Comparing characteristics and outcomes between hospitalized adults on a pea protein or dairy/soy protein formulas: initial findings

Enteral nutrition (EN), provided by tube or oral feedings, is a cornerstone of nutritional support for hospitalized patients who cannot meet their nutritional needs [1]. Carefully planned and monitored enteral feeding plans are essential to minimize malnutrition risk and optimize patient outcomes [2,3]. In the hospital setting, malnutrition elevates the risk of infections, pressure injuries, prolonged mechanical ventilation, anemia, compromised cardiac and respiratory functions, leading to increased length of hospital stay and mortality [4–8]. Thus, the availability of an enteral formula to meet nutritional requirements while optimizing tolerance and metabolic support is critical.

The high prevalence of malnutrition following hospitalization (20–45%) highlights the critical need for timely interventions by a registered dietitian to implement nutrition risk screenings, assessments and interventions during and after hospital stays [9]. EN initiated early, ideally within 24–48 h of admission, improves clinical outcomes, especially in critically ill patients [10,11]. Selection of an appropriate enteral formula, however, depends on multiple factors, including individual needs and preferences, underlying medical conditions and gastrointestinal (GI) tolerance [12]. Since the 1960s, dairy protein-based formulas (derived from cow’s milk proteins such as casein and/or whey) have been the standard of care for EN patients as they provide a complete protein source. Soy protein-based formulas later became the first widely used plant-based alternative, also offering a complete source of protein. Initially, these formulations were deliberately fiber-free or “low residue” in an attempt to minimize fecal output; despite updated clinical practice guidelines invalidating this, fiber-free formulas remain in use. Given the prevalence of food allergies and growing evidence of the potential negative environmental effects from heavy reliance on dairy and soy, yellow pea protein has emerged as a possible alternative. Recent studies have investigated a unique fiber-containing, yellow pea protein plant-based formula (PPPBF) and its effects on gastric emptying, GI tolerance, prebiotic activity, production of beneficial gut metabolites and patient and caregiver reported outcomes. This research suggests these formulations may improve GI tolerance, support patient adherence, enhance prebiotic activity and promote moderate gastric emptying (relative to casein- or whey-based formulas), while also providing dietary fiber and a low allergen profile – attributes that address issues historically associated with traditional EN formulas [13–15]. Tailoring nutrition therapy to individual patient needs and preferences is replacing the ‘one-size-fits-all’ approach to EN [16].

There is an increasing body of evidence of healthcare cost benefits related to the timely achievement of an individual’s EN goals. Patients who reach their EN goals may be transferred out of the intensive care unit setting and/or discharged from the hospital sooner resulting in lower costs for the healthcare system [17,18]. However, symptoms of GI intolerance can impede the successful delivery of the enteral formula, delay the patient’s attainment of the goal infusion rate/nutritional requirements, and postpone hospital discharge. The prevalence of enteral feeding intolerance (EFI) is substantial in the hospital setting, estimated to occur in 30.5–65.7% of hospitalized patients (more adults than pediatrics) [19,20]. GI related medications are used to help manage EFI symptoms, which add additional costs. Thus, efforts to find EN formulas to aid in GI tolerance have the potential to improve quality of life as well as minimize healthcare resource utilization (HCRU).

For decades, EN studies centered largely on dairy and soy protein-based formulas, given that these were essentially the main options available (apart from very specialized or elemental amino acid formulas). A gap, therefore, exists in research focused on pea protein plant-based enteral formula use. The primary objective of this study was to use real-world data to assess the differences in health economic outcomes between hospitalized adults who received a unique pea protein plant-based formula (PPPBF, Kate Farms™) and those who received a formula made with dairy and/or soy protein (DSP) formulas. The secondary objective was to compare patient and clinical characteristics and other HEOs between the PPPBF and DSP groups to identify differences to investigate in further analyses. This short report presents initial descriptive, unadjusted findings to inform further research.

Adult inpatients receiving a unique pea protein plant-based formula had a 2-day shorter LOS despite being more ill when compared (unadjusted) to a group of adult inpatients using dairy and/or soy protein formulas in our preliminary study analyses. Despite a higher CCI score, the PPPBF group also had a lower percentage of deaths (in unadjusted comparisons) than the DSP group during index visit and similar percentages of all-cause inpatient readmissions and all-cause outpatient medical visits within 90 days of hospital discharge. This study was an initial step in the research of the users and usage of PPPBF versus DSP formulas in hospitalized adults and associated HEOs. These preliminary findings identified, described and quantified unadjusted differences between groups for further study.

Hospital LOS and readmissions are vital qualitative checkpoints for healthcare systems, including the Hospital Readmission Reduction Program, which penalizes hospitals with high rates of readmissions for certain conditions within 30 days. Facilities can benchmark their performance against others using this metric and share it with the public [24]. Registered dietitians can help to optimize hospital performance metrics by the early identification, treatment and as possible, prevention of malnutrition. Studies have documented the benefits of EN support on improving nutritional status as well as helping to reduce LOS, readmission rates and reduce healthcare costs [25–28].

In the literature, a higher CCI score has been associated with a longer inpatient LOS in adults [29]. In this study, the PPPBF (versus DSP) group was observed to have a higher medical acuity based on a higher median CCI score and a higher prevalence of individuals reporting ≥1 GI intolerance symptom, diagnosis rates of malnutrition, weight loss, irritable bowel syndrome, inflammatory bowel disease and food allergies, and serious comorbidities like ‘malignancy’ and ‘metastatic solid tumor’. Yet this higher acuity in the PPPBF group was not accompanied by a higher percentage of in-hospital deaths, longer LOS or higher hospital readmission rates/outpatient medical visits when compared with the DSP group in preliminary analyses (descriptive, unadjusted comparisons). These outcome differences and potential economic benefits observed between formula groups in this study are noteworthy but should be interpreted with caution until the contribution of the baseline differences identified in this study are tested at the same time. Further analysis is in development.

The role of EN support has been undervalued in the treatment of both acute and chronically ill patients, even though global research has shown that malnutrition and nutrient deficiencies worsen conditions and lead to higher healthcare costs. When intake by mouth is tolerated, the use of oral nutrition supplements (ONS) can avoid the cost and burden of feeding tubes by the patient and healthcare system, and have demonstrated clinical and cost benefits by extending lives of malnourished older hospitalized patients at a low additional cost of $524 per life-year saved in the United States [30]. Additionally, ONS use among hospitalized Medicare patients aged 65 and older is associated with better outcomes and reduced healthcare expenses [31]. Yet in the outpatient/home setting in the United States, ONSs are not reimbursed or covered by the majority of government or private medical insurance policies despite demonstrated clinical and economic benefits [30,31]. EN support, and as possible ONS, are relevant options for providers looking for affordable, evidence-based nutrition strategies [31]. Enteral formula intake as ONS was higher in the PPPBF (versus DSP) group (Table 2), however, it was the primary route of intake for both groups. In the inpatient setting, insurance covers nutritional formula regardless of delivery route, so patient payer types were not relevant to formula ordering, recommendations or prescribing in this study.

Similar to other retrospective studies using hospital administrative databases, our study had some known and suspected limitations. For example, our dataset may be subject to hospital reporting practices, with outliers, missing or incomplete data (e.g., items not billable or reimbursable) and lack clinical details that could impact the characteristics and outcomes studied (e.g., frequency and involvement of registered dietitians, EN used prior to admission, readmissions/outpatient medical visits outside of the hospital system’s database parameters). Our preliminary results focused on outcome differences in the PPPBF versus DSP groups, but future research may perform sensitivity analysis to explore outcome differences in PPPBF versus DSP groups across different formulas (e.g., standard, peptide), age groups, and other relevant subgroups. The PPPBF group is noted to have a smaller and more heterogenous sample than the DSP group but this investigation was the first to evaluate a fiber-containing yellow PPPBF in hospitalized adults, so there was no scientific rationale to restrict analysis of the PPPBF cohort by geographic location; accordingly, the statistical plan included all available PPPBF data across sites. In parallel, strict data governance policies and safe data handling requirements prevented excluding data based on geography or hospital system, ensuring compliance while maximizing the use of all collected data. Finally, because hospital dietetic practices aim to follow uniform, evidence-based nutrition guidelines (which define criteria for initiating oral or tube feedings), one can expect minimal differences in clinical practice patterns between sites and across formula groups, further supporting the decision to analyze an unrestricted, comprehensive dataset.

The descriptive, unadjusted analyses reported in this initial study identified differences in patient characteristics between the PPPBF and DSP formula groups, but these characteristics were not adjusted for in the outcome comparisons. This is an important but expected limitation to these early study results. Further, smaller sample size in one arm may yield lower power in the statistical tests. Future studies involving adjusted or matched analysis will be able to address the difference in patient characteristics between the two groups, and a larger PPPBF sample may improve the power of the tests.

These preliminary study findings help fill a gap in this area of literature, inform further research and develop deeper understanding. These findings are intended to inform our next study which will test group differences identified for an impact on study outcomes and, as appropriate, be controlled for in adjusted analyses (for adjusted exponentiated coefficients, 95% confidence intervals). From this study, we also identified additional data points that would be useful for future prospectively designed studies and those which can help to further control for potential immortal time bias.

This descriptive study provides preliminary results that begin to address an important gap in the literature by highlighting demographic and clinical differences between adults who used a PPPBF as an oral supplement during a hospitalization compared with those who used a DSP formula as an oral supplement or via tube feeding. Additionally, these preliminary results provide a new insight into potential HCRU of PPPBF, used as an ONS, in hospitalized adults requiring nutrition support. Unadjusted comparisons between formula groups indicated that the PPPBF group had a 2-day shorter LOS than the DSP group, despite the PPPBF group presenting with higher indices of medical acuity. This unexpected finding is counterintuitive, given that more acute patients generally have longer hospital stays and has prompted additional research questions. Further research is underway.

V Millovich was responsible for substantial contributions to the design of this work, interpretation of data and writing/editing of the manuscript. IO Shingara was responsible for the interpretation of data, and writing/editing of the manuscript. GS Lopes was responsible for the analysis/interpretation of data, and critical reviewing/editing of the manuscript. M Tyagi was responsible for the analysis of data, and critical review of the manuscript. Z Cao was responsible for contributions to design of this work, analysis/interpretation of data, and critical reviewing/editing of the manuscript. CJ Valentine was responsible for substantial contributions to the design of this work, interpretation of data, and writing/editing of the manuscript. All authors are responsible for the final approval of the manuscript to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

The authors acknowledge Seth Emont PhD for his contributions to the initial design of this study.

Funding to conduct the study was provided by Kate Farms Inc (CA, USA). The publication of the study results was not contingent on the Sponsor's approval or censorship of the manuscript.

V Millovich, IO Shingara, CJ Valentine – salaried employee of Kate Farms Inc.; shareholder. IO Shingara received consultant and speaker fees (unbranded work) from Mead Johnson Nutrition (MJN) and was on the MJN Speakers Bureau. GS Lopes, Z Cao and M Tyagi are employees of Premier, Inc., and Premier, Inc. received payment from Kate Farms to conduct this research. The authors have no other competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was used in the review, writing, or editing of this manuscript.

In accordance with the HIPAA Privacy Rule, disclosed data from the Premier Healthcare Database (PHD) are considered de-identified per US Code of Federal Regulations (CFR) 45 Part 164.514(b)(1) through the ‘Expert Determination’ method. The study was not considered human subjects research and it was determined to be exempt from IRB oversight by a central IRB under US CFR 45 Part 46.106(d)(4).

This manuscript reports the results of a real-world evidence study. The protocol was not publicly registered, and the raw data are not available due to contractual restrictions. A predefined analysis plan is available upon request.

This work is licensed under the Attribution-NonCommercial-NoDerivatives 4.0 Unported License. To view a copy of this license, visit https://creativecommons.org/licenses/by-nc-nd/4.0/