Lessons for Program Design From a Food-as-Medicine Experiment

The spread of an innovative program piloted in Philadelphia is the result of persistence, rigorous study and evaluation, along with efforts to avoid common pitfalls.
June 27, 2018
Chicken noodle soup.
(Shutterstock)
Feather O'Connor Houstoun
By Feather O'Connor Houstoun  |  Contributor
A senior adviser to the Wyncote Foundation

Grandmother believed it: Feeding chicken soup to the ailing speeds their recovery. Now, with pilot programs evaluated and larger-scale studies underway, the Medicaid program may recognize as a medical expense free meals for patients with debilitating diseases where customized, nutritious meals may have life-saving impact.

On the strength of a pilot program in Philadelphia operated by the Metropolitan Area Neighborhood Nutrition Alliance (MANNA) with four Pennsylvania managed-care organizations, Medi-Cal, California's Medicaid program, has launched its own pilot with 1,000 patients with congestive heart failure and diabetes to test whether therapeutically-prescribed free meals can alter medical outcomes and costs.

MANNA's role in precipitating this innovation is an exemplar of persistence in one area of social services. But it also offers lessons that inform public administration and program design across the board.

MANNA is among a handful of large nonprofits that emerged during the AIDS epidemic and persist to offer intensive feeding support for individuals with acute nutritional risk due to life-threatening illnesses. Since 1990, it has served medically-prescribed prepared meals, 21 per week, to individuals whose illnesses demand special preparation and/or nutritional content. But despite myriad studies on interventions and outcomes within multiple disciplines, there was no research on whether free, specially-prepared meals reduced medical costs and improved outcomes in patients with high-acuity nutritional-risk conditions. "We were at a nutrition conference wondering why we didn't get any respect," laughs MANNA's CEO, Sue Daugherty. "We knew we needed to see if our field experience was confirmed by rigorous research."

MANNA raised funding and secured an experienced researcher to design a study that would simulate a controlled study of medical costs for patients receiving and not receiving nutritional support. But the study required the cooperation of a managed-care organization to release and supervise the use of retrospective cost data on the MANNA clients and a comparably ill comparison group. An area managed-care organization saw the potential and agreed to partner in the study.

The study, whose results were published in 2013 in the Journal of Primary Care and Community Health, took three years and was meticulous in its construction of a comparison group and the assembly of medical cost histories. It confirmed that the medically-prescribed meals had significant impact on medical costs and outcomes. Medical costs for the MANNA clients, all at severe nutritional risk, were about $12,000 a month less, and 85 percent of diabetes patients were better able to control their disease.

Seeing these positive results, Bill George, CEO of Health Partners Plans of Philadelphia, committed to full field trial of the program. Health Partners Plans determined an appropriate way to cover the cost through its capitation and provided the administrative support to help MANNA quality "food as medicine" at scale. In 2016, the program gained recognition as a proven value-based innovation from Pennsylvania's secretary of human services, Ted Dallas. That accelerated acceptance among the state's managed-care organizations, one of which operates statewide and has extended the program into rural areas.

Communication among peers also served to spread the idea. The nonprofit nutrition-service agencies that, along with MANNA, form the Food Is Medicine Coalition participated vicariously, monitoring the MANNA pilot and study as they progressed. Sharing among the agencies was a trigger for the Medi-Cal pilot and evaluation.

A final lesson is that risks lurk in success. Public administration is littered with successful pilots that go awry as they are expanded. Experience suggests that the most common pitfalls are two: First is failing to stick to the blueprinted program model by diluting the intensity or duration of the intervention to save money. Second is allowing the program to grow beyond the target group for which it is likely to be significantly beneficial or cost-effective.

The food-as-medicine experiment seems on track to avoid these pitfalls. The managed-care organizations in Philadelphia are the prescribers, and they are paying for a therapy whose financial and medical outcomes they can measure. On the delivery side, MANNA has formed an internal institute to develop protocols for its nutritional services, food and education, and the nonprofit is moving slowly on replication and expansion until it is confident that it can be done while retaining the program's effectiveness. Chicken soup never got that much attention.

NOTE: This column has been corrected to state that another managed-care organization in Philadelphia was MANNA's partner in the initial study. It also corrects the name of Health Partners Plans and clarifies the work it did in incorporating the program into its capitation program.