By Julie Knudson
For The Record
Vol. 26 No. 7 P. 22
Preventable readmissions are costly on several fronts, but technology and a dose of human touch can help lower rates.
The constant push to control costs and manage resources efficiently has many hospitals examining their readmission rates. With powerful data-crunching platforms and a bevy of HIT tools at their disposal, providers have more information than ever to help them determine who is most likely to be readmitted. However, turning the tide against excessive readmissions requires actionable data combined with a multilayered approach. It’s truly a case of technology meeting the human touch.
The scenarios vary, but it’s already been recognized that not every readmission is a negative event. “The CMS [Centers for Medicare & Medicaid Services] tries to distinguish between the preventable and the nonpreventable admissions—or as we call it, potentially preventable vs. the expected readmissions—by using a couple of codes,” says Kulleni Gebreyes, MD, MBA, a director at PricewaterhouseCoopers.
When patients are expected to return for additional treatment, such as oncology patients or those scheduled for surgery, there are specific codes providers can use to flag the case. “The CMS has exclusion criteria when they calculate the readmission rate so they can actually extract out all of those cases,” she explains.
The CMS also takes into account factors such as an organization’s historic performance, says Nell Buhlman, MBA, senior vice president of product strategy at Press Ganey Associates. Patient data, such as illness type and population demographics, are considered alongside provider data. “There isn’t necessarily a target rate of readmissions so much as there is a holistic kind of view that factors in the organization’s history, the historical performance, and the risks attached to the types of patients an organization sees,” she says, adding that the totals of acceptable and excessive readmissions are not hard and fast, so the CMS is more likely to consider ratios rather than straight rates.
Nevertheless, there’s enough historical data to set a workable baseline on readmission rates. “One thing we know is that readmissions across the country have been maybe 16% to about 26%, and that number hasn’t changed,” says Jyoti Kamal, PhD, chief data scientist at Health Care DataWorks. Much to the frustration of many in the industry, efforts to stem the tide largely have been unsuccessful. “People have tried to make efforts in trying to curb these readmissions, but the figure hasn’t really budged,” she says.
With only so many resources to go around, providers need to set achievable goals, says Christy Dempsey, MSN, MBA, CNOR, CENP, chief nursing officer at Press Ganey Associates. “Really, it’s the unplanned readmissions we have to tackle,” she notes. Similar to mortality rates, a feasible strategy for reducing unplanned readmissions must revolve around realistic results, Dempsey says, adding that no hospital is going to eliminate the problem entirely.
Where excessive readmission rates exist, penalties under the Affordable Care Act are designed to drive higher quality. “The national 30-day readmission rate for all causes is 16%, so those who are higher than that are penalized,” says Eric Heil, MBA, cofounder, president, and CEO of RightCare.
In 2014, the penalty rate was up to 2% of the total Medicare bill received by hospitals. “It’s starting to add up pretty significantly financially for them, and it’s only going to get worse,” he says, noting that in 2015, the figure will increase to 3% of the total Medicare bill, making reducing readmissions a top priority for many hospital executives.
Other penalties also may apply. For example, various states have implemented their own strategies to address potentially preventable events, with emphasis placed on avoidable readmissions. “There may be held-back payments as a penalty for that readmission rate being too high,” explains Charles Macias, MD, MPH, chief clinical systems integration officer and director of the Center for Clinical Effectiveness at Texas Children’s Hospital in Houston. “For the hospital, if services are provided that then don’t get reimbursed at either the same rate or at all, the hospital is absorbing the cost of that.”
The costs of excessive readmissions go beyond CMS- or state-imposed penalties. Looking holistically across the entire provider organization, above-normal readmission rates likely will spark several damaging effects downstream. For example, resources, often scarce in many hospitals, may be gobbled up unnecessarily when a patient is readmitted. “This could have been a resource utilized for another patient or a bed occupied by a patient who is not a readmission,” Kamal says. “That’s really what cost-effectiveness is—that you’re able to provide better care with less—and readmissions really stand out in that.”
Not only is resource utilization less efficient when readmission rates are high, but readmitted patients may consume more resources than patients entering the hospital for initial treatment. “Typically, that patient is coming back because they were not successful after discharge,” Buhlman says. “When they’re coming back, they’re either still sick or even sicker than they were before.” As a result, these patients may require more expensive care than typical patients. In some cases, readmissions cause hospitals to expend resources at a faster rate than their occupancy numbers may indicate.
Gebreyes says the billions of dollars spent annually on inpatient costs could be trimmed dramatically by just a 1% reduction in readmissions. “Hospital inpatient costs tend to be the highest costs for both the payer and for the provider, so even small reductions will lead to dividends just from the admissions perspective,” she says.
Beyond that, changes to how care is delivered also impact the bottom line. “The way you’re most likely to reduce readmissions is by providing quality care at that initial admission,” Gebreyes says. “It’s a multiplier effect.”
The fallout from excessive readmission rates extends beyond the financial ramifications. “More importantly, patients who are readmitted are often telling seven or eight friends and family of the bad experience of a readmit,” Heil says.
If that’s not enough, if readmissions continue to escalate, a hospital may find itself making headlines in the local media for all the wrong reasons. “As hospitals are focused on building market share and wanting to drive more quality outcomes for their institution, every readmit hurts,” Heil says. “Public perception is a powerful force, and in the highly competitive health care provider marketplace, a poor perception directly affects market share and revenue.”
Data Analytics to the Rescue
Identifying at-risk patients begins with data. Predictive models, which are becoming more widely used, must encompass many factors and information types. Some elements will influence readmission risk levels more than others. “One of the things we found is that if the patient was ever discharged from the hospital against medical advice, then that signal by itself increases threefold the risk for readmission,” says David Talby, PhD, senior vice president of engineering at Atigeo. Combined with other smaller indicators, each of these data points can help providers get a better overall sense of where their efforts likely will have the biggest influence.
Carefully paring the information sets often makes predictive modeling an even more useful tool, Talby says, adding that analytic models need to fit the demographics and other factors specific to each provider organization. “If you have a children’s hospital in Seattle and you have a VA hospital in Virginia or Florida, those are really very different populations,” he notes. “They’re going to see different distributions of patients and reasons for why people are at risk. In a sense, you have to localize the models.”
Because preventable readmissions are fairly rare in the grand scheme of health care events, developing highly predictive models requires information from a vast number of sources across a large patient population, says Matt Siegel, senior vice president of population health at Verisk Health. “While there’s a lot of hope and hype around Big Data and how it can be used to predict health care’s greatest cost drivers such as readmissions, we’re still in the early days,” he explains.
Siegel believes it would be ideal to leverage these data in near real-time to predict readmission risks at the point of care, but says many of the data sources aren’t yet standardized or longitudinally available.
Technology Ties It Together
HIT tools can aggregate population data into usable formats for providers. “[HIT] can help provide structure for creating analytic models and can help us understand who are the at-risk patients,” Macias says.
By pulling information such as socioeconomic factors into the discussion with clinical information, providers can access even more data to make care decisions and identify which patients may require support during discharge or various aftercare options. In the case of high-risk patients, this feature can prove invaluable. “A provider might act differently at that point to match the right kinds of support services—whether it’s social or whether it’s medical—to change the outcome,” he says.
As data analytic and HIT tools continue to gain momentum, everyone from providers to hospital leadership teams are better able to view and understand the key performance indicators affecting readmission rates. “They can then identify any gaps and understand the clinical and financial measures and outcomes,” says Lily Ariola, a senior in the advisory services practice at the consulting practice Ernst & Young, adding that it’s vital to get that information into the right hands at the right times. “Together, the organization’s HIT platforms and predictive modeling tools will give clinicians and managers the information necessary to identify the patients that need special attention.”
How to Handle At-Risk Patients
In today’s high-tech environment, providers have more opportunities to lower preventable readmission rates. Experts note that these efforts must begin long before patients are discharged. Oversight begins when the patient arrives at the hospital and continues through discharge and even beyond. “The minute that patient arrives in the hospital, if they fall into certain categories that would make them high risk, then that should trigger case management involvement,” Dempsey says, adding that risk factors can occur and evolve at any point along the continuum of care, making ongoing management crucial.
On admission, Macias says already there are factors present that can help providers estimate length of stay. The organization also can begin looking at the kind of resources the patient may need later to prevent a readmission. “When a patient gets discharged home, there is follow-up care that a patient may need or that the clinician may need to be attentive to prevent the readmission,” he says. “Factors in the community may influence that.” For example, the availability of local support services, from community support groups to faith-based organizations, should be considered when working to prevent readmissions, particularly in the case of chronic diseases where the continuum of care becomes crucial.
By flagging high-risk patients as soon as they arrive at the hospital, it enables case management staff to provide more targeted attention. “It triggers recommendations on what level of care best matches that patient and their specific needs much earlier,” Heil says. “It’s driving an action plan much sooner in the process.” Questions to consider include whether the patient would be better served in home health or at an inpatient facility and what sort of postdischarge support they may need.
Forming a bond with patients is key, Heil says. Inundated with information from all angles, patients can become overwhelmed. By collaborating with high-risk patients and their families, providers are better able to reduce the likelihood of readmission.
Patient compliance with postdischarge instructions is a crucial piece of the readmission puzzle. Unfortunately, it’s also a factor that sometimes may seem beyond the provider’s control, but that doesn’t need to be the case, Kamal says. “As part of care coordination, hospitals can be more proactive about educating not only the patient but the family and caregivers who are going to take care of the patient once the patient leaves the hospital premises,” she says.
For example, patients must understand their medication instructions. How often are drugs administered? Is filling a prescription convenient and affordable? Improved discharge planning and more robust patient education can help mitigate the risk factors associated with patients failing to adhere to a medication program.
Because readmits rarely are straightforward and often challenging when taken as individual situations, Ariola says a multifaceted approach to prevention garners the best results. “These include providing better and safer care during the patient’s course of hospitalization, completing the reconciliation of medications upon admission and discharge, putting together a comprehensive discharge plan which includes education with the patient and their caregiver regarding their follow-up care, and also improving care coordination through communication with the community and other health care providers,” she says.
In that regard, HIT and risk identification tools and software are key to integrating all these disciplines into a multilayered approach that spans the entire care setting.
— Julie Knudson is a freelance writer based in Seattle.