4

Benefits and Costs of Care

OVERVIEW

In her opening remarks, Session 3 moderator Wendy Johnson Askew¹ agreed with the earlier call for more health economic studies and also called for more culturally competent approaches to care. Without the latter, she predicted huge disparity issues around some outcomes. Without taking into account what is going on in a patient’s life, she said, “We are not going to achieve our ‘big hairy audacious goal’ of making this available to others.” The goal of Session 3 was to address financial challenges to reaching this goal, Johnson Askew continued. This chapter summarizes the Session 3 presentations and discussion.

First, Jim Lee discussed several key points to keep in mind when considering the cost of nutrition intervention benefits, including the complex and costly nature of measuring that cost. During the panel discussion at the end of the session, Lee called for more combined clinical outcome and budget impact studies, rather than full cost–benefit analyses.

Next, Brenna Shebel discussed ways that employers can help employees who have cancer or are serving as caregivers and how the National Business Group on Health (NBGH) is aiding in this effort. In collaboration with the National Comprehensive Cancer Network (NCCN), in 2013 the NBGH issued a series of recommendations, including one nutrition recommendation, to help employers align employee benefits with evidence from across the continuum of cancer care.

___________________

¹ Session 3 was co-moderated by Wendy Johnson Askew and Nico Pronk.

COST–BENEFIT CONSIDERATIONS²

Twenty-two years ago, Jim Lee began, a nutrition researcher was telling him about the struggle to determine the cost and benefits of nutrition intervention, to which Lee replied, “That is easy.” But 22 years later, he is still struggling to figure out not so much how to do it, but how to get people to act on the information. This is a particularly difficult challenge in the United States, he said.

He listed four key points to keep in mind when considering the cost of nutrition intervention benefits and then, for the remainder of his talk, discussed each in further detail:

1. Understand how malnutrition affects high priority outcomes.
2. Carefully consider co-morbidities and primary therapy effects. These are not simple cause and effect relationships, Lee explained.
3. Measure costs and benefits commensurate with the intervention and from multiple perspectives. Fortunately, nutrition interventions are not very costly, which Lee said needs to be understood when studies are designed.
4. Understand that measuring costs is complex, often costly, and best done within a clinical study. Because the United States does not have the benefit of a single payer system, the data get very messy very quickly, Lee remarked.

Understanding How Malnutrition Affects High Priority Outcomes

Lee emphasized the importance of considering outcomes, like mortality, that are important not just from a financial perspective, but also from a patient perspective, when examining the effects of malnutrition in oncology patients. An observational study of approximately 800 hospitalized patients in Singapore, including 55 oncology patients, showed that both costs and 1-, 2-, and 3-year mortality were significantly different between malnourished and well-nourished patients (after both statistical and risk adjustments) (Lim et al., 2012). Three-year mortality among malnourished patients was 48.5 percent, compared to 9.9 percent among well-nourished patients. Of the 55 oncology patients, 71 percent were malnourished within 48 hours of admission, based on a subjective global assessment of nutritional status. While these results do not imply causality, only correlation, Lee speculated that if you were to ask any patient if a five-fold difference in risk of mortality is an important factor, all would say yes.

___________________

² This section summarizes information and opinions presented by Jim Lee, M.S., Altarum Institute, Ann Arbor, Michigan.

Another high-priority outcome from both financial and patient perspectives is hospital readmission. Patients have busy lives, and few like to be readmitted. Moreover, hospitalizations are a major event not just for the patients, but also their caregivers and families. In the same Lim et al. (2012) study, while the difference in readmissions between malnourished and well-nourished patients was not statistically significant, nonetheless, in Lee’s opinion, the trend was clinically significant, with the higher readmission rates among malnourished patients.

Carefully Consider Comorbidities and Therapy-Related Effects

In addition to high priority outcomes, Lee emphasized the importance of also considering comorbidities and therapy-related effects when evaluating cost of care. In a study of what Lee considered a relatively inexpensive and generalized intervention, specifically a palliative care consultation within 2 days of admission among oncology patients, May et al. (2016) found that, compared to usual care, receipt of the palliative care was associated with 22 percent lower costs for patients who had comorbidity scores of 2-3 and 32 percent lower costs for patients with comorbidity scores of 4 or higher.

Lee emphasized the importance of understanding not just comorbidities but also treatment-related effects from a patient perspective. “We can get lost in the very important science and miss out on what is, in fact, most important to sometimes gravely ill patients,” he said. For example, sharing a meal together does not have just a nutritional effect. “It can be a very meaningful event,” Lee said.

Additionally, certain patients are at greater risk for higher costs due to completely independent factors, such as whether they have a spouse at home who will care for them. For example, Lee observed that elderly men often have a wife at home who will care for them, whereas elderly women do not. Where patients go after they leave inpatient care and who they have at home to support them are factors that often are not considered in economic analyses or outcome studies. Yet, Lee said, these can be very important determinants of risk of readmission, adherence with a care plan, and other outcomes.

While not designed for this type of risk adjustment, there are several methods available that can provide some risk adjustment in these types of situations. These include the Elixhauser Comorbidity Index, a method for measuring patient comorbidity based on International Classification of Diseases (ICD)-9 and ICD-10 diagnosis codes in administrative data (Elixhauser et al., 1998). Lee referred workshop participants to an Agency for Healthcare Research and Quality (AHRQ) website that pro-

vides software for generating Elixhauser Comorbidity Indices.³ Others are the Edmonton Symptom Assessment Scale (Bruera et al., 1991) and the Condensed Memorial Symptom Assessment Scale (Chang et al., 2004). These methods can be especially helpful for understanding why expected outcomes are not observed in some subsets of patients, according to Lee.

Measure Costs and Benefits Commensurate with the Intervention and from Multiple Perspectives

Lee challenged nutrition researchers to help payers understand the cost of nutritional interventions. He told the workshop audience how he used to joke with a colleague about diagnostic tests for $125,000 therapies that cost only a few hundred dollars and how they would refer to those few hundred dollars as “budget dust.” However, as Kathryn Phillips and colleagues at the University of California, San Francisco, have shown, often there is no evidence that these “budget dust” diagnostic tests were actually conducted (i.e., before therapy). Yet, Lee said, he gets “beaten up” over a $200 nutrition counseling session. He urged researchers to measure the costs of nutrition interventions. In many of the studies he has seen, there has been no attempt to at least try to measure some of the cost even when doing so would typically not be that difficult.

When costs also have financial benefits, however, measuring the overall costs becomes more complex, Lee continued. For example, using a medication can also lead to a better outcome (i.e., with financial benefits). Adding to the complexity, even in a single institution, there are typically multiple payers, each with different reimbursement methods for the same treatment and case management. Taking into account not just payer costs and financial benefits, but also employer costs and financial benefits (e.g., disability payments), as well as employee costs and financial benefits (e.g., out-of-pocket costs, lost wages), adds further to the complexity. Lee remarked that out-of-pocket costs for oncology patients were lower in 2012 compared to all other years Lee and colleagues had been measuring them. That was prior to the rapid expansion of high-deductible health plans, he noted. He did not expect to see such low costs again.

Finally, provider costs and financial benefits also need to be considered. Lee mentioned that he is looking forward to finding out if there is a nutrition component to the new Oncology Care Model being developed by the Center for Medicare & Medicaid Innovation (CMMI) and, if so, how the different incentive structure of the model (i.e., reimbursement being based not necessarily on treatment, but on how a treatment contributes to

___________________

³ See https://www.hcup-us.ahrq.gov/toolssoftware/comorbidity/comorbidity.jsp (accessed August 3, 2016).

outcome) will affect the costs and financial benefits of nutritional intervention. In his opinion, the model provides a tremendous opportunity to start measuring these costs and benefits in relatively carefully controlled studies.

When direct measures of costs and financial benefits are too costly or difficult to obtain, Lee suggested that proxy measures be used. For example, instead of trying to obtain highly confidential data from 12 different payers, a good proxy would be what Medicare would pay for a service. Additionally, he suggested conducting studies in places where there is a single payer. Most of his own research has been done in Canada, the United Kingdom, and other places where the data are much more readily available than they are in the United States.

Understand That Measuring Costs Is Complex and Often Costly

Finally, Lee emphasized the complexity and cost of measuring costs. Costs have very high variance, not only because treatments are different among practices but because prices are different. The statistical significance of cost is “incredibly difficult” to obtain because of sample size and the cost of a large study, according to Lee. That said, very rarely has he come across decision makers who insist on studies being designed for statistical significance of costs. Most understand that it is unlikely.

In addition to high variance requiring large sample sizes for statistical significance, exogenous factors often dominate the intervention effect. This is true even for very large sample sizes, Lee said. Exogenous factors include things like reimbursement, facility cost structure, and practice patterns. He noted that some of the work he had been involved with recently had engaged the two or three key payers (i.e., in either the institution or the state) from the beginning, including during the study design, to help control for these factors.

A final factor to consider is that even single-site studies may involve several payers, requiring intensive data collection and a tremendous amount of work even just to find out what happened (e.g., whether a patient was readmitted, the cost of readmission).

To close, Lee said, “Twenty-two years ago, I thought this was going to be easy. Now I stand here today and tell you, it is not. But it is well worth doing.”

PERSPECTIVES ON ONCOLOGY NUTRITION CARE: EMPLOYERS AND OTHER PURCHASERS⁴

Brenna Shebel and colleagues at the NBGH collaborated with the NCCN on a 3-year initiative to establish a set of criteria and recommenda-

___________________

⁴ This section summarizes information and opinions presented by Brenna Haviland Shebel, M.S., National Business Group on Health, Washington, DC.

tions for employers to adopt as a way to make cancer a priority within their workplaces. Shebel went on to describe some of these 56 recommendations, one of which was on nutrition care, and how the recommendations help employers to align their employee benefits with evidence. The Business Group, Shebel explained, is a membership organization of mostly large employers, including 70 of the Fortune 100.⁵ The employers Shebel serves work very closely with their health plans to ensure the benefits being offered to their employees meet certain evidence metrics. Because they are self-funded, NBGH members have a lot of flexibility within their benefit programs. So while serving a broad and diverse workforce is a challenge, this challenge is something Business Group members are “up for,” Shebel said.

How Employers Are Serving Employees Who Have Cancer or Are Serving as Caregivers

Shebel listed several reasons why employers are making cancer a priority. First is the very high incidence of cancer among the working population. About half of men and one-third of women are diagnosed with cancer at some point in their lives, and virtually everyone is touched by cancer in some way (e.g., being a caregiver, having a colleague who has cancer). Second, Business Group members are very concerned about variation in quality of cancer care and are becoming increasingly interested in matching their employees, particularly those with rare and aggressive cancers, with the best providers and with access to evidence-based information about cancer. Third, cost is a huge factor. While employers want the highest-quality care for their employees, they are also concerned with cost. The cost of cancer treatment is typically among the top three most costly medical conditions, representing, on average, 12 percent of total medical expenses for employers. Moreover, Shebel added, the cost of cancer treatments are rising faster than general medical costs. Lastly, employees want to return to work during treatment if they can, and employers want their employees to return to work. Returning to work is important for a patient’s sense of normalcy.

With respect to what employers are doing to address cancer among their employees, first is the provision of medical benefits. In fact, Shebel said, the bulk of recommendations put forth by the Business Group are in the area of medical benefits. These include ways to adapt evidence to an employer’s benefit design; ways to provide employees with access to evidence-based information and services, including behavioral services for the treatment of depression and anxiety; and recommendations relating to centers of excellence (COEs). COEs are subsets of in-network providers

___________________

⁵ The Fortune 100 is an annual list of the 100 largest public and privately held companies in the United States.

that provide condition-focused care and that employers either directly or indirectly contract with through their health plans, with participating providers having proven track records of positive outcomes. In a recent survey, the Business Group asked their members if they were or would be offering COEs for a variety of conditions. While cancer was not at the top of the list (transplants were), Shebel said, “It is definitely growing.” Twenty-nine percent of NBGH members offered a cancer COE in 2016, although only 3 percent used incentives (e.g., by waiving deductibles or making it less expensive for employees to seek care at a COE provider). Another one-third reported that they were considering offering cancer care COEs in the future.

The cost of medications, including parity of costs between oral oncolytics and infusion medications, is another key cancer care issue for many Business Group members and their employees. Cost should not be the main deciding factor for patients when choosing medications, Shebel said. Patients should be choosing the medications that are right for them. NBGH recommendations include ways to make medications more affordable for employees.

While many employers offer care management programs for their employees, including the provision of some kind of support for employees with cancer, the Business Group recommendations also include ways to bolster that support and the training of staff within those programs. A recent Business Group survey of employees indicated that, more often than not, employees want health information and support not so much from their employer, rather from their health insurer or another third party. This was not a surprising finding, Shebel said. From a privacy perspective, it is not as comfortable for employees to see that kind of support from their employers. When asked what type of support they want from their employer, flexibility was at the top of the list. That is, employees want to have flexibility around their work schedules so they can get to appointments, treatments, tests, and so on. Another important finding from this same employee survey was that getting a second opinion from a specialist is important for cancer patients.

In addition to medical benefits, pharmacy benefits, and care management programs, a fourth way that employers are addressing cancer among their employees is by offering short term disability (i.e., typically up to 6 months). Thus, that is another area where the Business Group provides a number of recommendations, including ways to work with treating physicians to determine duration of leave and ways to structure short-term disability to help people return to work when they are ready.

The Business Group also makes recommendations around family medical leave, which is protected leave that both patients and caregivers can take and which is often unpaid. Employees who need to take family medical leave are usually people who have cancer, have already faced very high medical costs, and need to be routed to the right providers, including providers of financial assistance.

Employee assistance programs are another way employers address cancer among their employees. While many NBGH employers offer such programs, the programs are highly underutilized, Shebel said. Employee assistance programs are designed to support employees around any type of life challenge or transition, whether it be financial, legal, related to child care, or something else.

Health improvement programs are another relevant benefit that employers provide, but one that is often misunderstood, according to Shebel. Because of their focus on cancer risk reduction, Business Group recommendations for health improvement programs cover strategies that support not just employees with cancer, but the entire working population as a way to reduce their risk of cancer.

All of these various benefits and programs are great for employees, Shebel said. But what if employees, when diagnosed with cancer, do not know these benefits exist? The Business Group has not only issued recommendations around employee communications, but the group has also actually written materials, with the help of NCCN, that employers can use to communicate with employees who are either newly diagnosed cancer patients or caregivers.

The Business Group’s Recommendation on Nutrition Care

More specifically (than stated in the quote above), the Business Group nutrition recommendation suggests that in-network providers be registered dietitians (RDs) who are board-certified specialists in oncology and that patients pay out-of-pocket as they would for other in-network services. The Business Group also recommends that employers adopt the NCCN Clinical Practice Guidelines in Oncology, which state that providers should conduct assessments to evaluate weight gain and loss and other nutrition-related conditions, consult or refer to nutritional experts in the case of any abnormalities, and manage deficiencies resulting from anorexia, diarrhea, nausea, and vomiting.

The genesis behind the nutrition recommendation, Shebel explained, was to help employees in treatment to maintain quality of life and ability to function both during and after cancer treatment. This is especially important for employees who want to continue working. Recognizing the data presented by previous workshop speakers, Shebel noted that the statistic she uses with employers is that approximately 50 to 60 percent of patients

diagnosed with cancer experience significant weight loss and poor nutrition during the course of their illnesses.

Future of the Employer Role

Based on her weekly work with corporations to improve their cancer benefits, as well as media interviews, Shebel speculated on several future employer roles in cancer care. First, as Lee had mentioned, consumer-driven health plans (CDHPs) and high-deductible plans are certainly the wave of the future, based on Business Group data. In fact, that future is already here, with 83 percent of Business Group members offering some sort of CDHP and one-third offering only CDHPs (i.e., only choice for their employees). For Business Group members, Shebel said, “what really keeps them up at night” are individuals with these plans who are facing major medical decisions and costs. With a lot of upfront or out-of-pocket costs, these individuals may not be able to meet other financial demands, like mortgages and bills. So some employers are looking into critical illness coverage to help with some of these financial issues.

In addition to offering critical illness coverage, employers are increasingly focusing on supervisors of employees as a source of support. Although employees do not need to disclose that they have cancer, because supervisors play key roles in helping employees to manage their schedules, employees often choose to disclose. As such, the Business Group is continuing to push their members to recognize the importance of training supervisors so that when a patient does discloses that he or she has a cancer diagnosis, the supervisor knows how to react, including avoiding the “wrong” thing to say, the “right” thing to say, and where the supervisor should direct the employee for benefits. Also increasingly, Shebel continued, employers are also looking at the evidence and wanting their providers to use evidence and to share outcomes.

Finally, many employers are providing concierge services to their employees. Concierge services provide employees with a single “advocate” who helps with all health care needs, whether that is enrolling in benefits, improving health (e.g., losing weight), or facing a major diagnosis. “You have one single person to work with,” Shebel said. In her opinion, it is a “fantastic” option because employees don’t have to figure out which number to call or who to talk to. Concierge advocates know exactly where to direct employees with respect to both providers and support.

In closing, Shebel noted that NBGH’s tools are open to the public, even though the group is a membership organization.⁶ Printed copies of the recommendations are also available upon request.

___________________

⁶ See http://www.businessgrouphealth.org/cancer (accessed August 3, 2016).

PANEL DISCUSSION WITH SPEAKERS: DATA GAPS IN OUTCOMES AND COSTS OF CARE

Following Shebel’s presentation, she and Lee participated in a panel discussion with the audience. This section summarizes the discussion that took place.

The Role of the Centers for Medicare & Medicaid Services (CMS), Reimbursement, and the Value of Nutrition Services

Pronk opened the discussion by asking the panelists if any policy work was being done to align interests of Medicare with investments that employers make in keeping people healthy before they retire.

Shebel observed that, with all the Medicare payment models being tested, employers now are looking to Medicare as the innovators and taking a “wait-and-see” approach to what emerges from all of the “exciting movement” around these models. She expressed hope that there is some place within CMMI’s work on oncology to address nutrition care. Lee added that the fact that the United States is fifth best in the world with respect to smoking rates among adults (i.e., fifth lowest) is a good example of how public policy can have a substantial effect on health, with both CMS and large employers being beneficiaries of that.

Observing that “the theme” among suggestions to “change the scope of things” with respect to reimbursement for nutrition services is to have a conversation with CMS, an audience member asked the panelists how they would suggest doing that. What kind of outcomes data are needed? Should data be collected nationally? What would be a good case to take to the CMS to make the necessary changes? Lee replied that CMS is primarily interested in examining different models of reimbursement, rather than adding more professional groups to a fee-for-service model, creating a challenge for those groups not currently part of the current fee-for-service reimbursement model. He would suggest approaching Medicare Advantage plans and others, rather than CMS directly, at least in the near term, or participating in some of the CMMI initiatives and demonstrating the value of nutrition care using what he described as “their experimental design.”

“I think if we are going to go up to ask for reimbursement and support, we have to go armed, which means we have to have the data,” Clinton remarked. But we also have to pick our battles, he added. This raises the question, where will nutrition support services have the greatest affect on outcome? After deciding what to fight for first, then organizations should conduct the necessary systematic reviews “to carry that battle up and down the spectrum of outcomes.” Gathering data on long-term outcomes among

cancer survivors will be a challenge. However, the same evidence used to develop the Dietary Guidelines for Americans and the American Institute for Cancer Research recommendations could also be used for this purpose. The greater question, Clinton opined, is, “How do we pay for that kind of quality intervention, which is labor intensive?” Four consults a year will not change someone’s diet. That amount of care would be “woefully inadequate” given the limited reimbursement dietitians currently receive for providing diabetes care.

While these new CMS models of reimbursement are going to be simpler, Clinton continued, in terms of eliminating paperwork and reimbursing a single set of money for the entire care process, which the hospital will have to figure out how to distribute among its providers (e.g., among the surgeon, the medical oncologist, the pharmacist, the RD, etc.), they will also have interesting effects. He predicted that this type of reimbursement will create internal battles, which “is not going to be helpful for our mission.”

Rather than trying to monetize return on investment, Pronk suggested that the focus should be on value. Oncology nutritional services are incredibly valuable, he said, but their return on investment is difficult to monetize.

Shebel commented on the team-based nature of the new payment models. For her, what is exciting about the models is that the patient is “absolutely at the center,” with shared savings among providers. Providers will not receive payment if the care provided is not evidence-based and the outcomes are not positive. Robien opined that the accountable care organization (ACO) model in particular is a “great place” for dietitians to become involved in these teams, especially in early cancer care. One of the reasons diabetes care is so successful, she said, is because dietitians can look at short-term biomarkers to evaluate how nutrition interventions modulate hemoglobin A1c levels over time. Oncology dietitians could do the same sort of short-term monitoring during early cancer care. She noted that most dietitians are not trained in this area, but that it is an area with which they need to become more familiar and soon. She encouraged dietitians to be more assertive about being part of this new model of care.

Evidence from Electronic Medical Records

Wendy Johnson Askew asked whether electronic medical records can help to facilitate calculating costs in some of these new benefit models and looking at long-term outcomes as a way to provide evidence for viable options. She recalled a frustrating time when dietitians did not have a systematic way to talk about the value of their services. “We knew anecdotally that dietitians saved money,” she said, “but we didn’t have any way to talk about it.” Now, with electronic medical records, she said, “It is unfortunate

that we haven’t been able to capture more of this.” She asked, “How can we organize ourselves in a way that we can capture this data?”

In Lee’s opinion, there is still potential for the electronic health record in the long-term. But currently, the U.S. health care system is so disparate that very seldom is one group taking care of all of a patient’s care. He said he often brings his own electronic health record up to date before he goes to the doctor. Interim solutions include, first, participating in professional registries and studies and, secondly, working with information technology groups to standardize data and move away from manual processes. It is going to take many steps, he said. That said, in his opinion, there are some good role models. He suggested looking at the Australian and Dutch models.

Clinton added that The Ohio State University and many other research institutions and comprehensive cancer centers have what are known as honest broker data warehouses. In his opinion, dietitians need to be trained on how to use those warehouses. At Ohio State, they will actually pull for free whatever you ask them to pull from the medical records. For example, you can obtain nutrition-related symptoms. This is very helpful, and he has trained a couple of their oncology dietitians to know how to do this. The information is de-identified and can be published.

Shebel suggested that ACOs, which she identified as one of the fastest growing trends for NBGH members, may someday be helpful in this effort.

Employer–Employee Relations

There was some brief discussion around employer–employee relations with respect to the provision of nutrition care for oncology patients. An audience member observed that, while many employers are supportive of their employees going through treatment, the opposite is also true, with patients struggling to make it through treatment because they feel they need to be at work so they do not lose their job or benefits. This raises the question, what conversations should employees be having with their supervisor to gain that support? Shebel agreed that this situation exists and that not every supervisor “is perfect” in this area, which is why it is so important that supervisors receive training in this area at least at a “baseline” level so they know what benefits exist when they have that first conversation with an employee facing a diagnosis. She emphasized that the culture of the organization also matters. It is not just the supervisor who should be involved with providing support for employees facing diagnoses, but also human resources, the employee assistance program, and maybe even the legal department.

A webcast participant asked Shebel what else employers could be doing to expand employee access to evidence-based nutrition care and what it will

take to get payers to incentivize the delivery of nutrition care. Currently, Shebel replied, payment and delivery methods are focused more broadly on cancer care. “We have a way to go,” she said, “to drill down all the way to nutrition care.” NBGH makes recommendations to its members to provide evidence-based information to its employees who are newly diagnosed. Specifically, they promote patient information from the American Cancer Society, NCCN, and the American Board of Internal Medicine Foundation’s Choosing Wisely campaign. Lee added that clinical guidelines can be very influential with payers. When he talks with payers, one of the first things they want to know is standard of care. They often turn to third-party sources to help make decisions about what to cover.

Gaps in Research and Communication

Pronk observed that both speakers emphasized the strength of the evidence, but neither touched much on research gaps. He asked which major research gaps would need to be closed for this area of the field to move forward in a productive manner. Lee identified the need to combine clinical outcomes and budget effects, rather than conduct full cost–benefit analyses. Employers want to know roughly how much something is going to cost to cover their employees and that a professional society has reviewed the care and determined it to be evidence based. He has seldom seen full cost–benefit analyses and did not think they should be the final arbiters.

Shebel agreed that clinical outcomes associated with providing services are very important. Recognizing the “great information” presented thus far in this workshop, she wondered whether the greater challenge is not gaps in research, rather that already existing data might not be getting to the hands of payers and employers. In addition to getting outcome data and information “out there,” she also pointed to the need for providers to provide data. For example, in reference to a question in an earlier session about oncology medical homes, she said, “We want to know more about that.”

This page intentionally left blank.