Repurposing Registries: Completeness of Real‐World Data for Regulatory and HTA Purposes in Three Cancer‐Focused Registries Authors: Manon C. Wilpshaar, Doranne Hilarius, Luisa Prada, Ansgar Weltermann, Christine Leopold Published: 2026 ABSTRACT Despite developments in supporting the identification and selection of fit‐for‐purpose registries, a translational gap remains between the original purposes of clinically focused registries and their current and future expected use in policymaking. This study aimed to assess the extent to which cancer‐focused registries collect essential data elements for regulatory and Health Technology Assessment (HTA) processes. Using literature and expert input, a reference dataset of essential data elements was developed. Completeness of data elements in this reference dataset was evaluated on the example of three European, cancer‐focused registries, varying in type, coverage, and health system: the Cancer Centre Upper Austria (CCUA) registry, the Dutch Medication Audit (DMA), and the Portuguese Oncology Institute of Porto (RIC IPO Porto) registry. The evaluated registries exhibit clear differences in inclusion criteria driven by their specific aims. Data elements related to comorbidities, concomitant therapies, patient‐reported outcome measures (PROMs), resource use, and safety outcomes were lacking partly or completely in all evaluated registries, highlighting the importance of developing disease‐specific core datasets to which registry holders can benchmark their data collection. This study also highlights that the initial aims and design of registries impact their strengths and weaknesses in supporting regulatory and HTA decision‐making. Optimally repurposing registry data requires capacity building for registry holders and regulatory/HTA stakeholders focusing on mutual understanding of aims, needs, and limitations. Additional research should address data interoperability, standardization, and data quality across registries. Study HighlightsWHAT IS THE CURRENT KNOWLEDGE ON THIS TOPIC?Patient or disease registries are increasingly acknowledged by regulatory and Health Technology Assessment (HTA) stakeholders as a Real‐World Data (RWD) source to inform their decisions. Specifically in oncology, their relevance is growing as many new therapies enter the market with limited clinical evidence. Despite several frameworks guiding fit‐for‐purpose assessments, the use of RWD in decision‐making is limited. This might be because many registries were originally designed to serve varying clinical purposes rather than decision‐making processes.WHAT QUESTION DID THIS STUDY ADDRESS?Currently, it is not known to what extent patient or disease registries collect data elements that regulatory and HTA decision makers consider essential. Based on guidance documents of regulatory and HTA bodies, a checklist of data elements that are considered essential for decision‐making was compiled. On the example of three European registries focused on cancer, we systematically describe registry key characteristics, data element completeness, and evolution of completeness.WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?Despite their shared focus on cancer, the three registries exhibit clear differences in organizational structure and in‐ and exclusion criteria, which have shaped their data completeness. The largest discrepancies between policymaker‐relevant data elements and those currently captured in these registries relate to comorbidities, concurrent therapies, resource use, patient‐reported outcomes (PROMs), and safety outcomes. Additionally, the reference dataset that was drafted in this study could be used by other registry holders to benchmark their registries to the needs of policymakers.HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?Our findings illustrate how registries can differ even within the same disease areas and underline the importance of defining disease‐specific core datasets to guide registry holders in leveraging their registries for regulatory and HTA purposes. Study Highlights WHAT IS THE CURRENT KNOWLEDGE ON THIS TOPIC? Patient or disease registries are increasingly acknowledged by regulatory and Health Technology Assessment (HTA) stakeholders as a Real‐World Data (RWD) source to inform their decisions. Specifically in oncology, their relevance is growing as many new therapies enter the market with limited clinical evidence. Despite several frameworks guiding fit‐for‐purpose assessments, the use of RWD in decision‐making is limited. This might be because many registries were originally designed to serve varying clinical purposes rather than decision‐making processes. WHAT QUESTION DID THIS STUDY ADDRESS? Currently, it is not known to what extent patient or disease registries collect data elements that regulatory and HTA decision makers consider essential. Based on guidance documents of regulatory and HTA bodies, a checklist of data elements that are considered essential for decision‐making was compiled. On the example of three European registries focused on cancer, we systematically describe registry key characteristics, data element completeness, and evolution of completeness. WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE? Despite their shared focus on cancer, the three registries exhibit clear differences in organizational structure and in‐ and exclusion criteria, which have shaped their data completeness. The largest discrepancies between policymaker‐relevant data elements and those currently captured in these registries relate to comorbidities, concurrent therapies, resource use, patient‐reported outcomes (PROMs), and safety outcomes. Additionally, the reference dataset that was drafted in this study could be used by other registry holders to benchmark their registries to the needs of policymakers. HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE? Our findings illustrate how registries can differ even within the same disease areas and underline the importance of defining disease‐specific core datasets to guide registry holders in leveraging their registries for regulatory and HTA purposes. A patient or disease registry is an organized system that longitudinally collects data on individuals with a particular disease, condition, or exposure. 1 Registries can include data originating from routine clinical practice, such as electronic health records, reimbursement claims, or patient questionnaires, referred to as Real‐World Data (RWD). During the last decades, a myriad of registries has been established, often capturing data from a broader and more diverse patient population compared with study populations in Randomized Clinical Trials (RCTs). The relevance of RWD is growing, particularly in the field of oncology, given that systemic anticancer therapies (SACTs) account for the majority of new marketing authorization applications, despite their varying levels of benefit and high costs. 2 , 3 The growing molecular specificity of new SACTs has led to clinical trials with smaller study populations or single‐arm study designs, limiting the available evidence at the time of evaluation. 4 This is reflected in the increasing trend of conditional marketing authorizations and subsequently also conditional reimbursement decisions often linked to managed entry agreements for anticancer medicinal products. 5 , 6 These conditional policy decisions come with requirements to submit additional evidence often from either RCTs or RWD from cancer registries. To generate this additionally required evidence, regulatory and HTA stakeholders increasingly acknowledge data from patient or disease registries. 7 , 8 , 9 , 10 However, due to several challenges, the contribution of RWD to (conditional) marketing authorizations and HTA decisions is still limited. 11 , 12 , 13 , 14 The challenges that currently limit the use of RWD include limited availability of essential variables, restricted data access, absence of standardized methodologies for data quality assessment, and a lack of consensus on the methodological standards for relative effectiveness assessments. Several efforts have been made to facilitate the use of RWD. Firstly, a growing number of frameworks and guidance documents have been developed to identify and assess the suitability of RWD sources for policy decision‐making, for example, the Registry Evaluation and Quality Standards Tool (REQueST). 15 Secondly, several initiatives at the European level have been established to facilitate the use of RWD, including the Data Analytics and Real‐World Interrogation Network (DARWIN)‐EU 16 for coordination of real‐world data sources, the GetReal institute 17 and the RWE4Decisions as multistakeholder platforms for collaboration in the discussion and potential adoption of RWD and the MetReal cluster of European Union‐funded projects studying the more effective use of RWD for regulatory and HTA decision‐making. 18 , 19 , 20 , 21 , 22 , 23 And thirdly, the new Regulation EU 2021/2282 establishes a legal basis for the harmonization of RWD in Europe, states that observational data must be considered to address trial evidence gaps, and advocates for an EU‐wide action plan for RWD collection. 24 Yet, beyond these international initiatives, the usefulness of RWD from registries ultimately depends on whether the data that are collected align with the needs of decision‐makers. The discrepancy between available RWD in registries and actual use in regulatory and HTA decision‐making might be explained by the fact that many registries were initiated to support clinical care, to derive epidemiological metrics, or to serve research goals of patient advocacy groups, rather than to support policymaking. Although a registry's suitability for evidence generation largely depends on the specific research question to be addressed, adequate coverage of data elements considered essential by regulators and HTA decision‐makers is a fundamental requirement for any use in evidence generation. While a growing body of literature has been published on stakeholders' opinions on the use of RWD in policy decision‐making, to the best of our knowledge, no study has looked at the extent to which clinical cancer registries actually collect specific data elements that regulators and HTA stakeholders consider essential (from here referred to as completeness). 7 , 8 Therefore, this study aims to explore the completeness of cancer‐focused registries that were originally initiated for clinical purposes and evaluate how evolutional and contextual factors have influenced their completeness. On the example of three European clinical cancer‐focused registries, we first systematically assessed the registries' evolutional and contextual factors, followed by an assessment of the registries' completeness based on a reference dataset of essential data elements for policy decision‐making and discussed this data completeness over time. METHODS Selection of registries Three registries were selected to achieve a representative geographical spread and coverage of different health systems. The selection of registries aimed to capture: (1) geographic variation, (2) differences in the healthcare system with regard to insurance and HTA structure, and (3) differences in the coverage of registries with regard to geography, population coverage, and reported cancers. 25 An overview of the selection criteria is presented in Table 1 . This sample consisted of the following registries: the registry of the Cancer Centre Upper Austria (CCUA), the Dutch Medication Audit (DMA) managed by the Dutch Institute for Clinical Auditing (DICA), and the registry of the Portuguese Oncology Institute of Porto (IPO Porto)—Repositório Integrado do Conhecimento (RIC). Table 1Selection criteria for registry selectionRegistrySelection criterionCCUADMARICGeographical featuresCountryAustriaThe NetherlandsPortugalGeographical coverageRegionalNationalLocal (single hospital)Healthcare systemPayer systemSingle payerMulti‐payerSingle payerBudgetSeparate in‐ and outpatient healthcare budgetSingle healthcare budget (including in‐ and outpatient care excluding long‐term care)Global hospital budget (covering in‐ and outpatient care)HTA bodyTwo National HTA bodies: A government‐appointed national commission (‘Nationales Bewertungsboard’) since 2025 and Austrian Social Insurance (for outpatient drugs only)One national agency, designated by government and giving binding recommendations on reimbursementOne national agency, designated by government and giving binding recommendations on reimbursementHTA remitFor all outpatient drugs and selected, cost‐intensive inpatient drugsBoth for in‐ and outpatient drugsBoth for in‐ and outpatient drugsAbbreviations: CCUA, Cancer Centre Upper Austria; DMA, Dutch Medication Audit; RIC, Repositório Integrado do Conhecimento. Selection criteria for registry selection Abbreviations: CCUA, Cancer Centre Upper Austria; DMA, Dutch Medication Audit; RIC, Repositório Integrado do Conhecimento. Systematic description of registry key characteristics To systematically describe the registries, a mixed‐methods approach was applied, including literature search, internal document search, and informal interviews with leading representatives of the registries (most of them co‐authors of this study). A data extraction sheet was developed by MW in agreement with all co‐authors to extract the key characteristics of the registries, which included: (1) initiation (year, profession of initiators, way of initiation, stakeholders), (2) organizational aspects (owner, funding and organizational structure), (3) coverage (geographical region, number of involved hospitals, tumor entities, in‐ and exclusion criteria for patients, type of inclusion), and (4) data collection and governance (data source(s), reporting mandate, end users, approval for external use, data access lag time, quality control procedures). Completeness of essential data elements By scoring the RWD completeness, we aimed to give an overview of the included cancer‐focused registries' completeness regarding regulatory and HTA expectations. The reference dataset used for this includes data elements which were extracted from two key documents: (1) the European Medicines Agency (EMA) Guideline on registry‐based studies 26 representing the opinions of regulators and (2) the final report case study of Programme Managing Registries for Expensive Medicinal Products 27 by the Zorginstituut Nederland (ZIN; national HTA agency in The Netherlands) representing the opinions of HTA stakeholders. Both documents describe data elements for all disease areas and are not cancer‐specific. In the Tables S1 and S2 , we present the list of the unique data elements from each document with an explanation on whether the element was included in the reference dataset. For the final scoring of policy‐RWD completeness of the registries, all unique extracted data elements were categorized into eight main domains (including administrative and patient data, comorbidities and concomitant therapy, diagnosis, disease‐related treatment, outcomes and monitoring, patient‐reported outcome (PROMs), resource use, and safety outcomes), excluding synonym terms. To show the completeness of the three registries, we used a traffic light system to indicate whether the registry was able to fully report the data element (green), to partially report the data element (orange), or not able to report the data element (red). Data elements were alphabetically ordered within each domain. Evolution of the availability of core data elements To assess the evolution of the availability of the core data elements over time in each registry, the availability of any data element in each of the eight domains was mapped in a timeline format (2009–2024, in three‐year intervals). These characteristics were examined at an aggregated level, aiming to identify overarching patterns across types of cancer rather than details on molecular, histological, or genetic subtypes, covering the WHO‐defined top five most 28 prevalent cancer types (lung, breast, colorectal, prostate, and stomach cancer) and for hematological neoplasms. Additionally, the numbers of patients registered in each registry were collected in the same three‐year interval. RESULTS Registries' key characteristics In general, the CCUA‐registry and RIC‐registry share more structural and operational similarities with each other than with the DMA (Table 2 ). This difference is mainly attributable to the design of the DMA, as it consists completely of data that have been registered for other purposes; it combines reimbursement data with data that were registered in one of the other registries managed by DICA. This is in contrast with the operationalization of the CCUA‐ and RIC‐registry, where the data are registered specifically for the registries. Table 2Key characteristics of the registriesDomainVariableCancer Centre Upper Austria (CCUA) 27 Dutch Medication Audit 28 Repositório Integrado do Conhecimento—IPO PortoInitiationYear of initiation201220171988Way of initiationClinician‐initiatedClinician‐initiated (Hospital pharmacists for the DMA; surgeons for other cancer‐specific DICA‐registries)Clinician‐initiatedInitial purposes of the registry Clinical decision‐making: supporting multidisciplinary tumor boards Measuring quality of treatment Providing epidemiological metrics Clinical decision‐making: benchmarking and identification of practice variations Measuring quality of treatment Providing epidemiological metricsOrganizationOwner of registryHospital operators of the CCUA (n = 5)Independent institute: Dutch Institute for Clinical Auditing (DICA)Hospital operator of IPO Porto (n = 1)FundingHospital operatorsHealthcare insurersHospital operatorsLevel of organizationCentrally for the regionHosted by DICAManaged by the epidemiology department and IT departmentNumber of hospitals involved (% of geographical coverage)a 13 (100%)54b 1 (100%)Involved staffCoordinators, IT, statisticians, documentation staffCoordinators, IT, statisticians, researchersCoordinators, IT, statistician, documentation staff, physiciansStakeholdersPhysicians, hospital management, hospital operators, regional Health AuthoritiesHospital pharmacists, physicians, healthcare insurersPhysicians, hospital management, hospital operators, regional health authoritiesCoverageLevel of geographical coverage (inhabitants)a Regional (1.5 million)National (17.9 million)Hospital‐based (catchment > 3.5million)c Total number of registered patients at 31‐12‐202484,123152,604188,193Tumor entities coveredAll tumor entitiesBreast cancer, colorectal cancer, gynecological cancer, head and neck cancer, lung cancer, upper gastrointestinal cancer, prostate cancerAll tumor entitiesType of inclusionDiagnosis‐basedTreatment‐basedDiagnosis‐basedInclusion criteriaAll newly diagnosed tumor diseases (both solid and hematological, both benign and malignant) treated in Upper AustriaAll patients that received high‐cost pharmacotherapy (designated by the government as eligible for separate reimbursement schemes 26 ) for one of the indications that have a linked tumor‐specific registry at DICA (see: “coverage of tumor entities”).All newly diagnosed tumor diseases (both solid and hematological, both benign and malignantd) treated at IPO PortoExclusion criteriaNone For registration: none For analyses: Referred patients and patients whose treatment started before data collection started in 2017 NoneData collection and governanceData source(s)Data forms (SQL‐based) filled by documentation staff or automated extraction from hospital Information Systems (HIS)Reimbursement data (automated extraction) linked to clinical data from other disease‐specific quality registries managed by DICA (filled out by data managers or healthcare professionals). Data forms (filled out by data managers) Automated extraction (through FHIR) of information of hospital sources (treatments (systemic treatments, RT, surgeries, transplants), pathological reports) Reporting mandate Compulsory by law: Epidemiological data of all cancer (National Cancer Registry); epidemiological and clinical data of breast cancer (National Breast Cancer Screening Program) Compulsory by internal agreement 29 Providing data to registries will become compulsory by law in 2026. 30 , 31 Reporting is mandatory to both the Regional Registry (RORENO) and the National Registry (RON). Additionally, an annual internal institutional report is produced covering survival rates and new case volumes.End users Physicians, (hospital) pharmacists, quality assurance officers, healthcare providers, scientists Data will be (pseudo‐)anonymized and/or aggregated in accordance with the role of the applicant and the purpose. Physicians, (hospital)pharmacists, quality assurance officers, healthcare payers (receive only data on aggregated levels)Physicians, (hospital) pharmacists, managers, investigatorsApproval for external useYes: data release for external use is subject to a formalized release request and processYes: data release for external use is subject to a formalized release request and processYes: data release for external use is subject to a formalized release request and processData access lag timeMaximum of 1 yearMaximum of 1 yearMaximum of 1 yearFollow‐up time from cancer diagnosisYearly follow‐up for 5 years, in case of relapse continued documentation until death, death date reported beyond 5 yearsUntil death or lost to follow‐up (including transfer to another healthcare institution)Until death or lost to follow‐up (including transfer to another healthcare institution)Quality control proceduresContinuous internal quality control process and yearly external auditsYearly quality control processYearly quality control process a Geographic coverage according to the total number of hospitals in the covered region. b Hospitals that signed a contract for the delivery of data on high‐cost medicine for specified indications. c Represents the reference catchment population of the IPO Porto, covering the Northern region, Aveiro‐North, and Southern Douro areas. d In cases of concurrent malignant and benign tumors, only the malignancy is recorded; benign tumors are reported only when occurring in isolation. Key characteristics of the registries Clinical decision‐making: supporting multidisciplinary tumor boards Measuring quality of treatment Providing epidemiological metrics Clinical decision‐making: benchmarking and identification of practice variations Measuring quality of treatment For registration: none For analyses: Referred patients and patients whose treatment started before data collection started in 2017 Data forms (filled out by data managers) Automated extraction (through FHIR) of information of hospital sources (treatments (systemic treatments, RT, surgeries, transplants), pathological reports) Compulsory by law: Epidemiological data of all cancer (National Cancer Registry); epidemiological and clinical data of breast cancer (National Breast Cancer Screening Program) Compulsory by internal agreement 29 Physicians, (hospital) pharmacists, quality assurance officers, healthcare providers, scientists Data will be (pseudo‐)anonymized and/or aggregated in accordance with the role of the applicant and the purpose. Geographic coverage according to the total number of hospitals in the covered region. Hospitals that signed a contract for the delivery of data on high‐cost medicine for specified indications. Represents the reference catchment population of the IPO Porto, covering the Northern region, Aveiro‐North, and Southern Douro areas. In cases of concurrent malignant and benign tumors, only the malignancy is recorded; benign tumors are reported only when occurring in isolation. The RIC‐registry is the oldest registry initiated in 1988, whereas the CCUA‐registry and the DMA were founded in 2012 and 2017, respectively. Although all three registries were initiated by clinicians, their purposes were different. The initial purpose of the CCUA‐registry was focused on improving care within its respective hospital network; the DMA was initiated to enable hospital benchmarking on a national level regarding high‐cost medicines in oncology, and the initial aim of the RIC‐registry was to provide epidemiological metrics such as prevalence and incidence. In terms of organizational aspects, the DMA is different as it is primarily funded externally, that is by healthcare insurers, contrary to the CCUA‐registry and RIC‐registry, which are funded internally by the hospitals that they are embedded in. This difference in funding is also reflected in the ownership of the registries: The DMA is owned by the Dutch Association of Hospital Pharmacists and managed by DICA, an externally governed institute, whereas the other two registries are owned and managed by the hospital(s) that they collect data from. In terms of patient inclusion, the CCUA‐registry and RIC‐registry are characterized by a comprehensive diagnosis‐based inclusion of tumor entities, covering both solid and hematological malignancies as well as benign tumors. By contrast, the DMA restricts inclusion to treatment with high‐cost medication (designated by the government as eligible for separate reimbursement schemes 29 ) for specific tumor types that are covered by a linkable DICA‐registry. Due to technical constraints, referred patients and patients whose treatment started before the start of data collection in 2017 are usually excluded from analyses done with DMA data. The differences in inclusion criteria are reflected in the growth trajectories of each registry over time (Figure 1 ). The RIC‐registry shows consistent growth of approximately 20,000 patients per three years, with roughly 188,000 patients registered at the end of 2024. The CCUA‐registry shows a comparable annual growth rate, albeit starting later with approximately 84,000 patients registered at the end of 2024. The DMA shows the largest growth rate with approximately 50,000 patients registered each three years, with a total of approximately 150,000 patients registered by the end of 2024. Title Total number of registered patients in each registry for the WHO top 5 cancer types and hematological cancers, from 2009 to 2024 in three‐year intervals. x‐axis: years, y‐axis: number of patients per year. CCUA, Cancer Center Upper Austria; DMA, Dutch Medication Audit; RIC, Repositório Integrado do Conhecimento. CCUA is indicated in orange; DMA is indicated in blue; RIC is indicated in green. All three registries employ a form of automated data extraction combined with manual data entry. In the CCUA‐registry, structured query language forms are filled either by data managers or automated extraction from the hospital information system. The DMA itself is a fully automatically extracted registry, as it reuses existing reimbursement data. Cooperation with other registries hosted by DICA allows linkages, which add manually collected data. For all three registries, there is a legal mandate for reporting: In Austria, reporting to the National Cancer Registry is mandatory, in addition to reporting epidemiological and clinical breast cancer data to the National Breast Cancer Screening Program by internal agreements. In Portugal, the RIC‐registry is subjected to compulsory reporting to the national cancer registry (Registo Oncológico Nacional, RON). The DMA does not have a legal mandate to report to other registries, but for hospitals, reporting to the DMA will become mandatory by 2026. The end users across all registries include physicians, hospital pharmacists, quality assurance officers, and researchers. Healthcare payers are among the end users of the DMA as well, but their access is restricted to aggregated data, whereas other end users receive more granular data for benchmarking. Procedures for external access are similar in the CCUA‐registry and the DMA and include a formalized release request. Completeness of essential data elements Table 3 presents an overview of completeness of the three included registries. For the administrative and patient data, varying completeness is found. Collection of functional status is limited, with no data in the CCUA‐registry and data depending on a linkage to other DICA‐registries for the DMA. Also, collection of length, weight, pregnancy data, and substance use is lacking or limited to specific disease areas. Data on comorbidities are available in two of the registries, but differentiation between past and current comorbidities is not registered or can be deduced partially. Data elements on the diagnoses are quite complete with all elements readily available or, in the case of the DMA, retrievable via linkage with one of the other DICA‐registries. Also, for data elements on the treatment of the disease, the CCUA‐registry and the DMA show full or partial coverage of all essential data elements. In the Portuguese registry, data elements on imaging and laboratory tests are collected, data on death is partially available, and data on progression is not available. Collection of PROMs is limited, with the CCUA collecting this only for hematological neoplasms and only a subset of the DMA‐linked DICA‐registries. As for resource use data, a clear difference is observed since the DMA collects all essential data elements, whereas the other registries collect none. Data collection on safety outcomes is lacking in all registries, except for the DMA. Table 3Coverage of core data elements by the three included registriesDomainIncluded data elementCancer Centre Upper AustriaDutch Medication Audita Repositório Integrado do Conhecimento—IPO PortoAdministrative and patient dataAge or birthdate (where permitted)ECOG/WHO PSGenderLengthName of centerPatient number/IDPregnancyReason for registry exitRegistry entry date or surrogateRegistry exit date or surrogateSubstance usea,b a,bb b , c WeightComorbidities and concomitant therapyComorbiditiesConcomitant therapyDifferentiation between past and current comorbiditiesDiagnosisAnatomical localizationa,b,c,dDiagnosis (international classification)Date of diagnosisGenomic informationMolecular diagnostics/biomarkersTNM‐stageDisease‐related treatmentBrand name [of systemic therapy]Dates of treatmentDose [of systemic therapy]Location of administration [of systemic therapy]RadiotherapyReason for discontinuation [of systemic treatment]ad Route [of systemic treatment]Schedule [of systemic treatment]Start and end dates (dates of prescription or dispensing)Substance (ATC‐code)SurgeryOutcomes and monitoringImagingb,c,d,eLaboratory test resultsb,c,d,eLaboratory testsOrgan functionCause of deathb,c,d,e e Date of death e Dates of confirmed progressionb,c,dPROMsCancer‐specific PROMsab,d,ePROMs in generalab,eResource useDates of encounters in clinical practiceNumber of days at intensive careNumber of days at nursing wardNumber of hospitalizationsNumber of surgical treatmentsNumber of visits emergency wardNumber of visits infusion therapy wardSafety outcomesAdditional adverse eventsAdverse events > grade 3 (CTCAE)c,dAdverse events that required therapy adjustmentsc,d Note: a‐only for lung cancer, b‐only for breast cancer, c‐only for colorectal cancer, d‐only for stomach cancer, e‐only for hematological cancers. Text in square brackets has been added for clarity.Abbreviations: ATC, anatomical therapeutic chemical; CTCAE, common terminology criteria for adverse events; ECOG/WHO PS, Eastern Cooperative Oncology Group/World Health Organization Performance Score; PROMs, patient‐reported outcome measures; TNM, tumor node metastases. a For the DMA, also the coverage of the linked DICA‐registries were assessed: for breast cancer the NABON Breast Cancer Audit (NBCA), for colorectal cancer the Dutch ColoRectal Audit (DCRA), for lung cancer the Dutch Lung Cancer Audit Lung carcinoma (DLCA‐L), for prostate cancer the Dutch Prostate Audit, and for stomach cancer the Dutch Upper GI Cancer Audit (DUCA), the DMA does not have a linkage to a registry specific for hematological cancer. b Only smoking status collected in terms of substance use. c Only alcohol intake collected in terms of substance use. d Only for immunotherapies. e Deaths occurring within the IPO are automatically migrated to the registry. Deaths occurring outside the institution must be reported externally (e.g., by family members). Coverage of core data elements by the three included registries Note: a‐only for lung cancer, b‐only for breast cancer, c‐only for colorectal cancer, d‐only for stomach cancer, e‐only for hematological cancers. Text in square brackets has been added for clarity. Abbreviations: ATC, anatomical therapeutic chemical; CTCAE, common terminology criteria for adverse events; ECOG/WHO PS, Eastern Cooperative Oncology Group/World Health Organization Performance Score; PROMs, patient‐reported outcome measures; TNM, tumor node metastases. For the DMA, also the coverage of the linked DICA‐registries were assessed: for breast cancer the NABON Breast Cancer Audit (NBCA), for colorectal cancer the Dutch ColoRectal Audit (DCRA), for lung cancer the Dutch Lung Cancer Audit Lung carcinoma (DLCA‐L), for prostate cancer the Dutch Prostate Audit, and for stomach cancer the Dutch Upper GI Cancer Audit (DUCA), the DMA does not have a linkage to a registry specific for hematological cancer. Only smoking status collected in terms of substance use. Only alcohol intake collected in terms of substance use. Only for immunotherapies. Deaths occurring within the IPO are automatically migrated to the registry. Deaths occurring outside the institution must be reported externally (e.g., by family members). Evolutional characteristics of completeness Table 4 gives a visual presentation, by registry and cancer type, of the domains in which data elements were collected over 2009–2024 in three‐year intervals. Several patterns are notable. The DMA was initiated most recently, showing the shortest longitudinal follow‐up in its dataset. It is also apparent that this registry relies on linkage to other registries, which do not always operate uniformly; for example, for some types of cancer, PROMs or safety outcomes are available, while for others, they are not. The RIC‐registry is the oldest registry and shows the most extensive data span. Data collection on comorbidities, disease‐related treatments, and outcomes/monitoring started more recently in 2018. However, the registry does not collect any data on PROMs, resource use, and safety outcomes. The CCUA‐registry shows a uniform pattern of coverage across cancer types, with the exception of PROMs collection since 2024 for hematological cancers. Also, no data related to comorbidities, concurrent therapies, resource use, or safety outcomes are available in this registry. Regarding the number of registered patients in each registry, the RIC‐registry covered the most patients with over 180,000 unique patients in 2024. The DMA covered over 160,000 by the end of 2024 and shows the largest yearly increase. The CCUA‐registry is the smallest registry in terms of patient numbers, with over 8,000 patients registered by the end of 2024. Table 4Visual summary of in which domains data were collected in each registry for each cancer type over 2009–2024 in three‐year intervalsDomainCancer typeBreast cancerColorectal cancerHematological cancersLung cancerProstate cancerStomach cancerYeara 091215182124091215182124091215182124091215182124091215182124091215182124 Dutch Medication Audit b Admin. + patient dataComorb. + concomitant th.DiagnosisDisease‐related treatmentsOutcome/monitoringPROMsResource useSafety outcomes RIC IPO Porto Admin. + patient dataComorb. + concomitant th.DiagnosisDisease‐related treatmentsOutcome/monitoringPROMsResource useSafety outcomes Cancer Centre Upper Austria Admin. + patient dataComorb. + concomitant th.DiagnosisDisease‐related treatmentsOutcome/monitoringPROMsResource useSafety outcomes a Years are shown in two‐digit format (e.g. ‘09’ = 2009). b For the DMA, also the coverage of the linked DICA‐registries were assessed: for breast cancer the NABON Breast Cancer Audit (NBCA), for colorectal cancer the Dutch ColoRectal Audit (DCRA), for lung cancer the Dutch Lung Cancer Audit Lung carcinoma (DLCA‐L), for prostate cancer the Dutch Prostate Audit, and for stomach cancer the Dutch Upper GI Cancer Audit (DUCA), the DMA does not have a linkage to a registry specific for hematological cancer. Coverage of domains is indicated by color: blue colored domains represent coverage (at least partial), and blank/white domains represent no coverage. Visual summary of in which domains data were collected in each registry for each cancer type over 2009–2024 in three‐year intervals Years are shown in two‐digit format (e.g. ‘09’ = 2009). For the DMA, also the coverage of the linked DICA‐registries were assessed: for breast cancer the NABON Breast Cancer Audit (NBCA), for colorectal cancer the Dutch ColoRectal Audit (DCRA), for lung cancer the Dutch Lung Cancer Audit Lung carcinoma (DLCA‐L), for prostate cancer the Dutch Prostate Audit, and for stomach cancer the Dutch Upper GI Cancer Audit (DUCA), the DMA does not have a linkage to a registry specific for hematological cancer. Coverage of domains is indicated by color: blue colored domains represent coverage (at least partial), and blank/white domains represent no coverage. DISCUSSION This systematic comparison of three cancer‐focused registries shows partial alignment with regulator and HTA data needs, and their completeness differs notably despite their shared focus on cancer. Data on the diagnosis and disease‐related treatments were mostly covered in the three registries. Regarding domains with limited alignment, data on comorbidities, concomitant medication use, resource utilization, and PROMs are often lacking, and when these data elements are captured at all, this is only recent or limited to specific indications. As for their mutual comparability, the registries differ in the range of cancer types covered, as the Austrian and Portuguese registries cover all cancer types, while the Dutch registry restricts data collection to a selection of types. So even though these registries share their focus on cancer, they differ in several aspects, which shape their usability for regulatory and HTA decision‐making. The usability of RWD from registries is, in fact, not only determined by the completeness of relevant data elements, but also by the governance around data access and the availability of sufficient patients to address the research question of interest. Regarding completeness, the frequent missing of comorbidity and concomitant medication data limits adjustments for prognostic confounders, which is important in establishing the comparability of registry populations with those in RCTs or other real‐world populations. The frequent missing of resource utilization and PROMs data is particularly problematic for HTA purposes, given their central role in economic evaluations and quality‐of‐life assessments. However, even when data completeness is adequate, poor governance, such as excessive lag time or restricted data access, results in the data being inaccessible regardless of their intrinsic quality. Addressing all three dimensions is therefore essential when evaluating registry suitability for regulatory and HTA decision‐making. The discrepancies in alignment with the regulator and HTA needs may be because these registries were initiated for clinical purposes rather than policymaking. The data element domains that showed the largest gaps with regulator and HTA needs, such as data on PROMs, comorbidities, safety outcomes, and resource use, often fall outside the scope of the original purposes of these registries. The CCUA‐registry was initiated for internal purposes, specifically to support tumor boards and improve the quality of care. The initial purpose of the DMA was to enable hospital benchmarking nationally regarding the use, outcomes, and costs of high‐cost medicines. The Portuguese registry was established primarily for providing epidemiological metrics, such as regional prevalence and incidence of cancer. The absence of the data elements that not align with the initial scope of a registry may also be attributable to the substantial registration burden that manual collection of these variables would impose on clinical staff in case automated extraction or retrieval via linkage to other sources is not possible. Therefore, retrospective harmonization or recategorization of already collected data elements, or prospective collection of additional data elements may be required before data from these registries can be repurposed for regulatory and HTA purposes. For several reasons, the recent increase in demand for RWD is expected to continue, stressing the importance of improvements in data completeness and quality. Firstly, as RCTs are increasingly limited by constraints in trial design or conduct, the need for mature RWD is additionally increasing as a rise in conditional approvals and reimbursement schemes is expected. Secondly, advancements in methodological guidance leverage the use of observational data, for example, in comparative effectiveness studies. The Target Trial Emulation and Estimand frameworks, for instance, provide a structured approach in the conduct of observational studies and defining objectives, and these are increasingly recognized by different stakeholders. 30 , 31 , 32 On the contrary, as part of the European Health Technology Assessment Regulation, the Member State Coordination Group published a guidance document on the scoping process including detailed explanations on relevant PICOs (population, intervention, comparator, outcome). 33 Various strategies can support the improvement of registry utility and anticipate the increasing demand for high‐quality RWD. Firstly, increasing the usability of registries in policymaking requires consensus on which data elements should be collected and in what format. To improve the completeness of existing registries, disease‐specific minimum datasets are needed, which can serve as a “gold standard” and guide registry holders in how to improve completeness. The current sources, including those used for the completeness assessment in this study, provide some guidance but lack disease‐specific granularity. Consensus on the data format is essential to increase interoperability. Enhancement of interoperability can be achieved both at the level of data elements, by applying international terminologies like the International Classification of Diseases and Health–related problems (ICD), and at the level of registry structure, by implementing a common data model, such as the Observational Medical Outcomes Partnership (OMOP). 34 , 35 Secondly, effective horizon scanning should not only include monitoring of emerging technologies but also identifying relevant registries and assessing their applicability for RWD collection required for conditional marketing authorizations or reimbursement schemes. Importantly, partial alignment with regulatory or HTA needs does not preclude a registry from being used for the data elements that are available. Therefore, assessments of registry applicability should include options for linkage with other data sources, such as other registries, electronic health records, or claims databases. The DMA illustrates this potential well, as its ability to link to other registries within DICA provides access to complementary clinical data, thereby expanding its utility without creating additional registration burden. Timely and comprehensive identification of relevant registries allows for tailoring of conditional approvals to the feasibility of obtaining RWD, particularly as data incompleteness in registries has been identified as a key challenge when these are considered as alternatives to post‐authorization RCTs. 36 Besides, this enables registry holders to timely adjust their data collection and linkage options to ensure availability of essential data elements. Ideally, a tool will be developed that combines the identification and assessment of RWD sources to support these processes, as proposed by Zebachi et al. 37 Frameworks for such assessments have already been proposed in the context of registry‐embedded clinical trials for regulatory purposes and may offer relevant guidance in the broader assessment of registry applicability for conditional marketing authorizations or reimbursement decisions. 38 , 39 Thirdly, national coordination of quality assurance and structural funding of available registries and RWD sources are important. This holds especially for HTA bodies, as they often make decisions with national implications. For example, the national HTA body in the Netherlands (Zorginstituut Nederland) has assessed all quality registries (focused on quality of care) for data content and governance. 40 The purpose of these assessments was twofold. On the one hand, to protect data contributors, often healthcare professionals, from an increasing registration burden by recognizing only registries of demonstrated value and minimizing overlap. On the other hand, the approved registries are guaranteed long‐term financial support and obtain a legal mandate for data collection, improving both completeness and coverage. An alternative for ensuring funding for registries could be to require manufacturers to provide financial support to collect the RWD that is required to assess the value of their newly registered products. Lastly, cross‐disciplinary learnings between registry holders and regulatory and HTA decision‐makers are essential to increase mutual understanding of each other's needs and limitations. Initiatives like the GetReal institute and RWE4Decisions play a key role in fostering this exchange. This study has both strengths and limitations. A strength of this study is that it was done in cooperation with representatives/initiators of the three registries that were included, allowing for the extraction of detailed information on the initiation and structure of the registries. Additionally, the three registries represent broad geographical coverage and are embedded in varying health systems. This increases the generalizability of the results. However, the broad scope of this study may also be a limiting factor, as most of the findings are on a high‐level basis as registries' capacities to generate evidence are highly dependent on the specific research questions to be answered. CONCLUSION This study demonstrates that the registries assessed are only partially aligned with the regulatory and HTA data requirements, with data coverage reflecting their original clinical purposes rather than policymaking needs. Also, these results highlight the urgent need for a consensus on standardized minimum datasets and the adoption of common data models to ensure interoperability, enhancing cancer registries' utility for decision‐making. Moreover, the identification and evaluation of relevant registries, combined with robust national coordination, are essential to predict future gaps in evidence for HTA purposes. Collectively, these measures will transform registries from being only clinical tools into vital resources, strengthening the contribution of registry‐RWD to evidence‐based regulatory and reimbursement strategies. FUNDING No funding was received for this work. CONFLICT OF INTEREST The authors declared no competing interests for this work. AUTHOR CONTRIBUTIONS All authors wrote the manuscript; all authors designed the research. MW performed the research. MW analyzed the data. Supporting information Table S1. Alphabetically ordered list of extracted and harmonized data elements, including the institution that included the data element in its guidance (EMA, ZIN, or both), decision on its inclusion in the reference list, comments on its inclusion and the final included data element. Table S2. Reference dataset of data elements categorized per domain.