Carroll A. Evidence Suggest That Meal Assistance Programs Do More Than “Sound Good”. JAMA Health Forum. Published online April 25, 2018. doi:10.1001/jamahealthforum.2018.0017
About a year ago, Mick Mulvaney, director of the Office of Management and Budget (OMB), got himself into a bit of trouble when he argued that cuts to the Community Development Block Grant program were justifiable because those programs were “just not showing any results.” The example he used—Meals on Wheels—as programs we can’t fund “just because they sound good” elicited a significant amount of pushback, even from me.
A new study argues that such programs are even better than we thought.
At the time he made the comments, the defense raised against Mulvaney’s assertions was that a large amount of data shows that interventions like Meals on Wheels do work. A 2014 systematic review, for instance, included 8 studies, 2 of which were randomized controlled trials, and reported that meal assistance programs increased the nutrient intake and quality of recipients’ diets, while reducing food insecurity and nutritional risk.
A 2015 review article found even more studies in support of such programs, but many of these were of lesser quality. Clearly more, and better, research is needed. This is especially true because, while such programs have the potential to also return economic benefits, their cost-effectiveness is not well understood.
Food insecurity is all too common, affecting affecting 14% of US residents. It’s believed to be associated with more than $77 billion in additional health care expenditures each year.
A recent study, however, makes the argument that meal assistance programs may be not only cost-effective but also may be cost saving as well.
The study looked at 2 types of programs, both of which were given to adults who were dually eligible for Medicaid and Medicare as members of the Commonwealth Care Alliance, a nonprofit health plan in Massachusetts. The first was a medically tailored meal program. Adults who needed this program had specific dietary needs because of chronic conditions like diabetes or renal insufficiency. The program, which delivered 5 days’ worth of lunches, dinners, and snacks, had 17 different tracks of tailored diets.
The second program was a nontailored program, more along the lines of Meals on Wheels. It delivered 5 days of lunches and dinners each week, usually on a daily basis.
Patients who received at least 6 months of one of these programs in 2014 through 2015 were eligible for the study. For each participant, random beneficiaries were chosen and matched to them as much as possible. The factors considered for matching included medical spending in the year before the study, insurance, comorbidities, risk score, age, ethnicity, sex, poverty, and prescriptions for drugs that indicated significant illnesses such as HIV infection or cardiovascular disease. Separate cohorts of matched controls were created both for the tailored and nontailored meal programs.
What makes this study different from many that have come before, however, is that it focused on utilization and spending. The researchers wanted to see if these tailored meal programs might prevent costly events. The main outcome of interest was the number of emergency department visits. Secondary outcomes included inpatient admissions and emergency transportation use. On top of these, data were also collected on medical spending as a whole.
The analysis of the tailored program included 133 participants and 1002 matched controls; for the nontailored program, it included 624 participants and 1318 matched controls. Many were followed up for 2 years or more.
Compared with controls, those who participated in the tailored meals programs were significantly less likely to visit the emergency department, be admitted to the hospital, or use emergency transportation services. Those in the nontailored meal programs were also significantly less likely to visit the emergency department or use emergency transportation, but no significant differences were seen for inpatient admissions.
Furthermore, participation in these programs was associated with lower medical spending. In the tailored program, participants spent $843, $570 less per month than a matched control group. Even in the nontailored program, medical spending was lower for participants, who spent $1007, $156 less per month than those in a control group.
Because the tailored meal program cost $350 a month, this meant that being in the program actually saved the Commonwealth Care Alliance $220 per participant. The nontailored program was cheaper to administer, at $146 per month, so it also saved the Alliance $10 per month. Further statistical analysis found that these results were robust.
As with any such study, limitations exist, and we should consider them when we interpret the generalizability of the findings. This was not a randomized controlled trial, and it’s possible that unmeasured factors are at work here. The researchers did, however, try to control for as many factors as they could and also conducted a number of analyses. All participants were also dually eligible for Medicare and Medicaid, and the results might not apply as well to the population as a whole. They all lived in relatively concentrated urban areas, and the programs might not be as cost-effective in rural areas of the country.
These results are compelling, though. In general, we seem more willing to spend money in the health care system on expensive treatments rather than on prevention. Even more often, prevention—while improving outcomes—does not save money. These programs seem to do both.
This was not a comprehensive cost-effectiveness analysis, taking into account all inputs and outputs. But it showed that nutritional assistance programs like these, especially for those who are medically fragile, might at worst break even and at best save money. They might do all that while preventing hospitalizations and emergency department visits, providing home-bound people with human contact, and making sure they get enough to eat. These programs don’t just “sound good.” They appear to “be good” as well.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.