Today, there are many treatment options available for the management of ulcerative colitis, creating challenges in selecting the most efficacious and cost-effective treatment sequences. Treatments in the anti-tumor necrosis factor alpha (TNFα) therapeutic class, as well as vedolizumab, are widely used and endorsed as first-line options according to treatment guidelines. The aim of this study was to compare treatment sequences involving vedolizumab and the anti-TNFα treatment adalimumab in terms of cost-effectiveness in the treatment of moderately to severely active ulcerative colitis in Italy. A cost-effectiveness model comparing treatment sequences within the Italian National Health Service in terms of costs and quality-adjusted life years (QALYs) with a lifetime horizon was developed. The analysis focused on the relative positioning of vedolizumab and adalimumab, leveraging the results of the landmark head-to-head VARSITY clinical trial as key inputs. The robustness of the results was investigated through a range of sensitivity and scenario analyses. The strategy of vedolizumab as a first-line advanced therapy followed by adalimumab was associated with higher costs and health benefits compared with first-line adalimumab followed by vedolizumab. The incremental cost-effectiveness ratio was €16,146/QALY, which was found to be robust to changes to inputs associated with areas of high uncertainty. This economic evaluation estimated a 94

A health economic model may include a set of related inputs whose true values are uncertain, but that can be assumed to follow a logical order. Various approaches are available for performing probabilistic sensitivity analysis while preserving the order constraint, one such approach is known as the difference method. The difference method approach appears to have many of the required properties, has been endorsed by good practice guidelines, and is likely to prove a popular approach. However, the proposed implementation of the difference method approach is cumbersome, requiring numerical estimation, which might present a barrier to its adoption. Furthermore, it is unclear whether the method can always be applied to three or more model inputs and whether it is unbiased across all possible input values. This study has investigated these three issues for ordered inputs bounded between 0 and 1. An analytic solution is given that allows for more straightforward and compact implementation. The difference method approach cannot always be applied to a set of three or more model inputs, and this depends on the relative size of the variances of the logit-transformed Beta distributions fitted to each variable. The approach can also produce samples with biased means and variances under certain combinations of input means and variances. It is recommended that the difference method approach be used where appropriate; however, an understanding of its limitations is necessary to identify such cases.

AIMS:There has been sustained growth in the prescribing of direct oral anticoagulants (OACs) in primary care in the UK. Given the different indications, properties and prices of OACs, variation between prescribers is expected; however, a high level of variation may be evidence of inappropriate or suboptimal prescribing. This study examined the variation in the relative use of OACs in primary care in Wales. METHODS:Data on total defined daily doses of all community-dispensed OACs in 2019 were linked at the GP practice level with disease registers, patient demographic data and GP and patient numbers. The relative use of each OAC, as a fraction of all OACs prescribed, was analysed using Dirichlet regression to quantify the association between prescribing patterns and practice and area-level characteristics. RESULTS:Across 417 GP practices, the mean (range) in the relative prescribing of warfarin was 37% (6%-64%), apixaban was 32% (2%-65%), rivaroxaban 23% (0%-66%), dabigatran 3% (0%-23%) and edoxaban 6% (0%-59%). Statistical modelling provided strong evidence that prescribing patterns are associated with a GP practice's health board and also their nearest major hospital. Compared to the null model, a model including health board resulted in a 15% fall in Akaike information criterion, increasing to 20% with the addition of nearest major hospital and 27% including further covariates. CONCLUSION:Systematic variation in OAC prescribing, by health board and based on nearest hospital, indicates that factors other than patient clinical characteristics and preferences may be influencing prescribing decisions.

The Bayesian decision-analytic approach to trial design uses prior distributions for treatment effects, updated with likelihoods for proposed trial data. Prior distributions for treatment effects based on previous trial results risks sample selection bias and difficulties when a proposed trial differs in terms of patient characteristics, medication adherence, or treatment doses and regimens. The aim of this study was to demonstrate the utility of using pharmacometric-based clinical trial simulation (CTS) to generate prior distributions for use in Bayesian decision-theoretic trial design. The methods consisted of four principal stages: a CTS to predict the distribution of treatment response for a range of trial designs; Bayesian updating for a proposed sample size; a pharmacoeconomic model to represent the perspective of a reimbursement authority in which price is contingent on trial outcome; and a model of the pharmaceutical company return on investment linking drug prices to sales revenue. We used a case study of febuxostat versus allopurinol for the treatment of hyperuricemia in patients with gout. Trial design scenarios studied included alternative treatment doses, inclusion criteria, input uncertainty, and sample size. Optimal trial sample sizes varied depending on the uncertainty of model inputs, trial inclusion criteria, and treatment doses. This interdisciplinary framework for trial design and sample size calculation may have value in supporting decisions during later phases of drug development and in identifying costly sources of uncertainty, and thus inform future research and development strategies.

Background Sciatica is a common condition reported to affect > 3% of the UK population at any time and is most often caused by a prolapsed intervertebral disc. Currently, there is no uniformly adopted treatment strategy. Invasive treatments, such as surgery (i.e. microdiscectomy) and transforaminal epidural steroid injection, are often reserved for failed conservative treatment. Objective To compare the clinical effectiveness and cost-effectiveness of microdiscectomy with transforaminal epidural steroid injection for the management of radicular pain secondary to lumbar prolapsed intervertebral disc for non-emergency presentation of sciatica of < 12 months’ duration. Interventions Patients were randomised to either (1) microdiscectomy or (2) transforaminal epidural steroid injection. Design A pragmatic, multicentre, randomised prospective trial comparing microdiscectomy with transforaminal epidural steroid injection for sciatica due to prolapsed intervertebral disc with < 1 year symptom duration. Setting NHS services providing secondary spinal surgical care within the UK. Participants A total of 163 participants (aged 16–65 years) were recruited from 11 UK NHS outpatient clinics. Main outcome measures The primary outcome was participant-completed Oswestry Disability Questionnaire score at 18 weeks post randomisation. Secondary outcomes were visual analogue scores for leg pain and back pain; modified Roland–Morris score (for sciatica), Core Outcome Measures Index score and participant satisfaction at 12-weekly intervals. Cost-effectiveness and quality of life were assessed using the EuroQol-5 Dimensions, five-level version; Hospital Episode Statistics data; medication usage; and self-reported cost data at 12-weekly intervals. Adverse event data were collected. The economic outcome was incremental cost per quality-adjusted life-year gained from the perspective of the NHS in England. Results Eighty-three participants were allocated to transforaminal epidural steroid injection and 80 participants were allocated to microdiscectomy, using an online randomisation system. At week 18, Oswestry Disability Questionnaire scores had decreased, relative to baseline, by 26.7 points in the microdiscectomy group and by 24.5 points in the transforaminal epidural steroid injection. The difference between the treatments was not statistically significant (estimated treatment effect –4.25 points, 95% confidence interval –11.09 to 2.59 points). Nor were there significant differences between treatments in any of the secondary outcomes: Oswestry Disability Questionnaire scores, visual analogue scores for leg pain and back pain, modified Roland–Morris score and Core Outcome Measures Index score up to 54 weeks. There were four (3.8%) serious adverse events in the microdiscectomy group, including one nerve palsy (foot drop), and none in the transforaminal epidural steroid injection group. Compared with transforaminal epidural steroid injection, microdiscectomy had an incremental cost-effectiveness ratio of £38,737 per quality-adjusted life-year gained and a probability of 0.17 of being cost-effective at a willingness to pay threshold of £20,000 per quality-adjusted life-year. Limitations Primary outcome data was invalid or incomplete for 24% of participants. Sensitivity analyses demonstrated robustness to assumptions made regarding missing data. Eighteen per cent of participants in the transforaminal epidural steroid injection group subsequently received microdiscectomy prior to their primary outcome assessment. Conclusions To the best of our knowledge, the NErve Root Block VErsus Surgery trial is the first trial to evaluate the comparative clinical effectiveness and cost-effectiveness of microdiscectomy and transforaminal epidural steroid injection. No statistically significant difference was found between the two treatments for the primary outcome. It is unlikely that microdiscectomy is cost-effective compared with transforaminal epidural steroid injection at a threshold of £20,000 per quality-adjusted life-year for sciatica secondary to prolapsed intervertebral disc. Future work These results will lead to further studies in the streamlining and earlier management of discogenic sciatica. Trial registration Current Controlled Trials ISRCTN04820368 and EudraCT 2014-002751-25. Funding This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment ; Vol. 25, No. 24. See the NIHR Journals Library website for further project information.

Background: The optimal treatment for sciatica secondary to a herniated lumbar disc (HLD) remains controversial with a paucity of evidence for non-surgical treatments such as transforaminal epidural steroid injection (TFESI) compared to surgical microdiscectomy (surgery). Methods: 163 patients with non-emergency sciatica secondary to a HLD were randomised to surgery or TFESI. Adults between 16 and 65 years, who failed simple conservative care and had symptoms for less than 12 months were recruited from 11 UK specialist spinal units. The primary outcome was the Oswestry Disability Questionnaire (ODQ) at 18 weeks post-randomisation. Secondary outcomes included visual analogue scores (VAS) for leg and back pain, modified Roland-Morris (MRM), Core Outcome Measures Index (COMI), and participant satisfaction. Cost-effectiveness was estimated from the EQ-5D-5L, Hospital Episode Statistics, medication usage and self-report resource-use data. Adverse event (AE) data were collected. Findings: No significant differences were found in ODQ at 18 weeks for surgery versus TFESI (mean improvement for surgery 26.7 points, TFESI 24.5 points; estimated treatment difference was -4.25 (95% confidence interval (CI) -11.09 to 2.59) p=0.22). No significant differences were found for secondary outcomes up to 54 weeks in ODQ scores, VAS-back, VAS-leg, MRM and COMI. There were 4 (3.8%) serious adverse events (SAEs) in the surgery group, and none in the TFESI group. Compared to TFESI, surgery had an incremental cost-effectiveness ratio of £38,737 per QALY gained, and a probability of 0.17 of being cost-effective at a willingness to pay threshold of £20,000 per QALY. Interpretation: For patients with sciatica secondary to HLD with duration less than 1 year, no statistically significant difference was found between surgery and TFESI. Surgery is unlikely to be a cost-effective alternative to TFESI. Trial Registration: The trial was granted Clinical Trial Authorisation (CTA reference: 21322/004/001) by the Medicines and Healthcare products Regulatory Agency (MHRA). ISRCTN04820368 and EudraCT: 2014-002751-25. Funder: Health Technology Assessment (HTA) programme of the National Institute for Health Research (NIHR; UK).Declaration of Interests: Anthony Marson reports grants from UCBpharma outside the submitted work. Paula Williamson was Director of the Clinical Trials Research Centre (CTRC) up to the end of December 2018. CTRC (now known as LCTC) was, and is, in receipt of NIHR CTU Support Funding. Ganesan Baranidharan reports grants and personal fees from Boston Scientific, grants and personal fees from Nevro Corporation, grants and personal fees from Abbott, grants and personal fees from Nalu Medical, outside the submitted work. PJ Hutchinson reports grants from the NIHR (Research Professorship) and the Royal College of Surgeons of England outside the submitted work. The remaining authors have no conflicts of interest to declare.Ethics Approval Statement: The trial protocol was approved by a research ethics committee (NRES Committee North West – Liverpool Central, reference: 14/NW/1219).

Market access and pricing of pharmaceuticals are increasingly contingent on the ability to demonstrate comparative effectiveness and cost-effectiveness. As such, it is widely recognized that predictions of the economic potential of drug candidates in development could inform decisions across the product life cycle. This may be challenging when safety and efficacy profiles in terms of the relevant clinical outcomes are unknown or highly uncertain early in product development. Linking pharmacometrics and pharmacoeconomics, such that outputs from pharmacometric models serve as inputs to pharmacoeconomic models, may provide a framework for extrapolating from early-phase studies to predict economic outcomes and characterize decision uncertainty. This article reviews the published studies that have implemented this methodology and used simulation to inform drug development decisions and/or to optimize the use of drug treatments. Some of the key practical issues involved in linking pharmacometrics and pharmacoeconomics, including the choice of final outcome measures, methods of incorporating evidence on comparator treatments, approaches to handling multiple intermediate end points, approaches to quantifying uncertainty, and issues of model validation are also discussed. Finally, we have considered the potential barriers that may have limited the adoption of this methodology and suggest that closer alignment between the disciplines of clinical pharmacology, pharmacometrics, and pharmacoeconomics, may help to realize the potential benefits associated with linked pharmacometric-pharmacoeconomic modeling and simulation.

Objective: We modeled the impact of changing Specialist Treatment Access Rates to different treatment pathways on the future prevalence of alcohol dependence, treatment outcomes, service capacity, costs, and mortality. Method: Local Authority numbers and the prevalence of people “potentially in need of assessment for and treatment in specialist services for alcohol dependence” (PINASTFAD) are estimated by mild, moderate, severe, and complex needs. Administrative data were used to estimate the Specialist Treatment Access Rate per PINASTFAD person and classify 22 different treatment pathways. Other model inputs include natural remission, relapse after treatment, service costs, and mortality rates. “What-if” analyses assess changes to Specialist Treatment Access Rates and treatment pathways. Model outputs include the numbers and prevalence of people who are PINASTFAD, numbers treated by 22 pathways, outcomes (successful completion with abstinence, successfully moderated nonproblematic drinking, re-treatment within 6 months, dropout, transfer, custody), mortality rates, capacity requirements (numbers in contact with community services or staying in residential or inpatient places), total treatment costs, and general health care savings. Five scenarios illustrate functionality: (a) no change, (b) achieve access rates at the 70th percentile nationally, (c) increase access by 25%, (d) increase access to Scotland rate, and (e) reduce access by 25%. Results: At baseline, 14,581 people are PINASTFAD (2.43% of adults) and the Specialist Treatment Access Rate is 10.84%. The 5-year impact of scenarios on PINASTFAD numbers (vs. no change) are (B) reduced by 191 (-1.3%), (C) reduced by 477 (-3.3%), (D) reduced by almost 2,800 (-19.2%), and (E) increased by 533 (+3.6%). The relative impact is similar for other outputs. Conclusions: Decision makers can estimate the potential impact of changing Specialist Treatment Access Rates for alcohol dependence. Objectif : Modéliser l’impact de la variation des taux d’accès aux différentes trajectoires de traitements spécialisés, sur la prévalence future de la dépendance à l’alcool, l’impact du traitement, le volume de services, les coûts et la mortalité. Méthode : Au sein des administrations régionales, les nombres et la prévalence de personnes ayant ‘potentiellement besoin d’être évaluées pour un traitement dans les services spécialisés en dépendance à l’alcool’ (PBÉTSSDA) sont estimés en fonction de niveaux de besoins dits légers, modérés, sévères et complexes. Les taux d’accès aux traitements spécialisés par personne ayant PBÉTSSDA sont estimés en fonction de 22 trajectoires différentes de traitements et sont classifiés à partir de données administratives. Les autres données intégrées dans le modèle incluent la rémission naturelle, la rechute après le traitement, les coûts de services et les taux de mortalité. Les analyses de différents scénarios permettent d’estimer les changements dans les taux d’accès aux traitements spécialisés et aux trajectoires de traitement. Les résultats du modèle incluent : le nombre et la prévalence des personnes ayant un PBÉTSSDA, le nombre de personnes traitées dans les 22 trajectoires, les résultats (avoir complété avec succès et abstinence, la réussite avec atteinte d’une consommation modérée non problématique, retourner en traitement dans les 6 mois, abandon, transfert, détention), les taux de mortalité, besoins en termes de capacité (nombre de personnes en contact avec les services dans la communauté, ou qui bénéficient des services dans un centre résidentiel ou en centre hospitalier), le coût total des traitements et les économies générales en soins de santé. Cinq scénarios sont illustrés : (a) pas de changement; (b) atteinte des taux d’accès au 70e percentile à l’échelle nationale ; (c) augmenter l’accès de 25%; (d) augmenter l’accès aux taux de l’Écosse; (e) réduire l’accès de 25%. Résultats : Initialement, 14 581 personnes présentaient un PBÉTSSDA (2,43% des adultes) et le taux d’accès aux traitements spécialisés est de 10,84%. L’impact sur 5 ans des scénarios sur le nombre de personnes présentant un PBÉTSSDA (par rapport à aucun changement) est : B) réduit de 191 (-1,3%); C) réduit de 477 (-3,3%); D) réduit de presque 2800 (-19,2%); E) augmenté de 533 (+3,6%). Un impact similaire est observé sur les autres résultats. Conclusion : Les décideurs peuvent estimer l’incidence potentielle de la variation des taux d’accès aux traitements spécialisés pour la dépendance à l’alcool. Objetivo: Modelado impacto del cambio de velocidades de acceso de tratamiento especializado para diferentes vías de tratamiento sobre la prevalencia del futuro de la dependencia del alcohol, los resultados del tratamiento, capacidad de servicio, costos, y la mortalidad. Métodos: Los números de la Autoridad Local y la prevalencia de personas que potencialmente necesitan evaluación y tratamiento en servicios especializados para la dependencia del alcohol (PINASTFAD) se estiman según las necesidades leves, moderadas, graves y complejas. La tasa de acceso de tratamiento especializado por persona PINASTFAD se estima y de 22 vías de tratamiento diferentes se clasifican a partir de datos administrativos. Otras variables del modelo incluyen la remisión natural, la recaída después del tratamiento, los costos de servicio y las tasas de mortalidad. Los análisis “¿y si?” Evalúan los cambios en las tasas de acceso al tratamiento y las vías de tratamiento del especialista. Los resultados del modelo incluyen: números y prevalencia de personas que son PINASTFAD, números tratados por 22 vías, resultados (finalización exitosa con abstinencia, consumo no problemático moderado con éxito, nuevo tratamiento dentro de los 6 meses, abandono, transferencia, custodia), tasas de mortalidad, requisitos de capacidad (números en contacto con los servicios de la comunidad, o estancias en lugares residenciales o de hospitalización), los costos totales de tratamiento y ahorros de salud generales.Cinco escenarios ilustran la funcionalidad: (a) sin cambios; (b) lograr tasas de acceso en el percentil 70 a nivel nacional; (c) aumentar el acceso por + 25%; (d) aumentar el acceso a la tasa de Escocia; (e) reducir el acceso por -25%. Resultados: Al inicio del estudio, 14,581 personas son PINASTFAD (2,43% de los adultos) y la tasa de acceso al tratamiento especialista es del 10,84%. El impacto de 5 años de los escenarios en los números de PINASTFAD (versus ningún cambio) es: B) reducir por 191 (-1.3%); C) reducir por 477 (-3.3%); D) reducir por casi 2800 (-19.2%); y E) aumento por 533 (+ 3.6%). El impacto relativo es similar para otros productos. Conclusión: Los responsables de la toma de decisiones pueden estimar el impacto potencial de cambiar las tasas de acceso de tratamiento especializado para la dependencia del alcohol.

Linked pharmacometric and pharmacoeconomic models provide a structured approach for assessing the value of candidate drugs in development. The aim of this study was to assess the utility of such an approach for identifying the properties of xanthine oxidase inhibitors (XOi) providing improved forgiveness to nonadherence and estimate the maximum reimbursement price. The pharmacometric and pharmacoeconomic models were used to simulate the time course of serum uric acid concentrations and estimate quality‐adjusted life years and costs for the XOi febuxostat and a range of hypothetical analogues. Compounds with reduced clearance or increased potency were more forgiving to missed doses, however, even following relatively large changes in these properties the predicted maximum reimbursement prices represented an increase of only 19% above febuxostat 80 mg. Linked pharmacometric and pharmacoeconomic modeling methods have the potential to inform early drug development by providing an indication of pricing options that may permit reimbursement.

Background: Dual urate-lowering therapy (ULT) with lesinurad in combination with either allopurinol or febuxostat is an option for patients with gout unsuccessfully treated on either monotherapy. Treatment failure is often a result of poor medication adherence. Imperfect adherence in clinical trials may lead to biased estimates of treatment effect and confound the results of cost-effectiveness analyses. Objectives: To estimate the impact of varying medication adherence on the cost effectiveness of lesinurad dual therapy and estimate the value-based price of lesinurad at which the incremental cost-effectiveness ratio is equal to 20,000 pound per quality-adjusted life-year (QALY). Methods: Treatment effect was simulated using published pharmacokinetic-pharmacodynamic models and scenarios representing adherence in clinical trials, routine practice, and perfect use. The subsequent cost and health impacts, over the lifetime of a patient cohort, were estimated using a bespoke pharmacoeconomic model. Results: The base-case incremental cost-effectiveness ratios comparing lesinurad dual ULT with monotherapy ranged from 39,184 pound to 78,350 pound/QALY gained using allopurinol and 31,901 pound to 124,212 pound/QALY gained using febuxostat, depending on the assumed medication adherence. Results assuming perfect medication adherence imply a per-quarter value-based price of lesinurad of 45.14 pound when used in dual ULT compared with allopurinol alone and 57.75 pound compared with febuxostat alone, falling to 25.41 pound and 3.49 pound, respectively, in simulations of worsening medication adherence. Conclusions: The estimated value-based prices of lesinurad only exceeded that which has been proposed in the United Kingdom when assuming both perfect drug adherence and the eradication of gout flares in sustained treatment responders.

BackgroundAs part of the UK National Institute for Health and Care Excellence (NICE) single technology appraisal process, independent evidence review groups (ERGs) critically appraise a company’s submission relating to a specific technology and indication.ObjectivesTo explore the type of additional exploratory analyses conducted by ERGs and their impact on the recommendations made by NICE.MethodsThe 100 most recently completed single technology appraisals with published guidance were selected for inclusion. A content analysis of relevant documents was undertaken to identify and extract relevant data, and narrative synthesis was used to rationalize and present these data.ResultsThe types of exploratory analysis conducted in relation to companies’ models were fixing errors, addressing violations, addressing matters of judgment, and the provision of a new, ERG-preferred base case. Ninety-three of the 100 ERG reports contained at least one of these analyses. The most frequently reported type of analysis in these 93 ERG reports related to the category “Matters of judgment,” which was reported in 83 reports (89%). At least one of the exploratory analyses conducted and reported by an ERG is mentioned in 97% of NICE appraisal consultation documents and 94% of NICE final appraisal determinations, and had a clear influence on recommendations in 72% of appraisal consultation documents and 47% of final appraisal determinations.ConclusionsThese results suggest that the additional analyses undertaken by ERGs in the appraisal of company submissions are highly influential in the policy-making and decision-making process.

AIMS Dual-urate-lowering therapy (ULT) with xanthine oxidase inhibitor and uricosuric medications is a treatment option for severe gout. Uricosuric agents can cause hyperuricosuria, a risk factor for nephrolithiasis and acute uric acid nephropathy. The aims of the present study were to simulate the relationship between suboptimal drug adherence and efficacy, and to quantify the risk of hyperuricosuria in gout patients receiving mono- and dual-ULTs. METHODS The impact of poor medication adherence was studied using two-compartment pharmacokinetic (PK) models based on published evidence, and a semi-mechanistic four-compartment pharmacodynamic (PD) model. The PKPD model was used to simulate mono and dual-ULT in gout patients with either under-excretion (lowered clearance) or overproduction of uric acid, with suboptimal adherence modelled as either a single drug holiday of increasing duration or doses taken at random. RESULTS Simulation results showed a surge in urinary uric acid occurring when dosing is restarted following missed doses. For under-excreters taking a 20-day drug holiday, the addition of 200 mg (or 400 mg) lesinurad to 80 mg febuxostat increased the percentage of patients experiencing hyperuricosuria from 0% to 1.4% (or 3.1%). In overproducers, restarting ULTs following drug holidays of more than 5 days leads to over 60% of patients experiencing hyperuricosuria. CONCLUSIONS Suboptimal medication adherence may compromise the safety and efficacy of mono- and dual-ULTs, especially in patients with gout resulting from an overproduction of uric acid. Clinicians and pharmacists should consider counselling patients with respect to the risks associated with partial adherence, and offer interventions to improve adherence or tailor treatments, where appropriate.

Evidence Review Groups (ERGs) critically appraise company submissions as part of the National Institute for Health and Care Excellence (NICE) Single Technology Appraisal (STA) process. As part of their critique of the evidence submitted by companies, the ERGs undertake exploratory analyses to explore uncertainties in the company’s model. The aim of this study was to explore pre-defined factors that might influence or predict the extent of ERG exploratory analyses.

This paper sets out the development of a methodological framework for detailed evaluation of public health strategies for alcohol harm reduction to meet UK policy-makers needs. Alcohol is known to cause substantial harms, and controlling its affordability and availability are effective policy options. Analysis and synthesis of a variety of public and commercial data sources is needed to evaluate impact on consumers, health services, crime, employers and industry, so a sound evaluation of impact is important. We discuss the iterative process to engage with stakeholders, identify evidence/data and develop analytic approaches and produce a final model structure. We set out a series of steps in modelling impact including: classification and definition of population subgroups of interest, identification and definition of harms and outcomes for inclusion, classification of modifiable components of risk and their baseline values, specification of the baseline position on policy variables especially prices, estimating effects of changing policy variables on risk factors including price elasticities, quantifying risk functions relating risk factors to harms including 47 health conditions, crimes, absenteeism and unemployment, and monetary valuation. The most difficult model structuring decisions are described, as well as the final results framework used to provide decision support to national level policymakers in the UK. In the discussion we explore issues around the relationship between modelling and policy debates, valuation and scope, limitations of evidence/data, how the framework can be adapted to other countries and decisions. We reflect on the approach taken and outline ongoing plans for further development.

BACKGROUND As part of the National Institute for Health and Care Excellence (NICE) single technology appraisal (STA) process, independent Evidence Review Groups (ERGs) critically appraise the company submission. During the critical appraisal process the ERG may undertake analyses to explore uncertainties around the company's model and their implications for decision-making. The ERG reports are a central component of the evidence considered by the NICE Technology Appraisal Committees (ACs) in their deliberations. OBJECTIVE The aim of this research was to develop an understanding of the number and type of exploratory analyses undertaken by the ERGs within the STA process and to understand how these analyses are used by the NICE ACs in their decision-making. METHODS The 100 most recently completed STAs with published guidance were selected for inclusion in the analysis. The documents considered were ERG reports, clarification letters, the first appraisal consultation document and the final appraisal determination. Over 400 documents were assessed in this study. The categories of types of exploratory analyses included fixing errors, fixing violations, addressing matters of judgement and the ERG-preferred base case. A content analysis of documents (documentary analysis) was undertaken to identify and extract relevant data, and narrative synthesis was then used to rationalise and present these data. RESULTS The level and type of detail in ERG reports and clarification letters varied considerably. The vast majority (93%) of ERG reports reported one or more exploratory analyses. The most frequently reported type of analysis in these 93 ERG reports related to the category 'matters of judgement', which was reported in 83 (89%) reports. The category 'ERG base-case/preferred analysis' was reported in 45 (48%) reports, the category 'fixing errors' was reported in 33 (35%) reports and the category 'fixing violations' was reported in 17 (18%) reports. The exploratory analyses performed were the result of issues raised by an ERG in its critique of the submitted economic evidence. These analyses had more influence on recommendations earlier in the STA process than later on in the process. LIMITATIONS The descriptions of analyses undertaken were often highly specific to a particular STA and could be inconsistent across ERG reports and thus difficult to interpret. CONCLUSIONS Evidence Review Groups frequently conduct exploratory analyses to test or improve the economic evaluations submitted by companies as part of the STA process. ERG exploratory analyses often have an influence on the recommendations produced by the ACs. FUTURE WORK More in-depth analysis is needed to understand how ERGs make decisions regarding which exploratory analyses should be undertaken. More research is also needed to fully understand which types of exploratory analyses are most useful to ACs in their decision-making. FUNDING The National Institute for Health Research Health Technology Assessment programme.