Introduction
This report aims to explore whether and how we can assess equity of access to high-cost drugs for the treatment of rare conditions in England. We set out to use relevant data sets already collected and accessible to the National Disease Registration Service (NDRS), namely Hospital Episode Statistics (HES) [1] and data from Blueteq [2] - a prior approval system that is used for high-cost drug applications. Blueteq is a software system with the purpose to manage and authorise high-cost drugs for various health conditions, and it is used by the National Health System (NHS) and private healthcare providers to organise and monitor the distribution of prescriptions.
The Department for Health and Social Care’s Rare Diseases Action Plan 2025 [3] underscored the importance of evidence-based research on access to high-cost drugs for the treatment of rare conditions, and this report aims to contribute to this action from a data-driven and methodological perspective.
This work aims to be a first methodological appraisal of whether it is possible to use HES-linked Blueteq data to explore equity of access to treatment for rare conditions. This proof of concept was intended to answer whether:
It is possible to look at equity of access to treatment for rare conditions; and
How scalable the analysis is to other rare conditions, e.g., how easily a reproducible pipeline can be developed.
This proof of concept is applied to a descriptive study of socio-demographic and geographic characteristics of patients who have been treated for the rare disease Systemic Lupus Erythematosus (SLE), widely known as lupus, with the high-cost drug belimumab, which was approved by the National Institute for Health and Care Excellence (NICE) in June 2016 [4].
SLE / belimumab was chosen as a 1:1 matching disease-drug to explore whether this is possible using the data that is routinely available at a national scale and not requiring further data collection.
Any patient given belimumab in England will have received this drug to treat SLE, and their applications for belimumab and related clinical characteristics will have been recorded in Blueteq starting from June 2016 onwards. Using Blueteq data is key for the proof of concept presented in this report, as it ensures all patients given belimumab are included. This, in turn, supports an initial step toward developing a minimum viable pipeline by starting with the simplest drug–condition mapping.
SLE is an autoimmune disease, mostly affecting women, with no cure [5]. Belimumab is recommended by NICE as a treatment option for active autoantibody-positive SLE in people with high disease activity. Belimumab is administered as an intravenous (IV) infusion, under expert supervision or, since 2021, subcutaneous (SC) injection which is self-administered at home. This drug is only available via Blueteq and approval of the Blueteq pre-authorisation form requires the decision to initiate treatment to have been made at, or in discussion with a relevant specialist multidisciplinary team (MDT) involving clinicians from an NHS trust on the provider eligibility list for specialised rheumatology services [6].
Guide to the contents presented
This report describes both how we identified patients who had the treatment, and the attempts made to identify the eligible cohort of patients who could receive the treatment. Both parts are needed to analyse equity of access to treatment.
In summary, we were able to present the distributions of key socio-demographic and clinical characteristics of SLE patients treated with belimumab in England between 2016 and 2025.
As further explained in the section Data specifications, below, this report highlights the challenges of clinical identification of a denominator population who are eligible for treatment with belimumab in the absence of:
SLE Disease Activity Index (SLEDAI) scores
patient comorbidity profiles which define suitability for belimumab
prescription data to assess interacting or antagonistic treatments; and
data on the treatment length and efficacy of belimumab.
Data availability and access issues have impeded the development of robust comparisons between patients treated and untreated with the drug belimumab, or any other drug suitable to the management and treatment of SLE. However, we were able to show the relevant characteristics for the main axes of equity of access for those SLE patients who were treated with belimumab, after linking Blueteq data to HES.
Further to this, to understand more about the overall SLE patients in HES and assess potential avenues to identify the cohort of SLE patients eligible for belimumab, we conducted exploratory analyses of hospital activity. We looked at intensity of inpatient admissions and outpatient appointments, as well as comorbidity, based on the Charlson comorbidity index, using HES.
Initially we focused on the belimumab treated population, looking at hospital activity (attendance and episodes) for each patient in relation to their belimumab treatment start date in Blueteq, including the mean and median number of attendances up to six months prior to receiving treatment. Our hypothesis was that there would be a distinct increase in the density of hospital activity in the time prior to belimumab, which would be a signature that could be used to differentiate who could be eligible. However, we found this approach was not sufficient to identify those eligible for treatment with belimumab in the overall SLE patients in HES as hospital episodes intensity is likely compounded with other factors, like age and comorbidity.
We then examined comorbidities to determine whether those not treated with belimumab had conditions that might have excluded them from being eligible. Although we observed some differences, such as a slightly higher burden of comorbidities among untreated patients, these factors alone were insufficient to reliably identify patients with high disease activity for SLE who would be suitable for belimumab treatment.
We present some of the results of the exploratory analyses to try to identify SLE patients eligible for belimumab treatment below. While these exploratory analyses provide some insights, these data and methods are not sufficient to assess equity of access to belimumab. More precise clinical data that allow direct coding of rare conditions and markers for aggravation, as well as advanced analytical methods that consider the complexities and less clear-cut characteristics of populations living with rare conditions, may allow progress in the assessment of equity of access to treatment for these conditions.
Analytical pipeline developed for this project
We were able to develop an analytical pipeline for the identification of SLE patients in England using the ICD-10 codes for SLE (M321, M328, and M329) to select patients from HES. We were also able to identify the SLE patients treated with belimumab in Blueteq and linked these to HES to analyse characteristics relevant to this report, which the pipeline fully covers.
Due to limited access to more extensive sources of patient clinical data, we were unable to identify the belimumab-treatment eligible subgroup within the HES SLE cohort, as HES records all patients with SLE but does not contain the clinical indicators required to determine eligibility.
The pipeline developed for this project is a starting point for future extensions to cover the analysis of treatments for patients with rare conditions. It could be scaled to cover conditions that are identifiable, at present, using data that are well-captured within HES, and largely using ICD-10 codes. The data sets used do not contain information on conditions that require more granular classifications, such as ORPHA codes and SNOMED CT, or more precise genomic data, however our pipeline could be expanded in the future to use other data sets which rely on these clinical coding systems.
The code developed for this work’s analytical pipeline is maintained on GitHub in the NHS England NDRS space at https://github.com/NHSE-NDRS. At present the repository is only accessible to NDRS analysts. We will aim to make it publicly available in the future.
Data specifications and challenges with defining a cohort of SLE patients eligible for treatment
Clinical identification of SLE patients eligible for treatment with belimumab: SLEDAI scores in Blueteq
A total of 924 SLE patients treated with belimumab were identified using Blueteq and were linked to patient records from HES Admitted Patient Care (APC), covering inpatient hospital attendances from January 2000 to August 2025 and to HES Out-Patient care (OP), covering outpatient appointments from April 2009 to August 2025.
SLE is a complex autoimmune disease and identifying the overall cohort of patients with active disease who are eligible for, but have not been treated with, belimumab is crucial to monitor and report on equitable access to treatment.
Using ICD-10 codes M321, M328, and M329 [7] allows for the identification of the SLE patients from HES. However, HES does not contain the clinical information necessary to determine eligibility for belimumab treatment.
Determining eligibility for belimumab requires specific clinical information, including comorbidity profiles, and in particular an SLE Disease Activity Index (SLEDAI) score [8], a clinical tool used to measure SLE disease activity. SLE patients’ SLEDAI scores are however captured when the clinician submits their prior approval request to treat with belimumab, which is available via Blueteq. SLE patients with SLEDAI scores equal to 10 and above are eligible for treatment with belimumab.
It is important to note that not all SLE patients with high disease activity as measured by SLEDAI scores will be eligible for belimumab treatment. Exclusions apply based on presence of other health conditions, related treatments already in progress, age, and duration of the disease [9].
In Blueteq, we identified 924 patients who, based on MDT recommendations and SLEDAI scores, were eligible to receive belimumab with their treatment planned to commence between June 2016 and August 2025.
SLE patients’ sociodemographic, geographic and hospital episodes data in HES
We linked the 924 SLE patients in Blueteq to HES data, in order to describe the axes of equity of access to treatment that we set out to analyse, namely gender, age, ethnicity, Index of multiple deprivation (IMD) [10], and travel time from a patient’s home to a specialised rheumatology treatment centre, as well as intensity of hospital episodes and comorbidity.
The SLE patients in England were identified through HES APC from January 2000 to 20 August 2025 and from HES OP from April 2009 to 20 August 2025 using the ICD-10 codes for SLE (M321, M328, M329). However, 18 patients with confirmed SLE from Blueteq were not initially captured in HES when selecting by the SLE ICD-10 codes, but were retrieved instead using codes for an admission or appointment for rheumatology (410 rheumatology) or renal medicine (361 Renal medicine).
In HES 72,308 SLE patients were identified and were alive when belimumab was approved for use by NICE (22 June 2016). Of the 72,308 patients, 71,384 (98.7%) patients were not known to have been treated with belimumab by the end of the observation window on 20 August 2025. A vital status trace for all the SLE patients in HES was carried out on 7 November 2025 using the Personal Demographics Service (PDS) [11].
Throughout the report, we refer to this as the HES SLE cohort.
Due to the difficulties in identifying a denominator population of all those, untreated and treated, eligible for belimumab treatment, the analyses presented here are exclusively descriptive; the 924 SLE patients treated with belimumab cannot be directly compared to the overall HES SLE cohort, as the latter contains the treated patients, patients not known to have been treated with belimumab or other drugs, and ineligible patients.
The consort diagram in figure 1 shows the process followed for the linkage between HES and Blueteq.
Figure 1. Consort diagram: Blueteq data linkage with HES APC and OP
Characteristics explored
This section presents characteristics of the SLE patients that are routinely available for the SLE treated and overall HES SLE cohort, divided by data source, i.e., Blueteq and HES.
As underscored in the sections above, these results are exclusively descriptive, with no intention to compare the SLE patients treated with belimumab and the overall HES SLE cohort, as the latter contains the treated patients, patients not known to have been treated with belimumab or other drugs, and ineligible patients.
The 924 SLE patients treated with belimumab: data from Blueteq
Belimumab treatment timeline
This figure reports the counts of patients treated with belimumab at treatment start date for the 924 SLE patients in Blueteq. Data points represent each month.
Figure 2. Belimumab treatment timeline - plot
Table 1. Belimumab treatment timeline - table
The uptake of belimumab increased over time with some variability, with accentuated increase from beginning of 2022. The change from IV to SC injection in 2021 may have affected belimumab uptake, as the latter is more easily administered.
SLEDAI scores
Figure 3 shows the distribution of SLEDAI scores in the belimumab-treated cohort, identified and reported in Blueteq data. SLEDAI scores are considered the most important disease-severity marker to prescribe treatment with belimumab. Without this clinical marker, identification of a suitable denominator population eligible for treatment becomes methodologically demanding.
As shown in the graph, Blueteq reports SLEDAI scores lower than 10 or missing for 149 (16%) patients treated with belimumab. This is likely due to two main causes: data entry issues and MDT approval of treatment with belimumab despite the SLEDAI score, depending on the comorbidity profiles of the patients.
Figure 3. SLEDAI Score - Treated with belimumab plot
Table 2. SLEDAI Score - Treated with belimumab table
The 924 SLE patients treated with belimumab: data from HES
This section of the report presents the distribution of key sociodemographic characteristics for the 924 patients in Blueteq, which is achievable via linkage with HES data. Had we been able to identify the population eligible for belimumab treatment from HES, we would be equipped to replicate these graphs below for this population, which would allow analysis of equity of access to treatment, and robust conclusions on differences between the treated and the untreated population.
This report is a first example of analysis on access to belimumab treatment for patients with severe SLE using Blueteq and HES data.
Gender
Gender was defined as person stated gender and the data only presented information for females and males. SLE patients are in large majority females, and this is reflected also in our data for those treated with belimumab.
Figure 4. Gender - Treated with belimumab plot
Table 3. Gender - Treated with belimumab table
Age
Using the data available, we could calculate age in two ways: 1) age at cohort entry, which corresponds to the first available record for the patient in HES (source = HES); 2) age at treatment, which corresponds to a patient’s age at the start of their belimumab treatment (source = Blueteq). Age was categorised in 5-year age bands, starting from 0-4 and ending with the band 90+.
Identifying the population eligible for treatment with belimumab would have allowed computations of age differences between the treated and the untreated patients.
Age at first SLE record in HES
Figure 5. Age at first SLE record in HES - Treated with belimumab plot
Table 4. Age at first HES record with SLE - Treated with belimumab table
Age at belimumab treatment
Figure 6. Age at belimumab treatment - Treated with belimumab plot
Table 5. Age at belimumab treatment - Treated with belimumab table
Ethnicity
Ethnicity was defined according to the most recent National Disease Registration Service (NDRS) ethnicity classification, with macro categories Asian (excluding Chinese), Black, Chinese, Mixed, White, Other, Unknown.
Ethnicity was available in HES for the overall SLE HES cohort, including the patients identified in Blueteq. We will be enabled to look at ethnic differentials in access to treatment with belimumab once we can identify a population eligible for the relevant treatment.
Figure 7. Ethnicity - Treated with belimumab plot
Table 6. Ethnicity - Treated with belimumab table
IMD
IMD in HES was computed using the English IMD 2019, based on lower super output area (LSOA) of residence at start of treatment with belimumab. It is coded from 1 – most deprived to 5 – least deprived, with equal number of LSOAs in each quintile.
Together with ethnicity, IMD represent one of the main axes of inequalities in socio-epidemiological studies assessing the wider determinants of health. Being able to adequately measure equity of access to treatment depends on the possibility to reliably define an eligible-for-treatment population, using clinical information and/or advanced statistical models to estimate this population, and then assess differences in the distribution of IMD and ethnicity between this and the treated patients.
Figure 8. IMD - Treated with belimumab plot
Table 7. IMD - Treated with belimumab table
Travel times
Using HES, postcode was selected at time of treatment for the 924 SLE patients who received belimumab. To assess travel times, we need to know both the patient’s postcode of residence, and where they were treated from. This can be identified from Blueteq. Having access to postcode data from HES, analyses can be mapped up to the geography of interest, including also ICBs, for instance.
For the travel times analysis, NDRS carried out postcode-to-postcode mapping of all postcodes in England using the Graphhopper Open-Source routing engine on maps from OpenStreetMap [12]. Using this table, we linked the patient’s postcode at the start of their belimumab treatment to the postcode of all 37 specialised rheumatology centres and selected the off-peak shortest driving distance a patient would need to travel to a centre.
Travel time data were missing for one patient. The mean and median travel times were, respectively just over 53 minutes and 42 minutes.
Figure 9. Travel times - Treated with belimumab plot
Travel times - Treated with belimumab summary table
Table 8. Travel times - Treated with belimumab table
Hospital episode activity
We investigated the intensity of hospital inpatient admissions and outpatient appointments up to 6-months prior to a patient’s belimumab treatment start date in Blueteq. The hypothesis was that patients would be undergoing more medical tests and consultations in the time leading to the prescription of belimumab, considering the drug is given for high disease activity. Future developments of this work will aim to use information on hospital activity, paired with other clinical data, to identify the population eligible for treatment and run analyses on equity of access to treatment.
This section of the report summarises hospital activity for each patient treated with belimumab for SLE. Hospital inpatient admissions and outpatient appointments were only included if patients had one of the three SLE ICD-10 codes (M321, M328, M329) or a main specialty of rheumatology or nephrology captured in HES. Two patients did not have any hospital activity six months prior to starting belimumab so were excluded.
Table 9: Summary of hospital activity six-months prior to a patient’s belimumab treatment start date (N = 922)
| Median | Mean | Min | Max | Q1 | Q3 | IQR | SD |
|---|---|---|---|---|---|---|---|
| 9 | 10.9 | 1 | 90 | 6 | 13.8 | 7.8 | 7.6 |
Table 9 and Figure 10 focus on SLE, rheumatology and nephrology inpatient admissions and outpatient appointments six months prior to a patient starting treatment with belimumab. Two patients did not have any hospital admissions/appointments 6 months prior to receiving belimumab.
This period was selected to explore whether patterns of utilisation might reflect episodes of increased disease activity. Because SLE is a complex condition, that typically requires frequent monitoring and multidisciplinary care, pre-treatment hospital activity can be substantial and may offer useful insight. However, utilisation alone is not a definitive marker of active disease. Without accompanying clinical data these patterns are exploratory rather than conclusive.
Figure 10: Activity density plot - hospital activity 6-months prior to their belimumab treatment start date by appointment type (N = 922)
Comorbidities and Charlson score
Same as for hospital episode activity, comorbidities were computed from HES for SLE treated patients with the hypothesis that eligibility for belimumab treatment depends also on the patient’s overall health status, in particular on presence of other severe conditions which would impede belimumab administration. The Charlson comorbidity score was used as a measure of the patient’s comorbid status, and the methodology was as in Quan et al. (2005) [13].
Differently from Quan et al., the SLE ICD-10 codes (M321, M328, M329) were removed from the codes used to identify ‘Connective tissue disorder’. This meant we did not include SLE as a comorbidity.
This section displays results on the Charlson score for the 924 SLE patients treated with belimumab.
Figure 11. Charlson Score - Treated with belimumab plot
Table 10. Charlson comorbidity score: SLE patients treated with belimumab (Teated cohort)
Discussion and conclusions
This report aims to be a starting point for population health surveillance focused on rare conditions and access to treatment, primarily by showing the limits of what can be achieved with what we currently have access to. For the first time we linked data on a rare disease cohort of patients who received treatment to population-level data from routinely collected electronic health records. We reported on treatment intakes among different population subgroups via descriptive analysis, however, we have shown that we can only describe equity of access to drugs by multiple axes of potential inequality if both the prescription information is available and the identification of a cohort eligible for treatment with specific drugs is possible.
As the data landscape evolves, including genomic data availability, this will improve our ability to identify rare conditions with sufficient specificity to accurately cohort. NDRS is progressing their work on the Rare Disease Data Set, which aims to collect data from highly specialised services to enhance efficiency, scalability, consistency and standardisation of rare disease registration (https://digital.nhs.uk/ndrs/data/data-sets/rdds). While able to identify populations affected by certain conditions, we currently lack the granularity of severity indicators and contraindications that allow us to refine the cohort to those that would be eligible for specific treatments.
Reflecting on the scalability of the approach proposed in this report, there are conditions for which assessing equity of access to treatment becomes incrementally harder, for instance for those with non-specific ICD-10 codes, common presentation, or treatment with drugs with multiple indications.
Equally as importantly, this report intends to start discussions about concrete and clinically sound data-driven plans to identify valid treatment cohorts, especially through:
Broadened access to rigorously validated, fine‑grained patient‑level clinical information and data;
Expanded analytical capacity, in terms of time spent developing the reproducible pipeline presented here, increasing capability of the minimum viable product, to scale analytical operations; and
Enhanced data‑processing capability through complex modelling and insight generation.
Data limitations and proposed solutions
This report clearly shows the necessity for high-quality clinical information on patients with SLE from administrative, routinely available electronic health data, which would allow the identification of a suitable cohort of SLE patients eligible to be treated with belimumab, i.e. a suitable denominator, and the comparison of this with the treated cohort identified in Blueteq.
Solutions to the issues presented in this report could be found in:
linkage to other data sets that would permit assessment of severity of disease, such as prescription and primary care data
appropriate data being made an integral part of administratively available data, such as severity markers and contraindications.
Having the right data available, in a timely way, to create the right intelligence is absolutely mission critical to improving the lives of those with these conditions.
Next steps
The present pipeline incorporates modular code for descriptive profiling of socio demographic and geographic characteristics of both the broader SLE population and the belimumab treated cohort. Below, we discuss approaches to expand the present capabilities.
Scaling the analytical approach
With respect to the analytical pipeline, we have delivered a minimum viable product demonstrating end-to-end linkage of Blueteq clinical data for a rare disease cohort (SLE) with routinely collected electronic health records in HES. The pipeline incorporates modular code for descriptive profiling of sociodemographic and geographic characteristics of both the broader SLE population and the belimumab-treated cohort.
Scaling this approach to support equity of access analyses across additional rare conditions requires addressing the same constraints identified for data augmentation — namely, the availability of valid, clinically coherent data for cohort definition, and sustained investment in analytical capacity. This includes the resource required to build and maintain underlying data infrastructure: robust clinical coding processes, systematic quality assurance, iterative testing, scalable data engineering practices, and the development of reproducible analytical frameworks.
Additional analytical resources
Subject to more accurate clinical data being available, additional analytical resources would be required to design, optimise, and validate analytical workflows. This would expand applications of data engineering and statistical modelling, developing reusable components that reduce duplication of effort and enable generalisation of analytical pipelines across multiple conditions.
Identification of an eligible population for treatment
More complex methods could be applied to the data to approximate an eligible rare disease cohort for treatment, reducing reliance on perfect clinical coding. When clinical coding or structured fields are insufficient to reliably identify an eligible patient cohort, analytical strategies that infer eligibility profiles can be used to build a probabilistic or approximate cohort. These methods allow analysts to work around incomplete, inconsistent, or non‑specific coding by leveraging patterns in routinely collected patient data. For example:
- Cluster analysis can be used to infer groups of patients with similar clinical characteristics when eligibility criteria are not explicitly coded. This approach leverages dimensionality reduction and unsupervised classification to detect patterns in combinations of diagnoses, treatments, episodes, and demographic variables.
Rare conditions often have heterogeneous presentations, leading to variable coding practices. Eligibility for specialised treatments may depend on combinations of comorbidities, severity markers, or prior treatment history that are not captured in a single code. Clustering helps uncover latent structures—groups of patients whose data patterns resemble those expected of an eligible cohort. Feature engineering would contribute to this, and we have shown that our pipeline can already accommodate for some of them (the Charlson Comorbidity Index, episode frequency), while further data are needed to identify acute flares, prescribing patterns or longitudinal trajectories (e.g., deterioration profiles). Unsupervised clustering algorithms offer a scalable approach for other rare diseases with similar coding gaps.
- Multilevel (hierarchical) models allow incorporation of multiple levels of data structure—for example, clinical episodes nested within individuals, and individuals nested within providers or regions. These models are suitable when eligibility depends on episode‑level data (e.g., flares, severity markers) that accumulate to define a patient’s overall profile. Many eligibility criteria for approximated cohorts are related to clinical progression: frequency of episodes, progression patterns, refractory disease, or sustained severity.
Generalised linear mixed models and specifically longitudinal trajectories based on episode-level data (e.g., hospitalisations, complications) provide a richer picture than static patient-level summaries. On top of this, multilevel models formally handle the dependency between repeated episodes within the same patient, reducing bias and improving predictive accuracy.
These methods reflect best practice in handling complex, noisy, real‑world health data and support scalable equity analyses across the rare condition landscape. The integration of these powerful statistical models with the latest data science developments have shown great potential in addressing complex questions such as risk of treatment delays and equity of access to care for the US context [14].
Conclusions
Different axes of equity of access to treatment for rare conditions could be explored reliably upon availability of information to derive patient’s eligibility for suitable treatment. Further work is required to improve data availability as well to test advanced analytical methods to robustly identify eligible-for-treatment populations.
NDRS’s commitment to contribute to the work carried out as part of the UK Rare Disease Framework would enable accurate national disease counts and population-level insights, to improve understanding of natural history of conditions, supporting policy, commissioning and research, towards burden reduction for public healthcare services.
References
[1] Data Services NHS England (2000-2025) Hospital Episode Statistics (HES) [Data set]. [Online]. Available: https://digital.nhs.uk/data-and-information/data-tools-and-services/data-services/hospital-episode-statistics. [Accessed 2026]
[2] NHS England (2026). Blueteq High-Cost Drugs Database [Data set]. Blueteq/NHS England. [Online]. Available: https://www.blueteq-systems.co.uk/nhse-hcd/default.aspx. [Accessed 2026].
[3] Department of Health and Social Care (2025). England Rare Diseases Action Plan 2025 [Online]. Available: https://www.gov.uk/government/publications/england-rare-diseases-action-plan-2025/england-rare-diseases-action-plan-2025-main-report. [Accessed 2026]
[4] National Institute for Health and Care Excellence (NICE) (2021). Belimumab for treating active autoantibody-positive systemic lupus erythematosus. Technology appraisal guidance. [Online]. Available: https://www.nice.org.uk/guidance/TA752. [Accessed 2026]
[5] National Health Service (NHS) (2023). Lupus. [Online]. Available: https://www.nhs.uk/conditions/lupus/. [Accessed 2026]
[6] National Health Service (NHS). Prescribed Specialised Services Tools. [Online]. Available: https://digital.nhs.uk/services/national-casemix-office/downloads-groupers-and-tools/prescribed-specialised-services-pss-tools. [Accessed 2026]
[7] Rutter, M (2023) Health outcomes in people with rare autoimmune rheumatic diseases before and during the COVID-19 pandemic and the effect of corticosteroids: whole population data from England. [Thesis]. [Online] Available: Edited Final Thesis Changes Accepted Megan Rutter.pdf. [Accessed 2026]
[8] Taraneh Tofighi, Eric F. Morand, Zahi Touma (2021). Systemic Lupus Erythematosus Outcome Measures for Systemic Lupus Erythematosus Clinical Trials, Rheumatic Disease Clinics of North America, Volume 47, Issue 3, Pages 415-426, ISSN 0889-857X, ISBN 9780323835565, https://doi.org/10.1016/j.rdc.2021.04.007.
[9] Tariq A, Sharif M, Ammar M. Association between disease duration and comorbidity burden in systemic lupus erythematosus: a multicentre study from Pakistan. (2025) Lupus Science & Medicine. 12:e001849. https://doi.org/10.1136/lupus-2025-001849 [10] Ministry of Housing, Communities and Local Government (2019). English indices of deprivation 2019. [Online]. Available: https://www.gov.uk/government/statistics/english-indices-of-deprivation-2019. [Accessed 2026].
[11] NHS England, The Personal Demographics Service (PDS) (2026). [Online]. Available from: https://digital.nhs.uk/services/personal-demographics-service. [Accessed 2026].
[12] Graphhopper contributors, “Graphhopper open source,” 2025. [Online]. Available: https://www.graphhopper.com/open-source/. [Accessed 2026].
[13] Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, Saunders LD, Beck CA, Feasby TE, Ghali WA (2005). Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 43(11):1130-9. doi: 10.1097/01.mlr.0000182534.19832.83. PMID: 16224307.
[14] Frosch ZAK, Hasler J, Handorf E, et al. (2023) Development of a Multilevel Model to Identify Patients at Risk for Delay in Starting Cancer Treatment. JAMA Netw Open. 6(8):e2328712. doi:10.1001/jamanetworkopen.2023.28712
About the NDRS
The National Disease Registration Service (NDRS) is part of NHS England (NHSE). Its purpose is to collect, curate, quality-assure and analyse on patients with cancer, congenital anomalies, and rare diseases. It provides robust surveillance to monitor and detect changes in health and disease in the population. NDRS is a vital resource that helps researchers, healthcare professionals and policy makers make decisions about NHS services and the treatments people receive.
The NDRS includes:
- the National Cancer Registration and Analysis Service (NCRAS)
and
- the National Congenital Condition and Rare Diseases Registration Service (NCARDRS)
Healthcare professionals, researchers and policy makers use data to better understand population health and disease. The data is provided by patients and collected by the NHS as part of their care and support. The NDRS uses the data to help:
- understand cancer, rare diseases, and congenital anomalies
- improve diagnosis
- plan NHS services
- improve treatment
- evaluate policy
- improve genetic counselling
For queries relating to this document, please contact ndrsenquiries@nhs.net
Patient acknowledgement
This work uses data that has been provided by patients and collected by the NHS as part of their care and support.