The Power of Item Banks to Address Measurement Challenges in Patient-Reported Outcomes

Despite the pervasive needs and desires to include patient-reported outcomes (PRO) in trials, many existing measures are inappropriate for a particular patient population, inadequately designed or lacking psychometric evidence to support their use.

With item banks, however, these challenges can often be overcome. Heather Gelhorn, senior research leader at Evidera – the provider of evidence-based solutions for PPD, part of Thermo Fisher Scientific – discusses how to utilize item banks to address common challenges of PROs in clinicals and shows how her team put item banks into practice to advance the clinical trial for a rare disease therapeutic.

Item banks are a repository of patient-reported outcomes (PRO) questions. They cover a wide range of concepts of interest and are gathered from existing, validated measures or incorporate items that are newly developed for the purpose of generating a comprehensive item bank.

With the growing role of patient centricity in health care intervention evaluation, researchers can find themselves challenged to identify measurement instruments for inclusion in clinical trials. But by leveraging concepts in existing item banks, researchers can create a measurement approach tailored to the patient population, therapeutic area, time span and measurement needs, thus overcoming issues such as data heterogeneity or a lack of existing research for a target population.

In effect, item banks offer a strong solution for researchers to create custom PRO measures and build an evidence-based approach to instrument development and patient-centered research – with the right expertise.

Prominent Item Banks

The item bank that researchers will utilize within a given study depends on the focus of research, the concepts of interest and the measurement challenges they encounter. Comprehensive in nature, item banks are put through rigorous qualitative and quantitative evaluations to support their contents and use. By leveraging concepts in existing item banks, researchers create customized short forms or implement computerized adaptive testing as part of their tailored measurement approach.

To supplement their studies, researchers often turn to one of several well-known item banks, including:

• The Patient-Reported Outcomes Measurement Information System (PROMIS), an item bank that covers many aspects of the global, physical, mental and social health of adults and children.
• The European Organization for Research and Treatment of Cancer (EORTC), a database of items used in fully and partially validated EORTC quality-of-life questionnaires.
• The Quality of Life in Neurological Disorders (NeuroQoL), which measures aspects of the physical, mental and social health for adults and children living with a neurological condition.

How Item Banks Overcome Clinical Trial PRO Measurement Challenges

Here are a few examples of common PRO measurement challenges researchers may face in a clinical trial and how item banks help them sidestep them:

• Heterogeneity: With some conditions, disease symptoms and impacts may manifest differently across individual patients or subgroups, and concepts can be difficult to measure across a wide range of abilities/severity levels and manifestations with existing PRO measures. Use of item banks enables information from a heterogenous population to be used together, which is particularly important in rare disease trials when subgroups may be small.
• Rare disease: Rare disease clinical research must often account for varied patient experiences within small populations. Item banks may enable researchers to conduct necessary qualitative work more efficiently with patients by providing an efficient starting point for the selection or development of patient-reported outcome measures.
• Timeline pressures: Item banks can jump-start the creation of a measurement instrument, enabling researchers to skip or minimize the time spent generating and retesting items with patients. In addition, the use of item banks may reduce the time and resources necessary later in a clinical trial to evaluate the psychometric properties of the scales.
• Lack of existing measures and/or research for the target patient population: For new diseases, or for those that have had little to no research to date, item banks provide a reliable starting point for the selection or development of patient-reported outcome measures.
• Need for greater depth and/or breadth of measurement of the concept of interest: When a scale needs to allow for the measurement of a wide range of within and/or between individual differences in severity, or when smaller changes need to be measured very precisely, item banks can be an incredibly powerful tool.

Item Banks in Practice: A Mixed Methods Approach Advances Study Objectives for a Rare Disease

Perhaps the easiest way to illustrate the power of item banks is through a case study.

I was recently part of a team supporting trials of a therapeutic developed to treat tenosynovial giant cell tumors (TGCT), a rare proliferative neoplasm involving the synovium and tendon sheaths of the upper or lower extremities that typically present in young and middle-aged adults.^1,2

Challenge

At the time, the standard of care was surgical resection. However, new therapeutic agents were under development. The research team at Evidera was tasked with going beyond the oncological endpoint of tumor volume by developing a PRO instrument to support the magnitude and relevance of the clinical endpoint from the patient perspective.

We learned that patients with TGCT commonly present with complaints of pain and stiffness that affect their physical functioning. However, because these tumors were heterogenous — occurring in knees, ankles, hips, elbows, shoulders and other joints across patients — and TGCT is very rare, no appropriate, validated PRO instruments existed to measure physical functioning in this population.

Approach

We needed a PRO measure that could quantify the impacts on physical functioning across all individuals, regardless of their tumor location and its effect on their physical functioning at the start of the trial. We began with the PROMIS Physical Functioning (PROMIS PF) item bank to create physical functioning short forms that were specific to upper and lower extremity tumors and covered a wide range of physical functioning ability.

Through a combination of patient interviews, physical functioning impact checklists, clinical expert input and a review of the existing PROMIS physical functioning parameters, we selected 15 of the most appropriate physical functioning items from the 122 items in the item bank. We ended up with two static short forms for inclusion in the clinical trial: one for patients with upper extremity tumors and one for those with lower extremity tumors. There were nine overlapping items across the forms, with two additional items administered only to those with upper extremity tumors and two additional items administered to those with lower extremity tumors.

Results

The data from the clinical trial enabled Evidera’s team to analyze the impact of the treatment on physical functioning.

• Thanks to the use of the item bank, the physical functioning information from all patients in this heterogeneous population could be analyzed together, regardless of tumor location or the extent to which their physical functioning was initially impacted. This is particularly important for rare disease trials, in which every patient data point matters and some subgroups may be small.
• Data from the clinical trial were also subsequently used to support the psychometric properties of the physical functioning short forms (e.g., the reliability, validity and responsiveness of the measures).
• Finally, existing evidence on the PROMIS PF item bank provided initial estimates of meaningful change in physical functioning, which the trial data later confirmed.

In summary, the use of the PROMIS item bank allowed us to move past the challenges presented when dealing with the combination of heterogeneity, a rare disease and a lack of an existing appropriate and/or validated PRO instrument. Through a mixed methods approach — qualitative interviews with patients supported by quantitative data from a PROMIS PF item checklist — Evidera’s research team advanced multiple study objectives including concept elicitation, content validity and item selection.

As with other item banks, PROMIS PF allowed us to identify well-established, high-functioning items using a smaller sample of patients – key for rare disease research – while simultaneously addressing heterogeneity through development of custom forms. Leveraging the power of an item bank to address measurement challenges in PRO, Evidera’s research team successfully demonstrated improvements in this key PRO for this trial.

The Growing Importance of Expert Instrument Development in Patient-Centered Data

As patient-centered data becomes increasingly important to demonstrate drug efficacy and safety, researchers cannot risk their trials being derailed by data or measurements that have inadequate design or lack supporting evidence to justify their use. Item banks offer researchers a valuable solution for PRO measurement to meet critical study objectives.

And with this insight, researchers gain a common understanding of patients and their experiences.

With the growing spotlight on high-quality, patient-centered data, an evidence-based approach to instrument development is imperative. Researchers need a partner that can seamlessly design and implement strategies for outcome evaluation – including PROs – to meet stakeholder demands, overcome hurdles and advance their clinical programs.