Medically tailored meals (MTM) may be beneficial to patients after hospital discharge.
To determine if 2 versus 4 weeks of MTM posthospitalization will improve patient outcomes.
Randomized unblinded trial.
Settings and Participants
Six hundred and fifty patients pending hospital discharge with at least one chronic condition.
One MTM a day for 2 versus 4 weeks.
Main Outcome and Measures
The primary outcome was a change from baseline to 60 days in the Hospital Anxiety Depression Scale (HADS). Secondary outcomes measured change in the Katz activities of daily living (ADLs), DETERMINE nutritional risk, and all-cause emergency department (ED) visits and rehospitalizations.
From baseline to 60 days the HADS anxiety subscale changed 5.4–4.9 in the 2-week group (p = .03) and 5.4–5.3 in the 4-week group (p = .49); the difference in change between groups 0.4 (p = .25). HADS changed 5.4–4.8 in the 2-week group (p = .005) and 5.3–5.1 in the 4-week group (p = .34); the difference in change between groups 0.4 (p = .18). ADL score changed from 5.3 to 5.6 in the 2-week group (p ≤ .0001) and 5.2–5.5 in the 4-week group (p ≤ .0001); the difference in change between groups −0.01 (p = .90). The DETERMINE changed in the 2-week group from 7.2 to 6.4 (p = .0006) and from 7 to 6.7 in the 4-week group (p = .19); the difference in change between groups 0.5 (p = .13). There was no difference in ED visits and rehospitalizations between groups or time to rehospitalization.
Different durations of short-term MTM did not affect patient-centered or utilization outcomes.
Hospitalized patients experience substantial physical and emotional stress due to acute illness and the hospital environment.1 Hospitalization, can further exacerbate social risk factors and impose short-term social challenges for patients when they return home.2 As a result of the 2020 CHRONIC Care Act,3 the Centers for Medicare and Medicaid Services has acknowledged the importance of social risk factors by authorizing Medicare Advantage to offer nonmedical supplemental benefits. Leaders of Medicare Advantage plans also recognized the importance of addressing social determinants to improve the health of their members and potentially decrease unnecessary health care utilization.4 One common supplemental benefit has been the provision of medically tailored meals (MTM) for those with chronic illness. However, evidence is lacking about basic programmatic decisions such as the number of meals and duration of meals to offer.5 For those who have been hospitalized, shopping for food and preparation of healthy meals may be particularly challenging. MTM are approved by a registered dietitian to adhere to evidence-based guidelines for nutritional support in individuals with chronic health conditions. Providing MTM after a hospitalization requires a partnership between a hospital or health care system and a community-based provider of MTM. Research on MTM interventions has shown some benefits of long-term MTM support, such as a reduction in symptoms and rehospitalizations.6-9 Little research has assessed how the duration of meals after a hospitalization might affect patient-centered outcomes such as depression and anxiety10 in a broad range of individuals with chronic health conditions.
In recognition of the trend to provide insurance coverage for social risk factors, and aligning with National Food is Medicine initiatives, we conducted a randomized trial comparing two different durations of MTM delivered to patients with chronic conditions posthospital discharge. Patients were randomized to a weekly delivery of one meal a day for 2 or 4 weeks. We hypothesized that those who received 4 weeks of meals would have reduced depression and anxiety 60 days after hospital discharge, compared to those who received MTM for 2 weeks. Secondary outcomes included rehospitalization and emergency department (ED) visits, functional status, and nutritional risk.
The trial design was an unblinded, randomized controlled trial with participants allocated to 2 versus 4 weeks of MTM in a 1:1 ratio after hospital discharge. A stratified, blocked randomization scheme was used with randomization stratified by hospital. As we explored the feasibility of the study with organizational partners, we learned that many of them felt uncomfortable with a no-meals control group and felt that the question regarding the duration of the intervention was important enough in their decision-making that there would be value in randomizing patients to two different durations of meal delivery (2 and 4 weeks). The Kaiser Permanente Colorado Institutional Review Board approved the study and all participants provided telephone informed consent before enrollment. Although study funding and design took place before the COVID-19 pandemic in Colorado, all patients were enrolled and followed during the pandemic.
Study participants were members of Kaiser Permanente of Colorado (KPCO), an integrated health care delivery system with more than 527,000 members in the rocky mountain west. Study participants received MTM meals from Project Angel Heart (PAH), a community-based organization in the Denver/Boulder metropolitan area that has been providing meals since 1991. Coordination between the three entities, KPCO, PAH, and the hospitals required weekly meetings between the three organizations, collaborative workflow development, multiorganization interaction for tracking participant flow through the study, and continuous feedback to assure of completion of study activities.
Heart failure, regardless of ejection fraction.
Cancer, defined as receiving active or suppressive treatment or as their admission diagnosis.
Chronic obstructive pulmonary disease (COPD).
Chronic liver disease/cirrhosis.
Chronic kidney disease defined as a glomerular filtration rate < 30 mL/min/1.73 m2.
Patients were excluded if: discharged to an institutional setting including a nursing home, assisted living facility, or another residence where meals were institutionally provided; hospitalized for ≤24 h; have a condition/food restriction/allergy that could not be accommodated by the meal provider; already receiving meals from a community-based meal provider; refused the MTM based on their condition(s); or outside the delivery area for the meal provider. Patients were also excluded if they had dementia, pregnant/peripartum, or discharged to hospice care.
Baseline data were collected by interview over the phone at the time of study enrollment. Sociodemographic characteristics were collected from medical records and by patient report. Survey data included specific questions to determine behavioral, social, and functional health. Baseline tools included the Hospital Anxiety Depression Scale (HADS) to assess anxiety and depression; the Katz ADL Survey to assess activities of daily living (ADLs); the DETERMINE Checklist to assess nutritional health; and the Lawton Instrumental Activities of Daily Living (IADLs) Scale to assess functional status.
Regular diet comprised of lean protein; healthy fats, whole grains, one to two cups of vegetables.
Renal diet comprised of low potassium, sodium, and phosphorus; no added sugar and mild spice.
Bland diet comprised of mild vegetables; no onion, garlic, peppers, tomatoes, or other strong seasoning/flavors.
Heart-healthy/diabetic diet comprised of low sodium (<300 mg sodium), reduced saturated fat, and no added sugar.
Comprehensive nutritional content of meals are provided in Supporting Information: Table A.
The diet was selected based on the patient's qualifying chronic condition by the research team. If there was any question regarding the appropriate diet or if the patient had multiple chronic conditions which made the decision difficult, the meal provider's registered dietitian assisted in choosing the most appropriate meal. The meal provider delivered seven frozen meals to each participant, and dependents when requested (one meal per day) for 2 or 4 weeks.
For this study, we included outcomes important to patients and clinicians such as depression, anxiety, and physical function as well as outcomes important to payers such as health care utilization. All outcome measures except hospitalization and ED visits were collected over the phone at 60 days posthospital discharge by the study team. Hospitalization and ED visits were collected from clinical and administrative records in the health care system.
Recognizing that the effects of medically tailored meal interventions are of great interest to health care systems and payers, we elected to define a patient-centered outcome, changes in anxiety and depression, as the primary outcome for the study. Our operational partners confirmed that reductions in depression and anxiety would be meaningful outcomes of an MTM intervention. Anxiety and Depression: The HADS is a well-validated measure for anxiety and depression in the hospitalized population. The HADS is designed for hospitalized patients by avoiding items related to somatic symptoms which can be related to both depression and anxiety and a medical condition and increases the utility for identifying depression and anxiety in medically complex and acutely ill patients.11, 12 The HADS assessed anxiety and depression at baseline and 60 days. The HADS has two subscales, depression: (HADS-D) and anxiety (HADS-A). Each item on the HADS ranges from 0 to 3, with 3 indicating the highest anxiety or depression. Each subscale is summed from 0 to 21 with 0–7 normal, 8–10 moderate, and scores of 11–21 denoting considerable symptoms of anxiety or depression. The sensitivity and specificity for both subscales are approximately 0.8, and validity and reliability have been well established.13 A 1.5–1.7 unit difference is considered a minimal clinically important difference.14, 15
Rehospitalization and ED visits: The utilization outcome for the study was defined as a composite of the total number of hospitalizations (including observational visits) and ED visits at all facilities from the hospital discharge date to 60 days were collected. Time to rehospitalization was defined as the time (in days) from the index hospital discharge date to the next hospital admission or ED visit date.
Functional status: Functional status was measured using the Katz ADL scale and collected at baseline and 60 days. The Katz ADL provides a standardized assessment of functional deficits, particularly for older adults who are at substantial risk of ADL loss with hospitalization. Evidence suggests that individuals can recover ADLs and if they do, they have improved outcomes.16, 17 The minimal clinically important difference for the Katz ADL is approximately 0.5.18
Nutritional health: Nutritional health risk was measured using the DETERMINE Checklist and collected at baseline and 60 days. Scores of ≤2 indicated good nutritional health, 3–5 denoted moderate nutritional risk, and ≥6 indicated high nutritional risk.19
We aimed to detect at least a clinically significant, 1.5-unit difference in change of the HADS over time between the 2- and 4-week meal groups. Sample sizes assumed 80% power, α = .05, a common baseline standard deviation of 4.1 for the anxiety scale, and 3.3 for the depression scale, and a correlation of .3 on within-patient scores. Assuming 320 patients per group, the minimal detectable difference in group changes was 1.1 and 0.9 for the anxiety and depression scales, respectively. Accounting for attrition, we estimated 270 patients per group, which allowed for the detection of differences in change of 1.2 and 0.9 for anxiety and depression, respectively.
For the 60-day readmission/composite outcome, assuming 80% power, α = .05, the two-sided test of relative risk, and a control rate of 31%, the comparison between the 2- and 4-week intervention groups with 320 patients could detect a relative risk of at least 0.661, after accounting for anticipated attrition (270 patients in each group).
Randomization and blinding
A stratified, block randomization scheme was used with randomization stratified by hospital. Randomization was generated by the statistician and concealed; blinding was not possible.
Analysis proceeded according to an intention-to-treat approach. We compared patients in the 2- and 4-week groups on age, sex, race/ethnicity, length of baseline hospitalization, using frequencies and means. For the HADS analyses, we assessed the normality of the distributions of scores and change scores. A linear mixed model was used to analyze the total HADS change scores, and the depression and anxiety subscales separately, assessing residuals for normality. We estimated an unadjusted model to assess the difference in scale score change between baseline and follow-up between intervention groups. Individual HADS scale items were imputed with the mean when a patient answered 50% or more of the items20 and then the individual items were imputed with the possible item values using 10 generated data sets using joint-modeling multilevel multiple imputations for accounting for repeated measurements per person.21 Unadjusted linear mixed models with treatment group, assessment time point, their interaction, and a random subject intercept were fitted to each imputed data set for the total HADS score, and the depression and anxiety subscales, and combined using Rubin's rules. To assess the difference in 60-day rehospitalization and ED encounters between 2- and 4-week meal delivery and control, we estimated Cox regression models using the time to first rehospitalization/ED encounter, censoring at 60 days, or other loss to follow-up event. We estimated unadjusted and adjusted models using patient characteristics for covariate adjustment, and a quick assessment of the proportional hazard assumption using martingale residuals was conducted for the intervention comparison. All analyses were conducted using SAS 9.4 (SAS Institute) and R (R Core Team, 2019).
A total of 4344 patients were approached and 3694 were excluded from the study (Figure 1). Randomization occurred for 650 patients, 325 randomized to each group. The majority of participants were aged ≥65 years and slightly more women than men. Approximately 30% of participants lived in areas of high social disadvantage22 and 23% reported food insecurity. The most common chronic condition in both groups was diabetes, followed by heart failure. The baseline characteristics of the two groups are provided in Table 1.
|Baseline participant characteristics||Meal duration|
|2 weeks, n = 325||4 weeks, n = 325|
|Age, year (mean, median)||65.4, 67.3||66.0, 67.5|
|Age categories, n (%)|
|18–64||136 (41.8)||143 (44)|
|65–74||107 (32.9)||93 (28.6)|
|75+||82 (25.2)||89 (27.4)|
|Female, n (%)||173 (53.2)||174 (53.5)|
|Race, n (%)|
|White||207 (63.7)||195 (60)|
|Black||23 (7.1)||36 (11.1)|
|Hawaiian/Pacific islander||1 (0.3)||2 (0.6)|
|Asian American||4 (1.2)||3 (0.9)|
|Native American||5 (1.5)||3 (0.9)|
|Other||17 (5.2)||21 (6.5)|
|Unknown||68 (20.9)||65 (20.0)|
|Ethnicity, n (%)|
|Hispanic||80 (24.6)||77 (23.7)|
|Not Hispanic||240 (73.8)||242 (74.5)|
|Unknown||5 (1.5)||6 (1.8)|
|Marital status, n (%)|
|Married/domestic, living with partner or in a serious or committed relationship but not living together||164 (50.4)||170 (52.4)|
|Single||63 (19.4)||51 (15.7)|
|Separated or divorced||42 (12.9)||55 (16.9)|
|Widowed||47 (14.5)||43 (13.2)|
|Highest grade or level of school, n (%)|
|8th grade or less, or some high school but did not graduate||17 (5.3)||27 (8.3)|
|High school graduate or GED||93 (28.6)||95 (29.2)|
|Some college or 2-year degree||107 (32.9)||108 (33.2)|
|4-year college graduate (BA, BS, etc.)||56 (17.2)||46 (14.2)|
|More than a 4-year college degree||43 (13.2)||43 (13.2)|
|Medical conditions, n (%)a|
|Liver disease||20 (6.2)||21 (6.5)|
|COPD||61 (18.8)||59 (18.2)|
|Cancerb||76 (23.4)||80 (24.6)|
|Chronic kidney diseasec||51 (15.7)||47 (14.5)|
|Heart failured||92 (28.3)||115 (35.4)|
|Diabetes mellitus||193 (59.4)||190 (58.5)|
|Covid infectione||39 (12.0)||40 (12.3)|
|Number of chronic condition/participant, n (%)a|
|1 condition||172 (52.9)||159 (48.9)|
|2 conditions||110 (33.8)||117 (36.0)|
|3 conditions||32 (9.8)||38 (11.7)|
|≥4 conditions||11 (3.4)||11 (3.4)|
|Length of hospital stay (mean, SD)||5.7 (4.3)||5.1 (3.7)|
|Home health care order, n (%)||38 (11.7)||31 (9.5)|
|CCI score—mean (SD)f||4.5 (2.6)||4.8 (2.8)|
|BMI—mean (SD)||31.2 (7.9)||31.4 (8.1)|
|ICU stay, n (%)||47 (14.5)||34 (10.5)|
|Medicare||212 (65.2)||202 (62.2)|
|Medicaid||19 (5.8)||22 (6.8)|
|Traditional HMO||18 (5.5)||21 (6.5)|
|DHMO (deductible HMO) or HDHP||66 (20.3)||74 (22.8)|
|Other||10 (3.1)||6 (1.8)|
|Area Deprivation Index (mean, SD)h||5.5 (2.4)||5.3 (2.4)|
|Categories of disadvantage, n (%)|
|Low (1–3)||77 (23.7)||76 (23.4)|
|Moderate (4–6)||134 (41.2)||150 (46.2)|
|High (7–10)||113 (34.8)||98 (30.2)|
|Adherence to intervention, n (%)i|
|All intervention (all meals delivered)||294 (90.5)||258 (79.4)|
|Partial intervention (some meals delivered)||16 (4.9)||57 (17.5)|
|None (no meals delivered)||15 (4.6)||10 (3.1)|
|Participants who also received meals for household members, n (%)||106 (47)||121 (53)|
|Which of the following best describes your current living situation?, n (%)|
|Live alone in my own home||80 (24.6)||75 (23.1)|
|Live in a household with other people||232 (71.4)||239 (73.5)|
|Temporarily staying with a relative or friend||1 (0.3)||5 (1.5)|
|Other||3 (0.9)||1 (0.3)|
|Are you a primary caregiver for, n (%)|
|Child under the age of 18||29 (8.9)||22 (6.8)|
|A person who is frail, chronically ill, or has a physical or mental disability||24 (7.4)||30 (9.2)|
|Both||0 (0.0)||1 (0.3)|
|Neither||272 (83.7)||272 (83.7)|
|HADS total (mean, SD)||10.8 (6.4)||10.8 (6.7)|
|Anxiety subscale||5.4 (3.8)||5.4 (4.0)|
|Depression subscale||5.4 (3.7)||5.3 (3.6)|
|Lawton independent activities of daily living, total (mean, SD)||7.1 (1.5)||7.2 (1.4)|
|Independent with the telephone, n (%)||303 (93.2)||315 (96.9)|
|Independent with shopping, n (%)||236 (72.6)||241 (74.2)|
|Independent with food preparation, n (%)||249 (76.6)||261 (80.3)|
|Independent with housekeeping, n (%)||216 (66.5)||220 (67.7)|
|Independent with laundry, n (%)||256 (78.8)||258 (79.4)|
|Independent with transportation, n (%)||252 (77.5)||257 (79.1)|
|Independent with medications, n (%)||290 (89.2)||301 (92.6)|
|Independent with finances, n (%)||281 (86.5)||287 (88.3)|
|Katz activities of daily living (mean, SD)||5.3 (1.1)||5.2 (1.4)|
|Independent in bathing, n (%)||259 (79.7)||268 (82.5)|
|Independent in dressing, n (%)||274 (84.3)||281 (86.5)|
|Independent in transfers, n (%)||269 (82.8)||278 (85.5)|
|Independent in toileting, n (%)||289 (88.9)||276 (84.9)|
|Continent, n (%)||253 (77.8)||240 (73.8)|
|Independent in feeding, n (%)||311 (95.7)||311 (95.7)|
|DETERMINE nutrition checklist (mean, SD)||7.2 (4.0)||7.0 (4.0)|
|I have an illness or condition that made me change the kind and/or amount of food I eat, n (%)||195 (60.0)||167 (51.4)|
|I eat fewer than two meals per day, n (%)||86 (26.5)||102 (31.4)|
|I eat few fruits or vegetables or milk products, n (%)||135 (41.5)||129 (39.7)|
|I have three or more drinks of beer, liquor, or wine almost every day, n (%)||12 (3.7)||11 (3.4)|
|I have tooth or mouth problems that make it hard for me to eat, n (%)||64 (19.7)||66 (20.3)|
|I don't always have enough money to buy the food I need, n (%)||51 (15.7)||59 (18.2)|
|I eat alone most of the time, n (%)||136 (41.8)||128 (39.4)|
|I take three or more different prescribed or over-the-counter drugs a day, n (%)||274 (84.3)||263 (80.9)|
|Without wanting to, I have lost or gained 10 pounds in the last 6 months, n (%)||181 (55.7)||166 (51.1)|
|I am not always physically able to shop, cook and/or feed myself, n (%)||96 (29.5)||96 (29.5)|
|Food insecure in past 12 months, n (%)j|
|Yes||75 (23.1)||78 (24.0)|
|No||241 (74.2)||242 (74.5)|
|How often do you feel lonely or isolated from those around you?, n (%)|
|Never/rarely||194 (59.6)||208 (64.0)|
|Sometimes||89 (27.4)||85 (26.2)|
|Often/always||33 (10.2)||26 (8.0)|
|Has the lack of transportation kept you from meetings, work, or getting things needed for daily living?, n (%)|
|Yes||34 (10.5)||40 (12.3)|
|No||282 (86.8)||280 (86.2)|
|How hard is it for you to pay for the very basics like food, housing, medical care, and heating?, n (%)|
|Very hard/hard||31 (9.5)||44 (13.6)|
|Somewhat hard||74 (22.8)||85 (26.2)|
|Not very hard/Not hard at all||211 (65.0)||191 (58.7)|
|In the past 12 months, was there a time when you did not have a steady place to sleep or slept in a shelter?, n (%)|
|Yes||8 (2.5)||7 (2.2)|
|No||308 (94.8)||311 (95.7)|
- Abbreviations: BMI, body mass index; CCI, Charlson Comorbidity Index; COPD, chronic obstructive pulmonary disease; GED, general equivalency diploma; HADS, Hospital Anxiety Depression Scale; HDHP, high deductible health plan; HMO, Health Maintenance Organization; SD, standard deviation.
- a Medical conditions: Every qualifying condition included for each participant.
- b Cancer inclusion criteria: If participant has cancer as admission diagnosis, is on active or suppressive treatment.
- c Chronic kidney disease inclusion criteria: lab glomerular filtration rate < 30 mL/min/1.73 m2.
- d Heart failure: All heart failure included regardless of ejection fraction.
- e COVID: Intervention group: Positive lab within 30 days of admission. Control Group: Positive lab within 30 days and/or COVID admission diagnosis.
- f CCI: The total score in the CCI is derived by summing the assigned weights of all comorbid conditions. Severity of comorbidity is categorized into three grades: mild, with CCI scores of 1–2; moderate, with CCI scores of 3–4; and severe, with CCI scores ≥5 score.
- g Insurance payer—indicates enrollment in a type of insurance coverage. Patients may have been enrolled in multiple coverages during the study. This field represents assignment according to a hierarchy where Medicare is assigned first if this coverage existed then Medicaid, and so on. Each patient would then only be assigned one type of coverage.
- h Area Deprivation Index: State rankings of neighborhoods by socioeconomic disadvantage in a region of interest, validated to the census block group level.
- i Adherence to intervention: All: All scheduled meals delivered, Partial: Some scheduled meals delivered, None: No scheduled meals delivered.
- j Food insecurity criteria: Responded “Often True” or “Sometimes True” to one or both of the following questions—(1) “Within the past 12 months, you worried that your food would run out before you got money to buy more.” (2) “Within the past 12 months, the food you bought just didn't last and you didn't have money to get more.”
In the 2-week group, 90.5% adhered by receiving all meal deliveries, compared to 79.4% in the 4-week group. In the 2-week group, 4.9% of the participants received some deliveries and 4.6% received no deliveries. In the 4-week group, 17.5% received some deliveries and 3.1% received no deliveries (between group adherence p < .0001).
Table 2 shows the difference between groups in the total HADS, HADS subscales, ADLs, and DETERMINE checklist from baseline to 60 days. Although the 2-week group experienced a statistically significant reduction in HADS score between baseline and 60 days, the reduction in HADS scores was not significant in the 4-week group, and the difference in change scores between groups was not significantly different. The proportion of participants who were independent in ADL increased slightly in both groups, but again the difference between groups was not significant. The 2-week group experienced a small decrease in nutritional risk as measured by the DETERMINE checklist. There was no difference in rehospitalizations or ED visits between groups or time to rehospitalization (Figure 2). Subgroup analyses according to diabetes, heart failure, and age (≥65) did not indicate significant differences between groups (data not shown).
|2-week group||4-week group||Modeled difference in changea||p Value for difference in change 2 versus 4 weeks|
|Outcomes||Baseline score||60 days score||Change||p Value||Baseline score||60 day score||Change||p Value|
- Abbreviations: ADLs, activities of daily living; HADS, Hospital Anxiety Depression Scale.
- a Modeled differences take into account the correlated nature of the data, due to repeated measurements per person, using mixed modeling. Missing values for HADS Total, anxiety subscale, and depression subscale scores have been imputed.
In this study of primarily older and socially disadvantaged individuals from an integrated health care system, we found no difference in outcomes between receipt of MTMs for 2 versus 4 weeks. Although most participants received the full course of meals for their group, the 4-week group demonstrated an increase in meal delivery cancellations, indicating that flexibility in meal duration after hospitalization may be warranted.
For individuals with chronic illness, observational studies and small randomized trials have found a correlation between the delivery of social services and reductions in avoidable hospitalizations.23, 24 Many studies have linked food insecurity and poor nutrition to increased health risks and chronic disease,10, 25-34 and proper nutrition to improved health outcomes.35-39 Studies of longer-duration MTM programs have shown benefits. A recent randomized trial found no difference in all-cause readmissions between individuals with chronic conditions randomized to 10 weeks of MTM or no supplemental meals. Subgroup analysis of patients with heart failure in that trial found a decrease in rehospitalizations and mortality for those receiving MTM.8 Subgroup analysis for our trial did not demonstrate a difference in outcomes for heart failure or other chronic conditions. A nonrandomized study of participants who were dually eligible for Medicaid/Medicare compared 6 months of daily meals in three groups—MTMs versus nontailored meals versus matched controls. Individuals in the MTM program showed a decrease in ED visits and hospitalizations compared with matched controls and a reduction in ED visits for MTM compared to nontailored meals.9
Food insecurity or evidence of poor nutrition are common among individuals at the time of hospitalization.40 These individuals may also experience other basic resource needs (inadequate or unsafe housing, transportation barriers, social isolation, etc.) that compound their difficulty in obtaining sufficient or healthy food. Thus, home-delivered meals may free up financial resources to address other social needs, reduce social isolation, and reduce transportation demands on the patient and their family. Home-delivered, MTM designed to address the dietary needs of patients with specific chronic health conditions may be even more beneficial in some circumstances.41 Hospital discharge is a vulnerable period for patients with chronic illness as they are recovering from acute illness, coping with increased functional impairments, and often returning to self-care in their home environment. Vaudin et al. demonstrated that in older adults, food-related anxiety significantly spiked during hospitalization and anxiety declined after receiving meals for 2 months. Some of the associated characteristics for those with food anxiety during hospitalization were Hispanic ethnicity, smoking, diabetes, unable to cook, and eating alone.10 Measurement of the HADS as the primary outcome allowed the trial to examine these intertwined issues.
The study demonstrated successful collaboration between health plan investigators, a contracted hospital system, and a community-based MTM provider to address a question of operational importance to health system leaders. The research team successfully integrated workflows between the three organizations to identify, track, and collect outcome data on enrolled patients. The COVID-19 pandemic did affect the conduct of the study, since it required all study activities to be conducted remotely, hampered the ability of hospital personnel to assist in identification of potential participants, and necessitated the meal provider to adopt a no-contact approach to meal delivery which minimized social contact with the participant.
This study had several limitations:
We did not enroll a comparison group that received no meals. Since organizational leaders found it difficult to envision randomization to a “no meals” option, we focused the study on meal delivery duration.
The baseline HADS scores were below the established cutoff for anxiety and depression (mean <7), which reduced our ability to detect significant differences between groups in the primary study outcome.
The MTM were prepared for in-home consumption, but there was no guarantee that the participant consumed the meals.
Participants represented a clinically heterogenous population of patients. Individuals discharged to an institutional setting (such as short-term postacute care) were excluded since the institution would be providing meals. This exclusion likely contributed to the cohort having higher baseline functioning, as measured by ADLs and IADLs at hospital discharge. Additionally, we excluded those already receiving home meal delivery services which may have decreased the neediest population of patients in the study. Certain subgroups of patients, such as those at higher risk of adverse behavioral health outcomes or rehospitalization, may have shown different results from the group as a whole.
Although the DETERMINE Checklist indicated that the study population was at increased nutritional risk (baseline ≥6), we did not objectively identify malnutrition among participants. It is unlikely that one meal a day in the short term would be sufficient to address malnutrition.
Finally, the COVID-19 pandemic began immediately after protocol development for this study. A rapid adjustment was required to conduct all aspects of the study virtually to ensure participant and research staff safety. The COVID-19 pandemic may also have affected study enrollment, the patient's experience of acute illness and hospital care, and needs surrounding food access and nutrition.
In conclusion, patients with one or more chronic health conditions who received 2 versus 4 weeks of MTM did not show a difference in self-reported or utilization outcomes. Determining the most appropriate duration of MTM after hospital discharge for patients with chronic illnesses will require further investigation.
The authors would like to thank members of the study team, including Ruth Bedoy and Courtney Ripley (also from the Institute for Health Research, Kaiser Permanente Colorado) for their extensive efforts in participant recruitment, enrollment, and the conduct of participant surveys for this study. We also acknowledge the generous support of SCL Health staff from the two participating hospitals and the team from Project Angel Heart. This work followed the Guidelines for the 2010 CONSORT Checklist (see Supporting Information).
CONFLICT OF INTEREST STATEMENT
This work was funded by Kaiser Permanente internal funds however Kaiser Permanente did not have a role in the conduct of the study including data collection, study management, analysis, interpretation of the data or manuscript preparation.
The Kaiser Permanente Colorado Institutional Review Board approved the study and all participants provided telephone informed consent prior to before enrollment.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
|jhm13084-sup-0001-CONSORT_2010_Checklist_JHM_v2.doc237 KB||Supporting information.|
|jhm13084-sup-0002-Supplement_File.pdf105 KB||Supporting information.|
Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
- 1. Post-hospital syndrome-an acquired, transient condition of generalized risk. N Engl J Med. 2013; 368(2): 100-102. doi:10.1056/NEJMp1212324
- 2, , , et al. Association between self-reported health-related social needs and acute care utilization among older adults enrolled in medicare advantage. JAMA Health Forum. 2022; 3(7):e221874. doi:10.1001/jamahealthforum.2022.1874
- 3H.R.1892—Bipartisan Budget Act of 2018, Pub. L. No. 115-123, (2017-2018). 2018. Accessed August 16, 2022. https://www.congress.gov/bill/115th-congress/house-bill/1892
- 4, , , et al. Perspectives of Medicare Advantage Plan representatives on addressing social determinants of health in response to the CHRONIC Care Act. JAMA Netw Open. 2019; 2(7):e196923. doi:10.1001/jamanetworkopen.2019.6923
- 5, , , , , . How Medicare Advantage plans use data for supplemental benefits decision-making. Am J Manag Care. 2022; 28(4): e132-e139. doi:10.37765/ajmc.2022.88866
- 6, , , et al. Nutritional intervention in malnourished hospitalized patients with heart failure. Arch Med Res. 2016; 47(7): 535-540. doi:10.1016/j.arcmed.2016.11.005
- 7, , , et al. Home-delivered meals postdischarge from heart failure hospitalization. Circ Heart Fail. 2018; 11(8):e004886. doi:10.1161/CIRCHEARTFAILURE.117.004886
- 8, , , et al. Effect of medically tailored meals on clinical outcomes in recently hospitalized high-risk adults. Med Care. 2022; 60(10): 750-758. doi:10.1097/mlr.0000000000001759
- 9, , , , , . Association between receipt of a medically tailored meal program and health care use. JAMA Intern Med. 2019; 179(6): 786-793. doi:10.1001/jamainternmed.2019.0198
- 10, . Food anxiety is associated with poor health status among recently hospital-discharged older adults. J Nutr Gerontol Geriatr. 2015; 34(2): 245-262. doi:10.1080/21551197.2015.1035825
- 11, . The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983; 67(6): 361-370. doi:10.1111/j.1600-0447.1983.tb09716.x
- 12. The hospital anxiety and depression scale. Health Qual Life Outcomes. 2003; 1:29. doi:10.1186/1477-7525-1-29
- 13, , , . The validity of the hospital anxiety and depression scale. J Psychosom Res. 2002; 52(2): 69-77. doi:10.1016/s0022-3999(01)00296-3
- 14, , , . Establishing the minimal clinically important difference for the Hospital Anxiety and Depression Scale in patients with cardiovascular disease. J Cardiopulm Rehabil Prev. 2019; 39(6): E6-E11. doi:10.1097/HCR.0000000000000379.
- 15, , , . The minimal important difference of the hospital anxiety and depression scale in patients with chronic obstructive pulmonary disease. Health Qual Life Outcomes. 2008; 6:46. doi:10.1186/1477-7525-6-46
- 16, , , et al. Hospitalization-associated change in gait speed and risk of functional limitations for older adults. J Gerontol A Biol Sci Med Sci. 2019; 74(10): 1657-1663. doi:10.1093/gerona/glz027
- 17, , , et al. Sequence of functional loss and recovery in nursing homes. Gerontologist. 2016; 56(1): 52-61. doi:10.1093/geront/gnv099
- 18, , , , , . Minimal important change and minimal detectable change in activities of daily living in community-living older people. J Nutr Health Aging. 2017; 21(2): 165-172. doi:10.1007/s12603-016-0797-8
- 19, , , . Nutrition and health risks in the elderly: the nutrition screening initiative. Am J Public Health. 1993; 83(7): 972-978. doi:10.2105/ajph.83.7.972
- 20, , , . Handling missing items in the Hospital Anxiety and Depression Scale (HADS): a simulation study. BMC Res Notes. 2016; 9(1): 479. doi:10.1186/s13104-016-2284-z
- 21, , . Jomo: a flexible package for two-level joint modelling multiple imputation. R Journal. 2019. 1 11(2): 205-228.
- 22, , , , . Introduction of an area deprivation index measuring patient socioeconomic status in an integrated health system: implications for population health. EGEMS (Washington, DC). 2016; 4(3):1238. doi:10.13063/2327-9214.1238
- 23, , , , . Effective management of high-risk medicare populations. Avalere Health LLC; 2014.
- 24, , , , , . Examining health care costs among MANNA clients and a comparison group. J Prim Care Community Health. 2013; 4(4): 311-317. doi:10.1177/2150131913490737
- 25, , , . Food insecurity is related to increased risk of type 2 diabetes among Latinas. Ethn Dis. 2011; 21(3): 328-334.
- 26, . Food insecurity and health outcomes. Health Aff. 2015; 34(11): 1830-1839. doi:10.1377/hlthaff.2015.0645
- 27, , , et al. Household food insecurity is associated with adult health status. J Nutr. 2004; 134(9): 2330-2335. doi:10.1093/jn/134.9.2330
- 28, , . Diet quality as assessed by the healthy eating index, alternate healthy eating index, dietary approaches to stop hypertension score, and health outcomes: an updated systematic review and meta-analysis of cohort studies. J Acad Nutr Diet. 2018; 118(1): 74-100.e11. doi:10.1016/j.jand.2017.08.024
- 29, , , , . Improvements in US diet helped reduce disease burden and lower premature deaths, 1999-2012; overall diet remains poor. Health Aff. 2015; 34(11): 1916-1922. doi:10.1377/hlthaff.2015.0640
- 30, , , et al. Diet quality and major chronic disease risk in men and women: moving toward improved dietary guidance. Am J Clin Nutr. 2002; 76(6): 1261-1271. doi:10.1093/ajcn/76.6.1261
- 31, , . Food insecurity is associated with chronic disease among low-income NHANES participants. J Nutr. 2010; 140(2): 304-310. doi:10.3945/jn.109.112573
- 32, , , et al. Effect of food insecurity on chronic kidney disease in lower-income Americans. Am J Nephrol. 2014; 39(1): 27-35. doi:10.1159/000357595
- 33, , , . Chronic physical and mental health conditions among adults may increase vulnerability to household food insecurity. J Nutr. 2013; 143(11): 1785-1793. doi:10.3945/jn.113.178483
- 34, , , , . Food insecurity is inversely associated with diet quality of lower-income adults. J Acad Nutr Diet. 2014; 114(12): 1943-1953.e2. doi:10.1016/j.jand.2014.06.353
- 35, , , et al. A pilot food bank intervention featuring diabetes-appropriate food improved glycemic control among clients in three states. Health Aff. 2015; 34(11): 1956-1963. doi:10.1377/hlthaff.2015.0641
- 36, , , et al. Intervention with delivery of diabetic meals improves glycemic control in patients with type 2 diabetes mellitus. J Clin Biochem Nutr. 2008; 42(1): 59-63. doi:10.3164/jcbn.2008010
- 37, , , . The effect of medical nutrition therapy by a registered dietitian nutritionist in patients with prediabetes participating in a randomized controlled clinical research trial. J Acad Nutr Diet. 2014; 114(11): 1739-1748. doi:10.1016/j.jand.2014.07.020
- 38, , , et al. Comprehensive and medically appropriate food support is associated with improved HIV and diabetes health. J Urban Health. 2017; 94(1): 87-99. doi:10.1007/s11524-016-0129-7
- 39. Meal programs improve nutritional risk: a longitudinal analysis of community-living seniors. J Am Diet Assoc. 2006; 106(7): 1042-1048. doi:10.1016/j.jada.2006.04.023
- 40, , , , . Association of social determinants of health and their cumulative impact on hospitalization among a national sample of community-dwelling US adults. J Gen Intern Med. 2022; 37(8): 1935-1942. doi:10.1007/s11606-021-07067-y
- 41, , , et al. Meal delivery programs reduce the use of costly health care in dually eligible Medicare and Medicaid beneficiaries. Health Aff. 2018; 37(4): 535-542. doi:10.1377/hlthaff.2017.0999