Article Text

Download PDFPDF

Original research
A systematic review of patient-reported outcome measures (PROMs) in cystic fibrosis
  1. Irushi Ratnayake,
  2. Susannah Ahern,
  3. Rasa Ruseckaite
  1. Department of Epidemiology and Preventive Medicine, Monash University, Clayton, Victoria, Australia
  1. Correspondence to Dr Rasa Ruseckaite; Rasa.Ruseckaite{at}monash.edu

Abstract

Background To determine patient-reported outcome measures (PROMs) which may be suitable for incorporation into the Australian Cystic Fibrosis Data Registry (ACFDR) by identifying PROMs administered in adult and paediatric cystic fibrosis (CF) populations in the last decade.

Methods We searched MEDLINE, EMBASE, Scopus, CINAHL, PsycINFO and Cochrane Library databases for studies published between January 2009 and February 2019 describing the use of PROMs to measure health-related quality of life (HRQoL) in adult and paediatric patients with CF. Validation studies, observational studies and qualitative studies were included. The search was conducted on 13 February 2019. The COnsensus-based Standards for the selection of health Measurement INstruments Risk of Bias Checklist was used to assess the methodological quality of included studies.

Results Twenty-seven different PROMs were identified. The most commonly used PROMs were designed specifically for CF. Equal numbers of studies were conducted on adult (32%, n=31), paediatric (35%, n=34) and both (27%, n=26) populations. No PROMs were used within a clinical registry setting previously. The two most widely used PROMs, the Cystic Fibrosis Questionnaire—Revised (CFQ-R) and the Cystic Fibrosis Quality of Life Questionnaire (CFQoL), demonstrated good psychometric properties and acceptability in English-speaking populations.

Discussion We found that although PROMs are widely used in CF, there is a lack of reporting on the efficacy of methods and timepoints of administration. We identified the CFQ-R and CFQoL as the most suitable for incorporation in the ACFDR as they captured significant effects of CF on HRQoL and were reliable and valid in CF populations. These PROMs will be used in a further qualitative study assessing patients’ with CF and clinicians’ perspectives toward the acceptability and feasibility of incorporating a PROM in the ACFDR.

PROSPERO registration number CRD42019126931.

  • public health
  • cystic fibrosis
  • mental health
http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Strengths and limitations of this study

  • As per our knowledge, this is the first systematic review evaluating patient-reported outcome measures (PROMs) in adult and paediatric cystic fibrosis (CF) populations.

  • This review involves a rigorous and extensive search of medical databases using clearly defined inclusion criteria and distinctly outlines how items will be selected and abstracted.

  • The study assesses the most relevant and acceptable PROM for the context of a CF clinical registry.

  • A limitation of this study is that the search was not conducted outside of medical databases, therefore may not capture studies examining PROM use in CF that are not published in peer-reviewed journals.

Introduction

Cystic fibrosis (CF) has undergone significant changes in the last few decades. In the mid-1900s, the majority of patients with CF did not survive beyond infancy. Now, over half of patients are adults1 and life expectancy exceeds 40 in most developed countries.1 The changing demographics of CF has led to new challenges in both disease management and clinical research. Treatment burden has increased2 such that treatments currently require 2–4 hours a day.3 The growing adult population encounters more difficulties balancing symptom and treatment burden of the disease with work, education or family demands.4 5 Therefore, there is an increasing requirement to examine and manage psychosocial impacts of CF.3 Another challenge is posed by the relative healthiness of the modern CF population resulting in traditional endpoints in clinical trials, such as forced expiratory volume in 1 s (FEV1) and frequency of pulmonary exacerbations, having reduced sensitivity.6

A possible solution to these challenges is to monitor and collect data on health-related quality of life (HRQoL).7 HRQoL is ‘an individual’s perception 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’.8 It encompasses physical health, social networks and relationships, psychological health and functional capacity.8 As HRQoL is subjective, it can be described using patient-reported outcome measures (PROMs).9 PROMs are standardised sets of questions completed by patients without clinician interpretation.9 PROMs have been used in a range of settings, from enhancing clinician–patient interaction to supporting health policy creation and economic analysis.10 They are widely used in research; in observational studies to describe the impact of a disease on daily functioning, as tools for cost analysis of medical interventions2 and the US Food and Drug Administration (FDA) have recommended HRQoL measures be used as outcomes in clinical trials.5

Australian Cystic Fibrosis Data Registry (ACFDR)

The ACFDR has been collecting data on Australian adults and children diagnosed with CF since 1998. In 2017, the ACFDR held records of 3151 patients,11 estimated to be over 90% of Australia’s CF population.4 The registry collects information on patients’ demographics, social functioning, physical health, treatments and mortality. In addition to increasing awareness about Australia’s CF population, the ACFDR has supported interventional and observational research and economic analysis.12 The ACFDR enables national and international benchmarking,12 which has transformed models of care worldwide.4

PROMs evaluating HRQoL have been incorporated in Australian and international clinical registries.13–15 In the USA, PROM information is used to support observational studies that assess the association between patient demographics, disease burden and HRQoL.16 In Sweden, the National Rheumatology Registry enters its PROM data into a database to which patients and clinicians have access, so that patients are empowered to monitor their HRQoL and shared decision-making is enhanced.15 In Australia, PROMs evaluating HRQoL are currently incorporated in a number of state and national registries.17 Information is used to monitor long-term quality-of-life outcomes of treatments and complications,17 to enable clinicians and health services to benchmark outcomes and ensure patient safety,14 and to influence changes in clinical practice.14

Integration of a PROM evaluating HRQoL into the ACFDR will reinforce the patient voice in data collection. PROMs in the ACFDR have the potential to be used for periodic review of aggregate HRQoL over time; to inform quality improvement for health services and clinicians; and for outcome measurement in registry-related clinical trials.10 In order to fulfil these functions, any PROM selected for integration must be comprehensive in capturing all effects of CF on HRQoL. It must also have demonstrated good psychometric properties, be feasible to incorporate in ACFDR data collection and be acceptable to patients.

Aims

The primary aim of this review was to identify PROMs used in adult and paediatric CF populations, to determine any that may be suitable for incorporation into the ACFDR. Secondary aims were to examine:

  • Contexts in which PROMs are currently being used in CF (eg, study design and setting).

  • Methods of administration of PROMs (eg, paper survey, electronic, interview and use of proxy respondents).

  • Assessed or stated psychometric properties of PROMs (eg, reliability, validity and responsiveness).

  • Acceptability of PROMs in adult and paediatric patient populations.

Methods

A protocol for this systematic review was created following the Preferred Reporting Items for Systematic review and Meta-Analysis (PRISMA) guidelines.18

Eligibility and inclusion criteria are described in table 1.

Table 1

Population, Intervention, Comparison, Outcome (PICO) Research Strategy for systematic review

Inclusion criteria

Articles were included according to the following criteria:

  • Study participants of all ages with a prior diagnosis of CF.

  • Inpatients and outpatients.

  • Study designs including quantitative (eg, cohort, longitudinal, prospective, retrospective and validation) and qualitative studies (eg, ethnography and case report).

Exclusion criteria

Articles were excluded according to the following criteria:

  • Published before January 2009.

  • No article available in the English language.

  • Conference abstracts.

  • Editorials and reviews.

  • Randomised control trials (RCT), as the same PROM was used for all and they provided limited additional information on secondary outcomes.

Reviewers searched MEDLINE, EMBASE, Scopus, CINAHL, PsycINFO and Cochrane Library databases on 13 February 2019. The search strategy was adapted to each database and included keywords: ‘patient reported outcome’ OR ‘patient reported outcome measure’ OR ‘self-report*’ OR ‘questionnaire’ OR ‘scale’ OR ‘perception’ OR ‘quality of life’ OR ‘QOL’ AND ‘cystic fibrosis.’ The search was restricted to English language, humans and last 10 years. Online supplemental file 1 describes the search strategy for each database.

Initial screening involved a reviewer reading titles and abstracts of all studies identified by the search. Any studies that clearly did not meet the inclusion criteria were removed. Full texts of remaining studies were then read one author. Another author reviewed each stage of study selection. A number of studies at each stage of the search were recorded using the PRISMA flow diagram.

A data extraction form was constructed to summarise selected studies in line with the outcomes of the systematic review. Information extracted included: type of study, mean age of participants, setting PROM(s) administered, method of administration, timepoints administered PROM(s) used, type of PROM(s), psychometric properties of PROM(s) and acceptability of PROM(s) to patients.

The COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) Risk of Bias Checklist was used to assess methodological quality of included studies. This tool was chosen as it was specifically created for studies using PROMs.19 One reviewer appraised studies using the tool. Items were rated on a 4-point scale denoted as very good, adequate, doubtful or inadequate. The results were summarised into a table presenting the lowest score for each property.19

A descriptive synthesis of results was undertaken, organised thematically by type of PROM and assessing context, administration, acceptability and reliability of each measure. A meta-analysis was not performed as included studies assess different outcomes.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Results

Search results

The search yielded 5671 results. A number at each stage is summarised in figure 1. A final number of 91 studies were included in the review. The data extraction table is presented in online supplemental file 2.

Figure 1

Flowchart of Study Identification and SelectionPROM: Patient Reported Outcome MeasureHRQoL: Health-related Quality of LIfeCF: Cystic FibrosisRCTs: Randomised Controlled Trials

Contexts in which PROMs were used

A large proportion (80%, n=73) of studies identified were of observational study design. Validation studies were the next most frequent, making up 15% (n=14) of all studies. The search also identified two non-RCTs, two qualitative studies and one study describing development of a PROM. Similar numbers of studies were conducted on adults (34%, n=31), children (37%, n=34) or both (29%, n=26) age groups.

Most studies recruited patients from a CF outpatient clinic (61%, n=56). Other studies used patient populations from: RCT data (8%, n=7), inpatients (7%, n=6), longitudinal cohort study data (5%, n=5) and national databases (4%, n=4). No study was conducted using clinical registry data. In 48% (n=44) of studies, PROM instruments were used in cross-sectional observational studies to evaluate whether there was an association between HRQoL and physical factors (eg, sleep and physical fitness), psychological factors (eg, self-esteem and illness perception), social factors (eg, stigma and employment status) or demographic factors (eg, age and gender). Other reasons for using PROMs were to assess HRQoL in a population (18%, n=16) or validate PROMs (18%, n=16).

Mode and method of administration

PROMs were commonly self-reported on paper in clinic for 19% (n=17) of studies. Many studies (14%, n=13) used multiple methods of administration, for example, paper and interview. Less commonly, data were collected using electronic methods for 8% (n=7) of studies. Many studies (55%, n=50) did not state mode or method of PROM administration.

For 43 studies conducted on young children below 13 years of age, the most common method of administration for 33% (n=14) was self-report using instruments specially designed for use in young children. Interviews were used in 28% (n=12) of studies and parents were used as proxy respondents in 23% (n=10) of studies completed on paediatric populations. When studies assessed the degree of agreement between child self-report and parent proxies, they found variable results. While some studies found a high level of agreement in parent–child reports,20 21 others found that parents were better able to report HRQoL in observable domains, such as physical symptoms.22–25 Two studies26 27 noted that parent–child agreement was better for younger children than older.

PROMs were administered once at the beginning of the study for the majority of studies (55%, n=50), which reflects the large proportion of cross-sectional studies. Several PROMs were administered twice (12%, n=11) and 15 (16%) studies applied PROMs longitudinally, between 5 and 12 times. The frequency of longitudinal administration varied from fortnightly28 to 2 yearly.29 Studies did not discuss the benefits of administering PROMs at their chosen frequencies. Dill et al30 applied the Cystic Fibrosis Questionnaire—Revised (CFQ-R) every 3 months and found individual variation in each domain. This was not seen in a study that administered the EuroQol-5 Dimension (EQ-5D) every 8 weeks.31 Abbott et al32 applied the Cystic Fibrosis Quality of Life Questionnaire (CFQoL) to the same patients over 12 years and observed a steady decrease of overall CFQoL Score at 1% per year, which correlated with the decrease in FEV1%.

Acceptability

Two studies assessing patient views towards PROMs found that parent caregivers were satisfied with the questionnaires.33 34 Salek et al3 observed that 76% of patients with CF in their study would be willing to complete the CFQoL at every clinic visit. Overall, as most studies did not report the patient burden of PROMs to their patient populations, this review has found limited information on acceptability of PROMs for patients.

PROMs identified

This review identified 27 different PROMs evaluating HRQoL. These were CF-specific, respiratory-specific, mental health-specific or generic PROMs. Some studies (25%, n=23) used two or more different PROMs. CF-specific PROMs were used more commonly than other types. The most common instrument used was CFQ-R, used in 54% (n=49) of studies.

CF-specific instruments

Table 2 summarises the characteristics of CF-specific PROMs identified in this review.

Table 2

CF-specific PROMs

CFQ-R was the most commonly used PROM in this review. It is widely used as it includes scales for children (6–11 years), adolescents (12–13 years), teens/adults (14+ years) and parents. This PROM is a revised version of the original CFQ.35 The CFQ was developed in France in 199736 and minor revisions were performed by Wenninger et al37 in 2003 due to inadequate psychometric properties found during validation of the German translation. The CFQ-R has been translated into 36 different languages.2 Gancz et al38 reported that the CFQ-R was generally completed in 10–30 min.

Studies demonstrated generally good psychometric properties of the CFQ-R. When considering only the studies in English, internal consistency evaluated by Cronbach alpha ranged from 0.62 to 0.9335 39–41 for adult and child questionnaires and 0.55–0.75 for parent questionnaires.42 Studies reported that the treatment burden, body image and school functioning domains were exceptions.25 35 39 41 Validity was demonstrated by the association between several CFQ-R domains and clinical parameters, in particular FEV130 35 43–47 and body mass index (BMI).46 47 Longitudinal studies have shown that CFQ-R is sensitive to changes to HRQoL with antibiotic treatment48 or over the course of a year.49 Authors suggested that it could predict survival50 and be a determinant for lung transplantation.51 Content validity was acceptable.25 52

The CFQoL was the second most commonly used PROM. It has only been developed for adult populations. Salek et al3 found an average 9-min completion time and that the majority of patients found the instrument acceptable for completion in every clinic appointment. Studies identified in our search described robust psychometric properties of the CFQoL. Reliability measured by Cronbach alpha ranged from 0.72 to 0.9532 53 for all domains. It was correlated with generic measures, Short Form Survey-36 (SF-36) and UK Sickness Impact Profile (UKSIP),3 32 and Shwachman-Kulczycki Score, a clinician reported outcome measure.54 Discriminant validity has been demonstrated by significantly worse CFQoL Scores in patients with CF than in controls.55 Studies demonstrated correlation between CFQoL domains and FEV1;3 32 56 however, one study did not find a significant correlation.57

Other CF-specific PROMs identified included the CFQ, which was the first CF-specific PROM developed and has child, teen/adult and parent versions.35 Studies demonstrated good internal consistency of most domains,27 58 with the exception of treatment burden domain in all versions, social functioning domain in child and adult and eating and digestion domains in adult and parent versions.27 The DISABKIDS-CF Module, which was developed for children, was used in two studies conducted in Brazil. Good internal consistency was demonstrated34 59 but one study found a ceiling effect and low test–retest reliability.59 Several CF-specific PROMs were developed or initially validated during the last decade. These included the CF Respiratory Symptom Diary (CFRSD),26 CF Symptom Progression Survey,33 CF Symptom Diary60 and the Respiratory Symptoms in CF.61

Respiratory-specific PROMs

Several HRQoL PROMs developed for chronic respiratory conditions were used in CF. These included the Leicester Cough Questionnaire (LCQ),61 62 St George’s Respiratory Questionnaire (SGRQ),63 64 the Sinus and Nasal Quality of Life Survey (SN-5),65 66 the Sino-Nasal Outcome Test (SNOT-22)67 and the Liverpool Respiratory Symptom Questionnaire (LRSQ).6 The SN-5 and SNOT-22 exclusively assess sinus symptoms.65–67 The other respiratory PROMs, LCQ, SGRQ and LRSQ were originally piloted in patients with asthma68 or chronic cough.69 The LCQ, SGRQ and LRSS demonstrated acceptable reliability6 58 70 and were found to correlate with CFQ-R domains61 62 and lung function tests.6 63 However, two studies found ceiling effects with the LCQ.61 62 Reliability of the SN-5 and SNOT-22 was not assessed, but SNOT-22 demonstrated floor effects67 and the validity of SN-5 has not been assessed in CF.66

Mental health-specific PROMs

The most common mental health-specific PROM identified was the Hospital Anxiety Depression Scale (HADS), which was used in eight observational studies in Europe and the USA. The instrument was reported to take 15–20 min to complete.71 Studies found good reliability assessed by Cronbach alpha.39 72 Yohannes et al71 found good test–retest reliability and correlation with CFQoL. The HADS was used to show increased anxiety and depression in patients with CF compared with the non-CF population.73 Other HRQoL surveys focused on mental health identified were the Patient Health Questionnaire-9, General Health Questionnaire and Generalised Anxiety Disorder-7. Each was used in one study and found to have acceptable reliability;64 74 however, validity was not assessed.

Generic instruments

Table 3 describes the characteristics of generic instruments included in this study.

Table 3

Generic PROMs

The most common generic instrument was the EQ-5D Questionnaire, which was developed to enable economic evaluations based on HRQoL Scores. It has five dimensions and includes EuroQol-5 Dimension-3 Level (EQ-5D-3L) version which has three response options, EuroQol-5 Dimension-5 Level (EQ-5D-5L) version which has five response options, and EuroQol-5 Dimension-Youth (EQ-5D-Y) which has been designed for children and adolescents. All three versions of the PROM were used in this review21 31 43 70 75–77 This review found EQ-5D-3L was reliable43 and correlated with CFQ-R.76 EQ-5D-5L distinguished HRQoL differences in CF and non-CF populations75 and was sensitive to change during pulmonary exacerbation.76 However, studies found a large proportion of patients reporting no problems with EQ-5D-3L and EQ-5D-Y,31 70 demonstrating that the PROMs may not be sensitive in collecting HRQoL data from patients with CF.

A similar finding was observed in the SF-36, which was used in four European studies on adult populations.50 55 63 64 The instrument demonstrated robust reliability with Cronbach alpha of 0.9564 and discriminated between CF and non-CF populations.55 64 However, Abbott et al50 found a high proportion of participants reporting no problems and that the instrument was less sensitive to clinical deterioration than the CFQoL.

The Paediatric Quality of Life Inventory (PedsQL) is a generic HRQoL instrument developed for children with paediatric cancers.78 The PedsQL demonstrated good internal consistency,20 discriminant validity comparing asthma and CF and correlated with BMI.48 Other generic HRQoL PROMs described in adult populations were World Health Organization Quality of Life Scale (WHOQOL-BREF),54 Core Outcome Measures Tool,40 UKSIP,3 KINDL and the Quality of Life Profile for the Chronically Ill.63 These instruments were each used in one observational study. Psychometric properties were not evaluated in included studies.

Risk of bias

The COSMIN Risk of Bias Checklist is designed to critically appraise studies evaluating the reliability or validity of PROMs. A number of studies in this review did not validate instruments for their study population and relied on previous reliability and validity statistics for the PROM used. Therefore, these studies were not critically appraised. The results of critical appraisal are summarised in online supplemental file 3.

Critically appraising articles using the COSMIN Checklist enables reviewers to discern whether psychometric properties have been evaluated using appropriate methodology. From this, reviewers can determine whether the information reported on psychometric properties of PROMs is trustworthy. For example, the second most commonly evaluated property ‘Internal Consistency’ frequently received optimal scores, demonstrating that researchers were in line with COSMIN recommendations and that ‘Internal Consistency’ reported is generally reliable. However, the most commonly reported property ‘Hypothesis Testing for Construct Validity’ received variable scores, demonstrating a lack of reliability in interpreting this statistic.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research

Discussion

Contexts in which PROMs were used

This review identified that PROMs are used in a variety of settings in CF. PROMs were most commonly used in observational studies, where they assessed the impact of physical, psychological, social or demographic variables on HRQoL. This review did not find studies describing implementation of a PROM in a clinical registry or which used clinical registry data.

Some studies were developing PROMs or undertaking validation of new PROMs. This may suggest that existing PROMs are not meeting researchers’ requirements. Limitations of existing PROMs may include the length of commonly used CF-specific PROMs, which could reduce patient compliance and increase data entry burden. Newly developed CF-specific PROMs identified in this study were substantially shorter,33 61 79 demonstrating that researchers require less burdensome CF-specific PROMs. Another limitation may be inadequacy of paediatric measures as currently, no validated PROMs exists to measure data in 0–6 year olds.26 This review identified researchers validating or developing PROMs for younger patient populations.26 33 59

Mode and methods of administration

The mode of administration of the selected PROM will be a major determinant of patient adherence and completion rates.9 Studies in this review used paper based methods most frequently. However, electronic or online administration is reported to have higher patient adherence,9 avoid the need for manual data entry and be more cost effective in the long term than paper methods.80

For paediatric populations, the most common method of administration was self-reporting, using instruments specially designed for use in children. Proxy reporting was uncommon and studies investigating the consistency of parent and child results found that it was better for observable symptoms22–25 and younger children.26 27 Edwards et al26 hypothesised this finding was because parents are more involved in care for younger children and therefore have a better understanding of their HRQoL.

This review demonstrated the advantages of longitudinal PROM collection, as associations between physical and sociodemographic characteristics and quality of life were seen in studies undertaken over a decade,29 32 which were not seen over 12-month or 18-month periods.30 However, where PROMs captured longitudinally, there was a range of frequencies of administration, demonstrating a lack of consensus on the most appropriate time required between PROM administration. Studies generally did not report information on the effectiveness of the frequency of administration in demonstrating changes in HRQoL. Further evaluation of the most useful and acceptable timepoints of administration must be conducted prior to incorporation of a PROM into the ACFDR.

PROMs identified

Our review identified that PROMs developed specifically for CF are more commonly used for patients with CF than generic PROMs. Generic PROMs, which ask about health domains relevant to everyone, have the advantage of applicability across all populations.14 Therefore, they were used to compare different diseases and in cost analysis studies.21 75 CF-specific PROMs include an assessment of CF symptoms that are not relevant in non-CF populations,14 therefore have comparatively limited uses in health policy. However, this review found that CF-specific PROMs are more responsive to changes in health9 and better correlated to clinical parameters22 81 compared with generic PROMs. Significant ceiling effects found using EQ-5D31 or SF-3650 suggest these generic instruments are not capturing problems faced by the CF population. Specific PROMs can, therefore, give more clinically relevant information than generic2 9 and better compare outcomes within CF populations.82

A number of symptom-specific PROMs were identified in our review that focused on respiratory symptoms or mental health. As CF affects all four domains of HRQoL, physical health, psychological health, social relationships and functional capacity, the use of these symptom-specific PROMs will not provide the comprehensive assessment of HRQoL required by the ACFDR. While it is important to assess depression and anxiety in CF, evaluating only these symptoms may give a limited understanding of the effect of CF on overall HRQoL.

Choosing a PROM for the ACFDR

The ACFDR was established to facilitate varying research methodologies and impart accurate information on the current outcomes of Australia’s CF population.4 One of its key functions, providing feedback on outcomes for clinicians and health services, is critical for the ongoing improvement of care.83 The inclusion of CF-specific domains in the chosen tool is, therefore, essential, as these domains will be most directly affected by changes in treatment and will be the most useful information to feedback to clinicians. Similarly, CF symptom information will be relevant for pharmaceutical companies or researchers following up the long-term outcomes of treatment and complications. In addition, ensuring that PROM data captures all domains of HRQoL will enable it to be widely used in research. Therefore, it is most appropriate to include a CF-specific PROM.

After evaluating PROMs based on the predetermined criteria for incorporation into the ACFDR; comprehensiveness, robust psychometric properties, feasibility and acceptability, the CFQ-R and CFQoL come closest to achieving this criteria. They are comprehensive as they include both general and CF-specific domains. This review establishes satisfactory psychometric properties for these two instruments.

A major limitation to incorporating either PROM into the ACFDR is the length of the instruments, which may dissuade patients from participating in data collection or completing the instrument. This poses a difficulty, as a large amount of missing data may cause collection of PROMs to become ineffectual. However, if patients believe that measuring HRQoL is useful to them, they may complete the instrument regardless of its length. At the Duke Cancer Institute in the USA, patients in solid tumour clinics have less than 5% missing data for a survey with median completion time of 11 min.81 Communication of the beneficial outcomes to patients, clinicians and researchers of HRQoL data collection may influence patients to regard completing the instrument as important to them.

Both selected CF PROM Tools are also the oldest specific instruments developed in CF.84 85 There is a possibility of longevity bias if these PROMs are most commonly used in CF because they are well known, rather than superior instruments. Another concern is that as the demographics and outcomes of CF have changed considerably since these instruments were first developed, their relevance to the current population may be limited. In addition, the PROM selected for the ACFDR must also be applicable to future populations, so that registry data collection remains consistent.81 However, both the CFQ-R and CFQoL demonstrated the most robust psychometric properties of all the PROMs and recent studies that used these instruments reported no requirement for modification,28 56 86 87 so it can be concluded they are currently relevant to the CF population.

Limitations of the review

This systematic review has several limitations. Researchers did not conduct a grey literature search, which may have resulted in reduced information extracted on PROMs use in CF registries. However, it may also occur because there is limited reporting on PROM incorporation in CF registries. Researchers excluded RCTs from this review, which limited our results on the extent of PROM use in CF research. Initial searches for this topic identified that RCTs only used the CFQ-R and did not report administration methods or psychometric properties of PROMs. Therefore, we felt that excluding RCTs enabled a focus on observational studies, which have data collection methods more closely resembling clinical registries and included more information on secondary outcomes of this study.

Another limitation is the lack of information identified on the views of patients with CF and caregivers on the relevance of PROMs, their clarity and structure, ease of use and whether completing PROMs was emotionally burdensome. Researchers found very few studies reported data on acceptability, such as response rates, administration time or qualitative perspectives of patients or caregivers on PROMs. Therefore, limited information on that outcome is described in this review. This information is important because symptoms and treatments are already emotionally and physically demanding, therefore a time-consuming and difficult questionnaire should not be imposed on patients. In addition, giving a questionnaire that is meaningful to patients and clinicians is essential to ensure compliance and guarantee complete data collection.

In order to overcome these limitations, researchers will conduct a further feasibility and acceptability study to identify patient and clinician perspectives toward incorporation of either the CFQ-R or CFQoL into the ACFDR.

Conclusion

This review aimed to identify whether existing HRQoL instruments are suitable for incorporation in the registry and to gain an understanding of the use of PROMs in CF. We found that PROMs are widely used in CF, but there is a lack of reporting on methods of administration and timepoints. We have identified two PROMs appropriate for ACFDR that will be used in a further qualitative study of patients with CF and clinicians, to gain their perspectives on the instruments and the feasibility of incorporating a PROM into the ACFDR.

References

Footnotes

  • Contributors All authors developed the protocol for this systematic review. IR conducted the screening of studies, data extraction and critical appraisal. RR reviewed each stage of study selection. All authors assisted in the interpretation and write up of results. All authors approved the final version to be published.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement Additional data are available upon reasonable request.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.