones.” He and colleagues Evan Rawley, a Wharton management professor, and Daniel Polsky, a professor at Penn Medicine, explore this subject in a new research paper titled, “Integration and Task Allocation: Evidence from Patient Care,” part of the National Bureau of Economic Research working paper series.
Their paper concludes that, rather than being a detriment to efficient operations, integration can solve coordination problems between organizations. The researchers focus specifically on timing problems in the process of transitioning patients from acute hospital care to follow-up care. “We find that vertically integrated hospitals are able to shift expensive patient recovery tasks downstream to lower-cost delivery systems by discharging patients earlier and in poorer health,” write the authors.
Moreover, they note, patients in vertically integrated hospitals do not receive substandard care. On the contrary, says David, “health outcomes are no worse when patients receive care from an integrated provider, and in some cases, integration leads to better outcomes. The evidence suggests that by improving the efficiency of the timing of patient transitions, integration solves coordination problems that would otherwise arise under pure market exchange.” The paper’s findings, he adds, “have important implications for the ongoing health care reform debate, as it pertains to both clinical and financial integration of various segments of the health care continuum.” The findings are also relevant to other industries, such as the manufacturing and distribution of personal computers.
The context underlying the research by David, Rawley and Polsky is the high cost of health care, especially the expense of hospital stays. Both hospitals and home health agencies are paid prospectively — i.e., insurers reimburse them with a lump sum for each patient, regardless of how long the patient is in the facility. “This means that the hospital has an incentive to get rid of you as soon as possible,” David says, adding that both sides in this equation would like the other to cover patient recovery to the greatest extent possible. “Hospitals gain from early patient discharge, while home health agencies gain from admitting patients later in the process,’” the authors write.
But if the hospitals and home health agencies are integrated, “they can figure out the optimal timing to shift the patient from one setting to the other,” according to David. “If firms are not integrated, they will bargain over this outcome.” This situation underlies a provision in the new health care bill that would penalize hospitals for re-admissions, because under the current system, “if you come back to the hospital after already being discharged, the hospital gets an extra reimbursement. They want to get you out quickly, but they welcome you back as well.”
Costs for health care come down once a patient is released from a hospital into a home health care facility. So from a business perspective, the question David and his colleagues analyzed was whether a hospital could move sicker patients more quickly into integrated home health care, thereby saving on expensive hospital care, and then invest some of those savings in improving the home health care recovery experience. Under those circumstances, the patient would spend the same total amount of time in the hospital/home health continuum. But because more of that time would be in the less expensive facility, the costs overall would be lower.
Going back to the article he read that sparked his interest in this topic, David suggests that its argument — integration is a bad idea because hospitals are going outside of their core business and besides, they don’t know how to manage those agencies — was not valid. “The hospitals don’t want to run the agencies, but they do want to influence whom they can send there and when. Once you think in those terms, suddenly the light bulb goes on and you understand that what they are trying to achieve is lower costs. Freeing up beds on the bigger revenue generating side — e.g. the surgical wards — by moving patients out to the lower-cost home health care side means the hospitals can increase their profits, with no downside for the patients. “This story is still consistent with home health agencies losing money, but it makes a lot of sense from a business perspective, and has nothing to do with the ability of hospitals to manage those agencies,” says David.
In their paper, the researchers state that they “focus on timing problems in market exchange that arise from the misallocation of tasks between two vertically distinct stages of production.” That misallocation “has performance implications,” they note, pointing to variations in cost structures and reimbursement rates between hospitals and post-acute care providers, which “ensure that tasks will not be efficiently assigned unless the hospital and downstream providers are vertically integrated.” Approximately one-third of nursing homes and home health agencies are vertically integrated into hospitals.
The researchers were able to track patients across organizations and observe — with a significant level of detail — the clinical procedures that patients receive in post-acute care. “The evidence shows that, on average, vertical integration leads to shorter hospital stays for six out of every 10 patients who are discharged to a skilled nursing facility or home health agency. We also find that patients received higher intensity of care from vertically integrated home health providers,” the authors write.
Basically, the researchers’ theory was that if hospitals and home health agencies are integrated, they will figure out exactly when to transition a patient between the hospital and home health. But “if they are not integrated,” David states, “they will have to bargain over this outcome, and the result won’t be efficient.”
As a contrast, David notes how the situation is different with the hospice industry, which deals with patients who are terminally ill. Hospices are paid on a per diem basis because “no one wants to incentivize them to shorten a patient’s length-of-stay, which in this case is synonymous to shortening the patient’s longevity,” says David. He points to the two biggest costs associated with hospice: first, when a new patient arrives, and second, when the patient is in his or her last hours of life. “There is a big incentive for hospices to get patients into hospice sooner rather than later because that extends the period between those two big costs. The hospice makes money each day a patient is alive.”
The hospital still has an incentive to get a patient out as soon as possible, David adds. “But the hospice, unlike the home health agency, has an incentive to bring the patient in as soon as possible. The incentives are aligned; the negotiations are efficient, and integration, which has its own costs, is not needed to achieve an efficient solution. Not surprisingly, integration between hospitals and hospices is quite rare.”
Paying for Health Care Reform
Whether health care reform survives court challenges — or the next election cycle — is a debatable question. But its uncertain future hasn’t stopped Wharton health care management professor Mark Pauly from raising questions about how the landmark legislation will be financed. He outlines his concerns in an article in the soon-to-be published book, The Economists’ Voice 2.0.
The article, “How Stable are Insurance Subsidies in Health Reform?” suggests that taxpayers do not fully understand how the expansion of health care will be financed. When they do, Pauly contends, voters may reject the entire program, including parts of the law that are popular, such as allowing adult children to be covered by a parent’s policy up to age 26.
Pauly favors the controversial national health plan, which could extend coverage to more than 20 million uninsured Americans when it takes full effect in 2014, but he says its financing mechanisms are “potentially worrisome.”
About two-thirds of the costs of subsidies to cover premiums for low-income Americans are expected to come from savings in the Medicare program for older Americans, Pauly notes. However, he adds, Medicare is already inadequately financed, especially as baby boomers begin to tap into the program. To cover predicted shortfalls in Medicare funding, Congress has already proposed cuts in the program. Now, those same cuts have been shifted to cover the health care expansion. Unless new changes to Medicare are enacted, Pauly projects that the amount of taxes paid to support Medicare as a percentage of GDP will double or triple between 2035 and 2050.
Financing for health care reform is based on reductions in private Medicare Advantage plan payments to hospitals as well as reduced payments to doctors under the so-called sustainable growth rate provision, which the Obama administration itself has asked Congress to adjust in doctors’ favor. In the original version of the article published in the electronic journal The Economists’ Voice in December 2010, Pauly suggested that one way around the financing shortfall would be “a substantial additional amount of funding [to] be raised by imposing taxes on worker compensation paid in the form of benefits.”
“It is likely that there will need to be some sort of compromise at a lower level of generosity of benefits and subsidy,” the article states. “The time to start the conversation about what really is the highest priority and highest value is now.”
The health care reform legislation “was sold on the promise that nobody would pay for it,” according to Pauly. “Politicians did what you would expect them to do — emphasized the benefits of the program and deemphasized the taxes. But the problem is that the strategy of paying for health reform out of the savings in Medicare leaves Medicare in worse shape.”
Pauly says he is concerned that when true price of reform becomes apparent to voters and taxpayers, they may balk, forcing a costly unraveling of the parts of the program already in place. “I would have preferred a more straight-forward approach,” Pauly notes.
Citing prior research, Pauly points out that even those who have medical insurance benefit when others also have coverage. Aside from ethical — what Pauly calls “clean conscience” — arguments to support expanded access to health insurance, those with health coverage might have less-than-altruistic reasons to support the program. Pauly has done research showing that increasing coverage in a health care market area improves the quality of care for all of those in the same region. “I would have hoped those would be good reasons to bite the bullet and say we’re willing to pay more taxes,” Pauly says. “Now it’s going to come as a surprise.”
December 20, 2011 in Knowledge@Wharton