The Gehr Center/Slow Medicine series "Disruption or Distraction: A Critical Look at Healthcare Delivery Reform" unpacks the data behind healthcare delivery system interventions, highlighting well-intended reforms that have yet to achieve the impact many have assumed. The series is edited by Slow Medicine columnists , and . This second article in the series explains why targeted programs for high needs, high cost patient populations may not be achieving their intended goals. Links to all articles in the series are available on the and websites.
When Jeffrey Brenner, MD, took notice more than a decade ago of the spotty care afforded certain high-risk, high-need patients in Camden, N.J., neither he nor anyone else could have guessed that the rather modest, prevention-oriented treatment program he would soon begin would grow into the burgeoning, nationwide movement it has become today.
The paradigm -- referred to as "hot-spotting" -- has sprung up in dozens of places in the U.S. Brenner's defines healthcare hot-spotting as "the strategic use of data to reallocate resources to a small subset of high-needs, high-cost patients."
By providing follow-up consultation and care to such patients before they might otherwise seek further in-hospital care, "hot-spotting" aims not only to cut healthcare costs associated with repeated hospitalizations, but ultimately to improve the health of those most in need.
Brenner's program, the , is widely considered the prototype for the country's growing number of programs for high-need and high-cost populations. Its model utilizes "care teams" made up of health professionals with expertise in such disciplines as nursing, pharmacy, health coaching, counseling, social work, and community health work.
In 2011, a widely circulated by Atul Gawande reported that the first enrollees in the program experienced a 40% drop in hospital and ER visits, and their collective hospital bills dropped from $1.2 million to just over $500,000 per month -- a 56% reduction.
In response to Camden's initial purported success, healthcare systems around the nation scrambled to
But upon closer inspection, careful analysis of the Camden program and others like it raises concerns that these apparent jaw-dropping improvements may be overstated.
It turns out that we easily can be tricked by simple analyses of programs targeting high-risk populations because of a well-established – but counter-intuitive – statistical principle known as . The tenets of regression to the mean dictate that if a variable is extreme on its first measurement, it will, on its own, tend to be closer to the average on its second measurement.
As a result, many "super-users" of hospital facilities today are likely to stop being high-volume users in the future, regardless of whether they receive intensive targeted services – often referred to as "complex care management" – or not.
Without rigorous research to determine the exact impact of such programs, practitioners and patients alike cannot assume that complex care management is the primary reason for any improvement of such high-use patients, says Steven Asch, MD, professor of medicine at Stanford University and an expert in the field.
"As a researcher, I hold [complex care management] programs to very high standards. Very few of these programs have reached the level [of success] I would consider convincing," he said.
The ideal study design for evaluating hot-spotting approaches, says Asch, is a randomized control trial. Even when a randomized trial is not feasible, however, the rigor of a complex care management program evaluation can be enhanced by the inclusion of a matched contemporaneous control population. For example, a program could be piloted in one clinic and compared with a nearby similar control clinic.
Few complex care management programs have undertaken controlled evaluations, however.
One that did was the , a program in which 15 healthcare organizations from around the country were given the opportunity to develop programs for improving care coordination among the most complex populations. Most programs in the demonstration focused on patients with chronic conditions who had been hospitalized within the previous year and assigned them to a care coordinator (typically a registered nurse) who assessed them and developed a plan of care. Because most coordinators had caseloads of well under 100 patients, they were able to contact their assigned patients frequently – often several times a month or more. Nearly everyone involved anticipated success, and indeed enrolled patients experienced substantial improvements in health and reductions in hospitalization rates.
But when independent researchers conducted a pre-planned evaluation of the program in which they compared enrolled patients with similar controls, the data told a different story. According to , patients assigned to the control group improved just about as much as those receiving the intensive services. The study concluded that, relative to controls, "these programs had favorable effects on none of the adherence measures and only a few of many quality of care indicators examined," and did not favorably impact the costs of care.
Many of the other rigorous studies of "hot-spotter" approaches have come to similar disappointing conclusions. With this in mind, the dramatic results reported about programs like the Camden Coalition seem much less impressive. While emergency room visits, hospitalizations, and overall costs among enrolled patients may have dropped substantially, there is no telling whether patients would have experienced these same improvements even without the program.
To their credit, the Camden Coalition has acknowledged that the dramatic data initially reported about their program may be overblown. To rigorously assess their model, the Camden Coalition is now participating in a . While it is too early to report data or to draw conclusions from that study, these results should provide insights about the program's true impact.
Though enthusiasm for "hot-spotting" may be over-hyped, it is important to highlight that some programs have shown success, even in rigorous controlled evaluations. for its "highest risk" individuals with complex and chronic conditions -- those representing the highest 10-20% of projected future cost.
Boeing funded a specialized intensive outpatient clinic for these employees with a dedicated physician and nurse care manager, who oversaw the care of 150 to 200 high-risk patients. After an initial intake visit, the team contacts enrolled patients at a minimum of twice monthly.
When measured against a control group receiving care from outside providers, per-capita spending for enrollees in the pilot program Patient-reported missed worked days dropped by over 56% compared to baseline measures, patient self-reported function improved by approximately 15%, and access to care "as soon as needed" improved 18%.
Another example of a successful complex care management effort was a program for high-risk older adults in Pennsylvania led by an organization called Health Quality Partners as part of the same Medicare demonstration described above. Unlike other less successful programs in that demonstration, the . The coordinators also focused their efforts on hospital-to-home care transitions and medication management. The controlled analysis of this program, unlike most of the others in the demonstration, showed not only reductions in avoidable emergency room and hospital visits but also reduced mortality among those enrolled.
"In this field, people are truly trying to improve care," Asch says. "So you want to make sure [these programs] are implemented properly." Because so few controlled trials have been undertaken, he says, "the jury is still out."
Even Brenner counts himself among those calling for more stringent research measures.
"Unfortunately, our knowledge is way behind where it should be," he said. "We haven't invested enough in the science of healthcare delivery. We don't understand what drives some patients to over-utilize healthcare services and others to under-utilize care."
"We need a Manhattan-style project with anthropologists, sociologists, psychologists, data scientists, nurses, social workers, physicians, and others to build a new field and movement for complex care," he continued. "We're just at the beginning of this process. Chemistry has the periodic table, physics has Newtonian equations, biology has the theory of evolution. What do we have in complex care? A pile of meaningless terms and poorly tested ideas."
It would be hard to say it better. Recent efforts by the Camden Coalition and other similar-minded innovators have inspired a much-needed renewed focus on the most complex patients in our healthcare system.
Yet it is clear these efforts are falling short of expectations, and much more work – and rigorous evaluation -- will be needed to deliver on the promise of better, more efficient care for this population.
Joel Greenberg has been a science journalist for four decades. He was science and medicine editor of the Los Angeles Times and Editor-in-Chief of Science News magazine. He was also a science writer for the Miami Herald and the New York Times.