Most non–inborn errors of metabolism (non-IEM) medical foods (MFs) do not meet the regulatory MF definition and lack scientific evidence for safety and efficacy. Non-IEM MFs are not yet ready for reimbursement by public insurers.
Objectives: The use of medical foods (MFs) specifically for non—inborn errors of metabolism (non-IEM) is rising. Concomitantly, evidence for the safety and efficacy of these non-IEM MFs is lacking. We examined the current use and costs of non-IEM MFs and determined whether the scientific evidence supporting their effectiveness and medical utility is adequate to warrant public reimbursement.
Study Design: We employed a qualitative literature review analysis.
Methods: PubMed and MEDLINE databases were searched for all years using relevant keywords, including names of non-IEM MFs identified in the California Workers’ Compensation System (CAWCS) claims dataset from 2011 to 2013. The quality of extracted data was scored with the Delfini Evidence Tool Kit.
Results: Only 2 (3.2%) of 62 studies were conducted with scientific rigor. These 2 studies were for dietary management of Alzheimer disease, which does not have a distinctive nutritional requirement necessitating an MF. Seventy-one percent of the studies of MFs used by patients in the CAWCS were considered to have uncertain validity.
Conclusions: Most reviewed non-IEM MFs lack evidence to support their safety and efficacy. These non-IEM MFs do not abide by FDA draft guidance, as they do not address a distinct nutritional requirement for a disease and yet often have a National Drug Code or “Rx only” label. Consequently, these products do not meet the statutory definition of an MF. We recommend that CAWCS and other payers not provide insurance coverage for non-IEM MFs until more scientific evidence supports their safety, efficacy, and use for nutritional need of a disease.
Am J Manag Care. 2018;24(Spec Issue No. 8):SP273-SP278Takeaway Points
Global sales for medical foods (MFs) are estimated at more than $9 billion and expected to exceed $2 billion in the United States.1,2 This projected growth in estimated sales will likely increase the financial burden in the California Workers’ Compensation System (CAWCS), which currently reimburses for MFs.3 The statutory definition and regulations of MFs were originally defined in the Orphan Drug Act, 21 USC 360ee(b)(3) (1983), to promote the easy accessibility of MFs for those with inborn errors of metabolism (IEMs); IEMs are inherited biochemical disorders in which a specific enzyme defect interferes with metabolism of protein, fat, or carbohydrates. MFs are indispensable for these individuals, and most insurers cover their costs.4 However, the current rise in insurance claims for MFs is primarily for non-IEM MFs, and uncertainty surrounds whether the MFs that are the subjects of these claims actually meet the requirements of an allowed MF and whether they should be reimbursed. Insurers and the public need to understand what MFs are and whether the products currently labeled as MFs actually fit the MF definition, are used appropriately to meet a medical nutritional requirement of a disease, and have scientific evidence of safety and effectiveness.
An MF is defined under the Orphan Drug Act as “a food which is formulated to be consumed or administered enterally under the supervision of a physician and which is intended for the specific dietary management of a disease or condition for which distinctive nutritional requirements, based on recognized scientific principles, are established by medical evaluation.”5 This statutory definition is considered by the FDA to narrowly constrain the types of products that fit within this MF category,6 and the regulations of MFs fall somewhere between those of drugs and foods. MFs are not subject to FDA drug regulatory requirements and are exempt from the labeling requirements for nutrient content claims, but they are mandated to adhere to the requirements concerning the misbranding of foods and food allergen labeling. Although any facility manufacturing or processing MFs must register as a food facility with the FDA, the FDA does not keep a list of MF products. MFs require evidence of a “distinctive nutritional requirement for the dietary management of a disease or condition” and “should be consumed or administered enterally under the supervision of a physician” in order to distinguish them from an overall diet to manage symptoms or reduce disease risk. MFs also must be “specially formulated and processed” as opposed to being naturally occurring food used in a natural state. Additionally, MFs are for patients who require a product as a major component of a disease or condition that requires a specific dietary management not met by ordinary diet alone. Finally, although some evidence of effectiveness, safety, and medical need, which is based on “recognized scientific principles,” must be provided, clinical trials are not required as evidence. The quality of this evidence is unknown.
Our objective is to examine if the quality of scientific evidence supporting the effectiveness and medical need for non-IEM MFs is adequate to warrant public payment. Our specific aim is to identify non-IEM MF use by patients in the CAWCS (P/CAWCS) and to conduct an evidence review to assess the extent and quality of the scientific studies for MFs used by P/CAWCS and of all additional non-IEM MF studies that have claimed to support the need for and use of non-IEM MFs.
To identify how much evidence exists to support the need for and use of non-IEM MFs, we conducted a 2-stage systematic literature search. First, to gain insight into which non-IEM MFs are frequently used, we identified the CAWCS utilization and costs of paid MF claims, using Workers Compensation Information System complete claims datasets from 2011 to 2013 for both physician- and pharmacy-dispensed non-IEM MF claims. Non-IEM MFs were first identified by the National Drug Code (NDC), product name, and RedBook description that indicated MF status (see Table 1 for highlighted MFs). We also identified MFs based on product websites and package insert descriptions. We then coded the MFs by indication.
We then searched the PubMed database across all years using the MF names prescribed to P/CAWCS. Next, we combined this search with a second, broader systematic literature search using both PubMed and MEDLINE databases with the keyword medical food. Other similar terms were avoided to exclude medical supplements or foods.
Study eligibility and inclusion and exclusion criteria were the same for both search stages. The inclusion criterion was the use of a non-IEM MF as an intervention in an experimental comparative study. Experimental comparative studies included both randomized controlled trials and pre- and postintervention studies. Exclusion criteria included (1) no mention of an MF or the term MF, (2) MFs for IEM, (3) animal studies, (4) nonexperimental comparative studies or those not addressing an MF directly (eg, laboratory, assay, or MF composition study), (5) descriptive or opinion article, (6) patient MF case study, or (7) not in English. Two reviewers independently reviewed the full texts of the eligible studies.
We used the Delfini Evidence Tool Kit (DETK), checklists, and evaluation scheme to grade research designs, evidence, findings, and study conclusions (Table 2).7 The DETK assessed the study validity and usefulness of results and identified the strength of the scientific evidence, applied for our purposes here to non-IEM MFs. Checklists determined (1) if study type influenced results, (2) if methods increased chance findings, and (3) whether clinical outcomes were meaningful. The strength of the evidence grade was based on the Agency of Healthcare Research and Quality and Effective Health Care Program group methodology, assessing risk for bias, consistency, directness, and precision between the interventions and examining whether safety, tolerability, and efficacy were discussed. A final Delfini Evidence grade was given for each study, using the Validity and Usability Grading Scale of A, B, B-U, or U, from each reviewer (Table 2). Grade A indicates that evidence is strong and sufficient to make healthcare decisions; grade B indicates that evidence is potentially strong and probably sufficient; grade B-U indicates possible to uncertain usefulness; and grade U indicates sufficient uncertainty to caution against use in clinical decisions. The 2 reviewers discussed any nonmatching scores and came to an agreement for a single score.
MF Product Identification
Non-IEM MFs represent approximately 154,000 out of 20 million total prescriptions and $15.5 million paid by CAWCS across 3 years, or about 51,000 prescriptions and $5.2 million annually. CAWCS claims data indicated that non-IEM MFs were most commonly used for pain or inflammation (52.77%), including Theramine, Trepadone, Proteolin, and Limbrel; and for sleep disorders (31.46%), including Sentra PM, Gabadone, Somnicin, and Sintralyne. Also common were MFs for fatigue (10.53%), primarily Sentra AM; appetite suppression (2.68%), primarily Apptrim; depression (0.37%), primarily l-methylfolate (Foltx, Foltanx, and Metanx); and other indications (2.19%), primarily VSL#3, a probiotic. Also used were Glucerna and Ketonex to increase appetite, Hypertensa for blood pressure, Lipochol for liver health, Lister-V for viral infections, Metafolbic for Alzheimer disease, Percura for neuropathic pain, Pulmona for pulmonary artery pressure, Vayacog for early memory impairment, Virilex for premature ejaculation and impotence, and Cerefolin for memory loss and depression (Table 1).
Strength of Scientific Evidence
The initial literature search was by the product names, described above, of non-IEM MFs that were used by P/CAWCS. This search identified 102 unique studies, and 84 (82%) of the found studies were ultimately excluded: 25 were descriptive or opinion articles, 19 had no MF mentioned, 15 involved only animal testing, and 25 had other exclusions (Figure). Only 18 studies met our inclusion criteria. Many of the non-IEM MFs did not have any published studies to attempt to demonstrate their effectiveness or safety; these non-IEM MFs without support from studies included 2 of the most frequently prescribed MFs in the CAWCS data, Somnicin and Sentra AM. Our findings led to a second, broader search of the keyword medical food identified 157 total unique studies as of March 2016, with only 44 meeting our inclusion criteria. In this search, we only found 3 studies on non-IEM MFs used by P/CAWCS that met our inclusion criteria. Overall, 95 studies (61%) did not fulfill inclusion criteria, leaving 62 total articles (18 articles from the initial search plus 44 articles from the second search) for further evaluation (eAppendix Table [eAppendix available at ajmc.com]).
Evidence for non-IEM MF use in P/CAWCS. We first identified the preliminary evidence strength of the 36 non-IEM MFs used by P/CAWCS. Twelve of the 36 non-IEM MFs (33%) were evaluated by at least 1 study: Deplin, Gabadone, Glucerna, Hypertensa, Limbrel (flavocoxid), l-methylfolate, Metanx, Pulmona, PhenylAde60, Sentra PM, Theramine, and VSL#3. Only 6 (17%) non-IEM MFs used by P/CAWCS had at least 1 study that met our inclusion criteria: Deplin (l-methylfolate), Limbrel (flavocoxid), Metanx, Sentra PM, Theramine, and VSL#3. In our Delfini grade evaluation of just the CAWCS drugs, none of the 21 studies received grade A or B, and only 11 (27%) received a grade B-U. The majority of studies of drugs used by P/CAWCS (73%) received a grade of U, indicating they were of uncertain validity or usefulness.
Scientific evidence from systematic review of all MF studies. Our second literature review found studies that evaluated either a specific MF, other single or combinations of non-IEM MF ingredients, or a nonspecified MF. For example, Faber et al8 evaluated “a medical food, high in protein and leucine, and enriched with fish oil and oligosaccharides” and Sapwarobol9 evaluated grape seed extract without naming a particular MF.
Only 34 (54.8%) of the 62 studies evaluated were randomized controlled studies; the rest were proof-of-concept, open-label, or pilot studies that provided preliminary results. Among the 62 studies, the sample sizes ranged from 8 to 1067 subjects; 29 studies did not have a representative sample or enough subjects. One study did not provide any descriptions of their subjects. Twenty studies (32.3%) either did not have equal treatment of groups or did not have control groups. Finally, only 21 studies (33.9%) fully discussed safety, tolerability, and efficacy.
In our Delfini grade evaluation, none of the 62 studies received grade A and only 2 (3.2%) received grade B. These 2 studies evaluated Souvenaid, a patented combination of fatty acids, vitamins, and other nutrients derived from milk, fish, and soy for dietary management of Alzheimer disease. Although the strength of the evidence was strongest in these 2 studies, no literature exists that states Alzheimer disease has distinctive nutritional requirements. Likewise, the Alzheimer’s Association does not support the use of MFs to treat Alzheimer disease.10
A grade B-U was given to 23 studies (37.1%), primarily due to inadequate sample sizes and other study issues that raised uncertainty. For example, Faber et al,11 which evaluated FortiCare, a nutritionally complete milkshake enriched with n-3 fatty acids for dietary management of cachexia in oncology patients, received a B-U grade because the study was exploratory, lacked a control group, and was inconclusive regarding the MF’s effectiveness.
The 37 studies (60%) that received grade U were proof-of-concept pilots or poorly designed. The strength of the evidence was low due to insufficient sample size, unequal treatment between groups, or lack of a control group (eAppendix Table). For example, Bibiloni et al12 evaluated VSL#3, a probiotic used for ulcerative colitis and irritable bowel syndrome in an open-label study. The sample size was small (34 subjects) and the authors discussed only safety and efficacy, not tolerability.
When evaluating the levels of bias, consistency, directness, and precision (see Table 2 for rubric), we found that none of the studies received a rating of “high,” which indicated low bias and consistency. Only 7 (11%) studies received a “moderate” rating, which indicated that a possible risk of bias, some inconsistency, and some uncertainty existed in the study conclusion. Nineteen (31%) studies were rated as “borderline,” which suggested that “moderate” was not precise enough to address evidence of borderline usefulness. Thirty-six (58%) studies were “inconclusive,” which suggested that either clinically useful evidence was unavailable or a conclusion could not be reached. In summary, there was insufficient scientific evidence to support the use of these non-IEM MFs.
We found high use and cost of non-IEM MFs within CAWCS, supporting the 2010 and 2011 study findings showing increasing use of MFs from 2.3% of medication expenses in 2006 to 12% in 2009.13,14 The high cost of these medical foods is alarming. For example, Theramine, an amino acid blend to reduce inflammation for chronic low back pain, accounts for 45% of non-IEM MF utilization but 63% of total non-IEM MF payments made by CAWCS. Given these high costs, it is crucial to expect adequate scientific evidence underlying these non-IEM MFs. Looking at studies supporting only those drugs used by P/CAWCS, we found, however, that the majority were of uncertain validity or usefulness and the remainder were only of possible to uncertain usefulness. We also demonstrated across our full review that the majority of non-IEM MFs do not meet the FDA definition and labeling requirements of MFs, which were updated in a final Guidance for Industry, confirming and further narrowing constraints on the types of products that are defined as MFs. FDA warning letters15 to manufacturers of MFs also provide direction on the interpretation of this guidance. We found 3 main themes in these warning letters that were supported by our study results: (1) inadequate scientific evidence, (2) no distinctive medical requirement for a disease, and (3) misbranding.
Lack of Scientific Evidence of Safety and Efficacy
Scientific evidence required to meet the current definition of an MF was absent for all of the 36 MFs used by P/CAWCS. Only 21 (33.9%) of 62 studies fully discussed safety, tolerability, efficacy, and potential interaction effects. For example, Somnicin, a combination of natural ingredients to promote effective sleep, was among the most frequently dispensed (4.83%) non-IEM MFs used by P/CAWCS (Table 1). The product contains melatonin, 5-hydroxytryptophan, l-tryptophan, vitamin B6, and magnesium. Because the ingredient 5-hydroxytryptophan is identified as an MF, Somnicin would be classified as an MF.16 However, there are no scientific studies supporting the efficacy of Somnicin. If a patient requires prescription medication for sleep, there are a number of prescription sleep medications that are supported by evidence provided through randomized clinical trials. These medications include but are not limited to eszopiclone, ramelteon, temazepam, triazolam, zaleplon, and zolpidem.
In light of this finding, we recommend that CAWCS not reimburse any of the currently billed non-IEM MFs. If physicians and patients want to use a non-IEM MF that does not provide scientific evidence of need, safety, or effectiveness, that product should be categorized as a medical supplement and paid for out of pocket, as medical supplements are. Alternatively, if the non-IEM MF is intended to cure or mitigate a disease, it should be branded as a drug and be subject to full requirements for FDA drug approval.
Lack of Evidence for Distinctive Nutritional Requirement for Disease States
There is a lack of evidence on whether different diseases have a nutritional need that cannot be provided by diet. Using Somnicin as an example, the Official Disability Guidelines do not recommend the use of Somnicin for pain-induced insomnia.17 An independent medical review in a warning letter concluded, based on the California Medical Treatment Utilization Schedule, that Somnicin is neither medically necessary nor appropriate for the treatment of sleep-induced pain because the condition lacks distinctive nutritional requirements as required for an MF.18 For another example, Souvenaid is marketed as an MF for Alzheimer disease, a condition that does not have distinctive nutritional requirements.10
Mislabeling and Other Violation of Requirements
If an MF does not meet the statutory definitional requirements of an MF, it must be considered either a supplement or a drug. Several FDA warning letters cite products that have been mislabeled and should be called drugs.15 For example, Accera, the manufacturer of Axona (purportedly for Alzheimer disease), was cited in December 2013 for both misbranding and making inappropriate therapeutic claims.19 If the FDA responded to all violations, the manufacturers of all 36 reimbursed MFs within CAWCS would likely receive citations for mislabeling. Given this evidence, we again suggest that non-IEM MFs should not be reimbursed by health systems without more evidence.
We also found that manufacturers of non-IEM MFs do not follow other FDA requirements for the legal designation and sale of a product as a non-IEM MF.20 All 36 of the non-IEM MFs used within CAWCS were designated with an NDC. Identification by an NDC obfuscates the status of MFs and may confuse consumers, physicians, and payers. In addition, 14 of the 36 non-IEM MFs used by P/CAWCS had “Rx only” on the box or container. Finally, according to the FDA, MFs are required to be used under the supervision of a physician, meaning that patients should be receiving ongoing medical supervision.20 Many of these non-IEM MFs are offered on manufacturer websites and are available for sale without requesting physician information, raising questions of whether patients are properly supervised by physicians and receiving ongoing care as they use these products.
We recognize that the 36 non-IEM MFs in the studied sample are targeted toward medical indications, such as pain, depression, and insomnia, specific to a workers’ compensation population who were injured at work. Nevertheless, this sample allowed us to provide an overview of currently available non-IEM MFs by name, indication, and frequency of use within a healthcare system.
To ameliorate the current deficiencies surrounding non-IEM MFs, we suggest implementing an updated evaluation framework, as was implemented for the study of supplements in the past. First, the FDA should make publicly available a list of all MFs that tracks each MF’s utilization, costs, and scientific evidence supporting its efficacy and need. Currently available information, such as the presence of an NDC, “Rx only” label, and promotional information, is confusing to patients, physicians, and payers. The distribution of such a list could accompany educational materials provided to patients, physicians, and payers about non-IEM MFs and their definitional requirements. Second, the FDA should continue to monitor and enforce the misused and mislabeled non-IEM MFs to protect the population from unknown safety risks and from using products that lack scientifically proven efficacy.
Third, health systems should be encouraged to influence the use of non-IEM MFs through coverage rules. We suggest that CAWCS should not reimburse the costs of any of the 36 MFs, all of which currently lack scientific evidence of safety, efficacy, and need. An MF should be reimbursed by a health system only if the following conditions are met: (1) the MF is supported by scientifically rigorous studies; (2) the MF is used to treat a disease state that requires management of distinctive nutritional requirements not met by diet; (3) testimonials and other materials do not indicate that the non-IEM MF can cure or mitigate a disease, and is therefore acting as a drug, without proper approval or labeling; and (4) a physician information is requested to ensure that the MF is used under physician supervision.
Finally, additional independent studies that investigate the efficacy, safety, risks, and benefits of non-IEM MFs are needed. More education materials about non-IEM MFs’ definitional requirements should be provided to patients, physicians, and payers. Given these critical deficiencies of non-IEM MFs, we recommend that (1) the FDA implement a framework for evaluation, and (2) CAWCS and other payers not reimburse costs of non-IEM MFs until more independent scientific evidence supports their safety, efficacy, and nutritional need.Author Affiliations: University of California, San Francisco (LW, TKL, AO, VC), San Francisco, CA.
Source of Funding: California Workers’ Compensation, Department of Industrial Relations.
Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Authorship Information: Concept and design (LW, TKL, AO, VC); acquisition of data (LW, AO, VC); analysis and interpretation of data (LW, TKL, AO, VC); drafting of the manuscript (LW, TKL, VC); critical revision of the manuscript for important intellectual content (LW, TKL, AO); statistical analysis (LW, TKL, VC); provision of patients or study materials (LW, VC); obtaining funding (LW); administrative, technical, or logistic support (LW, TKL); and supervision (LW, TKL).
Address Correspondence to: Leslie Wilson, PhD, University of California, San Francisco, 3333 California St, Ste 420, San Francisco, CA 94118. Email: firstname.lastname@example.org.REFERENCES
1. Stephens G, Campbell S. Scientific requirement in support of medical food claims. Nutraceuticals World website. nutraceuticalsworld.com/issues/2015-09/view_columns/scientific-requirements-in-support-of-medical-foods-claims. Published September 8, 2015. Accessed July 2016.
2. Ghosh D. Medical foods: opportunities in an emerging market. Nutraceuticals World website. nutraceuticalsworld.com/issues/2013-04/view_features/medical-foods-opportunities-in-an-emerging-market. Published April 1, 2013. Accessed February 10, 2015.
3. Ireland J, Swedlow A. The cost and utilization of compound drugs, convenience packs and medical foods in California workers’ compensation. CWCI Research Notes 2010. California Workers’ Compensation Institute website. cwci.org/document.php?file=1337.pdf. Published August 2010. Accessed March 24, 2018.
4. Berry SA, Kenney MK, Harris KB, et al. Insurance coverage of medical foods for treatment of inherited metabolic disorders. Genet Med. 2013;15(12):978-982. doi: 10.1038/gim.2013.46.
5. Orphan Drug Amendment of 1988, PL 100-290, 100th Cong (1988).
6. Code of Federal Regulations Title 21. FDA website. accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/cfrsearch.cfm?fr=101.9. Published 2002. Accessed November 21, 2016.
7. Tools & educational library: critical appraisal tools. Delfini website. delfini.org/delfiniTools.htm#catool. Accessed March 24, 2018.
8. Faber J, Berkhout M, Fiedler U, et al. Rapid EPA and DHA incorporation and reduced PGE2 levels after one week intervention with a medical food in cancer patients receiving radiotherapy, a randomized trial. Clin Nutr. 2013;32(3):338-345. doi: 10.1016/j.clnu.2012.09.009.
9. Sapwarobol S, Adisakwattana S, Changpeng S, Ratanawachirin W, Tanruttanawong K, Boonyarit W. Postprandial blood glucose response to grape seed extract in healthy participants: a pilot study. Pharmacogn Mag. 2012;8(31):192-196. doi: 10.4103/0973-1296.99283.
10. Medical foods. Alzheimer’s Association website. alz.org/documents_custom/statements/Medical_Foods.pdf. Published 2015. Accessed July 2016.
11. Faber J, Uitdehaag MJ, Spaander M, et al. Improved body weight and performance status and reduced serum PGE2 levels after nutritional intervention with a specific medical food in newly diagnosed patients with esophageal cancer or adenocarcinoma of the gastro-esophageal junction: early intervention in esophageal cancer with a specific medical food. J Cachexia Sarcopenia Muscle. 2015;6(1):32-44. doi: 10.1002/jcsm.12009.
12. Bibiloni R, Fedorak RN, Tannock GW, et al. VSL#3 probiotic-mixture induces remission in patients with active ulcerative colitis. Am J Gastroenterol. 2005;100(7):1539-1546. doi: 10.1111/j.1572-0241.2005.41794.x.
13. Swedlow A, Auen E. Current trends in compound drug utilization and cost in the California workers’ compensation system. California Workers’ Compensation Institute website. Published February 2013. Accessed March 24, 2018.
14. Wynn BO. Use of compound drugs, medical foods, and co-packs in California’s workers’ compensation program. RAND Corporation website. rand.org/pubs/working_papers/WR828.html. Published January 2011. Accessed November 21, 2016.
15. Warning letters. FDA website. fda.gov/ICECI/EnforcementActions/WarningLetters/default.htm. Updated February 20, 2018. Accessed August 14, 2016.
16. Maximus Federal Services, Inc. Independent Medical Review final determination letter. State of California Department of Industrial Relations website. dir.ca.gov/dwc/IMR/IMR-Decisions/IMR-Decisions/IMR%20Decisions/IMR2013_10001-20000/CM13-0018109.pdf. Published December 24, 2013. Accessed August 14, 2016.
17. Medical fee dispute resolution findings and decision. Texas Department of Insurance website. tdi.texas.gov/medcases/medfee15/m4152935.pdf. Published May 2015. Accessed November 21, 2016.
18. Medical Treatment Utilization Schedule (MTUS) chronic pain medical treatment guidelines and opioids treatment guidelines. State of California Department of Industrial Relations. dir.ca.gov/dwc/DWCPropRegs/MTUS-Opioids-ChronicPain/MTUS-Opioids-ChronicPain.htm. Published July 28, 2016. Accessed November 24, 2016.
19. Warning letters: Accera, Inc. FDA website. fda.gov/ICECI/EnforcementActions/WarningLetters/2013/ucm381320.htm. Published December 26, 2013. Accessed August 14, 2016.
20. Guidance for industry: frequently asked questions about medical foods; second edition. FDA website. fda.gov/downloads/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/UCM500094.pdf. Accessed March 20, 2018.