Anchoring Patient Voices to Shape the Future of Patient-reported Outcomes

Patients with interstitial lung disease (ILD) frequently experience symptoms such as cough, dyspnea, and fatigue. These symptoms not only affect the patient’s quality of life (QOL) (1) but also have prognostic significance in predicting mortality (2). Immunosuppressive treatments for inflammatory ILD may improve dyspnea (3), whereas disease-modifying antifibrotic medications slow disease progression in fibrotic ILD without significantly improving QOL (4, 5). From a patient perspective, this can mean taking medications that do not alleviate symptoms and may cause considerable side effects. To date, the only interventions shown to improve QOL in ILD are ambulatory oxygen (6), pulmonary rehabilitation, and morphine therapy for cough (7), none of which prolong survival in this progressive disease.

Recently, there has been a shift in the research community to ensure studies not only assess therapeutic efficacy on objective clinical outcomes, such as forced vital capacity, but also incorporate patient voices and priorities. A recent mixed-methods study involving patients, caregivers, and healthcare professionals identified symptom relief and QOL improvement as critical research priorities for ILD (8). Indeed, key recommendations have been put forth to ensure that patient-reported outcomes (PROs) and patient-reported measures are at the forefront of ILD research and patient care. As such, PROs have now become an essential outcome measurement in clinical trials assessing medications and other healthcare interventions. There are countless PROs that measure symptom severity, symptom burden, and QOL. Furthermore, a PRO or patient-reported measure can also be used as an intervention itself; for example, their integration into patient care has been shown to improve survival and QOL in patients with cancer, most likely because of better symptom management and medication tolerability (9).

PROs are only valid if their scores lead to the correct conclusion about the participant of a study. In addition, understanding the meaningfulness of these outcomes to patients is essential. A meaningful change in a PRO score should reflect a meaningful change in how a patient feels or functions, and the minimal threshold is the smallest change in the score that a patient regards as important. Meaningful score differences (or minimal [clinically] important differences/changes) are typically determined using distribution-based methods, which are purely statistical, and/or anchor-based methods. Anchors are external criteria that are meaningful to the patient and measure the same construct or are closely related to the PRO of interest. Functional measurements (e.g., 6-min-walk distance), other PROs or patient impression can be used as anchors (10). Patient global impressions of severity (PGISs) are simple, single-item assessments developed in the 1970s, with the advantage of directly incorporating the patient’s perspective (11). PGISs have also been proposed by regulatory agencies as anchors to validate PROs as clinical trial outcomes (12). In patients with ILD, meaningful changes in the clinically relevant PGISs of cough, dyspnea, and fatigue have not been determined—until now.

In this issue of AnnalsATS, Swigris and colleagues (pp. 1670–1677) provide a deep insight into how patients with ILD describe and rate different categories of symptom severity and what change they consider relevant (13). Using qualitative interviews and quantitative survey analysis they provide minimal meaningful changes for PGISs and, beyond this, succeed in putting patient voices forward for future PRO validation. In the interviews, 18 patients describe in their own words the daily life situations, activities, coping mechanisms, and emotions they attribute to different degrees of symptom severity. The authors provide insightful quotes and concrete statements through explicit analysis, deliberately avoiding additional interpretations on the underlying, unspoken meanings of the quotes, which occurs during latent or thematic analysis. In the survey, the 298 participants were then asked by how many points shortness-of-breath and cough frequency scales would need to differ over a 3-month period (improve or worsen) for them to consider a change meaningful. Interestingly, there was a tendency for patients to rate worsening more impactful than improvement, and answers were slightly different if numeric scales (0–4) versus ordinal scales for dyspnea (not at all, mild, moderate, severe, very severe) were used. Patients on long-term oxygen therapy reported worse symptoms and more frequent increases over the preceding 3 months, yet they rated the dyspnea scales similarly to those not on oxygen therapy. Agreement between patients was moderate to good and concordant between oxygen users and nonusers. These analyses, which include patients both with and without oxygen therapy, ensure that a broad spectrum of disease severities is represented, further supporting the face validity of the PGIS concept.

Overall, most patients deemed a 1- or 2-point improvement or worsening over 3 months meaningful, and the authors suggest using PGISs of 1 and 2 as anchors to assess meaningful changes in shortness of breath and cough. This study also confirms challenges with present-state bias: patients with more severe shortness of breath at the time of the survey were more likely to state that their shortness of breath had worsened over the last 3 months. Consequently, the authors do not recommend using patient global impressions of change to validate PROs. Although the authors provide valuable thresholds for meaningful changes in PGIS shortness of breath and cough frequency, the meaningfulness of changes in cough severity and in fatigue has not yet been established. In addition, further research is needed to determine PGISs and meaningful change thresholds for other important symptoms, such as appetite, mood, and chest pain. It will also be interesting to explore how the PGIS findings from this study relate to symptom visual analog scales, which are increasingly common PRO measures in respiratory research (14).

To effectively implement these data, continued collaboration with patient registries and industry partners is essential. It remains uncertain whether PGISs will be sufficient as stand-alone endpoints in ILD clinical trials or if they will primarily serve as anchors to validate PROs. However, continued industry engagement in integrating these tools into research and providing researchers access to the data will enable further refinement and clarity. In addition, web-based platforms and smartphone-enabled registry participation may facilitate a broader patient cohort, improving data collection. Beyond single-point PRO assessments, future technologies could enable real-time ecological momentary assessments capturing patients’ experiences in their natural environment (15). With the evolving landscape of ILD treatments and diagnostic classifications, new PROs may be needed to better capture the diverse sociocultural backgrounds, complex disease management, and geographical challenges faced by patients. Fortunately, the presented data can guide the development of any such new PROs, advancing ILD patient management with precision medicine and treatable traits in mind (16).

Bridging the divide between patient voices and objective scientific data involves harmonizing patient-reported experiences with quantifiable data, ensuring that subjective symptoms and QOL measures are considered alongside rigorous scientific evidence to drive more comprehensive and meaningful healthcare solutions. We congratulate the authors for underscoring that PROs, and the meaningful score differences they reflect, should be anchored in what truly matters to patients.