Conceptual frameworks for the selection of PROMs in clinical trials

Institutions and authors have adopted a range of terminologies to describe conceptual frameworks for PROMs. In this resource, we utilise definitions provided by the Food and Drug Administration (FDA) (1) and the International Society for Quality of Life Research (ISOQOL) (2) due to their recommended use by authors of the SPIRIT-PRO (3). Such recommendations have been guided by the widely cited Wilson and Cleary model (1995), which presents a taxonomy of patient outcomes related to health-related quality of life (HRQL).

‘Concepts’ (also known as ‘constructs’) are distinct aspects of patient experience, attitudes or beliefs that are ‘latent’ – i.e. can only be measured by asking the patient.

ISOQOL recommends that researchers define focal outcome concepts and how these will be measured in the intended population and intervention contexts (2).

This can be depicted by a conceptual framework, which includes two important representations: 

1. A conceptual model, which depicts the expected relationships between the concepts of interest (e.g. pain, fatigue, treatment convenience).
2. A measurement model, defined by a representation of how PROMs will generate a score(s) to measure the concepts of interest.

To illustrate, Figure 1A is a conceptual model depicting 18 concepts identified from interviews and surveys to be relevant and important to patients with hormone-refractory prostate cancer (HRPC). These concepts were categorised into five general domains. The conceptual model then informed a sample measurement model (Figure 1B) for a hypothetical treatment for HRPC (4). Both models establish the theoretical basis for the intervention, thus providing the foundation for the desired labelling claim.

_{Figure 1 (A) Conceptual model of relevant outcomes issues from hormone-refractory prostate cancer (HRPC) and (B) measurement model for a hypothetical treatment of HRPC. From “Constructing a Conceptual Framework of Patient-Reported Outcomes for Metastatic Hormone-Refractory Prostate Cancer,” by D. T. Eton et al., 2010, Value in Health, 13(5), p. 613-623. Copyright 2010 by International Society for Pharmacoeconomics and Outcomes Research (ISPOR).}

Clinical trial protocols often just list PROs or other outcomes, without justifying why each has been chosen, nor providing hypotheses regarding expected effects on each outcome or relationships between them (5). This can lead to:

• omission of outcomes that are most important to patients and/or most likely to differ between groups;
• unnecessary burden on participants and researchers where outcomes overlap;
• apparently conflicting results that are difficult to interpret; and
• reduced confidence in positive effects due to uncertainty regarding mechanisms and/or multiple hypothesis testing.

According to the SPIRIT-PRO Extension (3):

“The justification for the selection of PRO instrument(s) is required in the trial protocol. This will help trial personnel and participants understand why specific measures are being used and how they directly address the trial objectives and stakeholder needs.”

Thus, a conceptual framework can:

• inform the choice of which concepts to measure and their position in the outcome hierarchy;
• facilitate hypothesis generation and testing regarding mechanisms and role of contextual factors, which be used to develop an endpoint model;
• reduce unnecessary burden on participants and researchers; and
• inform your statistical analysis plan.

Summarise any relevant theoretical frameworks that will be used to build rationales for: 

• which patient-related concepts are expected to be affected;
• expected directions and magnitudes of effects;
• expected relationships between different patient-related concepts;
• expected relationships between patient-reported concepts and other measures.

Review the literature for pre-established frameworks or qualitative studies that have explored relevant concepts. If the topic of interest is unexplored, consider conducting a qualitative study to identify and inform key concepts. This will help you establish whether the concepts you have listed are necessary and sufficient to test key parts of your framework, and highlight any additional concepts that should be measured but are not currently listed.

                Organise concepts as ‘proximal’ or ‘distal’

Consistent with the Wilson and Cleary (1995) framework, the identified concepts may be classified as ‘proximal’ in nature – that is, directly impacted by the disease and/or treatment (e.g. symptoms such as pain, fatigue, nausea, rash and anxiety) – or more ‘distal’ flow-on effects (e.g. functional status and global quality of life) (6). Distal concepts are more likely to be affected by factors beyond the intervention, and are therefore less likely to be responsive than more proximal concepts that are influenced more directly.  Figure 2 provides examples of proximal and distal effects of disease and treatment on patient symptoms and quality of life.

_{Figure 2 Conceptual model of proximal and distal effects from ovarian cancer and chemotherapy. From “SPIRIT-PRO Extension explanation and elaboration: guidelines for inclusion of patient-reported outcomes in protocols of clinical trials,” by M. Calvert et al., 2021, BMJ Open, 11(6), e045105. Copyright 2021 by Authors (or their employer(s)).}

Based on the conceptual model, determine which PROM(s) best assess these concepts. Examine the specific items in each measure and how the items are combined into scales to deliver scores.

The ISOQOL checklist for selecting PROMs may support your decision-making regarding the choice of PROM.

We have developed a template for investigators to create their own conceptual framework, which can then be inserted as a figure into a study protocol. Instructions and an example of a resulting conceptual framework can be accessed through the file below.

 CONCEPTUAL FRAMEWORK TEMPLATE (PPTX, 2.54 MB)

It is possible to quantitatively assess the conceptual framework to inform refinements using established techniques (7, 8). An analysis plan for doing this could be developed. Techniques to understand the measurement relationships within the framework include correlations (to understand convergence and divergence of concepts), exploratory factor analysis (to explore main concepts used to generate a model) and structural equation modelling (to describe the structure of concepts within the overall framework).

After identifying key concepts, PROMs and their associated scale scores of interest from your conceptual framework, these can be used to specify clear PRO objectives or hypotheses in your clinical trial. 

The SPIRIT-PRO extension offers guidance on the information that should be included to clearly define and answer PRO-specific research questions in clinical trial protocols. For more information on the SPIRIT-PRO and other PRO methodological guidance, please visit the PROTEUS Consortium website.

The estimand framework was introduced in a new addendum to the International Committee for Harmonisation (ICH) E9 guidelines as a systematic approach to ensure alignment among clinical trial objectives, study design, endpoints, analysis and interpretation of results (9).

Your conceptual framework can help to identify different estimands of interest, based on strategies taken to handle intercurrent events. Specification of estimands should then enable clearer definition of PRO objectives in protocols and analysis plans. Lawrance et al. (10) describes the application of the estimand framework to establishing PRO objectives when designing clinical trials. 

The Setting International Standards in Analysing Patient-Reported Outcomes and Quality of Life Endpoints Data (SISAQOL) Consortium has also presented recommendation to facilitate standard approaches for PRO analysis in cancer clinical trials (5). These recommendations outline a taxonomy to develop well-defined PRO research objectives, which can then be matched with appropriate statical methods.

1.            Food and Drug Administration (FDA). Patient-focused drug development: Selecting, developing, or modifying fit-for-purpose clinical outcome assessments. Guidance for Industry [draft] 2022 [Available from: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/patient-focused-drug-development-selecting-developing-or-modifying-fit-purpose-clinical-outcome.]

2.            Reeve BB, Wyrwich KW, Wu AW, Velikova G, Terwee CB, Snyder CF, et al. ISOQOL recommends minimum standards for patient-reported outcome measures used in patient-centered outcomes and comparative effectiveness research. Qual Life Res. 2013;22(8):1889-905.

3.            Melanie C, Madeleine K, Rebecca M-B, Olalekan A, Derek K, Anita S, et al. SPIRIT-PRO Extension explanation and elaboration: guidelines for inclusion of patient-reported outcomes in protocols of clinical trials. BMJ Open. 2021;11(6):e045105.

4.            Eton DT, Shevrin DH, Beaumont J, Victorson D, Cella D. Constructing a Conceptual Framework of Patient-Reported Outcomes for Metastatic Hormone-Refractory Prostate Cancer. Value Health. 2010;13(5):613-23.

5.            Coens C, Pe M, Dueck AC, Sloan J, Basch E, Calvert M, et al. International standards for the analysis of quality-of-life and patient-reported outcome endpoints in cancer randomised controlled trials: recommendations of the SISAQOL Consortium. The Lancet Oncology. 2020;21(2):e83-e96.

6.            Wilson IB, Cleary PD. Linking clinical variables with health-related quality of life: a conceptual model of patient outcomes. JAMA. 1995;273(1):59-65.

7.            Williams B, Onsman A, Brown T. Exploratory factor analysis: A five-step guide for novices. Australasian journal of paramedicine. 2010;8:1-13.

8.            Schreiber JB, Nora A, Stage FK, Barlow EA, King J. Reporting structural equation modeling and confirmatory factor analysis results: A review. The Journal of educational research. 2006;99(6):323-38.

9.            International Council for Harmonisation. Addendum on estimands and sensitivity analysis in clinical trials to the guideline on statistical principles for clinical trials E9(R) 2019 [Available from: https://database.ich.org/sites/default/files/E9-R1_Step4_Guideline_2019_1203.pdf.]

10.         Lawrance R, Degtyarev E, Griffiths P, Trask P, Lau H, D’Alessio D, et al. What is an estimand & how does it relate to quantifying the effect of treatment on patient-reported quality of life outcomes in clinical trials? Journal of Patient-Reported Outcomes. 2020;4(1):68.