Care Redesign
Integrating Mental, Social, and Physical Health

Prescribing Food as a Specialty Drug

Article · April 10, 2018

“The last time I looked in my textbook, the specific therapy for malnutrition is food.” That response was given by Jack Geiger, who helped create the community health center model in rural Mississippi in the 1960s. He had been challenged for buying food for patients and charging it to his center’s pharmacy budget. His retort became a rallying cry for social needs activists, who loved Geiger’s combination of rebelliousness and common sense.

Nevertheless, in most health care organizations, Geiger’s argument for meeting patients’ social needs has been more rhetoric than reality — that is, until recently. Despite political uncertainty from Washington, health care is moving away from its focus on fee-for-service and toward value, i.e., meeting patients’ needs as efficiently as possible. To do so, many providers are developing systems to detect and meet patients’ unmet needs.

Concentrating the Research Focus

It’s new work, and it’s hard work. It’s usually unfunded work. But it works.

At least it can work, when pursued with the same discipline with which researchers might test a new drug: focusing on a well-defined population with a good chance of showing benefit if the intervention is effective. At Geisinger Health System, we understood the risks of trying to “boil the ocean,” and instead tried to see what was involved in meeting just one social need in just one population. We decided to concentrate on diabetes, a condition that affects 11.3% of the population in Pennsylvania, which is higher than the national average. We wanted to work out administrative procedures in which we provided healthy food as if it were a drug and started a program in 2016 called Fresh Food Farmacy.

We focused on food because research suggests that food insecurity, or lack of access to nutritionally adequate food, is one of the most important risk factors for developing type 2 diabetes. Food-insecure adults are two to three times more likely to have diabetes than adults who are food secure. And food insecurity is widespread; more than one in eight American adults and one in six children is food insecure. People with severely limited incomes often turn to inexpensive, easily accessible food that is rich in calories and poor in nutrients, which can cause and exacerbate diabetes.

Directly Dealing with Food Insecurity

The key steps in our program: First, we screen for food insecurity. Then, we give away free, nutritious food. Along the way, we teach our patients about their disease and how to manage it.

Geisinger based its program upon two optimistic assumptions: first, that eating healthfully improves health, and second, that people want to eat better but may not know how to or have the means to do so.

The target patients eligible for Fresh Food Farmacy are adults whose type 2 diabetes is not well controlled. They must also answer “yes” to at least one of two questions that screen for food insecurity: (1) “Within the past 12 months I/we worried whether our food would run out before we got money to buy more,” and (2) “Within the past 12 months the food I/we bought just didn’t last, and we didn’t have money to get more.” Data were collected via our EMR searching for diagnosis of type 2 diabetes with elevated sugars out of control typified by a blood test measurement of hemoglobin A1c (HbA1c) 8. Patients were screened for food insecurity electronically (via the patient portal), by phone, or in person when at a clinic visit. Of 458 patients screened, 128 patients were found eligible for the Fresh Food Farmacy, and 95 patients were enrolled.

Providing Free Food as a Treatment for Diabetes Yields Improved Outcomes for Patients While Reducing Cost of Care - Fresh Food Farmacy - Geisinger program for addressing food insecurity in patients with diabetes

  Click To Enlarge.

Exploring Program Components

The “Farmacy” for this program is a food pantry built in a Geisinger clinical center in Shamokin, Pennsylvania. It provides enough fresh fruits and vegetables, whole grains, and lean proteins to feed program participants and their entire household two healthy meals 5 days per week each. Patients are also given a weekly menu and recipes to help them learn to use ingredients that may be unfamiliar to them.

Once enrolled, participants are required to attend 15 hours of group classes. They learn basic concepts about their disease, including: What is diabetes? How do you get it? What is blood sugar? And how does my diet affect my blood sugar? The goal is to give patients a better understanding of their diabetes, so they increase their ability to care for themselves.

Participants also have access to a care team that includes a nurse, primary care physician, registered dietitian, pharmacist, health coach, community health assistant, and administrative support personnel. The team offers services to help patients sustain their lifestyle improvements, including direct medication-management assistance; nutrition counseling; health coaching; and ongoing case management to address transportation, family care, and other challenges that can make engagement difficult.

Our initial pilot was limited to six patients. Within 9 months, we scaled up to include 50 patients. As of March 2018, we are serving 112 patients/households, feeding an average 336 people per week, which amounts to 3,360 meals per week or 174,720 meals per year. One of our goals is to reduce the meal gap in our community, and extending free food to those in the patient’s household is helping with that.

We currently fund the program through grants (40%), in-kind reciprocal contributions with Geisinger Health System (30%), and private donations (30%). Geisinger provides in-kind administrative support and rent-free space for the food pantry and clinical areas, though we anticipate that, as we expand, it’s likely that the program will pay rent.

Delivering Clinical and Financial Results

In 18 months, this approach has led to a drop in HbA1c levels from an average of 9.6% before enrolling in the program to 7.5%. To put this change in perspective, diabetes patients who take two or three medications can expect their HbA1c to drop between 0.5 and 1.2 percentage points. The average 2.1 percentage point drop in HbA1c levels is a better outcome and corresponds to a more than 40% decrease in their risk of death or serious complications.

Biometric Outcomes for Fresh Food Farmacy Enrollees - Geisinger program for addressing food insecurity in patients with diabetes

  Click To Enlarge.

Because many of the participants are insured by Geisinger Health Plan, health care spending data are available for 37 of our patients. Thus far, claims data shows costs for our pilot patients dropped by 80%, from an average of $240,000 per member per year, to $48,000 per member per year. These data come from a small sample of patients, of course, but the enormous baseline annual cost suggests that the combination of factors that defined eligibility (poorly controlled diabetes, screening positive for food insecurity, and willingness to enroll in the program) identifies an extraordinarily sick and costly population. We assume that Geisinger case managers were most relentless in encouraging their sickest patients to undergo screening for the program, thus further skewing the population toward high costs, so some of the decrease in costs could reflect “regression toward the mean.”

Nevertheless, the large drop suggests considerable potential to improve health via a program with operational costs of about $2,400 per patient per year. Is free healthy food the driver of this improvement in health and reduction in costs? Or is it the education and case management that come with that food? We don’t know, but we suspect that there is synergy and would not want to separate them.

From a financial perspective, one could argue that most of the program cost (i.e., the care management team) is consistent with standard of care traditional diabetes disease management, and those costs are consistent with traditional care. The acquisition costs to the program’s “specialty drug” — free healthy food — averages about $6 per person per week. (Yes, that’s about 60 cents per meal for 10 meals. We purchase about 60% of the food through the local food bank, and about 40% through retail providers for fresh fruits, vegetables, and fish.) If a new diabetes drug became available that could double the effectiveness of glucose control, it would likely be priced considerably higher than $6 per week (and if it wasn’t, the pharmaceutical firm’s stockholders would be in revolt).

There are ancillary costs to the program, but ancillary benefits as well. Fresh Food Farmacy is feeding entire families because, well, feeding only the participants is impractical when they eat with their families. Patients have had significant improvements in their cholesterol, blood sugar, and triglycerides and many are asking for help in other areas, such as becoming more physically active and quitting smoking. The health benefits to family members have not been part of the evaluation to date, but they will be assessed in the future.

Some might call the Fresh Food Farmacy a nontraditional approach to diabetes, but if one considers fresh healthy food to be the equivalent of a drug covered by insurance and provided by the health care system, then this is essentially a disease management program — just more successful than most. And, without systems for its appropriate, effective, and efficient use, fresh healthy food is no better than any medication. Our experience suggests that by detecting the patients most likely to benefit from the program and meeting their needs, we can reduce hunger in our communities while improving the health of our patients.

New call for submissions ­to NEJM Catalyst

Now inviting longform articles

Connect

A weekly email newsletter featuring the latest actionable ideas and practical innovations from NEJM Catalyst.

Learn More »

More From Care Redesign
Summary of Comprehensive Approach to Physician Behavior and Practice Change

Engaging Stakeholders to Produce Sustainable Change in Surgical Practice

How an initiative designed to improve patient outcomes and satisfaction while containing costs led to sustainable change in surgical practice and physician behavior.

Myths and Realities of Opioid Use Disorder Treatment.

Primary Care and the Opioid-Overdose Crisis — Buprenorphine Myths and Realities

There is a realistic, scalable solution for reaching the millions of Americans with opioid use disorder: mobilizing the primary care physician (PCP) workforce to offer office-based addiction treatment with buprenorphine, as other countries have done.

Coffey02_pullquote family-centered care in medical and surgical procedures

What If Family-Centered Care Were Extended to Medical and Surgical Procedures?

Though the concerns are valid, early experiences suggest that family member engagement may be an effective tool for improving the value of care.

Evidence Needed for Health Systems Change to Address Social Determinants of Health and Obesity and Diet-Related Diseases in Turn

Better Clinical Care for Obesity and Diet-Related Diseases Requires a Focus on Social Determinants of Health

To more effectively treat the problems of obesity and diet-related conditions, health systems need to restructure the traditional medical model of care delivery to address the social determinants of health.

People Living with Dementia Around the World - Value-Based Chronic Illness and Dementia Care

Value-Based Care Must Strengthen Focus on Chronic Illnesses

To effectively control costs and improve value, new models must address our increasingly older patients and chronic care patients, especially those with Alzheimer’s and related dementias.

The Barriers to Excellent Care Vary Widely Across Geographic Regions - both Rural Health Care and Urban Health Care

Survey Snapshot: Rural Health Innovations Born from Challenges

According to NEJM Catalyst Insights Council members, every health system has to develop its own definition of what is meant by “rural” health.

Same-Day Breast Biopsy Workflow at Baylor College of Medicine

How Care Redesign and Process Improvement Can Reduce Patient Fear

Seeing how clinicians take care of their own when they are in frightening situations was the epiphany that led to a same-day breast biopsy program.

Rural Health Care Is Rated Comparable or Worse Across Quadruple Aim Aspects

Care Redesign Survey: Lessons Learned from and for Rural Health

Although care delivery models in rural and urban/suburban areas are distinct, by virtue of geographic density and resource availability, each locale affords lessons for the other.

Comprehensive Intervention Review at Lurie Childrens Hospital - improving patient flow and length of stay

Reducing Length of Stay in the ED

A comprehensive redesign of triage and ED care.

Pumonary Nurse Post-Discharge Follow-Up Note for Patients with COPD

TOPS: Telephonic Outreach in the Pulmonary Service at VA Boston Healthcare System

A nurse-directed intervention targeting veterans who had been hospitalized for COPD resulted in improved access to ambulatory care and a reduced rate of readmissions.

Connect

A weekly email newsletter featuring the latest actionable ideas and practical innovations from NEJM Catalyst.

Learn More »

Topics

Social Needs

88 Articles

Better Clinical Care for Obesity and…

To more effectively treat the problems of obesity and diet-related conditions, health systems need to…

A Successful Pilot to Improve Access…

Actionable data and modest financial incentives can help motivate clinicians to adjust their behavior around…

Coordinated Care

129 Articles

The Evolution of Primary Care: Embracing…

Primary care must leverage disruptive innovations to ensure that patients receive first-access, comprehensive, coordinated, continuous…

Insights Council

Have a voice. Join other health care leaders effecting change, shaping tomorrow.

Apply Now