In the era of patient-centricity, is the use of PROMs optimized?—a narrative review based on AI models

Introduction

Background

Patient-reported outcome measures (PROMs) are tools designed to gather information about a patient’s health status, quality of life, and the outcomes of healthcare interventions from the patient’s perspective. These instruments typically take the form of questionnaires or surveys that patients complete themselves. By capturing subjective data, PROMs provide essential insights into how patients perceive their symptoms, the impact of their condition on daily life, and their overall satisfaction with the care they receive (1).

PROMs began to be used in the 1970s as part of a broader effort to integrate patients’ voices into healthcare evaluations (2). Initially, these measures were primarily employed in clinical research, but their utility has since expanded into routine clinical practice. The core rationale for using PROMs is to obtain a direct assessment of the patient’s experience, which can differ significantly from clinical or physiological measures. This patient-centric approach is designed to enhance the quality of care, improve patient satisfaction, and support shared decision-making between patients and healthcare providers (3).

In addition to PROMs, several other types of clinical outcome assessments (COAs) exist, including clinician-reported outcome measures, observer-reported outcome measures, and performance outcome measures. Clinician-reported outcome measures are based on clinical judgments of a patient’s condition, observer-reported outcome measures involve reports from individuals other than the patient or clinician (such as caregivers), and performance outcome measures assess a patient’s function or performance through standardized tasks. While each type of COA provides unique insights, PROMs are particularly valuable as they reflect the patient’s personal experience and perceived well-being, making them a critical component of patient-centered care (4).

Rationale and knowledge gap

Hidradenitis suppurativa (HS) is a chronic inflammatory skin condition characterized by the formation of painful nodules, abscesses, and scarring, primarily in areas where skin rubs together, such as the armpits, groin, and under the breasts. This debilitating disease significantly impairs the quality of life of affected individuals. Epidemiologically, the reported prevalence estimates vary between 0.0003% and 4.1%, and data from various geographical regions are still to be collected (5). This disease is more common in early adulthood (late teens to early twenties) and is diagnosed more frequently in females (6,7).

HS is influenced by several risk factors, which can increase the likelihood of developing the condition. A significant risk factor for HS is genetic predisposition; having a family history of the condition suggests a genetic component to the disease, with studies indicating that it can run in families in approximately 30–40% of cases (7). Obesity is another strong risk factor, as a higher body mass index is associated with increased severity of HS due to friction and sweating in skin folds, exacerbating the condition (6,7).

Smoking is also a major risk factor for HS. Although the exact mechanism is unclear, smoking is believed to worsen inflammation and may impair the skin’s immune response. Hormonal factors play a role as well; hormonal changes, particularly those related to androgens, are implicated in HS. The condition often starts or worsens after puberty and can fluctuate with menstrual cycles, indicating a hormonal link, in terms of gender (6,7).

Furthermore, there is a higher prevalence of HS in individuals with certain medical conditions such as metabolic syndrome, inflammatory bowel disease (particularly Crohn disease) (8), polycystic ovary syndrome (9), and other inflammatory and autoimmune disorders.

The impact of HS on patients is profound, extending beyond physical discomfort to psychological distress and social stigmatization. Patients often experience a diminished quality of life, facing limitations in daily activities, work, and social interactions. The chronic nature of the disease, combined with frequent flare-ups and visible lesions, contributes to heightened levels of anxiety, depression, and a reduced sense of well-being (10).

The European S1 guideline suggests that the disease should be treated based on its individual subjective impact and objective severity.

The expert group recommends treatment based on the Hurley severity grade, following a structured treatment algorithm. Usually, before considering surgical options for HS, medical therapy is typically the first choice, particularly for extensively spread lesions. This therapy may include antibiotics such as clindamycin combined with rifampin or tetracyclines, acitretin, and biologics like adalimumab and infliximab. Despite these approaches, medical treatments often provide only partial relief, with frequent recurrences and varying responses among patients. If medical therapy does not achieve sufficient control, surgical options such as classical surgery or LASER techniques may be considered for localized, recurring lesions (5). Pain management, treatment of superinfections, weight loss, and tobacco abstinence should also be considered in the treatment plan for HS.

Objective

This article aims to explore the application of PROMs within the context of HS. Specifically, it examines 177 clinical trials conducted from 2005 to 2024, focusing on the COAs utilized in HS research. The objective is to identify the prevalence and role of PROMs in capturing patient experiences and to highlight the need for integrating tailored PROMs more comprehensively into HS research and clinical practice. This study seeks to address current gaps in assessing patient-centric outcomes and underscores the importance of collaborative efforts among stakeholders to advance HS management and improve patient care through evidence-based approaches. We present this article in accordance with the Narrative Review reporting checklist (available at https://jmai.amegroups.com/article/view/10.21037/jmai-24-218/rc).

Methods

To address these objectives, we generated a comprehensive COA report. This extraction focused on clinical trials written in English, regardless of the study design or phase, related to HS, and conducted from January 1, 2005, to May 31, 2024, sourced from ClinicalTrials.gov. The collected data points included URL links, titles, brief summaries, dates, study types, study phases, study statuses, countries, drugs, drug comparators, conditions, study populations, primary and secondary COA names, and COA descriptions (Table 1).

Data selection was meticulously carried out using ClinicalTrials.gov as the primary source. We employed a two-phase extraction approach to ensure that only trials specifically addressing HS were included. In the first phase, we focused on titles, keywords, and conditions. In the second phase, we reviewed the full text and the inclusion criteria of the studies. This process was guided by the MeSH ontology labeled “Hidradenitis Suppurativa” to ensure precise pathology classification. In the data harmonization phase, we employed several artificial intelligence (AI)-powered methodologies to ensure consistency and accuracy. First, we investigated and compared the drugs involved in the trials using an in-house ontology (an ontology based on multiple sources such as TTD, NCIT, and MeSH, refined by our data scientists and medical team, and tested for its sensitivity and specificity). This process helped in accurately standardizing and categorizing the investigational and comparator drugs. Next, we used AI to segment and group similar aspects and tools used as COAs. This was followed by standardization to ensure uniformity across the data points: To do so, we use a biomedical embedding model, UMAP for dimensionality reduction, HDBSCAN for unsupervised clustering and our COA ontology for standardization.

Attributing types to the COAs involved using an natural language processing (NLP) model, which facilitated the categorization and classification of the various COAs employed in the studies. Additionally, we utilized an NLP model to extract and homogenize patient details (using a trained sentence classifier model), ensuring that patient-related data was consistent and comparable across different studies. This step was crucial in generating COA-drug and COA-patient links, providing a comprehensive understanding of the relationships between COAs, treatments, and patient outcomes.

Data enrichment further enhanced the dataset by incorporating additional layers of information. We defined COAs using generative AI (ChatGPT 4), providing detailed and standardized definitions. Original and MCID references were added to enhance the reliability of the data.

Statistical analysis followed the data standardization process. We conducted systematic sampling and employed descriptive statistics to analyze frequency distributions and percentages. Cross-tabulation and stratified analysis were performed to explore relationships and differences across various strata within the data. Statistical tests like chi square, logistic regression were used in the sub-analysis. Statistical analysis was conducted using statisty.app. and P<0.05 is considered as significant.

Quality control was an integral part of our methodology: We implemented a sampling method to verify the standardization of drugs, COAs, type attribution, and the data enrichment process. This step involved cross-checking the accuracy and consistency of the standardized terms and ensuring that the enrichment processes were correctly applied.

Results

Descriptive analysis

In this study, 177 clinical trials were examined, employing a total of 1,431 COAs to evaluate HS. Clinician-reported outcomes were the most frequently observed, accounting for 48.5% of all COAs (694 occurrences). Patient-reported outcomes were utilized 436 times, comprising 30.47% of the total, while performance-reported outcomes and PROM or observer-reported outcomes were noted 14 and 2 times, respectively, constituting 0.98% and 0.14% of the total. Biomarker-related COAs were documented 231 times, representing 16.15%, with other types making up 3.42% (49 occurrences).

To quantify the impact of HS on quality of life, various PROMs have been utilized. Specifically tailored to HS, these include the Hidradenitis Suppurativa-Quality of Life [17], Hidradenitis Suppurativa Symptom Diary [14], Hidradenitis Suppurativa Symptom Assessment [9], Hidradenitis Suppurativa Impact Assessment [6], Hidradenitis Suppurativa Quality of Life Score [1], Quality of Life (HiSQOL) in Hidradenitis Suppurativa (QUALIVER) [1], Hidradenitis Suppurativa Symptom Questionnaire [1], Hidradenitis Suppurativa Burden of Disease [1], and Hidradenitis Suppurativa European Research Group [1], totaling 51 out of 438 PROMs. This indicates that 11.7% of the PROMs used are specific to HS. On the other hand, general PROMs related to quality of life, dermatology, treatment adherence, and other aspects are employed, such as the Dermatology Life Quality Index [87], Skin Pain Numerical Rating Scale [62], Participants Attrition, Treatment Modality, Preference, Adherence, and Discontinuation [22], Pain Visual Analog Score [21], Patient Global Assessment [18], Hospital Anxiety and Depression Scale [13], and the European Quality of Life-5 Dimensions [10].

Additionally, ClinROs used to assess HS include both specific measures tailored for HS and more generalized assessments. Specific HS ClinROs such as the Hidradenitis Suppurativa Clinical Response [139], International Hidradenitis Suppurativa Severity Score System [53], and Hurley Classification [23] target the unique aspects of HS symptoms, severity, and treatment response. In contrast, general ClinROs like Adverse Events Occurrence [119], Physician’s Global Assessment [54], and measures for depression, wound healing, and skin condition are applied across various diseases and conditions. The ratio of specific ClinROs used compared to the total number of ClinROs is 55%, showing a significant increase in specificity compared to the proportion of PROMs tailored specifically for HS (P<0.05 for chi-square).

Sub-analysis

Per study type

This graph (Figure 1) shows that the use of PROMs is well established in observational studies, whereas in interventional studies, there is less emphasis on quality of life and a greater focus on clinical aspects [χ²(16)=104.18, P<0.001 for chi-square].

Figure 1 Frequency of each COA type across different study types. COA, clinical outcome assessment; ClinRO, clinician-reported outcome; PRO, patient-reported outcome; PerfO, performance outcome; ObsRO, observer-reported outcome.

Per drug investigated

Figure 2 shows the difference in the strategies for the top 10 most investigated drugs in HS. Notably, adalimumab, the first monoclonal antibody approved for the treatment of HS, has adopted a strategy where patient-reported outcomes constitute an important part in the endpoints studied. This is more pronounced in the cases of guselkumab and bermekimab. On the other hand, povorcitinib and bimekizumab have adopted a different strategy, where clinician-reported outcomes are heavily used to study the efficacy of the treatments compared to PROMs.

Figure 2 Distribution of COA types across the top 10 most investigated drugs in HS. COA, clinical outcome assessment; ClinRO, clinician-reported outcome; PRO, patient-reported outcome; ObsRO, observer-reported outcome; HS, hidradenitis suppurativa; AI, artificial intelligence.

Over time

Pharmaceutical interest in treating HS intensified notably in 2016 and has continued to rise through 2024. A notable trend is the fluctuating use of PROMs for evaluating treatments: while PROMs were prominently featured in 2016, their use decreased from 2017 onward in favor of ClinROs (Figure 3). Although the PRO/ClinRO ratio varied over the years, it underscores the increasing importance of ClinROs in evaluating HS treatments (R=0.17, SE =2.35, P<0.001 for logistic regression after 2016).

Figure 3 Distribution of COA types across clinical trials treating HS over the last 20 years. COA, clinical outcome assessment; HS, hidradenitis suppurativa; ClinRO, clinician-reported outcome; PRO, patient-reported outcome; PerfO, performance outcome; ObsRO, observer-reported outcome.

Discussion

Importance of PROMs in HS research

HS presents a unique challenge in healthcare due to its significant impact on quality of life despite not affecting life expectancy. This underscores the critical role of Patient-Reported Outcomes in assessing the true burden of the disease from the patient’s perspective. Unlike clinician-reported outcomes which focus on clinical signs and symptoms, PROMs capture subjective experiences such as pain, emotional well-being, and social functioning which are pivotal in understanding the holistic impact of HS.

The study highlights that while a variety of COAs were utilized across 177 clinical trials, PROMs accounted for 30.47% of assessments. This proportion signifies recognition within the research community of the importance of integrating patient perspectives into evaluating treatment efficacy. However, despite this proportion, there remains a significant gap in the comprehensive use of PROMs across all stages of clinical trials, particularly in earlier phases, where they could inform more patient-centered trial designs (11). Specifically tailored PROMs for HS, such as the Hidradenitis Suppurativa-Quality of Life and Symptom Diary, provide insights into aspects of the disease that are not captured by generic quality of life measures or ClinROs focused on disease severity alone.

Disease-specific PROMs: tailoring assessments to HS

The significance of disease-specific PROMs lies in their ability to capture nuances specific to HS, such as chronic pain, malodor, and impaired mobility, which profoundly impact patients’ daily lives. The study identifies that 11.7% of PROMs used were specifically designed for HS, indicating a growing recognition of the need for targeted assessments in this patient population. This is particularly important as recent evidence suggests that disease-specific PROMs can also improve patient adherence to treatment by providing more relevant and understandable measures of progress (12). These instruments not only enhance the sensitivity of clinical trials to detect meaningful changes but also empower patients by validating their experiences in research settings.

Strategic implications for market differentiation

In the competitive landscape of pharmaceutical development for HS treatments, the adoption of disease-specific endpoints and PROMs can serve as a crucial differentiating factor. Drugs like adalimumab, guselkumab, and bermekimab have strategically incorporated PROMs into their clinical endpoints, aligning their development strategies with patient-centric outcomes. This trend is increasingly seen as essential not only for market approval but also for achieving premium pricing and market access, as payers and health technology assessments (HTAs) are increasingly demanding evidence of patient benefit beyond clinical efficacy (13). This approach not only meets regulatory requirements but also resonates with patients and clinicians seeking therapies that address the broader impact of HS beyond clinical symptoms.

Conversely, drugs emphasizing ClinROs over PROMs may miss opportunities to demonstrate comprehensive treatment benefits, potentially limiting their market appeal. The fluctuating use of PROMs observed over time highlights evolving strategies in response to both scientific insights and market dynamics, underscoring the adaptability required to meet patient needs effectively. Furthermore, the incorporation of real-world evidence (RWE) alongside PROMs is becoming a powerful strategy to demonstrate long-term benefits of treatments in actual clinical settings, thereby enhancing market differentiation (14).

Addressing unmet needs and enhancing patient care

The findings emphasize the evolving role of PROMs in reshaping HS research and clinical practice, moving beyond traditional endpoints to prioritize outcomes that matter most to patients. By integrating robust methodologies, including AI-powered data extraction and harmonization, the study ensures rigor in evaluating COAs across diverse clinical contexts. This methodological approach not only enhances data reliability but also supports informed decision-making in drug development and healthcare policy.

Ultimately, the integration of PROMs in HS research represents a paradigm shift towards patient-centered care, advocating for treatments that not only alleviate symptoms but also enhance overall quality of life. As pharmaceutical innovation continues to evolve, leveraging disease-specific PROMs remains pivotal in addressing unmet patient needs and driving meaningful advancements in HS management.

Limitations and future directions

The current study offers valuable insights into the use of PROMs in HS research, yet it has certain limitations that warrant further exploration. First, the reliance on data from clinicaltrials.gov may introduce selection bias, as not all relevant studies may be registered or accessible through this platform. Additionally, the study’s temporal scope, limited to 2005–2024, might not capture recent advancements or trends in PROM utilization. Variability in COA definitions and methodologies across studies also poses challenges in ensuring data consistency and comparability.

Future research should aim to expand the data sources and consider a broader temporal range to capture emerging trends in HS research. Furthermore, there is a need to develop standardized and validated disease-specific PROMs that can be consistently applied across studies to enhance the accuracy of patient-centric outcome assessments. Collaborative efforts among researchers, clinicians, and patient advocacy groups will be crucial in refining these instruments and ensuring their integration into routine clinical practice, ultimately advancing HS management and improving patient care.

Conclusions

In conclusion, the study’s findings underscore the critical need for increased utilization of tailored PROMs in HS research and clinical practice. The observed proportions highlight a gap in the adoption of disease-specific PROMs, reflecting missed opportunities to comprehensively capture the burden of HS on patients’ lives. This study was conducted using ArcaScience AI models, which enabled a clear assessment of the state of the art in HS treatment and provided robust evidence regarding the use of clinical endpoints in HS treatment. Therefore, it is imperative for ongoing collaboration among researchers, clinicians, regulators, and patient advocacy groups to prioritize the development and validation of PROMs specifically tailored to HS. With the comprehensive evidence and support provided by advanced AI models, such collaboration will be facilitated, equipping healthcare workers with the necessary information for better decision-making. This collaborative effort is essential to address current limitations, drive meaningful advancements in HS management, and ultimately improve the quality of life for individuals affected by this challenging disease.

Acknowledgments

This article was written with the assistance of ChatGPT-4, an AI language model developed by OpenAI.

Funding: None.

Footnote

Reporting Checklist: The authors have completed the Narrative Review reporting checklist. Available at https://jmai.amegroups.com/article/view/10.21037/jmai-24-218/rc

Peer Review File: Available at https://jmai.amegroups.com/article/view/10.21037/jmai-24-218/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jmai.amegroups.com/article/view/10.21037/jmai-24-218/coif). All authors are employed by ArcaScience. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Fayers PM, Machin D. Quality of life: the assessment, analysis and interpretation of patient-reported outcomes. New Jersey: John Wiley & Sons, 2013.
2. Fiona K. ‘Patient reported outcome measures’: we have barely started to walk let alone run. In: Medical Leadership. Routledge, 2018; 176-188.
3. Pereira L, Jeronimo C, Salgado A, et al. Patient centricity as strategy to improve quality of service in healthcare management. International Journal of Healthcare Technology and Management 2023;20:1-15. [Crossref]
4. Dirven L, Armstrong TS, Taphoorn MJ. Health-related quality of life and other clinical outcome assessments in brain tumor patients: challenges in the design, conduct and interpretation of clinical trials. Neurooncol Pract 2015;2:2-5. [Crossref] [PubMed]
5. Bouazzi D, Andersen RK, Vinding GR, et al. The Global Hidradenitis Suppurativa Atlas Methodology: Combining Global Proportions in a Pooled Analysis. Dermatology 2024;240:369-75. [Crossref] [PubMed]
6. Kromann CB, Deckers IE, Esmann S, et al. Risk factors, clinical course and long-term prognosis in hidradenitis suppurativa: a cross-sectional study. Br J Dermatol 2014;171:819-24. [Crossref] [PubMed]
7. Miller IM, McAndrew RJ, Hamzavi I. Prevalence, Risk Factors, and Comorbidities of Hidradenitis Suppurativa. Dermatol Clin 2016;34:7-16. [Crossref] [PubMed]
8. Zhang M, Chen QD, Xu HX, et al. Association of hidradenitis suppurativa with Crohn's disease. World J Clin Cases 2021;9:3506-16. [Crossref] [PubMed]
9. Phan K, Charlton O, Smith SD. Hidradenitis suppurativa and polycystic ovarian syndrome: Systematic review and meta-analysis. Australas J Dermatol 2020;61:e28-e33. [Crossref] [PubMed]
10. Dessinioti C, Katsambas A, Antoniou C. Hidradenitis suppurrativa (acne inversa) as a systemic disease. Clin Dermatol 2014;32:397-408. [Crossref] [PubMed]
11. Mullins CD, Vandigo J, Zheng Z, Wicks P. Patient-centeredness in the design of clinical trials. Value Health 2014;17:471-5. [Crossref] [PubMed]
12. Ahmed S, Berzon RA, Revicki DA, et al. The use of patient-reported outcomes (PRO) within comparative effectiveness research: implications for clinical practice and health care policy. Med Care 2012;50:1060-70. [Crossref] [PubMed]
13. van Overbeeke E, Forrester V, Simoens S, et al. Use of Patient Preferences in Health Technology Assessment: Perspectives of Canadian, Belgian and German HTA Representatives. Patient 2021;14:119-28. [Crossref] [PubMed]
14. Maruszczyk K, McMullan C, Aiyegbusi OL, et al. Paving the way for patient centricity in real-world evidence (RWE): Qualitative interviews to identify considerations for wider implementation of patient-reported outcomes in RWE generation. Heliyon 2023;9:e20157. [Crossref] [PubMed]