A Scoping Review of Patient‐Reported Measures Administered to Adults Receiving Home Enteral Tube Feeding and Contextual Factors Surrounding Their Use

ABSTRACT

Home enteral tube feeding (HETF) faces significant challenges across micro, meso and macro levels of healthcare. Individual and aggregate data generated from a patient‐reported outcome (PROM) or experience measure (PREM) can be used to tackle multilevel healthcare problems. This review sought to comprehensively describe the current landscape of PROMs and PREMs administered to adults receiving HETF and report on relevant psychometric evaluations. A scoping review was conducted sourcing evidence from five bibliographical databases, two clinical trial registries, a comprehensive grey literature search and citation chaining up to January 2025. Eligible studies reported on administration of at least one PROM or PREM to adults receiving HETF. Data was extracted in relation to study characteristics, population attributes, setting and tool details. The search yielded 14,377 unique studies assessed for eligibility, of which 314 reports were included which corresponded to 256 studies. This consisted of 167 publications and clinical trials reporting results, 70 clinical trial registrations without results and 19 reviews and guidelines. The top focus areas of primary studies were cancer (50%), nutrition support (19%) and neurology (19%) and across studies reporting results 45% were conducted in a population where HETF constituted more than 50% of the sample. There were 184 unique instruments (68 disease‐specific, 43 treatment‐specific, 40 symptom‐specific, 33 generic) of which relevant measurement property evaluations in cohorts with HETF patients were conducted for 22 instruments. Reviews and guidelines suggested instruments with limited consideration of their psychometric testing in HETF cohorts. This review provides a comprehensive overview of the range of PROMs and PREMs administered to adults receiving HETF and explores key considerations when selecting or developing an instrument for use in this cohort.

Key points

• Most PROMs and PREMs administered to adults receiving home enteral tube feeding (HETF) have not been validated in HETF populations, including instruments recommended by clinical practice guidelines.

• No PROM or PREM administered to adults receiving HETF provided details pertaining to evaluation of all measurement properties in this cohort.

• Select HETF‐specific instruments with content validity were amongst the most promising to capture experiences and outcomes unique to HETF, although opportunities remain for novel tool development and evaluation.

1. Introduction

Home enteral tube feeding (HETF) is a method of delivering nutrition to an individual through a feeding tube beyond an inpatient setting, to meet nutritional needs that cannot be achieved orally [1]. HETF enables individuals to receive essential nutrition while remaining in their own homes, but its implementation is associated with multiple challenges, including treatment‐related complications, inadequate patient and caregiver education, fragmented care transitions and limited outpatient support [2, 3]. At a system level, disparities in funding, workforce shortages, and lack of standardised guidelines further hinder effective service delivery [4, 5, 6].

Patient‐reported outcome measures (PROMs) and patient‐reported experience measures (PREMs) can capture outcomes and experiences of care directly from patients without interpretation from a researcher or clinician. PROMs are standardised questionnaires about outcomes relevant to patients, while PREMs collect information on patients' experiences of health care processes [7]. Routine collection of PROMs and PREMs in healthcare systems can support communication between patients and clinicians, help assess service performance, inform quality improvement initiatives, and guide decisions on resource allocation [8]. In the context of multi‐level challenges with delivery of HETF, aggregated patient‐reported outcomes and experiences could play an important role in strengthening care delivery. When seeking a potential PROM or PREM, Valderas and Alonso proposed considering three independent questions: what is being measured (construct), whose experiences are being measured (population), and how those experiences are quantified (the measurement model employed) [9]. PROMs may measure a range of distinct constructs, ranging from individual symptoms, functional status (i.e. one's ability to perform tasks), perceptions related to health, health‐related quality of life (i.e. aspects of quality of life specific to health) and overall quality of life (QOL) [9, 10]. For PREMs, constructs could include the experience of relationships during care (relational) or experiences related to accessibility of care (functional) [7]. PROMs and PREMs are heavily intertwined as healthcare experiences can influence outcomes, while outcomes may also shape perceptions of care [11]. This interconnectedness has also been reported within HETF, where experiences of care have been reported to influence QOL [1]. Careful consideration prior to use of a PROM or PREM is necessary as inappropriate or poorly targeted measures can produce uninformative or misleading data, so ensuring that PROMs and PREMs accurately capture the constructs of interest is critical to avoid research waste and generate findings of value to clinical practice and policy [12]. The Consensus‐based Standards for the Selection of Health Measurement Instruments (COSMIN) has provided guidance and tools around appropriate evaluation of outcome measurement instruments including their validity, reliability and responsiveness properties [13, 14, 15, 16, 17].

People receiving HETF constitute a clinically heterogeneous cohort as multiple conditions may result in the need for HETF [18]. This could add an additional layer of complexity when considering what outcomes or experiences (constructs) are important to measure across this entire cohort. A scoping review of tools used to measure QOL identified 22 instruments used in home enteral and parenteral nutrition cohorts [19]. Treatment‐specific burdens of HETF may not be captured within existing QOL constructs and the evolving perspectives associated with long‐term HETF can complicate interpretation of QOL data [20, 21]. Identifying additional constructs of potential relevance to HETF cohorts and what instruments are useful to measure these constructs would be of value. As relevant instruments may be used in studies specific to HETF or mixed cohorts where individuals receiving HETF comprise only a subset of participants, a broad search strategy is required to capture relevant measures across diverse populations and contexts.

This review aimed to scope the evidence to identify which PROMs and PREMs have been administered to adults receiving HETF and the contextual factors surrounding their use in clinical practice and research settings. This review further sought to report psychometric evaluations of the identified instruments in cohorts including adults receiving HETF, as described by included studies and their cited sources. The findings are intended to support researchers and clinicians working with adults receiving HETF by informing decision‐making regarding selection, application and development of patient‐reported instruments and ultimately enhance care.

2. Materials and Methods

This review was informed by JBI methodology [22] and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) Extension for Scoping Reviews [23]. A protocol for this review was registered on Open Science Framework [24] and the completed PRISMA‐ScR checklist is provided in Supporting Information Table S1.

2.1. Eligibility Criteria

Published and unpublished sources of evidence available in English were considered. This included primary studies, reviews and guidelines. Abstracts without corresponding complete records were also considered if they contained adequate information to meet the specified eligibility criteria. We did not place a restriction on evidence sources or a limit on year of production as this is a novel area of research in a relatively small and under‐researched population group. In doing‐so we ensured any PROMs and PREMs used in suitable contexts would be captured.

To be eligible for inclusion, primary studies' population had to include adults (≥ 18 years old) receiving HETF for a minimum of 6 weeks. This HETF cohort could constitute any proportion of the study's population. The concept was administration of at least one PROM or PREM to the HETF cohort and the context included any geographic or clinical setting. Reviews and guidelines had to include a recommendation or commentary for at least one PROM or PREM that could be used in adults receiving HETF to be eligible for inclusion. A comprehensive list of eligibility criteria as relates to population, concept and context for each evidence source is available in Supporting Information Table S2.

2.2. Evidence Sources

Five bibliographical databases (MEDLINE, EMBASE and PsycInfo [OVID interface], CINAHL and Scopus) were searched. The search strategy was developed for Medline with librarian consultation and translated across the remaining four databases. Clinicaltrials.gov and the World Health Organisation International Clinical Trials Registry Platform (ICTRP) search portal were also searched using a strategy modified from the bibliographical search to meet the registries' interface requirements as previously recommended [25]. Grey literature was also sourced through a systematic search of relevant websites identified by content experts on the research team and an advanced Google search. Choice of grey literature sources and search methods were informed by insights of prior methodology reviews [26, 27] and preliminary searches of potential sources collated in the Cochrane technical supplement [28] and recommended by librarian consultation. GreySource, OpenGrey (Archive), World Health Organisation library database, the Analysis and Policy Observatory (APO–Health Section) and dissertations and theses databases did not yield relevant results from a preliminary search thus were not included as evidence sources. Relevant reviews and guidelines identified through the searches were also reviewed in full to assess the eligibility of studies cited within them. Searches across all sources were conducted up to October 17th, 2022, and then updated on January 16th, 2025. Additional reports relating to included studies identified during full‐text review and data extraction were also included. A summary of all evidence sources and search strategies used is provided in Supporting Information Table S3.

2.3. Study Selection

All bibliographical records were imported into the reference management software EndNote (Version 20, Clarivate Philadelphia) for removal of duplicates [29] and remaining records were managed using Covidence Systematic Review Software (Veritas Health Innovation, Australia) for screening. All clinical trial registrations and grey literature sources were managed using Microsoft Excel (Microsoft Corporation Versions 16.77.1–16.101.2). Duplicates were removed from clinical trial records using Excel conditional formatting of record IDs as previously described [25]. Unique records were then filtered based on their target population age group and those targeting paediatric populations were excluded.

One reviewer (AK) conducted title and abstract screening and full‐text assessment of the 2022 and 2025 search results. A second reviewer independently conducted title and abstract screening (CH, LB) and full‐text assessment (CH) of the 2022 search results with high inter‐reviewer agreement (96%, Cohen's κ 0.72). Conflicts that could not be resolved through discussion were resolved by a third reviewer (MAF).

The first 100 results of the advanced Google searches were screened based on their title, excerpt and URL landing page (AK). Webpages that could not be excluded were collated in an Excel spreadsheet. This spreadsheet was also populated with websites identified from content experts on the research team. All the websites assembled in this spreadsheet were then systematically searched in their entirety as outlined in Supporting Information Table S3c (AK) and content meeting eligibility criteria was included. Records identified as involving the same sample within and across evidence sources were merged into single studies for data extraction.

2.4. Data Extraction and Charting

Data extraction and charting was completed using forms created in REDCap electronic data capture tool, Version 15.3.3, hosted by The University of Sydney. The forms were finalised following piloting by two reviewers (AK, CH) on a subset of eligible studies of varied evidence type (Supporting Information Table S4). Data extracted broadly related to study characteristics, population attributes, setting details and tool details. Tool details included attributes, contextual factors and measurement properties. One reviewer extracted data for all studies (AK) and a second reviewer (CH) extracted data for 20% of studies. Agreement between reviewers for duplicate extractions was 87%. Conflicts that could not be resolved through discussion were resolved by a third reviewer (MAF) who also checked all final extractions pertaining to tool measurement properties.

For primary studies, study characteristics included year of publication/last update, authorship, study type, aim and eligibility criteria. Population attributes included sample size, age, disease characteristics and HETF‐related aspects of the sample (proportion fed via tube, additional feeding routes, types of tubes used and duration of feeding). Population attributes were only available for primary studies reporting results (i.e. not all clinical trial registrations). Setting included the country where the study was conducted and the type of healthcare environment in which PROM or PREM data collection took place. Attributes of the tools extracted from included studies were the name and version used (including language) or whether it was a novel, study‐specific tool. Additional fields included the construct measured, item and domain count, response format, scoring algorithm, duration to complete and recall period. These additional fields were also charted from studies cited directly by the included study when describing the tool (e.g. developmental studies or prior studies employing the tool). These cited studies were not required to meet this review's eligibility criteria and are not reported as included reports as they were used to obtain collateral information on the tools employed when this was not available within the included study. When the versions described in primary and cited studies contradicted each other, the additional fields regarding the latest version were extracted. Contextual factors extracted from primary studies were justification(s) for selecting the tool, the format of the tool used (e.g. paper, digital, interview), whether respondents were provided any assistance for administration of the tool and whether the tool was completed by the participant or a proxy (e.g. carer). Methods and results pertaining to measurement properties of the identified tools as defined by the taxonomy proposed by COSMIN [15] were also charted from included studies. When provided in studies cited by an included study, this data was only extracted if psychometric testing was conducted in an eligible cohort (i.e. included adult HETF recipients). For synthesis papers, study characteristics (year of publication/production, authorship, study type and aim), tool names and commentary relating to use of the tools was extracted.

In some studies, HETF recipients formed a subgroup within a broader clinical cohort and were not the primary focus of the study. To distinguish between such studies, a primary study focus was assigned to each study informed by the study's aim and eligible cohort. A classification was also assigned to all identified PROMs and PREMs with respect to the population they were intended to be administered to. Disease‐specific instruments were defined as those intended for individuals with a specific diagnosed condition (e.g. cancer, inflammatory bowel disease, motor neurone disease). Treatment‐specific instruments were defined as those intended for individuals receiving a specific healthcare treatment or intervention irrespective of condition (e.g. enteral or parenteral nutrition, chemotherapy). Symptom‐specific instruments were defined as those intended for individuals reporting a particular symptom, functional problem or symptom cluster without a specified underlying condition (e.g. dysphagia, constipation, fatigue). Generic instruments were defined as those intended for use in the any patient population, not restricted by disease, treatment or symptom. Categorising instruments this way facilitated identification of the constructs administered across diverse HETF populations.

2.5. Data Synthesis

Select data was presented for each study and each unique tool identified. Study characteristics, setting details, tool attributes and contextual factors were aggregated across all included primary studies, and population attributes were aggregated from primary studies reporting results. Extracted data was presented in figure and tabular format and accompanied by a narrative summary. Descriptive statistics including frequencies, proportions and ranges were reported. All analyses were carried out using Microsoft Excel.

3. Results

3.1. Study Selection

The database and register search identified 23,121 records from which 8995 duplicates were removed. The remaining 14,126 records were screened with 862 reports and 3523 registrations assessed for eligibility. An additional 251 records were identified from the grey literature search and citation chaining, of which 239 were assessed for eligibility. A total of 314 reports were included which corresponded to 256 studies. This process is outlined in Figure 1. Fifteen of these studies were identified from the 2025 search, corresponding to 6.2% of the studies identified from the original search.

Figure 1.

PRISMA flow diagram adapted from Page et al. [30]. ^†686 clinical trial registrations excluded based on eligible age.

3.2. Study Characteristics

An overview of characteristics for all primary studies (n = 237) is presented in Figure 2. Primary studies were conducted across 32 countries with most from the USA (47/237, 20%), the UK (21/237, 9%) and France (18/237, 8%). The earliest study was conducted in 1994 and 169/237 (71%) of studies were published or registered since 2015. Most studies were interventional 128/237 (54%) and the predominant primary focus was cancer research 119/237 (50%), followed by nutrition support 45/237 (19%). Of studies reporting results (n = 167), the sample size ranged from 1 to 1180 and the proportion of participants who received HETF was reported in 151 (90%) studies. Patients receiving HETF constituted more than 50% of the sample in 75/167 (45%) and 100% of the sample in 49/167 (29%) studies. The duration of HETF ranged from 6 weeks to 19.5 years. Participants received HETF exclusively in 9/167 (7%) studies, partially in 34/167 (20%) studies and both in 39/167 (23%) studies. Additional feeding routes could not be ascertained in the remaining studies. Between one and seven PROMs and PREMs were employed with 163/237 (69%) using one or two tools. A comprehensive summary of each included study is available in Supporting Information Table S5.

Figure 2.

Characteristics of all primary studies (n = 237). Numbers present on the map are the total number of studies conducted in each country, with a darker gradient representative of a higher number of studies conducted in that country. All remaining numbers and accompanying percentages are the sample sizes that correspond to the respective characteristic. “Sample receiving HETF” summates findings of only studies reporting results (n = 167).

3.3. Overview of Tools

Across all primary studies, 184 unique PROMs and PREMs were identified of which 45 (24%) were developed specifically for the study in which they were used and were not employed by any additional studies included in this review. QOL was the most prevalent construct measured by these tools 60/184 (33%). One hundred and twenty‐seven (69%) tools were administered in an outpatient setting, 79 (43%) in a community setting (e.g. domiciliary, primary care clinic) and 29 (16%) in an inpatient setting. Setting was either not reported or unclear for all administrations of 16 tools. Fifty‐seven (31%) tools were administered across more than one setting type. For at least one administration, clinician or researcher assistance was reported for 61 (33%) tools and no assistance was required for 42 (23%) tools. No information regarding additional assistance was available across all administrations of 103 (56%) tools. Thirty‐seven (20%) patient outcome or experience measures were completed by a proxy in at least one study. Seven of these tools were designed for proxy reporting in their standard form, predominantly for individuals with advanced dementia, severe cognitive impairment, or end‐of‐life care needs. Paper form was the most prevalent administration method (47/184, 26%), followed by digital (34/184, 18%) and verbal (29/184, 16%). More than one format was used for 29 (16%) tools. The method of administrations was not reported for 110 (60%) tools. A language other than English was specified for 56 (30%) tools. Attributes of each tool and contextual information related to their administration aggregated across all studies is available in Supporting Information Table S6.

3.4. Validation of Tools in HETF Cohorts

Of the 184 unique PROMs and PREMs, 26 (22 unique) instruments were accompanied with details of at least one psychometric evaluation in a relevant HETF cohort (Table 1). This information was provided by 43 of the 237 primary studies and the most common construct measured by these tools was QOL (11/22). Six of the 26 tools were tested in a population of which more than half (n = 12–72) received HETF (the Functional Assessment of Cancer Therapy‐Enteral Feeding [FACT‐EF], the Dutch Swallowing Quality of Life [SWAL‐QOL], the Brazilian and English NutriQoL and two study‐specific questionnaires) [36, 46, 55, 56, 58, 59]. The MD Anderson Dysphagia Inventory (MDADI) was tested in the largest HETF cohort (n = 378) which constituted 44% of the sample [42]. Hypothesis testing for construct validity (19/26) and internal consistency (17/26) were the most reported measurement properties across all tools. Detailed reporting of the psychometric testing methods and results are available in Supporting Information Table S7.

Table 1.

Summary of measurement properties reported for tools tested in Home Enteral Tube Feeding cohorts. ^a

Tool (version used in validation)	Construct measured	Frequency used b (cited validation)	Validation population	Content validity	Criterion validity	Construct validity		Internal consistency	Measurement error and reliability	Responsiveness	Other
						Structural	Hypothesis testing
Other versions used
Disease‐specific
EORTC QLQ‐C30 (v3.0) [31] v2.0, select items/domains	QOL (Cancer)	56 (1)	Laryngeal cancer following resection ± radiation therapy. No additional HETF details.				C, D, KGC
EORTC QLQ‐H&N35 [32, 33] H&N43, select items, modified	QOL (HNC)	31 (8)	HNC. No additional HETF details.				C, D, KGC	[33]		[32]
EORTC QLQ‐OES24 [34] EORTC QLQ‐OES18	QOL (Oesophageal cancer)	11 (4)	Oesophageal cancer (multiple treatment types). HETF n = 14 (11%), exclusive and partial.
FACT‐HN (v2.0) [35] v3.0, v4.0	QOL (HNC)	9 (2)	HNC (during/after chemoradiotherapy or ≥ 3 months after surgery). No additional HETF details.				C, KGC
FACT‐EF [36] previously known as QOL‐EF	QOL (Cancer receiving enteral feeding)	2 (1)	HNC for curative (chemo)radiotherapy. HETF n = 72 (100%), exclusive and partial, duration 1.9 (0.2–8.7) months.				C, KGC
UW‐QOL (v4) [37] modified	QOL (HNC)	8 (1)	HNC, post‐treatment. Percutaneous tube. No additional HETF details.				C
MDASI [38]	Symptom burden (Cancer)	1 (1)	HNC (mostly advanced) post‐treatment. HETF n = 11 (41%), gastrostomy, at least 8 weeks.
MDASI‐HN [39]	Symptom burden (HNC)	4 (2)	HNC with dysphagia. HETF n = 37 (18%).				C, D, KGC
PCI (v1) [37] modified	Concerns (HNC)	2 (1)	HNC, post‐treatment. Percutaneous tube. No additional HETF details.				C
HNQOL [40]	QOL (HNC)	1 (1)	HNC. HETF n = 32 (13%), gastrostomy tube.				C, KGC
ALSAQ‐40 [41]	QOL (ALS)	3 (1)	ALS for 39.1 ± 32.8 months (range:1–122), Japan. HETF n = 3 (10%), gastrostomy.				C, D
ESAS (adapted for cancer cohort) [38]	Symptom severity (Palliative conditions)	2 (1)	HNC (majority advanced), 8.2 ± 11.2 months post‐treatment, experiencing ≥ 3 nutritional issues. HETF n = 11 (41%), gastrostomy ≥ 8 weeks.
Symptom‐specific
MDADI [42]	Impact of dysphagia on QOL	21 (5)	HNC for modified barium swallow. HETF n = 378 (33%), exclusive and partial.				KGC				I
SWAL‐QOL [43, 44, 45]	QOL (Swallowing‐related)	13 (2)	Dysphagia (Cancer, vascular disease, neurological, obstructive respiratory disease, trauma, chronic medical conditions, dementia). HETF n = 57 (18%).				C, D, KGC
SWAL‐QOL (Dutch) [46]	QOL (Swallowing‐related)	1 (1)	Oral/Pharyngeal cancer (curative treatment). HETF n = 72 (71%).				C, D, KGC				I, F, TP
SWAL‐QOL (Turkish) [47]	QOL (Swallowing‐related)	1 (1)	Stroke, Multiple Sclerosis, ALS, Myasthenia Gravis, Parkinson's disease. HETF n = 42 (40%).				C, D				TP
EAT‐10 (Turkish) [48] English	Dysphagia severity	13 (2)	Neurogenic dysphagia (Stroke, Multiple Sclerosis, ALS, Myasthenia gravis, Parkinson's disease). HETF n = 42 (40%), exclusive and partial.				C				TP
SHI [49]	Speech impairment	2 (1)	Oral and Pharyngeal Cancer. HETF n = 13 (13%), gastrostomy.				C, KGC				I
Treatment‐specific
NutriQoL (v2, Spanish) [50, 51, 52, 53, 54]	QOL (Home Enteral Nutrition)	6 (4)	Spain, oncologic, neurologic and malabsorptive diseases. HETF n = 77 (44%), exclusive and partial. Percutaneous (unspecified) or NGT, 1–12+ months (54, 55). Spain, oncologic, neurologic and malabsorptive diseases. HETF n = 62 (44%), exclusive and partial. Percutaneous (unspecified) /NGT, Cohort 1 3–6+ months, Cohort 2 mean (SD) 21(14) days (56, 57). Spain, HETF n = 33 (42%), exclusive and partial. G‐tube/NGT/J‐tube, 1–2+ months (58).	[50]		[51]	C, D, KGC [53, 54]	[52, 53]	[52]	[52]	I [52]
NutriQoL (v2, Brazil) [55]	QOL (Home Enteral Nutrition)	1 (1)	Brazil. HETF n = 12 (100%), exclusive. G‐tube/unspecified, 3+ months.								TP
NutriQoL (v2, English) [56] adapted	QOL (Home Enteral Nutrition)	2 (1)	New Zealand, HETF n = 42 (100%), exclusive and partial, G‐tube/NGT/NGJT, mean (SD) duration 86.2 (88) months				C, D, KGC
‘PEG‐specific questionnaire’ [57]	Experience	1 (1)	Advanced HNC, received treatment. HETF n = 10 (20%), exclusive and partial. Gastrostomy.
‘Self‐administered questionnaire’ (French) [58]	Tolerance and impact of HETF	1 (1)	Head and Neck/Oesophageal cancer for treatment. HETF n = 39 (100%) exclusive and partial. G‐tube/NGT, ≥ 4 weeks				C, D
‘Study‐specific questionnaire’ (Swedish) [59]	Experiences and symptoms	1 (1)	Living independently (HNC, ALS, Neurological disease, Crohn's disease and other gastrointestinal disorders). HETF n = 40 (100%). G‐tube/NGT, 2 weeks ‐ 2+ months.
Generic
Equity and patient experience questionnaire [56]	Experience	1 (1)	New Zealand, HETF n = 42 (100%), exclusive and partial, G‐tube/NGT/NGJT, mean (SD) duration 86.2 (88) months
Agnew Relationship measure ‐ 5 Item Version (Dietitian and Patient) [60]	Experience ‐ therapeutic alliance	1 (1)	HNC. HETF n = 71 (29%), exclusive and partial. G‐tube.

Abbreviations: ALS, amyotrophic lateral sclerosis; ALSAQ‐40, Amyotrophic Lateral Sclerosis Assessment Questionnaire; C, convergent validity; C30, Core Questionnaire; D, discriminant validity; EAT‐10, Eating Assessment Tool; EORTC QLQ, European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire; ESAS, Edmonton Symptom Assessment Scale; FACT, Functional Assessment of Cancer Therapy; FACT‐EF, Functional Assessment of Cancer Therapy–Enteral Feeding; FACT‐HN, Functional Assessment of Cancer Therapy–Head and Neck; F, feasibility; G‐tube, gastrostomy tube; H&N35, Head and Neck Module; HETF, home enteral tube feeding; HNC, head and neck cancer; HNQOL, Head and Neck Quality of Life; I, interpretability; J‐tube, jejunostomy tube; KGC, known‐group comparisons; MDADI, MD Anderson Dysphagia Inventory; MDASI, MD Anderson Symptom Inventory; MDASI‐HN, MD Anderson Symptom Inventory–Head and Neck Cancer Module; NGT, nasogastric tube; NGJT, nasogastro‐jejunal tube; OES24, Oesophageal Cancer Module; PCI, Patients Concerns Inventory; PEG, percutaneous endoscopic gastrostomy; QOL, quality of life; SD, standard deviation; SHI, Speech Handicap Index; SWAL‐QOL, Swallowing Quality of Life; TP, translation process; UW‐QOL, University of Washington Quality of Life Questionnaire.

^a Shaded cells indicate the measurement properties reported for the respective tool.

^b Frequency used refers to all versions of the tool.

3.5. Disease‐Specific

When studies were stratified by proportion of the sample receiving HETF, disease‐specific tools were the most prevalent across all quartiles (Figure 3). There were 68 disease‐specific tools, of which 31 (46%) related to cancer. Across all tools the most frequently used (55/237, 23% studies) was the core module (C30) of the European Organisation for Research and Treatment of Cancer (EORTC) QOL Questionnaire. Ten tools available as part of the EORTC suite of questionnaires were used across 64 studies. EORTC questionnaires used in studies with > 50% of the sample receiving HETF were the ‐C30, head and neck module (‐H&N35), oesophageal cancer modules (‐OES18/24), abridged core module for palliative care (‐C15‐PAL) and oesophago‐gastric cancer module (‐OG25). Four records of psychometric testing for three EORTC tools in cohorts that included HETF recipients [31, 32, 33, 34] were provided by 12 studies, although minimal additional detail pertaining to the HETF subgroup was provided for the validation population. Cancer‐specific QOL questionnaires of the Functional Assessment of Chronic Illness Therapy (FACIT) suite were also used by 15 studies. Relevant psychometric testing of the FACT‐EF questionnaire, developed from a cohort of patients with malignancies of the head and neck that required tube feeding, was

Figure 3.

Distribution of types of patient‐reported outcome and experience measures administered in (A) primary studies reporting results (n = 167) stratified by the proportion of the sample receiving Home Enteral Tube Feeding, (B) all primary studies (n = 237) stratified by primary study focus.

reported for five of the seven measurement properties extracted [36]. Thirteen of the 31 cancer‐related tools focused on malignancies of the head and neck, with the Head and Neck Quality of Life (HNQOL), MD Anderson Symptom Inventory ‐ Head and Neck Cancer Module (MDASI‐HN), Patients Concerns Inventory (PCI) and University of Washington Quality of Life Questionnaire (UWQOL) having relevant testing of one to four measurement properties. Fourteen tools related to neurological conditions, predominantly amyotrophic lateral sclerosis and dementia. Of these, the Amyotrophic Lateral Sclerosis Assessment Questionnaire (ALSAQ‐40), used to measure QOL in three studies, had been psychometrically tested in a Japanese cohort of which 10% received HETF [41]. Ten disease‐specific tools related to disorders of the gastrointestinal tract which measured QOL (four tools) or symptom attributes such as severity or frequency (six tools). No relevant measurement properties were provided for any of these tools.

3.6. Treatment‐Specific

A total of 43 treatment‐specific tools were identified which related predominantly to nutrition support, cancer treatment, inpatient care and swallow rehab. When stratified by the proportion of the sample receiving HETF, treatment‐specific tools were most prevalent in the top quartile of studies, constituting 23% of all tool administrations (Figure 3). The most frequently used tools were the NutriQoL, Percutaneous Endoscopic Gastrostomy Quality of Life Questionnaire (PEG‐Qu), Patient‐Reported Outcomes version of the Common Terminology Criteria for Adverse Events (PRO‐CTCAE), Hospital Anxiety and Depression Scale (HADS) and Oral Mucositis Questionnaire. The Spanish NutriQoL which was developed for adults receiving home oral and enteral nutrition support was the tool reporting the most psychometric testing in a cohort of which 42‐44% received HETF [50, 51, 52, 53, 54]. The translation process to Brazilian Portuguese and hypothesis testing of an English version were also reported [55, 56]. Thirty‐two of all 45 study‐specific tools were treatment‐specific and the majority measured acceptability, experience and satisfaction with treatment (Supporting Information Table S6). Three of these study‐specific tools provided data pertaining to measurement properties [57, 58, 59]. All three reported on content validity, one on hypothesis testing for construct validity and one on internal consistency.

3.7. Symptom‐Specific

There were 40 symptom‐specific tools, primarily used to measure presence, frequency or severity of a single or set of symptoms, the overall burden experienced by these or their impact on a broader construct such as QOL. There were 17 tools which focused on upper aerodigestive symptoms such as swallow, speech, voice or breathing impairments. Eight tools measured symptoms related to mental wellbeing including anxiety, depression and distress. The most frequently used tools were the MDADI, Patient Generated Subjective Global Assessment (PG‐SGA) and SWAL‐QOL. The MDADI measures the impact of dysphagia on QOL and construct validity in a HETF‐containing cohort [42] was cited by five of the 21 studies employing the tool. The SWAL‐QOL measures swallowing‐related QOL and was used by 15 studies, with data pertaining to all measurement properties but responsiveness reported across the identified versions [43, 44, 45, 46, 47]. The PG‐SGA which is a mixed patient‐ and clinician‐rated tool was used by 19 studies. The patient‐reported component of this tool measures nutrition‐related symptoms and no HETF‐relevant validations were identified.

3.8. Generic

There were 33 generic tools used across all studies, measuring QOL, quality of care, mental wellbeing, independence, diet, physical activity and sleep. The EuroQol Group questionnaire (EQ‐5D) was the most used generic tool and the second most frequently used tool across all studies (32/237, 14% studies). Generic tools constituted 32% (12/37) of all tools that were also completed by a proxy. Generic tools were the most commonly administered (44%, 24/54 administrations, by 17 studies) tool type within the top quartile of studies for which nutrition support was the primary focus. This included the EQ‐5D, Short Form Health Surveys (SF‐36 and SF‐12) and World Health Organisation Quality of Life Survey ‐ Brief Version (WHOQOL‐BREF). No measurement property data in a HETF population was provided for any of these tools. Two of all generic tools, the Equity and patient experience questionnaire and Agnew Relationship Measure each reported on a single measurement property ‐ hypothesis testing and internal consistency data, respectively [56, 60].

3.9. Reviews and Guidelines

There were 19 reports which either recommended or provided commentary pertaining to PROMs or PREMs that could be used in HETF cohorts. These included six clinical practice guidelines [61, 62, 63, 64, 65, 66], six systematic or scoping reviews [4, 19, 67, 68, 69, 70], four narrative reviews [71, 72, 73, 74] and three practice support documents developed by professional organisations—an optimal care pathway [75], a toolkit [76] and a competency framework [77]. The most referenced tools were the EQ‐5D, NutriQoL, the EORTC suite and SF‐36/‐12. Generic tools such as the EQ‐5D and SF‐36/‐12 were suggested to assess QOL by some reports [64, 66, 75, 76, 77]. Reviews and guidelines highlighted extensive psychometric testing of generic tools and discussed the disadvantages of their use to include poor sensitivity and responsiveness to address HETF challenges. Disease‐specific tools like the EORTC suite were described as more appropriate for their respective clinical subgroups and they were described as not designed to assess the impact of HETF on QOL [74]. Psychometric testing and suggestion by a professional society were described as reasons in support of using the NutriQoL [61, 64, 72, 75] and some reports highlighted suboptimal testing results for responsiveness and known‐group comparisons [19, 64, 69]. An additional 20 PROMs and PREMs were mentioned in these reviews and guidelines which were not identified as having been used in a HETF cohort.

4. Discussion

This scoping review mapped the landscape of PROM and PREM use among adults receiving HETF, offering a comprehensive overview of 184 distinct tools applied across 237 primary studies. This included disease‐specific, treatment‐specific, symptom‐specific and generic tools measuring simple to multi‐dimensional constructs. A summary is presented of the contextual factors surrounding use of each type of PROM and PREM, and the psychometric properties tested in HETF cohorts. Thirty of the identified tools were intended specifically for enteral tube feeding populations, although measurement property details were only reported for four. Many of the tools identified were designed for use in specific clinical cohorts which commonly include HETF‐recipients, although when psychometric testing of the tools was provided HETF was often not reported. In presenting this information we provide readers with insights to guide future selection and development of PROMs and PREMs across HETF research and clinical practice.

At an international level, PROM and PREM use in HETF cohorts was most reported across Europe and North America, which was in keeping with the distribution of published HETF studies identified by Ojo et al. in a 2015 review [2]. Across Oceania and Asia, 75% of studies identified were published in the last 10 years, mirroring the upward trend in health expenditure of nations of the Western Pacific over the same time period [2, 78]. PROMs and PREMs need to be linguistically and culturally adapted to the populations in which they are administered [79]. Many studies, conducted in non‐English speaking populations did not specify using a locally adapted version of the tool they employed. Some individuals requiring HETF may have an underlying clinical condition resulting in the need for a proxy respondent. Of the tools providing evaluation in a HETF‐containing cohort, only the NutriQoL, Equity and patient experience questionnaire and EAT‐10 were completed by a proxy for at least one of their administrations. The NutriQoL was the only tool for which responses from proxies and patients were compared [52]. Satisfactory cross‐cultural validity and agreement between proxy and patient reporting is a key consideration when selecting a PROM or PREM for use in HETF.

The majority of evidence identified pertained to interventional studies, which is unsurprising given that PROMs and PREMs have historically been confined to clinical trials, with initiatives to integrate them into routine clinical practice emerging in recent years [80]. The rise in use of PROMs and PREMs has been accompanied by a wider movement towards value‐based health care [81]. The report of these instruments in five case studies/series highlights how they may already constitute part of routine clinical care in some instances [82, 83, 84, 85, 86]. HETF care is typically delivered in community, outpatient and inpatient care settings [64]. PROMs and PREMs were administered to this cohort in all three settings, highlighting the potential to capture relevant data using such instruments across the entire continuum of care.

There was an abundance of studies measuring QOL or health‐related QOL. Chen et al. identified 27 unique tools which also included clinician‐rated instruments such as the Karnofsky Performance Scale in a 2022 scoping review of instruments measuring QOL in home enteral and parenteral nutrition [19]. Contrary to the current review, Chen et al. reported substantially more psychometric testing of the identified tools alongside a form of quality assessment. The source of reported psychometric evaluations was unclear and when considered alongside the limited evaluations identified in the present review, the ratings by Chen et al. were likely of testing conducted in populations where HETF was absent. Content validity refers to whether an instrument's content captures the intended construct [15]. To ensure relevance and comprehensiveness of the content, qualitative studies such as focus groups, Delphi studies or interviews are typically employed as a first step in instrument development [87]. Most of the tools identified by Chen et al. and reported in the current review, despite their widespread use in cohorts in which HETF patients predominated, did not consult HETF subgroups as part of their development process. There are unique treatment‐specific burdens related to HETF which may impact QOL [69, 88, 89] thus measuring a construct not developed with HETF recipients' involvement diminishes the relevance of the tool for this group. Even amongst the four HETF‐specific instruments identified in this review, the extent of patient involvement in developing instrument content varied considerably, with as few as two participants involved in one study. Ojo et al. identified 14 studies assessing the effect of tube feeding on QOL and reported a net positive effect despite acknowledging the use of non‐specific tools as a significant limitation of the included studies [69]. Their review also included four studies that involved single administration of a tool which limits interpretation in terms of longitudinal changes. Contrastingly, a 2007 review concluded QOL to be lower in adult HETF patients when assessed using non‐specific tools [74]. Inconclusive findings when non‐specific tools were applied in HETF cohorts may reflect the lack of HETF‐relevant content validity of these tools. The EORTC and FACIT suites use a modular approach to measure QOL where cancer type modules are added on to a core cancer‐specific module [90, 91]. The FACT‐EF was developed as an add‐on module for cancer patients receiving tube feeding and was the only disease‐specific tool to report content validity in a majority HETF population with an appropriate degree of patient consultation during development. In disease‐focused services where HETF constitutes only a portion of clinicians' caseload, using an adequately tested HETF‐specific module as an add‐on to a wider disease‐specific tool may be an appropriate option.

Some prior reviews of evidence suggested periodically or regularly measuring QOL using a non HETF‐specific instrument to monitor or evaluate treatment [64, 66, 75, 76, 77]. On a macro level, using results from inappropriate measures to prioritise and plan services can become particularly problematic. The tools identified here were also predominantly used in a research capacity and when transferred to a healthcare facility, instruments measuring QOL solely to monitor treatment as opposed to being integrated into treatment planning and evaluation, are less likely to influence clinical decision making [92]. Responsiveness reflects an instrument's ability to detect change over time and interpretability is required to determine whether such change is meaningful from a clinical or patient perspective [15]. Improving or maintaining QOL is a key goal of HETF treatment [64]. Responsive instruments with established interpretability, including minimal clinically important difference estimates, can help determine whether changes in QOL over time are meaningful to patients. This information may support clinical decision‐making by helping clinicians evaluate whether modifications to HETF treatment and care are associated with meaningful improvements in patient outcomes. The NutriQoL, which measures health‐related QOL of patients receiving home enteral nutrition was tested for responsiveness and interpretability, although there were methodological limitations. NutriQoL scores increased over time, although these changes were not evaluated against predefined hypotheses, and factors occurring between assessments that may have influenced health‐related QOL in HEN were not described. Authors also calculated minimal clinically important difference using a patient‐reported anchor reflecting global perceived health change (a five‐point scale from “much worse” to “much better”) although did not report correlation between the anchor and NutriQoL scores [52, 93]. Qualitative studies report aspects of HETF treatment and care that patients describe as influencing their wellbeing [88, 89]. Responsiveness may therefore be more meaningfully evaluated against predefined HETF‐relevant changes expected to influence the construct of interest. Responsiveness to such clinically meaningful time‐points (e.g. a more/less disruptive feeding regime) was not identified for any of the tools that reported psychometric evaluation in a HETF cohort. Regardless of the clinical context in which a PROM or PREM is administered, thoughtful consideration of what is sought to be measured, when it should be collected and how the data will be used is essential to avoid research waste and potential harm to future patients [12]. Concerns identified by this study with regards to use of patient‐reported instruments in HETF reflect issues with the use of PROMs and PREMs more broadly.

4.1. Strengths, Limitations and Future Directions

To our knowledge this review is the first to provide an overview of how PROMs and PREMs have been used within the HETF landscape. A major strength of this review is its comprehensive search strategy and high exclusion thresholds which yielded a substantial number of included studies that contributed a thorough synthesis of the evidence base. By not restricting eligible studies to those with a 100% HETF cohort, findings are of use to clinicians working with a varied HETF case load which is reflective of the diverse care arrangements that exist for HETF recipients. As the search strategy was centred on HETF, it is possible that relevant studies focused on other clinical populations that included adults receiving HETF (e.g. dysphagia) may not have been identified. This review also had some limitations as regards study selection of results from the 2025 search update having been conducted by a single reviewer, although a high agreement was achieved between the two reviewers for results of the first search. A good agreement was also reached for data extraction which was completed 20% in duplicate, and remaining extractions by a single reviewer were checked by a second. As we employed a scoping review, quality appraisal of included studies was not conducted, thus evidence presented is of varied quality. Future research could include development of a core outcome set as set out by COSMIN that specifies the constructs of relevance to adults receiving HETF. A core outcome set, together with the findings of this review would then elucidate the need for systematic reviews of specific measurement instruments.

5. Conclusions

The value of PROMs and PREMs in clinical practice across all levels of healthcare has been widely reported. This scoping review provides a synopsis of the PROMs and PREMs that have been administered to adults receiving HETF across the world, alongside the contextual factors surrounding their use. Disease‐, treatment‐, symptom‐specific and generic tools were all widely used across the continuum of care, although there were limited psychometric evaluations of the tools identified in relevant HETF cohorts. When selecting a non‐HETF‐specific instrument to assess outcomes in people receiving HETF, clinicians and researchers should consider the conceptual fit of the instrument and the population for which it was originally developed. In doing so, this would minimise the risk of research waste with use of poorly targeted instruments and maximise the value of the data generated. Few HETF treatment‐specific tools were identified, of which those with strong content validity are amongst the more promising options, although clinicians and researchers should ensure additional validity, reliability and responsiveness measures are appropriately tested prior to their routine use. The frequent use of proxy reporters also highlights the importance of considering proxy reporting during instrument development and evaluation, including the assessment of patient–proxy agreement. The overall scarcity of relevant, well‐developed tools suggests opportunities exist for novel instrument development and evaluation.

Author Contributions

Andriana Korai, Sharon Carey, Margaret Allman‐Farinelli contributed to conceptualisation and developed methodology. Juliana Chen, Sharon Carey and Margaret Allman‐Farinelli provided supervision. Andriana Korai ran the search strategy. Andriana Korai and Carrie Ka Wai Ho completed study screening and assessment. Andriana Korai, Carrie Ka Wai Ho, Margaret Allman‐Farinelli contributed to data extraction. Andriana Korai carried out synthesis and visualisation and drafted the original manuscript and Carrie Ka Wai Ho, Juliana Chen, Sharon Carey, Margaret Allman‐Farinelli were involved in review and editing. All authors read and approved the final article.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not‐for‐profit sectors. AK was supported by a Research Training Programme Stipend Scholarship from the Australian Government and the King and Amy O'Malley Trust Foundation Postgraduate Research Scholarship.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1

Supporting File 2

Data Availability Statement

Data available in article supporting information.