Background/aims The purpose of this systematic review was to identify the frequency and type of patient-reported outcome measures (PROMs) used in recent randomised controlled trials (RCTs) for age-related macular degeneration (AMD).
Methods The authors conducted a systematic search between January 2010 and November 2013 in MEDLINE, EMBASE, Scopus, Cochrane Library (Central) and the clinical trials registries (http://www.controlled-trials.com and http://www.ClinicalTrials.gov) according to defined inclusion criteria (RCTs on AMD in English). Two independent reviewers evaluated studies for inclusion. One reviewer extracted data of included studies, and a second masked reviewer assessed 10% to confirm accuracy in data collection. Reference lists of included papers and appendices of relevant Cochrane systematic reviews were scanned to identify other relevant RCTs. Information collected on extracted outcomes was analysed using descriptive statistics.
Results Literature and registry search yielded 3816 abstracts of journal articles and 493 records from trial registries. A total of 177 RCTs were deemed to have met inclusion criteria. Of the 858 outcomes reported, 38 outcomes were identified as PROMs (4.4%). Of the 177 RCTs examined, PROMs were used in 25 trials (14.1%). The National Eye Institute Visual Function Questionnaire-25 was the most frequently used PROM instrument (64% of RCTs with PROMs included).
Conclusions This review highlights that a small proportion of AMD RCTs included PROMs as outcome measures and that there was a variety in the instruments used.
Trial registration number The systematic review was registered in the PROSPERO database for systematic reviews, registration number CRD42014010040.
Statistics from Altmetric.com
Age-related macular degeneration (AMD) is a serious eye disorder prevalent in older adults.1 Recent advances in treatment over the past decade have led to a significant increase in the number of randomised controlled trials (RCTs) of interventions for AMD, and currently patients have a range of therapies available.1 Treatments have a potential to improve lives of AMD sufferers, yet that does not determine a full recovery. Thus, the same emphasis should be assigned to quality of life (QoL) in AMD RCTs as in other chronic diseases or cancer trials where cure is also often impossible to achieve.2 With the guidance of the Food and Drug Administration (FDA) approving use of QoL outcomes for development of medical products,3 more AMD trials should take them into consideration. However, there is no consensus whether QoL outcomes should always be included, and if so, which instrument should be used. Using patient-reported outcomes in health economic modelling is a significant part of the National Institute for Health and Care Excellence approval for new drugs in the UK. QoL is traditionally captured by questionnaires known as patient-reported outcome measures (PROMs). These ‘outcomes reported by patients’ are based on physical, psychological and social aspects of their experience of the disease.4 Various authors have highlighted that it is important to assess disease management from a patient's perspective, especially when a chronic condition is considered.5 PROMs may concern supplementary aspects of complex medical care upon a patient. The ability to perform daily activities and a patient's sense of independence are considered as important indicators that treatment is successful.6
AMD is known to influence QoL to a significant extent. Severe AMD impairs visual function, limits the range of activities a patient can do and causes emotional distress.7 Depression and falls are also associated with visual impairment, and patients with AMD suffer stress and lower satisfaction.7 Comprehensive evaluation of the impact of reduced visual acuity (VA) is critical as individuals with similar changes in the macula and similar VA may often experience difficulties in performing visual tasks to different extent.8 Thus, it is expected that AMD treatment should help to maintain both vision and QoL. A variety of vision-specific psychometric tools have been developed due to the increased interest in evaluating QoL in AMD.9 It has been also suggested that QoL measures should be implemented in everyday clinical practice,10 and routinely within RCTs.11 However, it is not known to what extent PROMs are used within AMD RCTs or what specific instruments are employed. The aim of this study was to systematically identify the frequency and type of PROMs reported in recent AMD RCTs.
We registered our systematic review in the PROSPERO (http://www.crd.york.ac.uk/PROSPERO/) after our searches were done when we were at the stage of collecting the data and preliminary analysis. The systematic review was assigned with registration number: CRD42014010040. We set up two objectives for the systematic review to identify the frequency and type of clinical outcomes and PROMs in AMD trials. This paper addresses the second objective. The data on the first objective has been submitted for publication elsewhere.
We systematically reviewed recent clinical trials of treatments for AMD with restriction to studies published in English since January 2010 until November 2013. Four databases—Ovid MEDLINE Full, Ovid EMBASE, SCOPUS and Cochrane Library (Central)—were searched. A modified sensitive search strategy with subject subheadings and text terms was adapted to find abstracts connecting AMD and RCTs. In conducting the search, two domains of terms were searched: AMD or equivalents (eg, ARMD, age-related macular degeneration) and RCTs and equivalents. The results from each domain were combined with an AND operator. The detailed databases search strategy is illustrated in supplementary appendix 1. Citations and abstracts were exported to Refworks, and a combined folder with results from the different databases was constructed. An additional search based on the clinical trials registries (http://www.controlled-trials.com and http://www.clinicaltrials.gov) was implemented. The term ‘macular degeneration’ and the above time brackets were used to obtain the desired study records. Publications from RCTs that had been designed prior to 2010 were included, and outcomes data from original study design was elicited regardless of the year of publication.
Inclusion and exclusion criteria
Two independent reviewers reviewed all abstracts. Inclusion criteria were defined; only RCTs on AMD published in English were eligible for inclusion. Apart from RCTs publication and trials registries, we searched for and identified additional RCTs from references and appendices of included Cochrane Systematic Reviews. The abstracts and the records from the registries were organised into the following categories: ‘include’, ‘uncertain’, ‘exclude’. If there were any disagreements in categorising them between the reviewers, it was resolved by the way of a discussion. The full papers and the full records of ‘uncertain’ and ‘include’ were reviewed by one investigator.
Data extraction strategy
Online data capture forms were used for data extraction (http://www.maculasenilis.com ‘Downloads’ tab or http://maculaesenilis.com/Core/Downloads/forms.zip). The forms helped in collating relevant information on the type of AMD for the participants recruited into the trial, study design (eg, intervention information) and outcomes.
Ten per cent of RCTs were randomly selected and appraised by a second reviewer to assess accuracy in data extraction.
In total, 3816 abstracts were obtained through our search. The number of RCTs found was 103. An additional 493 records were identified from the registries of clinical trials, from which 74 RCTs were included. As a consequence of both searches, a total of 177 RCTs were retrieved (figure 1). Agreement between masked observers in data extraction was 100% in identifying primary outcomes measures. In one case, the first investigator interpreted outcome information as one primary outcome, whereas the second one divided it into primary and ‘other primary outcome’.
PROMs were titled after the name of questionnaire or the tool, which was implemented to describe PROMs/QoL. Whenever the name of PROM questionnaire or the specific name of a measure was unknown, we labelled them as ‘undetermined’, for example, questions to the participants regarding subjective changes in visual function such as whether patients were able to read faster or smaller sizes than previously.12 A total of 858 clinical and PROM and adverse outcomes were reported. PROMs were identified 38 times, which constitutes 4.4% of all outcomes. Twenty-five (14.1%) RCTs used PROMs within their outcome measures. The National Eye Institute Visual Functioning Questionnaire-25 (NEI-VFQ-25) was the most commonly used PROM. It was reported as an outcome in 16 of the 177 RCTs (n=16, 64% of RCTs with PROMs). The second most common category of outcome was European Quality of Life-5 Dimension (EQ-5D) and ‘undetermined’ (n=3, 12% of RCTs with PROMs). Other types of questionnaires and their frequency of use are presented in table 3.
PROMs can be classified as generic and disease-specific ones.13 In our study, we presented the reported PROMs according to domains (eg, general health, daily activities), from which they consist of (table 4).
The Activity Inventory, used in one trial, was not included in the table as it is based on individually tailored questions concerning a visual function history as well as estimation of patients' visual ability.14 Also, Time-Trade-Off, used in one trial, was not included in the table as it is not a questionnaire but a variable that mirrors a patient's current state of health by reporting the number of remaining expected years of life, which the patient would trade for a cure.15
Only one RCT included a PROM as a primary outcome (the PROM in this trial was the Activity Inventory16), otherwise PROMs were always reported as secondary outcomes.
Measuring PROMs to evaluate effectiveness of interventions is important because it helps in translating clinical outcomes of treatment into complementary patient-perspective outcomes.11 PROMs illustrate disease activity, impact and efficacy of treatment in a unique way.
Previous studies indicate that there is insufficient attention drawn to the impact of AMD on QoL.10 ,17 It has also been shown that response to disease management should be assessed not only from an objective clinical measure perspective but also from a patient-related perspective.7 Our review extends these observations by demonstrating that a minority of recent RCTs included PROMs. Although Covert et al18 stated that using PROMs in clinical trials has become more popular, our study suggests that this is still not widely accepted among AMD researchers.
There are some potential explanations why QoL measures are not consistently used during the conduct of AMD research. First, ophthalmic professionals, researchers and the wider public often underrate everyday difficulties faced by patients with AMD.7 ,9 It may be due to the fact that VA may be confused with QoL. The majority of clinical studies primarily attempt to capture the impact of AMD and its treatment using VA,7 and it is recognised that there is a strong correlation between VA and QoL based on statistical analysis.8 ,19 However, VA only captures the ability of the patient to detect and resolve high-contrast stimuli using only approximately 1 degree of the retina, whereas, for example, visual function encompasses many other important factors such as sensitivity to low-contrast objects, dark adaptation or movement detection, which are not assessed by VA and yet are important in facilitating vision and quality of life. Second, another explanation why some AMD RCTs neglect to use PROMs may be that they were designed for pilot data only or to make comparisons between interventions already well established. Even in such studies PROMs bring valuable information about studied interventions from the perspective of a sufferer of a chronic condition, and moreover, they do not significantly increase the burden for participants of a study as questionnaires are an efficient and economical means in order to retrieve data.20 ,21 Third, it may be argued that when there is only one ‘study eye’ and the fellow eye retains relatively good vision PROMs do not have to be included. QoL is indeed correlated predominantly with the better seeing eye,8 but it cannot be denied that it also reflects burden that new medications being tested bring to patients, which in turn is important when, for example, FDA approval is sought.
We also noted that variety of QoL instruments were used in AMD RCTs. To reduce this variety, ideally one validated instrument should be nominated as a standard and then used uniformly in each AMD RCT. Such an instrument would be included in a core outcome set (COS) for AMD RCTs.6 ,22 For instance, NEI-VFQ-25 is perceived as a potential core outcome measure in terms of PROMs in AMD trials. It was the most frequently used PROM instrument among the evaluated studies included in our review. This questionnaire presents good reliability and validity in measuring vision-related functioning and QoL,8 ,23 yet other instruments (Macular Disease Quality of Life questionnaire, EQ-5D) should also be considered. To decide which instrument is appropriate to be incorporated in the AMD COS, further investigation in a context of domains of life affected by AMD is needed.
Strengths of this study include the fact that the review was conducted in a systematic way. It was carried out according to Preferred Reporting Items for Systematic Reviews and Meta-analyses recommendations (http://www.prisma-statement.org).
Identified studies included different AMD populations and interventions. Potential limitation is that we took into consideration only published and ongoing RCTs. It means we omitted the 'grey literature'. We also only considered research published in English. Some may also argue that our searches are only current up to November 2013, yet given that we included records from registries of ongoing trials, we were able to capture trials that would be published well after November 2013.
In summary, we demonstrated that PROMs are not often used in AMD trials. If using PROMs was a common practice in AMD RCTs, then an assessment of interventions would be more holistic, addressing issues derived from a patient's standpoint. In addition, if the number of AMD outcomes was reduced to a selection of a core standard group of outcomes, including a PROM, it would help in obtaining high-quality, easy-to-compare medical evidence on the effects of interventions.
We thank Sebastian Krezel for his contribution to the design of the online system for our systematic review and Richard Fallis for his support with creating our search strategy.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Collaborators Sebastian Krezel Richard Fallis.
Contributors All authors have contributed adequately to the planning and designing of this work or the analysis and interpretation of the data, and the writing of the manuscript
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
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.