Dietary assessment methods for measurement of oral intake in acute care and critically ill hospitalised patients: a scoping review

Abstract

Quantification of oral intake within the hospital setting is required to guide nutrition care. Multiple dietary assessment methods are available, yet details regarding their application in the acute care setting are scarce. This scoping review, conducted in accordance with JBI methodology, describes dietary assessment methods used to measure oral intake in acute and critical care hospital patients. The search was run across four databases to identify primary research conducted in adult acute or critical care settings from 1st of January 2000-15th March 2023 which quantified oral diet with any dietary assessment method. In total, 155 articles were included, predominantly from the acute care setting (n = 153, 99%). Studies were mainly single-centre (n = 138, 88%) and of observational design (n = 135, 87%). Estimated plate waste (n = 59, 38%) and food records (n = 43, 28%) were the most frequent assessment methods with energy and protein the main nutrients quantified (n = 81, 52%). Validation was completed in 23 (15%) studies, with the majority of these using a reference method reliant on estimation (n = 17, 74%). A quarter of studies (n = 39) quantified completion (either as complete versus incomplete or degree of completeness) and four studies (2.5%) explored factors influencing completion. Findings indicate a lack of high-quality evidence to guide selection and application of existing dietary assessment methods to quantify oral intake with a particular absence of evidence in the critical care setting. Further validation of existing tools and identification of factors influencing completion is needed to guide the optimal approach to quantification of oral intake in both research and clinical contexts.

Introduction

Malnutrition is both a cause and consequence of ill health and is a significant issue in healthcare settings worldwide⁽¹⁾. Adequate provision of nutrition is an accepted component of the prevention and treatment of malnutrition yet insufficient food intake has been estimated to occur in 47–76% of patients admitted to an acute care setting^(2,3) . It has been hypothesised that certain populations, including acute and critically ill patients, are at an increased risk of suboptimal intake due to the presence of additional disease-related barriers such as fatigue, weakness and altered appetite^(4,5) . Recent studies indicate significant nutrition deficits both within the intensive care unit (ICU) and on transfer to the acute care ward in patients consuming an oral diet^(6,7) .

The causes of suboptimal food intake in the acute and critical care setting are complex and multi-faceted, involving patient and system factors^(1,8) . Patient characteristics including age, length of stay, appetite, clinical symptoms and prescription of therapeutic diets have all been associated with reduced food intake^(8–10). Compounding this are system factors associated with the hospital environment including mealtime interruptions, inadequate feeding assistance, dissatisfaction with meals, and restrictive mealtimes^(1,8,11) .

The relationship between malnutrition, suboptimal food intake and related increased morbidity and mortality has led to an emphasis on nutritional monitoring within contemporary healthcare safety and quality standards^(12–14). Yet accurate measurement of oral intake, a core component of monitoring, remains a significant challenge. Multiple tools including food records, ready reckoners and plate waste diagrams have been developed to quantify intake in an acute care setting but concerns regarding accuracy persist⁽¹⁵⁾. Knowledge deficits, time pressures and competing priorities for healthcare staff responsible for completing these tools, combined with missing data and impaired patient recall, have been found to result in inaccuracies and compromise completion^(16–18). Errors in the quantification of oral intake may impact research quality, as well as adversely affect timely escalation of care, malnutrition identification and prioritisation of healthcare resources in the clinical setting⁽¹⁵⁾. Despite the perceived importance of accurately measuring oral intake, there has been no systematic exploration of dietary assessment methods that are used to measure oral intake in the acute and critical care setting.

The primary objective of this scoping review was to map and describe dietary assessment methods used to measure oral intake in acute and critical care hospital settings. Secondary objectives were to:

1. 1. Describe the characteristics of the dietary assessment methods used;

2. 2. Report the number of studies, the dietary assessment methods used and the population within which validation processes were completed;

3. 3. Document factors reported to influence completion of the dietary assessment methods;

4. 4. Identify existing evidence gaps and future research priorities aimed at developing accurate but feasible dietary assessment methods for the measurement of oral intake in acute care and critically ill hospital patients.

Methods

Protocol and registration

This scoping review was conducted in accordance with the JBI Manual for Evidence Synthesis and reported in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR)^(19,20) . The protocol was registered a priori on Open Science framework on 6 May 2022; available from https://osf.io/k6m7y. In accordance with JBI process, minor amendments were made to the protocol as the review progressed (listed in Supplementary Table 1, Appendix I).

Eligibility criteria

Original primary research including observational and experimental designs published in English from any geographical location from 1st of January 2000-15th of March 2023 were considered for inclusion. Date restrictions were applied to produce a feasible search result representing modern practice. Sources of evidence were included if they met the following criteria:

Eligibility

Population. Included adults (≥ 18 years) consuming an oral diet. Consumption of an oral diet was defined as ingestion of any oral food or fluids via the mouth with exception of fluid only diets.

Context. Completed in the acute care setting, including critical care but excluding maternity, pre-operative assessment, day-surgery, inpatient rehabilitation or outpatient services.

Concept. Reported on the application of a dietary assessment method to quantify oral diet and included at a minimum calculation of energy intake. Calculation of energy intake was defined as quantification of energy intake from all macronutrients (kilocalorie); studies that included a global estimate of meal consumption were included even though intake was not reported in kilocalories on the basis that such estimates provide an indicator of dietary intake adequacy.

Exclusion. All forms of grey literature were excluded; the original protocol included theses, but due to the size of the final search, a decision was made to also exclude these sources. Additionally, studies were excluded if they:

• quantified intake retrospectively prior to acute hospital admission;

• enrolled patients receiving exclusive enteral or parenteral nutrition with no concomitant consumption of oral diet or included patients receiving oral intake and/or enteral or parenteral with no distinction made regarding quantification of oral intake;

• reported on malnutrition screening tools where dietary intake was estimated as a component of screening;

• included a mixed population where results were not presented separately for acute care and/or critically ill patients; or

• were based on secondary reporting of data. Where multiple published studies reported the same data only the data from the original primary study was included.

Supplementary Appendices II and III Table 2 further outline the eligibility criteria and key definitions used in this review.

Information sources and search

The search strategy was conducted in accordance with the JBI Manual for Evidence Synthesis⁽¹⁹⁾. Following the development and piloting of the search strategy within Medical Literature Analysis and Retrieval System (MEDLINE) via OVID and Cumulative Index of Nursing and Allied Health Literature (CINAHL) via EBSCO, the final search was conducted in consultation with a trained research librarian across four databases: MEDLINE Epub, ahead of print, in process, in-data-review and other non-indexed citations, daily and versions; Excerpta Medica Database (Embase Classic+Embase) (OVID 1947 to date); Emcare (OVID 1995 to date); and Cumulative Index of Nursing and Allied Health Literature (EBSCOhost1937 to date). The search strategy for MEDLINE can be found in Supplementary Appendix IV. The final search was conducted on 15 March 2023. Systematic, scoping and narrative reviews were reviewed only to identify additional primary studies eligible for inclusion in the review.

Selection of sources of evidence

Search results were exported to EndNote (version 20.2.1) and a single author (C.F.) removed duplicates and articles that did not meet the inclusion criteria according to the article title, as per pre-defined criteria (Supplementary Appendix V). Remaining articles were exported to Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia, available at www.covidence.org. Prior to commencing formal screening processes, piloting of title and abstract screening was completed by four reviewers on ten randomly selected articles (C.E.F., O.A.T., J.N.A., I.M.H.). Title and abstract screening were independently completed by two reviewers (C.E.F., I.M.H. or L.M.) with discrepancies resolved by consensus. Articles deemed eligible for full text review were screened independently by two reviewers (C.E.F. and either I.M.H. or L.M.) and conflicts resolved by a third reviewer (O.A.T. and/or E.R.).

Data charting process

A data charting tool was developed and piloted on two articles by three reviewers (C.E.F, O.A.T, L.M.) (supplementary Appendix VI). Data was independently charted by two reviewers (C.E.F and L.M.) with discrepancies resolved via consensus by a third reviewer (O.A.T and/or E.J.R). Details of modifications made to the tool during data charting are listed in Supplementary Appendix VII.

Data items

Article characteristics including location, study design, population, characteristics of the dietary assessment method (method of assessment, format of data collection instruments (automated versus interviewer administered 24 h recalls, paper versus electronic food record and estimated plate waste forms), person responsible for applying the tool and nutrient/s quantified), validation (if completed including reference method and nutrient/s quantified) and factors influencing completion were extracted (supplementary Appendix VI).

Critical appraisal of individual sources of evidence

In line with the JBI Manual of Evidence critical appraisal of the evidence was not completed.

Synthesis of results

Findings for acute care and critically ill patients are reported together. Publication details and information pertaining to validity are presented in a tabular format. Information on the frequency of each dietary assessment method was reported in the literature, and the nutrient/s measured are presented graphically. A narrative summary accompanies the results summarizing the findings in relation to the scoping reviews aims. Where possible, findings are summarised using number (n) and percentage (%).

Results

Selection of sources of evidence

The search identified 12422 articles with an additional ten articles identified from screening reference lists of reviews. Following removal of duplicates, 6161 articles underwent title and abstract screening and 670 underwent full-text screening with 155 articles included in the review (Figure 1).

Fig. 1. PRISMA diagram

Characteristics of sources of evidence

Study characteristics are presented in Table 3 (supplementary Appendix VIII). The majority of studies were completed in an acute care setting (n = 153, 99%) with only two (1%) including patients admitted to an ICU^{(12,21–174)} . The largest number of studies originated from Australia (n = 25, 16%), followed by Denmark (n = 20, 13%) and the UK (n = 18, 12%)^{(12,25–27,37,45–47,49–51,53,58,62–64,67,70,72,73,85,87–90,92,93,96,101,105–111,113,114,116–119,122,126,127,130,132–135,147,156,157,159,161,162,166–168,170,172,173)} . Most were single-centre (n = 138, 88%) and predominantly observational designs (n = 135, 87%), with cross sectional being the most common design overall (n = 79, 51%). A total of fifteen studies (10%) were randomised controlled trials^{(26,31,45,48,59,62,68,74,107,132,137,142,143,155,164)} . The sample size was reported in 150 studies (97%) and ranged from 9 to 1012 participants with the remaining five papers reporting the number of meals or meal trays rather than number of participants^{(72,82,147,157,166)} .

Synthesis of results

Which dietary assessment methods are used to measure oral intake and how have they been applied within acute and critical care hospital patients?

Estimated plate waste (n = 59, 38%), followed by food records (n = 43, 28%) and then 24 h recall (n = 23, 15%), were the most frequently reported assessment methods with the remaining studies using a variety of approaches to quantification (Figure 2)^{(12,21–174)} . Estimated plate waste was predominantly collected using paper-based forms (n = 40, 68%), with six studies (10%) using an electronic form and the remaining studies providing inadequate detail to enable classification of the recording approach (n = 13, 22%)^{(12,22,25,28–30,32,35–38,41,54,56,70–72,75–80,82–84,86,87,95,97–101,103,106,111–113,118,121,124,127,128,131,134,138,141,143,147,152,153,157,160,165–168,174)} . Similarly, food records were mainly completed using paper-based forms (n = 30, 70%) with two (5%) studies reporting on the use of an electronic form and eleven (25%) studies providing insufficient data to enable classification^{(23,24,27,31,37,42,44,45,48–50,53,57–59,62,65–68,73,85,88,91,104,105,107–110,114,119,122,126,129,136,144–146,155,162,171,174)} . Within studies that used a 24 h recall to quantify dietary intake, recall was primarily collected using an interviewer-administered approach (n = 20, 87%), with one (4%) study using a self-administered computer-guided recall and the remaining two (9%) studies providing inadequate detail to enable classification^{(21,26,33,34,40,52,69,74,90,92,102,115,120,130,139,142,148,149,151,164,169,172)} . In total, six studies (4%) investigated novel technologies with 4 (2%) studies using either artificial intelligence or software programs to automate dietary intake estimation, and the remaining two (1%) studies investigated the use of mobile/tablet applications^{(94,116,117,123,125,135)} . Application of the dietary assessment method was completed by a range of individuals with researchers and dietitians being the most common professional groups (Figure 3)^{(12,21,23,25,26,37,39,40,43,46,50,52,56,62,63,68,69,73,74,76,80,81,85–87,89,92,95,99,100,102,103,106,111,112,115,118–121,130,132–134,137–139,141–149,151,154,156–161,163–168,170,171,175)} . The majority of studies (n = 94, 60%) did not specify the type of oral diet that was quantified; where this was specified, regular texture (n = 31, 19%) was the most commonly quantified diet type^{(27,28,36,46,50,51,54,64,77,81,87,102,103,105,107–109,116,128,130,132,139,142,150,159,160,165,172,174)} . Energy and protein (n = 81, 52%) were the main nutrients quantified^{(12,17,25–28,30,31,38,39,41,43,44,46,49–51,59–64,67,69,72,73,76–78,82,85,86,90,91,93,94,96,97,100,101,105–110,115–117,121,122,125–127,130,131,134–139,143,146–148,150,153,154,156,158,162,164,165,167,170,172,173)} . Micronutrients were quantified in eighteen studies (12%) in combination with energy and protein or all macronutrients (Figure 4)^{(21,40,52,53,66,74,81,87,92,102,113,120,141,142,145,149,151,161)} .

Fig. 2. Dietary assessment methods used to quantify intake in the acute care setting

Fig. 3. Profession responsible for applying the dietary assessment method

Fig. 4. Percentage of tools quantifying different nutrient/s of interest

Which dietary assessment methods have undergone a validation process and in which populations were these conducted?

In total, twenty-three (15%) studies reported on validation of the reported method in comparison to another dietary assessment method in the acute setting (criterion validity) (Table 1)^{(24,28,30,36,41,44,53,54,72,82,87,94,98,117,122,123,125,135,140,147,153,160,174)} . Estimated plate waste was the most common dietary assessment method which underwent validation, with fifteen (60%) of the twenty-three studies reporting on the validity of this method in comparison with a reference tool. In total 15 (65%) studies reported using an objective (weighed) method as the reference method. However, intake was only weighed pre- and post-consumption in six (26%) studies with the other nine (39%) studies calculating intake on the basis of comparison of a standard portion to the weight of the food remaining after consumption^{(24,30,36,41,53,54,72,82,87,117,122,147,153,160,174)} . The remaining eight (35%) studies assessed validity in comparison with reference methods reliant on estimation of consumption including estimated food records (n = 1, 4%), estimated plate waste (n = 5, 22%) or 24 h recalls (n = 2, 4%)^{(28,44,94,98,123,125,135,140)} . Researchers (n = 13, 56%) and dietitians (n = 6, 26%) were predominantly responsible for applying the reference method with validity assessed in the majority of studies (n = 14, 61%) via comparison of energy and protein estimates^{(28,30,36,44,53,54,72,82,87,94,98,117,122,125,135,147,153,160,174)} . Most studies compared nutrient estimates via comparison of either one or two meals (n = 10, 43%)^{(17,24,28,30,36,72,82,87,123,147)} . The remaining studies looked at average intake from either a single day of intake data (n = 6, 26%) or multiple days of intake data (n = 7, 30%)^{(41,44,53,54,94,98,117,122,125,135,140,153,160)} .

Table 1. Validation processes

¹ Refers to a modified approach to weighed food records and plate waste where the weight of standard serves was used with intake calculated as the difference between these weights and the weight of food remaining on the plate after consumption

What are the reported reasons for non-completion and were any strategies to enhance completion of the reported dietary assessment methods reported?

A quarter (n = 39, 25%) of studies reported completion rates for the dietary assessment method of interest^{(27,30,41,43–46,50,56,61,67,73,82,88,89,91,93,95,97,98,101,107,111,116,122,124,125,127,129,131,134–136,140,144,151,163,170,139)} . Definitions of completion varied with twenty-five (63%) reporting this as the number of participants with complete dietary intake data and the remaining studies (n = 14, 36%) defining this as the number of complete dietary intake registrations recorded using the assessment method of interest. In total, four studies (2.5%) reported on factors influencing completion of the dietary assessment method with three (2%) studies looking at patient-related factors, and the remaining study (n = 1, 0.5%) exploring the influence of staff training on rates of missing data^{(98,122,125,135)} . Patients reported symptom burden/ illness, ease of use, technological familiarity and confidence, and tool design as factors influencing completion^(98,125,135) . Design characteristics identified as aiding completion included provision of detailed instructions, addition of word cues to aid with quantification, incorporation of realistic visual diagrams and provision of free text space to record food consumed between meals⁽⁹⁸⁾. Staff training was identified in one study as influencing completion with higher rates of missing data observed in food records completed by nursing staff as part of routine care compared with weighed food records completed by dietitians⁽¹²²⁾.

Discussion

This is the first scoping review summarising the literature on dietary assessment methods used to quantify oral intake in adult inpatients within acute and critical care settings. The literature on this topic was broad, with 155 studies completed over the last decade across a range of geographic locations. Studies were mainly single centre with only a small number of randomised controlled trials. Two key themes emerged from the literature: (1) a lack of high-quality evidence and validation of tools in the acute care setting (including ICU) and (2) concern regarding validation processes, and lack of consensus on completion definitions combined with insufficient evaluation of factors influencing completion of dietary assessment methods.

Across all studies, the most common methods used to quantify dietary intake were estimated plate waste and food records^{(12,22–25,27–32,35–38,41,42,44,45,48–50,53,54,56–59,62,65–68,70–73,75–80,82–84,86–88,91,95,97–101,103–114,118,119,121,122,124,126–129,131,134,136,138,140,143–147,152,153,155,157,158,160,162,165–168,171,173)} . Traditional paper-based tools were the most common methods used to capture data across all studies that reported the use of food records and estimated plate waste. However, validation of these methods within the reported studies was limited and there was an absence of literature in critical care. Methods reliant on estimation, such as food records and estimated plate waste, provide a practical approach to the quantification of intake at the bedside. Compared with approaches such as weighed food records, which have typically been used in research settings, such tools are quick and low cost, representing a feasible approach to intake quantification; however, they are also prone to bias⁽¹⁷⁶⁾. Patient recall, inaccurate portion size estimation and high rates of missing or inadequate data have been found to compromise the accuracy of these tools^(16,177,178) . Moreover, there is also a lack of standardisation with food record forms, typically varying by site and plate waste recorded using a range of different scales and approaches to estimation (whole meal versus meal component method)^(178–180). This absence of standardisation has implications for the generalisability of study findings, making interpretation of the existing literature challenging.

Missing or incomplete data is known to be an important factor influencing measurement accuracy. Given the significance of this source of error, an understanding of completion rates is essential. Yet this review found limited measurement of completion reported, with only a quarter of studies quantifying this and a lack of consensus on how to define ‘completion’^{(27,30,41,43,46,50,56,61,67,73,82,88,89,91,93,95,97,98,101,105,107,111,116,122,124,125,127,129,131,134,136,140,144,151,163,171)} . Moreover, only four studies evaluated characteristics influencing completion, with the majority focusing on completion from a patient perspective^{(98,122,125,135)} . Factors identified by patients as influencing completion included technological literacy and confidence, nutrition education, tool design, incorporation of real-time feedback and feeling too unwell^(98,125,135) . Previous studies in long term care and rehabilitation in contrast have highlighted the importance of speed, level of effort, diet type and attitudes and knowledge of healthcare staff as influencing completion; however, whether the same factors apply in an acute care setting remains unknown^(181,182) . Additionally, staff training was identified as influencing completion, with higher rates of missing data occurring when food records were completed by nursing staff as part of routine care without prior training in comparison with weighed food records completed by dietitians⁽¹²²⁾. Monitoring of dietary intake in the clinical setting is reliant on healthcare staff and patients who may lack prior nutrition training. Accordingly, there is a clear need for the development of tools which are feasible and incorporate appropriate training and support of patients and staff to enable accurate quantification of dietary intake within an acute and critical care setting.

Existing guidelines emphasise the importance of using validated tools when measuring dietary intake with attention also paid to the quality of validation completed⁽¹⁷⁶⁾. Yet, of the fifteen studies which reported validating the tool of interest in comparison with an objective reference method (weighed food records or plate waste), only six of these studies actually calculated intake on the basis of weights of food measured pre- and post-consumption^{(24,41,53,54,72,160)} . The remaining nine studies calculated intake on the basis of the difference between standard portions and the weight of food remaining post-consumption, such an approach has the potential to introduce bias compromising the criterion validity of the reference method^{(30,36,82,87,117,122,147,153,174,176)} . Several studies also attempted validation using methods which are not considered as accepted reference methods, including estimated food records and plate waste^{(28,44,94,98,123,125,135,140)} . Comparing a new tool to an existing tool with similar characteristics increases the likelihood of correlated error arising due to inherent bias present in both the assessment and reference method⁽¹⁷⁶⁾. Additionally, the majority of studies reporting on validation processes used short time frames, predominantly comparing nutrient estimates from individual meals raising questions about the ability of such tools to accurately capture intra-individual variations in nutrient intake, vital in both a clinical and research context^{(17,24,28,36,72,82,87,98,123,147)} .

Substantial gaps in our understanding of the optimal way to quantify oral intake in the acute and critical care settings remain, with several priority research areas emerging from this review. Interest in the role of nutrition across the continuum of care and evidence of significant nutritional deficits in critically ill patients receiving an oral diet has resulted in increasing attention being paid towards methodologies used to quantify intake this patient cohort^(6,7,170) . Yet this review found an absence of literature in critically ill patients with only two studies completed in either the ICU or the post-ICU phase^(170,171) . Further research is urgently needed in the critical care setting to evaluate which dietary assessment methods are capable of accurately quantifying oral intake at the bedside, both within the ICU and following transfer to the ward, in critically ill adults. The role of technology to aid with dietary intake quantification in an acute care setting remains relatively unexplored and is another area for future development, with only six studies reporting on the application of such solutions^{(94,116,117,123,125,135)} . Given the recent adoption of technology in some hospital food service systems, future research focusing on the integration of food intake monitoring within these systems is warranted⁽¹⁸³⁾. Specifically, implementation of electronic bedside menu (eBMOS) systems presents a promising innovation to engage patients in their nutrition care and enable real-time monitoring of intake. Such systems allow patients, their caregivers or healthcare staff to enter intake data at the bedside, with automated calculation of nutrient intake⁽¹⁸⁴⁾. However, existing research to date has focused on aspects such as food waste, costs and ordering satisfaction with limited investigation of the capabilities and validity of these technologies with respect to dietary intake quantification⁽¹⁸⁴⁾. Additionally, whilst other technological innovations such as mobile applications have shown promise in other settings, the applicability of such findings in the acute care setting remains unclear^(185,186) . It is plausible that factors influencing the application, acceptability and completion of novel technologies are influenced by characteristics specific to an acute care setting, for example, higher patient symptom burden, varying levels of staff and patient technological literacy or increased time pressures on healthcare staff. Additional research is required to explore whether technological innovations can overcome inherent limitations of traditional tools to improve quantification of oral intake and to determine factors which influence completion unique to an acute care setting.

Strengths and limitations

This is the first review to describe the dietary assessment methods used to quantify oral intake in acute and critical care settings and provides valuable information to inform clinicians and researchers working in this field. Strengths of this scoping review include the rigorous methodology, including prospective registration of the protocol, completion of the review in accordance with the JBI Manual for Scoping Reviews and development of the search in consultation with an experienced research librarian. Another strength of the review is its breadth, with 155 studies included. Limitations include the restriction of this review to primary research articles published in English resulting in exclusion of potentially relevant literature. Exclusion of grey literature and studies enrolling patients receiving supplementary enteral or parenteral nutrtition in combination with an oral diet, which is common practice in acute and critical care settings, may have compromised the comprehensiveness of this review. Moreover, details regarding the characteristics of each dietary assessment method were frequently limited within the identified literature. Consequently, we are unable to extract data on whether assessment methods were used as part of a dedicated research project versus as part of routine clinical practice which has implications for the generalisability of the findings presented here. Detailed reporting of dietary assessment method characteristics should be a consideration for further research.

Conclusion

Traditional paper-based methods remain the most common approach for the quantification of oral diet in an acute setting despite significant concerns existing regarding their accuracy. Overall, this review found a lack of high-quality evidence regarding the optimal approach to dietary intake quantification with a particular absence of literature in the critical care setting. Evidence regarding factors influencing completion of dietary assessment methods and the validity of existing tools is lacking. Further high-quality research is urgently needed to inform clinician decision making and enable selection of the most appropriate tool for quantification of oral diet in both a research and clinical context.