Background

Patient-reported outcome measures (PROMs) can be used to assess patients' health-related quality of life (HRQoL) from the patients' perspective.1 The Food and Drug Administration of the US Department of Health and Human Services defines PROMs as "any report of the status of a patient's health condition that comes directly from the patient, without interpretation of the patient's response by a clinician or anyone else".2^(p.2) Therefore, PROMs can be used to measure patients' own perceptions of their health status,1,3 HRQoL1,3 and individual treatment effects.4 However, self-reporting can be problematic or impossible for some patients.5 In these situations, proxy-reported outcome measures may be used for the assessment of the patient's HRQoL as the proxies can provide relevant information about how they think the patient would report on his or her HRQoL. Patient-reported and proxy-reported outcome measures are also used in the context of enteral feeding.6,7 Enteral feeding is an alternative feeding strategy to be used in patients with insufficient oral food intake,8 such as patients with severe dysphagia and other conditions.9-11 Although patients may benefit from this feeding strategy in terms of nutritional status,9,12 the impact on patients' HRQoL is equally important.12 Therefore, PROMs or proxy-reported (i.e. when self-reporting is complicated) outcome measures should be evaluated in addition to conventional clinical assessments and instrumental investigations.13 As a means to select the most suitable PROM and proxy-reported outcome measure for assessing HRQoL among patients receiving enteral feeding, it is important to evaluate the psychometric properties and the clinical utility of these measurement instruments.14

Quality of Life (QoL) is defined by the World Health Organization as follows: "individuals' perceptions of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns. It is a broad-ranging concept incorporating in a complex way the person's physical health, psychological state, level of independence, social relationships, personal beliefs and their relationships to salient features of the environment".15^(p.3) There is still a lack of consensus among researchers about the definition of QoL that can explain the different constructs or concepts that are selected for their measurement instruments.16,17 Quality of life varies between populations, time periods and diseases.18-20 Also, the concept is subjective as it is based on a perception by the individual around his or her wellbeing.12 Individuals from the same environment seem to have a similar conception about QoL (e.g. "reaching happiness" and "satisfaction in life" are important constructs of QoL in individuals living in the Western countries).16 Health-related quality of life is distinct from QoL as it is directly associated to a health state of an individual.16,21 Various definitions of HRQoL exist.19,22 However, agreement is reached in that HRQoL is a multidimensional concept (i.e. consists of physical, psychological and social domains of health), it is based on the individual's perception of his or her wellbeing, encompasses the current health situation and the future perspective and is not restricted to the disease.16,19

Assessment of HRQoL is useful for measuring the outcome of disease and treatment effectiveness.23 The most common way to measure HRQoL issues is by patient-reported outcome (PRO) instruments, using a standardized questionnaire or an interview.4,13,24-26 Patient-reported outcome measures can be generic instruments, disease-specific instruments or battery of scales. Generic PROMs can be used in diverse patient populations and are suitable for comparing HRQoL and the efficiency of different treatment strategies across these populations.24,27,28 Generic instruments can examine a broad range of domains useful to various health conditions and diseases.19,24 In contrast, disease-specific PROMs assess one specific patient population in which a particular disease manifests, measure the HRQoL domains relevant to that specific patient population and the impact of specific treatments on that population group.19,24,28 Battery of scales comprises multiple scales that measure one particular HRQoL domain (i.e. social functioning).24 Such unidimensional HRQoL instruments will assess a relevant domain in depth, but does not allow to benchmark populations. Moreover, the burden on patients to complete multiple scales can become onerous and time consuming, which may impede results.24,29 In this review, we are interested in multidimensional PROMs, such as the generic and disease-specific scales. These multidimensional scales are intended to measure the constructs of interest and consist of subscales that measure different aspects of health.27 To select the most appropriate PROMs for measuring HRQoL among patients receiving enteral feeding, it is important that these PROMs measure the relevant constructs or concepts that are crucial for patients receiving enteral feeding.30 Clear definitions about HRQoL in enteral feeding and the constructs being measured should have been provided in the studies that have established the psychometric properties of the PROMs of interest. We will collect these definitions (cf. data extraction).

Among patients with cognitive, speech and language disorders, or in some other populations (i.e. infants), self-reported methods may be inappropriate or impossible.23,24 Proxies, defined as healthcare providers, spouses, parents or relatives who are closely involved with the patient,31 can act as alternative sources to obtain valuable information about patients' HRQoL.23,32 The proxies are asked to assess the patient's HRQoL in the manner that they believe the patient would report on this.5 The use of proxy-reported outcome measures allows to reveal information about the patient's HRQoL that may otherwise be lost.33 However, these proxy-reported outcome measures are not the same as PROMs as the PRO information is obtained by proxies.21,34 Furthermore, relying on proxy-provided information is only useful if proxies are able to report on the different HRQoL domains and if their ratings are sufficiently reliable.33 The reliability of the proxy ratings can be evaluated by examining patient-proxy agreement. This requires that the PROM and the accompanying proxy-reported outcome instrument measure the same constructs with identical items and that these instruments and items are valid and reliable.35,36 As various terms are available for these proxy versions (i.e. proxy-reported outcome measures, parent proxy-report instruments),34,35 we will make use of the term 'proxy-reported outcome measures' in this systematic review. Considering the patient-proxy agreement, a review of Sneeuw et al.23 revealed that proxy raters tend to report more HRQoL problems than patients do themselves. However, the subsequent underestimation of patient's HRQoL should be tempered as discrepancies in patient-proxy agreement were only found significant in a minority of cases.23

There is a growing interest to take into account HRQoL measures in patients receiving enteral tube feeding in hospital and among outpatients as well.12,37 Enteral tube feeding refers to the delivery of complete nutrition into the stomach. The tube can be placed through the nose via the esophagus into the stomach (nasogastric feeding or NG-feeding) or directly into the stomach through the skin on the abdomen (percutaneous gastrostomy (PEG) feeding or percutaneous endoscopic gastrostomy-feeding).11 Enteral tube feeding is an alternative for patients with insufficient oral food intake8 and is most commonly (64.1%)9 used in patients with dysphagia.9,11 Dysphagia or swallowing disorder refers to any dysfunction in the swallowing process as a result of anatomical or physiological deficits in the mouth, pharynx, larynx and/or esophagus.38,39 Enteral feeding therapy can benefit patients with dysphagia, who cannot meet their nutritional needs through any other intervention.9 Although most commonly used in patients with dysphagia, also other conditions can benefit from enteral tube feeding, such as patients with gastrointestinal dysfunction, patients with mental health disorders, patients with psychological problems and patients with increased nutritional requirements.11

The provision of enteral feeding can yield nutritional benefits11,40 and allow weight maintenance9 in all populations (infants,41 children,42 adults43 and the elderly9). Some patients with rapidly progressive diseases can also benefit from enteral tube feeding in earlier stages of their disease progress. For example, enteral tube feeding can relieve pressure to eat at mealtimes and can reduce mealtime duration.18 It may also reduce accelerated weight loss due to poor caloric intake.44 However, enteral feeding may also have downsides, for example enteral feeding can result in dependency upon the feeding tube.9 The subsequent psychological and social aspects9,43 of this dependency may affect the patient's HRQoL in a negative way.43 For instance, patients receiving enteral feeding sometimes report feelings of discomfort while dressing and washing, as well as uncomfortable feelings about their body image.12 Also, loss of social interaction associated with feeding (e.g. not visiting their family and not going out because of tube feeding) can take place.9 In a study of Roberge et al.,45 feelings of depression were reported in some patients receiving enteral feeding. Though the feeding technique is typically physically well tolerated,45 some consequences of enteral feeding may occur such as local wound complications, leakage around the insertion site, tube occlusion and increased reflux.46,47 Aspiration pneumonia, which refers to the inhalation of oropharyngeal material into the larynx and lower respiratory tract,48 can also be a common complication.49 Finally, in some populations, enteral feeding is associated with an increased mortality rate, both in hospital50 and in the community.51 Evaluating the impact of enteral feeding on the patients' HRQoL is therefore essential and should be taken into account when considering enteral feeding.

A literature review related to the topic of this systematic review has been previously published.12 Brotherton et al.12 reviewed studies that measured the impact of enteral tube feeding on adult patient's HRQoL to identify feeding-related factors that influence patient's HRQoL. This literature review exposed a lack of specific HRQoL assessment tools.12 Our systematic review will evaluate the psychometric properties of PROMs and proxy-reported outcome measures that assess HRQoL among patients receiving enteral feeding. Based upon an initially limited search in MEDLINE, the Cochrane Database of Systematic Reviews and JBI COnNECT+ in February 2016, to our knowledge there are no systematic reviews that have reviewed the psychometric properties and the clinical utility of PROMs and proxy-reported outcome measures assessing HRQoL among patients receiving enteral feeding.

The objective of this systematic review is to evaluate the psychometric properties (i.e. validity, reliability and responsiveness) and the clinical utility of PROMs and proxy-reported outcome measures (questionnaires or structured interviews) that assess HRQoL among patients receiving enteral feeding to make recommendations for its use in clinical practice and research.

Definitions

To fully understand the psychometric properties, some terms should be defined52,53: (Note that these descriptions are also applicable for the proxy-reported outcome measures.)

Validity52 can be defined in terms of:

* Content validity refers to the extent to which the different domains of the PROM are comprehensively sampled by the items in the questionnaire.

* Criterion-related validity measures the extent to which scores on a particular PROM are related to a gold standard. For most PROMs, it is not possible to measure criterion validity because there is no gold standard.34 A gold standard only exists when a shorter version of the PROM is tested against its complete version. This entire version of the PROM is then considered the gold standard.30

* Construct validity examines the extent to which the subscales on a particular PROM are correlated with subscales from other PROMs that assess similar constructs. These comparisons are based on theoretically derived hypotheses concerning the constructs that are being measured. These constructs and hypotheses should be defined in the studies of the PROMs being reviewed. Therefore, the constructs will be used as defined in the studies to evaluate construct validity.

* Floor and ceiling effects: In the case of floor and ceiling effects, extreme items are missing in the lower or upper end of the scale. Consequences of these effects are reduced reliability (no ability of the questionnaire to distinguish between patients with highest and lowest scores) and limited responsiveness (changes in subjects cannot be measured).

Reliability encompasses "reproducibility"52 and "internal consistency":53

Reproducibility can be defined in terms of "agreement" and "reliability":52

* Agreement refers to the degree to which the scores on repeated measures are close to each other, more specifically the absolute measurement error. The following estimates of agreement are frequently used:Standard error of measurement (SEM) = measurement error.Systematic differences should be considered as part of the SEM because they should be distinguished from real changes (i.e. due to treatment). Therefore, SEMagreement is preferred.Standard error of measurement could be converted in the smallest detectable change (SDC). The SDC is the smallest within-person change in score that - with P < 0.05 - can be interpreted as a "real change" in one individual (SDCind) with SDCind = 1.96 x [square root]2 x SEM or for a group SDCgroup with SDCind/[square root]n.Limits of agreement (LOA): "equal the mean change in scores of repeated measurements >= (mean change) +/- 1.96 x SDchange".54(p.37) SDchange = standard deviation of these changes.Minimal important change (MIC): "the minimal amount of change in the (sub)scale that is considered to be important".54(p.37)

* Reliability refers to the degree to which patients can be distinguished from each other, despite measurement errors (relative measurement error). Reliability is defined as "the proportion of the total variance in the measurements which is because of true differences among patients".55(p.743) The "true" does not refer to the accuracy of the score, but to its consistency. As previously reported,53 it is important to note that "true" should be seen in the context of classical test theory, as one of the two components of any observation, namely the "true score" and the "error associated with the observation".55(p.743) "True" is defined by Mokkink et al. as "the average score that would be obtained if the scale were given an infinite number of times".55(p.743) The preferred estimates of reliability are the intraclass correlation coefficient (ICC) for continuous measures33,52,53 and weighted Cohen's kappa for ordinal measures.52,53

Internal consistency refers to the extent to which items in a (sub)scale are intercorrelated and thus measuring the same concept. Factor analysis is applied to assess whether the items form one scale/dimension or more than one. After assessing the number of (homogeneous) (sub)scales, each (sub)scale should be measured by Cronbach's alpha.52

Responsiveness52 assesses the ability of a PROM to detect clinically important changes over time, even if these changes are small. Responsiveness is a measure of longitudinal validity and should therefore be examined by testing predefined hypotheses about expected correlations between changes in measures or expected differences in changes between known groups. The PROM should be able to distinguish clinically important change from measurement error. The following estimates of responsiveness are frequently used:

* Responsiveness ratio (Guyatt's RR): The degree in which the clinically important change (MIC) is related to the between-subject variability in within-subject changes in stable subjects (SDchange). Terwee et al.52 recommend the RR to be at least 1.96. "At the value of 1.96, the MIC equals the SDCind (= 1.96 x SDchange)".54(p.37)

* Area under the receiver operating characteristics curve or AUC: The ability of a questionnaire to distinguish between patients who have changed and who have not, based on an external criterion.

Clinical utility is defined in terms of "interpretability" and "feasibility" to complete the PROM and the proxy-reported outcome measure. It is considered as an important characteristic of an instrument for its use in research or clinical practice.56

* Interpretability is the extent to which a qualitative meaning can be assigned to quantitative scores. Interpretability can be determined by mean and SD for various subgroups and by determining MIC.52

* Feasibility is the facilitation in completing the PROM, the way of administration, time duration, the patient's burden to complete the PROM and/or the proxy's burden to complete the proxy-reported outcome measure.

Inclusion criteria

Types of participants

The current review will consider all studies including patients of all ages (infants, children, adults and elderly), who receive enteral tube feeding as a consequence of dysphagia or because of other conditions, and their proxies.

Intervention/phenomena of interest

This review will consider primary studies that report on:

* Multidimensional PROMs that assess HRQoL among patients receiving enteral feeding.

* Multidimensional proxy-reported outcome measures that assess HRQoL among patients receiving enteral feeding.

Batteries of scales will be excluded because these are not multidimensional HRQoL instruments.

Outcomes

The current review will consider studies that report on the clinical utility and psychometric properties concerning validity, reliability, responsiveness of the PROMs and the proxy-reported outcome measures of interest. It is likely that not all of these outcomes will be present in each of the included articles. This review will include studies that report on at least one of the following outcomes involving validity, reliability, responsiveness and/or clinical utility.

For validity, studies reporting on at least one of the following outcomes will be included:

* Content validity: Purpose of the instrument: definition of HRQoL related to enteral feeding and the constructs being measured, target population: adults/children/infants/elderly, mode of administration: self-administered/interview based, number of domains, number of items per scale/domain, scales (single-item and multiple-item scales), items related to enteral feeding, item selection (target population and experts or investigators were involved in item selection), range of scores, response options and interpretability of the items (clearly, simple items; recall time = the time period to which the questions refer).

* Criterion-related validity: Pearson's or Spearman correlation coefficient or other appropriate statistics.

* Construct validity: Constructs and hypotheses are specified in advance, correlation coefficients (i.e. Spearman or Pearson's correlation coefficient) or appropriate statistics according to the stated hypotheses.

* Floor and ceiling effects: Percentage (%) of respondents that achieve the lowest or highest possible score.

For reliability, studies reporting on at least one of the following outcomes will be included:

* Reproducibility: As defined in the background section, reproducibility encompasses "agreement" and "reliability". Therefore, studies reporting on at least one of the following outcomes of agreement and/or reliability will be included:Agreement: SDC (SDCind when applied in individuals and SDCgroup when applied in groups), LOA, convincing arguments that agreement is acceptable, and MIC.Reliability: ICC for continuous measures and weighted Cohen's kappa coefficient for ordinal measures.

* Internal consistency: Factor analysis, sample size and Cronbach's alpha.

For responsiveness, studies reporting on at least one of the following outcomes will be included: SDC, MIC, LOA, RR and AUC.

Clinical utility is defined in terms of interpretability and feasibility to complete the PROM and the proxy-reported outcome measure. Studies that report on at least one of the following outcomes will be included:

* Interpretability: Mean, SD and MIC.

* Feasibility: Time to administer, mode of administration, facilitation required to complete the PROM, patient's burden to complete the PROM and proxy's burden to complete the proxy-reported outcome measure.

Types of studies

The current review will consider any study that reports on multidimensional PROMs and proxy-reported outcome measures that assess HRQoL among patients receiving enteral feeding and addresses validity, reliability, responsiveness and clinical utility.

Search strategy

The search strategy aims to find both published and unpublished studies. A three-step search strategy will be utilized in this review. An initial limited search of MEDLINE and CINAHL will be undertaken followed by analysis of the text words contained in the title and abstract, and of the index terms used to describe the article. A second search, using all identified keywords and index terms, will then be undertaken across all included databases. Third, citations and the reference list of all selected studies and relevant reviews will be checked to identify any studies that were missed in our searches.

Four concepts will be used to determine the search terms: (1) enteral feeding, (2) PROM, (3) QoL and (4) validity. For each concept, both keywords (e.g. MeSH terms for PubMed) and free-text words will be used. All search terms within each concept will be combined with "OR". Then, all concepts will be combined by means of "AND". The full search strategy in PubMed is represented in Appendix I.

The databases to be searched include:

MEDLINE (via PubMed), Embase, CINAHL and PROQOLID.

The search for unpublished studies will include:

Google Scholar, MedNar and relevant homepages of patient-reported and proxy-reported outcome tools. Homepages of international and national societies for PROMs, such as the International Society for QoL Research, will be searched. Dissertations and theses will be hand searched; more specifically, three libraries of university colleges for dieticians (VIVES University College Bruges, HoGent campus libraries and Thomas More University College) will be hand searched. The search for dissertations and theses will also be performed by using the ProQuest database.

The search for unpublished studies and screening references will be independently performed by two reviewers (I.S. and A.W.).

Selecting studies

Selecting studies will also be independently performed by the two reviewers (I.S. and A.W.). The selection will be conducted in two stages: first, both reviewers will screen titles and abstracts against the inclusion criteria. All articles that potentially fulfill all criteria will be retrieved. Subsequently, the full texts will be screened using the same inclusion criteria. Studies published in English, German, French and Dutch will be considered for inclusion in this review, as these languages are mastered by the review team. All studies published prior to the start of this systematic review will be considered for inclusion in this review. Another search, as an update, will be conducted between the period of starting the review and finishing it.

Differences in opinion will be resolved by discussion. If no consensus is reached, a third reviewer will be asked to judge whether or not the article fulfills the inclusion criteria. A PRISMA flow chart will be used to present included and excluded studies.57

Assessment of methodological quality

Papers selected for retrieval will be assessed by two independent reviewers for methodological quality prior to inclusion in the review.

Study quality will be assessed using the COnsensus-based standards for the selection of health status Measurement INstruments (COSMIN) checklist.14,58 This is a validated and internationally accepted tool for the assessment of the methodological quality of studies.14 The checklist focusses on Health-Related Patient-Reported Outcomes (HR-PROs). The COSMIN tool consists of 12 separate sections ("boxes"), of which 10 will be used to evaluate whether a study meets the standard for good methodological quality. Of these, nine boxes represent psychometric properties, and one box focusses on the methodological quality assessment of studies on interpretability of HR-PRO tools.14 The remaining two boxes contain general requirements (i.e. Generalizability box) and will not be used in this review.

The COSMIN tool is a modular tool, which means that the number of boxes to fill out depends on the number of psychometric characteristics reported in a study. It may not be necessary to complete the whole checklist when assessing the methodological quality of a particular study. When a study, for example, only assesses responsiveness and criterion validity, two boxes need to be scored. Concerning the validation assessment of an instrument in different language groups, the same box may need to be scored multiple times in case the design of the study was different among countries.14

The COSMIN tool uses a 4-point scale (excellent, good, fair and poor) to rate each psychometric property for a particular study.58 A score can be obtained per box by taking the lowest rating of any item in a box ("worse score counts"). Finally, an overall score, which is determined by the lowest box score, can be obtained to rate the methodological quality of the study. The Interpretability box should be completed only for studies that investigate the interpretability of an instrument. The Interpretability box is also recommended to be used in systematic reviews of psychometric properties as a data extraction tool to extract all available information on interpretability issues (i.e. MIC), and therefore, no scoring was developed for that box.58 Appendix II outlines the COSMIN tool to be used.

Any disagreements concerning assessment of study quality that arise between the reviewers will be resolved by discussion with a third reviewer.

Studies with low quality, defined as "poor" scores in all boxes, will be excluded from the review.

Data extraction

Data will be extracted from papers included in the review by means of an attached self-developed data extraction tool (Appendix III). The data extracted will include specific details about the study characteristics, instrument characteristics and the outcomes of significance to the review question and specific objectives. The "Generalizability" box from the COSMIN tool will also be included in the data extraction tool.14

The following data will be extracted from included studies:

* Study characteristics: Authors, title and year of publication, aim of the study, study design, patient population characteristics (age, gender, medical history, etiology and/or comorbidities, method of enteral feeding, time on receiving enteral feeding and underlying etiology of receiving enteral feeding), characteristics of controls (age and gender), characteristics of proxy (type of proxy, age and gender) and sample size (= size of the population [patient, controls and proxy] in which the instrument was applied).

* Descriptive characteristics of the instrument: Name of the instrument, purpose of the instrument (definition of HRQoL related to enteral feeding and definition of the constructs), study/target population, mode of administration (self-administered or interview based), facilitation required to complete the PROM, number and type of scales/domains, number of items per scale/domain, total number of items, response options, range of scores and time to administer, and items related to enteral feeding.

* Outcomes: The relevant data on the psychometric properties and the clinical utility will be extracted, as previously reported, under the "Inclusion criteria" and the "Background".

Two reviewers will extract all data independently. Any differences in opinion will be discussed and resolved in a consensus meeting between the two reviewers.

Authors of primary studies will be contacted by electronic mail in case of missing information or to clarify unclear data.

Data synthesis

To judge the psychometric properties of the PROMs and the proxy-reported outcome measures, the quality criteria proposed by Terwee et al.52 will be used. These criteria are based on the criteria of the "Scientific Advisory Committee of the Medical Outcomes Trust"59 and have explicit criteria for what constitutes good psychometric properties. The checklist was applied by Barten et al.60 for the psychometric evaluations of both questionnaires and semi-structured interviews. All psychometric properties of the PROMs will be rated as "+" (positive rating), "-" (negative rating), "?" (indeterminate rating) and "0" (no information available).52

For validity, the following quality criteria will be used:

* Content validity: Purpose of the instrument: definitions of HRQoL in enteral feeding and the constructs being measured, target population: adults/children/infants/elderly, mode of administration: self-administered/interview based, multiple domains, multiple items per scale/domain, items related to enteral feeding, item selection: target population and experts or investigators were involved in item selection, large range of scores, interpretability of the items (clearly, simple items), recall time (= the time period to which the questions refer) is stated and justified.

* Criterion-related validity: Good validity if correlation with gold standard >=0.70. Convincing arguments that gold standard is gold should be described.

* Construct validity: Constructs and hypotheses are specified in advance, correlation coefficients (i.e. Spearman or Pearson's correlation coefficient) or appropriate statistics according to the stated hypotheses. Good validity if statistics (i.e. Spearman's rho) show that at least 75% of the results are in correspondence with these hypotheses, in (sub)groups of at least 50 patients or 50 proxies.

* Floor and ceiling effects: Absence of floor and ceiling effects: <=15% of respondents achieve the lowest or highest possible score in a sample of at least 50 patients or 50 proxies.

For reliability, the following quality criteria will be used:

* ReproducibilityAgreement: Good agreement if MIC < SDC or MIC outside the LOA or convincing arguments that agreement is acceptable. A sample size of at least 50 patients or 50 proxies is needed to assess agreement.Reliability: Good reliability if ICC or Weighted Cohen's kappa is at least 0.70 in a sample size of at least 50 patients or 50 proxies.

* Internal consistency: Good internal consistency if factor analysis is first applied on adequate sample size (7 x number of items and >100) AND Cronbach's alpha(s) calculated per dimension AND Cronbach's alpha(s) between 0.70 and 0.95.

For responsiveness, good responsiveness is obtained if SDC < MIC OR MIC outside the LOA OR RR > 1.96 OR AUC >= 0.70.

Clinical utility is defined in terms of interpretability and feasibility to complete the PROM and the proxy-reported outcome measure. Therefore, the following criteria will be used:

* Interpretability: Interpretability is good if mean and SD of scores for at least four relevant subgroups of patients are presented AND MIC is defined in a sample size of at least 50 patients or 50 proxies.

* Feasibility: Based on the arguments of the studies, it will be clear if the PROM and the proxy-reported outcome measure is feasible in terms of time to administer, mode of administration (interview based or questionnaire), facilitation required to complete the PROM, patient's burden to complete the PROM and proxy's burden to complete the proxy-reported outcome measure.

The data will be presented in a narrative text including the psychometric properties and the contextual data, both presented in the included studies. A meta-analysis will not be performed, because it is inappropriate for determining the psychometric properties of the PRO-tools.

The results of the review will be presented for each review question separately. Tables will be used in the presentation of the results.

Appendix I: Search strategy

Appendix II: Critical appraisal tool (COSMIN checklist)

(Terwee et al.58)

Appendix III: Data extraction tool

References