Metabolic medicine is changing faster than the evidence systems built to track it. New therapies for type 2 diabetes and obesity are reaching more patients, across more indications and through more channels than ever before. For HEOR and Medical Affairs teams, that raises a pressing question: how do you build a credible picture of real-world value when so much of what matters never reaches a database?
Real-world evidence (RWE) is central to the answer. Regulators, health technology assessment (HTA) bodies, and payers increasingly expect it. But in the broad metabolic space, generating RWE that is truly decision-grade is harder than it looks.
Why metabolic RWE is uniquely difficult
- The right patients are hard to find. BMI is not routinely recorded in health records (nor necessarily an indicator that treatment is required) and obesity is frequently left undiagnosed, so studies based on diagnosis codes miss many eligible patients[1]
- Care is fragmented. Patients move between insurance providers, primary care, multiple specialists, pharmacy, weight management services and telehealth providers. This can lead to fragmented patient records and no single dataset that follows the whole patient journey, and linking sources is slow and tightly governed.
- Treatment patterns distort the picture. Persistence on newer metabolic therapies is often low[2]. Supply constraints and coverage rules mean prescribing can reflect therapy availability rather than clinical choice, and private, cash-pay and direct-to-consumer use may not appear in claims data at all.
- Key outcomes take time to emerge, and many are never recorded. Cardiovascular events, kidney decline, and diabetes onset take years to emerge, while quality of life, stigma, treatment burden and productivity are rarely captured in routine data.
- Behaviour is largely invisible. Diet, activity and adherence shape outcomes profoundly, yet this sort of information usually sits entirely outside structured data. Wearable (e.g., fitness trackers, glucose-monitors, etc.) data helps but remains siloed and skewed towards more engaged patients.
- Equity remains a blind spot. The populations carrying the highest metabolic burden are often under-represented in commercial datasets and digital cohorts, leaving payers and HTA bodies asking whether the evidence reflects the people who most need treatment.
Where primary research fits
Structured data tell you what happened, but they don’t often tell you why, which is why primary research is crucial for developing an effective RWE strategy: not as a substitute for claims, EHR or registry data, but as the layer that provides the data that those areas do not. For HEOR or Medical Affairs, it can give strategic direction in five ways.
- Explaining the patterns. When real-world data show patients stopping or switching treatment, primary research with patients and healthcare professionals (HCPs) can uncover the reasons, whether it is related to side effect, cost, unmet needs, stigma, barriers to care, or gaps in support.
- Defining key outcomes. Pre-survey development qualitative interviews with patients, HCPs and payers can identify which outcomes resonate with each group, informing endpoint selection, and survey design.
- Mapping the real care pathway. Understanding who initiates the treatment discussion, manages the treatment, journey and identifying where care fragmentation occurs , can answer important questions and lead to improved care outcomes.
- Prioritising evidence gaps. Research with payers, HTA advisers, and specialist HCPs reveals what evidence will help inform decisions, and what will not. Primary research helps teams focus limited budgets where they will have most impact.
- Inviting under-represented voices into the conversation. Targeted recruitment can reach less visible and underserved patient groups, helping teams test whether findings generalise and strengthen the equity case that decision-makers increasingly expect.
Seeing inside the consultation
Some of the most consequential moments in metabolic care happen in the office between the patient and the HCP. How are treatment options presented? How are weight, side effects and long-term commitment discussed? What do patients take away? Is there shared decision making? These conversations shape initiation and persistence, yet routine data collection approaches almost never capture them.
KJT’s DialogueDx methodology is designed to close that gap. It captures authentic patient–provider conversations through in-office observation, then builds on them with patient video diaries, provider interviews and quantitative validation. Generative AI is used to analyse these interactions at scale.
Because it observes real conversations rather than relying on recall or separate participant interviews or surveys, DialogueDx gives teams a real-world view of where communication supports or undermines treatment goals, from the first prescribing discussion to the conversations that come before discontinuation. Those insights can sharpen evidence priorities, empower patients with appropriate questions to ask, inform HCP education and scientific communication, and point observational studies towards the questions that matter most.
Making it work in practice
The value comes from integration. Primary research can generate important real-world evidence when it is designed as part of the wider evidence plan, uses consistent definitions across data sources and studies, and informs both study design and field medical engagement. Its contribution is strongest when considered alongside other RWE sources, with appropriate attention to sample design, context, and the limitations inherent in observational evidence.
The bottom line
Metabolic care will remain one of the most scrutinised therapy areas for years to come. The Medical Affairs teams that make the strongest case for value will be those that combine what the data can measure with what only patients and HCPs can explain.
[1] Suissa K, Schneeweiss S, Lin KJ, Brill G, Kim SC, Patorno E. Validation of obesity-related diagnosis codes in claims data. Diabetes Obes Metab. 2021 Dec;23(12):2623-2631. doi: 10.1111/dom.14512. Epub 2021 Aug 18. PMID: 34338404; PMCID: PMC8578343 https://pubmed.ncbi.nlm.nih.gov/34338404/
[2]Xie L, Anazco D, Chancay AH, et al. Glucagon-Like Peptide-1 Receptor Agonist Switching and Treatment Persistence in Adults Without Diabetes. JAMA Netw Open. 2026;9(3):e261272. doi:10.1001/jamanetworkopen.2026.1272 – https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2846157
Dziewierz A, Siudak Z. Adherence and Persistence with GLP-1-Based Therapies: International Real-World Evidence and the Role of Nutritional and Lifestyle Support-A Narrative Review. Nutrients. 2026 May 30;18(11):1761. doi: 10.3390/nu18111761. PMID: 42280404; PMCID: PMC13259006 – https://pmc.ncbi.nlm.nih.gov/articles/PMC13259006
Author
Danny Wolin
Real World Evidence, SVP
Danny Wolin has over 25 years of RWE and regulatory affairs experience in the pharmaceutical industry. He has a proven track record of collaborative partnership in the design and leadership of qualitative and quantitative RWE studies across multiple therapeutic areas that help inform important healthcare decisions.
Download the Info Sheet to Learn More About DialogueDx
