Search

Objectives: Public health decision making is a complex process based on thorough and comprehensive health technology assessments involving the comparison of different strategies, values and tradeoffs under uncertainty. This process must be based on best available evidence and plausible assumptions. Causal inference and health decision science are two methodological approaches providing information to help guide decision making in health care. Both approaches are quantitative methods that use statistical and modeling techniques and simplifying assumptions to mimic the complexity of the real world. We intend to review and lay out both disciplines with their aims, strengths and limitations based on a combination of textbook knowledge and expert experience.Methods: To help understanding and differentiating the methodological approaches of causal inference and health decision science, we reviewed both methods with the focus on aims, research questions, methods, assumptions, limitations and challenges, and software. For each methodological approach, we established a group of four experts from our own working group to carefully review and summarize each method, followed by structured discussion rounds and written reviews, in which the experts from all disciplines including HTA and medicine were involved. The entire expert group discussed objectives, strengths and limitations of both methodological areas, and potential synergies. Finally, we derived recommendations for further research and provide a brief outlook on future trends.Results: Causal inference methods aim for drawing causal conclusions from empirical data on the relationship of pre-specified interventions on a specific target outcome and apply a counterfactual framework and statistical techniques to derive causal effects of exposures or interventions from these data. Causal inference is based on a causal diagram, more specifically, a directed acyclic graph (DAG), which encodes the assumptions regarding the causal relations between variables. Depending on the type of confounding and selection bias, traditional statistical methods or more complex g-methods are needed to derive valid causal effects. Besides the correct specification of the DAG and the statistical model, assumptions such as consistency, positivity, and exchangeability must be checked when aiming at causal inference.Health decision science aims for guiding policy decision making regarding health interventions considering and balancing multiple competing objectives of a decision based on data from multiple sources and studies, for example prevalence studies, clinical trials and long-term observational routine effectiveness studies, and studies on preferences and costs. It involves decision analysis, a systematic, explicit and quantitative framework to guide decisions under uncertainty. Decision analyses are based on decision-analytic models to mimic the course of disease as well as aspects and consequences of the intervention in order to quantitatively optimize the decision. Depending on the type of decision problem, decision trees, state-transition models, discrete event simulation models, dynamic transmission models, or other model types are applied. Models must be validated against observed data, and comprehensive sensitivity analyses must be performed to assess uncertainty. Besides the appropriate choice of the model type and the valid specification of the model structure, it must be checked if input parameters of effects can be interpreted as causal parameters in the model. Otherwise results will be biased.Conclusions: Both causal inference and health decision science aim for providing best causal evidence for informed health decision making. The strengths and limitations of both methods differ and a good understanding of both methods is essential for correct application but also for correct interpretation of findings from the described methods. Importantly, decision-analytic modeling should be combined with causal inference when developing guidance and recommendations regarding decisions on health care interventions. Zielsetzungen: Die Entscheidungsfindung im Gesundheitswesen ist ein komplexer Prozess, der auf gründlichen und umfassenden Bewertungen von Gesundheitstechnologien beruht und den Vergleich verschiedener Strategien, Werte und Kompromisse unter Unsicherheit beinhaltet. Dieser Prozess muss auf den besten verfügbaren Erkenntnissen und plausiblen Annahmen beruhen. Kausalinferenz und Entscheidungsanalyse sind zwei methodische evidenzbasierte Ansätze, die Informationen liefern, um die Entscheidungsfindung im Gesundheitswesen zu unterstützen. Bei beiden Ansätzen handelt es sich um quantitative Methoden, die statistische Techniken und Modellierungstechniken sowie vereinfachende Annahmen verwenden, um die Komplexität der realen Welt zu imitieren. Wir beabsichtigen in dieser Publikation, auf der Grundlage von Lehrbuchwissen und Expertenerfahrung beide Disziplinen mit ihren Zielen, Stärken und Grenzen darzustellen.Methoden: Um das Verständnis und die Unterscheidung der methodischen Ansätze der kausalen Inferenz und der gesundheitswissenschaftlichen Entscheidungsanalyse zu erleichtern, haben wir beide Methoden mit dem Schwerpunkt auf Ziele, Forschungsfragen, Methoden, Annahmen, Grenzen und Herausforderungen sowie verfügbare Software untersucht. Für jeden methodischen Ansatz setzten wir eine Gruppe von vier Experten aus unserer eigenen Arbeitsgruppe ein, um jede Methode sorgfältig zu begutachten und ihre Charakteristika zusammenzufassen. In den darauffolgenden strukturierten Diskussionsrunden und schriftlichen Überprüfungen waren Experten aus allen Disziplinen einschließlich Health Technology Assessment (HTA) und Medizin beteiligt. Die gesamte Expertengruppe diskutierte Ziele, Stärken und Grenzen der beiden methodischen Bereiche und mögliche Synergien. Abschließend wurden Empfehlungen für die weitere Forschung abgeleitet und ein kurzer Ausblick auf zukünftige Trends gegeben.Ergebnisse: Methoden der kausalen Inferenz zielen darauf ab, kausale Schlussfolgerungen aus empirischen Daten über die Beziehung zwischen vorher festgelegten Interventionen und einem bestimmten Zielendpunkt zu ziehen. Es werden ein kontrafaktischer Rahmenansatz und statistische Techniken angewandt, um kausale Auswirkungen von Expositionen oder Interventionen aus diesen Daten abzuleiten. Die kausale Inferenz basiert auf einem Kausaldiagramm, genauer gesagt auf einem directed acyclic graph (DAG), der die Annahmen bezüglich der kausalen Beziehungen zwischen Variablen darstellt. Je nach Art des Confounding und des Selektionsbias sind traditionelle statistische Methoden oder komplexere g-Methoden erforderlich, um gültige kausale Effekte abzuleiten. Neben der korrekten Spezifikation des DAG und des statistischen Modells müssen Annahmen wie Consistency , Positivity und Exchangeability überprüft werden, wenn man kausale Schlüsse ziehen möchte.Die gesundheitswissenschaftliche Entscheidungsanalyse zielt darauf ab, die politische Entscheidungsfindung in Bezug auf Gesundheitsmaßnahmen zu unterstützen, wobei mehrere konkurrierende Ziele einer Entscheidung auf der Grundlage von Daten aus verschiedenen Quellen und Studien berücksichtigt und abgewogen werden. Diese Studien umfassen z. B. Prävalenzstudien, klinische Studien und langfristige Routinebeobachtungsstudien zur Wirksamkeit sowie Studien zu Präferenzen und Kosten. Die Entscheidungsanalyse bietet einen systematischen, expliziten und quantitativen Rahmen, um Entscheidungen unter Unsicherheit zu strukturieren. Entscheidungsanalysen basieren auf entscheidungsanalytischen Modellen, die den Krankheitsverlauf sowie die Aspekte und Folgen der Intervention nachbilden. Je nach Art des Entscheidungsproblems werden Entscheidungsbäume, Zustands-Übergangs-Modelle (z. B. Markov-Modelle), diskrete Ereignissimulationsmodelle, dynamische Übertragungsmodelle oder andere Modelltypen verwendet. Die Modelle müssen anhand von Beobachtungsdaten validiert werden, und es sind umfassende Sensitivitätsanalysen durchzuführen, um die Unsicherheit zu bewerten. Neben der angemessenen Wahl des Modelltyps und der validen Spezifikation der Modellstruktur muss geprüft werden, ob die Einflussparameter für die daraus resultierenden Auswirkungen als kausale Parameter im Modell interpretiert werden können. Andernfalls werden die Ergebnisse verzerrt.Schlussfolgerungen: Sowohl die kausale Inferenz als auch die gesundheitswissenschaftliche Entscheidungsanalyse zielen darauf ab, die beste kausale Evidenz für eine informierte Entscheidungsfindung zu liefern. Die Stärken und Grenzen beider Methoden sind unterschiedlich, und ein gutes Verständnis beider Methoden ist für die korrekte Anwendung, aber auch für die korrekte Interpretation der Ergebnisse der beschriebenen Methoden unerlässlich. Wichtig ist, dass entscheidungsanalytische Modellierungen mit kausalen Inferenzmethoden kombiniert werden, wenn es um die Entwicklung von Leitlinien und Empfehlungen für Entscheidungen über Interventionen im Gesundheitswesen geht.

Objective: The goal of this review was to identify decision-analytic modeling studies in early health technology assessments (HTA) of high-risk medical devices, published over the last three years, and to provide a systematic overview of model purposes and characteristics. Additionally, the aim was to describe recent developments in modeling techniques.Methods: For this scoping review, we performed a systematic literature search in PubMed and Embase including studies published in English or German. The search code consisted of terms describing early health technology assessment and terms for decision-analytic models. In abstract and full-text screening, studies were excluded that were not modeling studies for a high-risk medical device or an in-vitro diagnostic test. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram was used to report on the search and exclusion of studies. For all included studies, study purpose, framework and model characteristics were extracted and reported in systematic evidence tables and a narrative summary.Results: Out of 206 identified studies, 19 studies were included in the review. Studies were either conducted for hypothetical devices or for existing devices after they were already available on the market. No study extrapolated technical data from early development stages to estimate potential value of devices in development. All studies except one included cost as an outcome. Two studies were budget impact analyses. Most studies aimed at adoption and reimbursement decisions. The majority of studies were on in-vitro diagnostic tests for personalized and targeted medicine. A timed automata model, to our knowledge a model type new to HTA, was tested by one study. It describes the agents in a clinical pathway in separate models and, by allowing for interaction between the models, can reflect complex individual clinical pathways and dynamic system interactions. Not all sources of uncertainty for in-vitro tests were explicitly modeled. Elicitation of expert knowledge and judgement was used for substitution of missing empirical data. Analysis of uncertainty was the most valuable strength of decision-analytic models in early HTA, but no model applied sensitivity analysis to optimize the test positivity cutoff with regard to the benefit-harm balance or cost-effectiveness. Value-of-information analysis was rarely performed. No information was found on the use of causal inference methods for estimation of effect parameters from observational data.Conclusion: Our review provides an overview of the purposes and model characteristics of nineteen recent early evaluation studies on medical devices. The review shows the growing importance of personalized interventions and confirms previously published recommendations for careful modeling of uncertainties surrounding diagnostic devices and for increased use of value-of-information analysis. Timed automata may be a model type worth exploring further in HTA. In addition, we recommend to extend the application of sensitivity analysis to optimize positivity criteria for in-vitro tests with regard to benefit-harm or cost-effectiveness. We emphasize the importance of causal inference methods when estimating effect parameters from observational data. Zielsetzung: Unser Ziel war es, entscheidungsanalytische Studien zur frühen Technologiebewertung von Medizinprodukten aus den letzten drei Jahren zu identifizieren und einen systematischen Überblick über den Zweck der Modelle und die Modellcharakteristika sowie über aktuelle Entwicklungen in der Modellierungstechnik zu geben.Methoden: Für diesen Scoping Review wurde eine systematische Literatursuche in den Datenbanken PubMed und Embase nach Studien in englischer und deutscher Sprache durchgeführt. Der Suchcode setzte sich aus Begriffen für die frühe Technologiebewertung und Begriffen für ein entscheidungsanalytisches Modell zusammen. Im Screening von Zusammenfassung und Volltext wurden Studien ausgeschlossen, bei denen es sich nicht um ein Modell für Hochrisiko-Medizinprodukte oder diagnostische Tests handelte. Für alle eingeschlossenen Studien wurden Zweck und Rahmen der Studie sowie Modellcharakteristika extrahiert und zusammenfassend in sytematischen Evidenztabellen sowie in narrativer Form dargestellt.Ergebnisse: Aus 206 Studien wurden neunzehn Studien in den Review eingeschlossen. Die Studien betrafen entweder hypothetische Medizinprodukte oder existierende Produkte, nachdem sie bereits auf dem Markt erhältlich waren. Keine Studie nutzte die Extrapolation technischer Daten aus frühen Entwicklungsstufen, um den potentiellen Wert für die Gesellschaft zu evaluieren. Mit Ausnahme einer Studie schlossen alle Studien Kosten in die Evaluation ein. Zwei Studien waren Budget-Impakt-Analysen. Die meisten Studien zielten auf die Aufnahme in den Leistungskatalog und die Kostenerstattung der Medizinprodukte ab. Die Mehrheit der Studien betraf In-Vitro-Diagnostika für personalisierte und zielgerichtete Interventionen. Zeitgesteuerte Zustandsautomaten, ein Modelltyp, der bisher nicht im HTA genutzt wurde, wurden zur Beschreibung von komplexen, individuellen klinischen Pfaden und von Interaktionen dynamischer Systeme eingesetzt. Nicht alle Unsicherheitsquellen bei In-vitro-Tests wurden explizit modelliert. Das Einholen von Expertenwissen und Expertenbeurteilungen wurde als Ersatz für fehlende empirische Daten verwendet. Die Analyse der Unsicherheiten stellte den größten Vorteil der entscheidungsanalytischen Modellierung im frühen HTA dar, aber kein Modell wandte Sensitivitätsanalysen an, um den Schwellenwert für Testpositivität hinsichtlich der Nutzen-Schaden-Abwägung oder der Kosteneffektivität zu optimieren. Die Value-of-Information-Analyse wurde selten eingesetzt. Die Anwendung von Methoden der kausalen Inferenz zur Schätzung von Effektparametern aus Beobachtungsstudien wurde in keiner Studie erwähnt.Schlussfolgerung: Unser Review gibt einen Überblick über Ziele und Modelleigenschaften von neunzehn Studien zur frühen Bewertung von Medizinprodukten. Der Review zeigt die wachsende Bedeutung der personalisierten Medizin und bestätigt frühere Empfehlungen der sorgfältigen Modellierung der mit diagnostischen Tests verbundenen Unsicherheiten und eines vermehrten Einsatzes der Value-of-Information-Analyse. Zeitgesteuerte Zustandsautomaten könnten eine lohnenswerte Erweiterung der Modelltypen im Rahmen der Technologiebewertung sein. Zusätzlich empfehlen wir die Anwendung von Sensitivitätsanalysen zur Optimierung des Schwellenwerts für Testpositivität hinsichtlich der Nutzen-Schaden-Relation oder der Kosteneffektivität. Wir betonen die Wichtigkeit der Anwendung von Methoden der kausalen Inferenz für die Schätzung von Effektparametern aus Beobachtungsdaten.

Objectives Psoriatic arthritis (PsA) has a strong genetic component, and the identification of genetic risk factors could help identify the ~30% of psoriasis patients at high risk of developing PsA. Our objectives were to identify genetic risk factors and pathways that differentiate PsA from cutaneous-only psoriasis (PsC) and to evaluate the performance of PsA risk prediction models. Methods Genome-wide meta-analyses were conducted separately for 5,065 patients with PsA and 21,286 healthy controls and separately for 4,340 patients with PsA and 6,431 patients with PsC. The heritability of PsA was calculated as a single-nucleotide polymorphism (SNP)–based heritability estimate (h2SNP) and biologic pathways that differentiate PsA from PsC were identified using Priority Index software. The generalizability of previously published PsA risk prediction pipelines was explored, and a risk prediction model was developed with external validation. Results We identified a novel genome-wide significant susceptibility locus for the development of PsA on chromosome 22q11 (rs5754467; P = 1.61 × 10−9), and key pathways that differentiate PsA from PsC, including NF-κB signaling (adjusted P = 1.4 × 10−45) and Wnt signaling (adjusted P = 9.5 × 10−58). The heritability of PsA in this cohort was found to be moderate (h2SNP = 0.63), which was similar to the heritability of PsC (h2SNP = 0.61). We observed modest performance of published classification pipelines (maximum area under the curve 0.61), with similar performance of a risk model derived using the current data. Conclusion Key biologic pathways associated with the development of PsA were identified, but the investigation of risk classification revealed modest utility in the available data sets, possibly because many of the PsC patients included in the present study were receiving treatments that are also effective in PsA. Future predictive models of PsA should be tested in PsC patients recruited from primary care.

SARS-COV-2 vaccines have been shown to be efficacious primarily in healthy volunteer populations and population level studies. Immune responses following SARS-CoV-2 vaccination are less well characterised in potentially immune vulnerable patient groups, including those with immune-mediated inflammatory and chronic diseases (inflammatory arthritis [IA] incorporating rheumatoid arthritis [RA] and psoriatic arthritis [PsA]; ANCA-Associated Vasculitis [AAV]; inflammatory bowel disease [IBD]); hepatic disease (HepD), end stage kidney disease requiring haemodialysis (HD) without or with immunosuppression (HDIS); solid cancers (SC) and haematological malignancies (HM), and those that have undergone haemopoietic stem cell transplant (HSCT). The OCTAVE trial is a multi-centre, multi-disease, prospective cohort that will comprehensively assess SARS-CoV-2 vaccine responses within and between the abovementioned disease cohorts using common analytical platforms in patients recruited across the United Kingdom (UK). The majority of subjects received either COVID-19 mRNA Vaccine BNT162b2 (Pfizer/BioNTech) or ChAdOx1 Vaccine (AstraZeneca formerly AZD1222) as part of the UK National COVID19 vaccination programme. As of 13 th August 2021; 2,583 patients have been recruited. We report herein the humoral and T cell immune response results from the first 600 participants recruited where serology data are available at baseline, pre-second vaccine dose (boost) and/or 4 weeks post second dose. We also include in the analysis, data obtained from 231 healthy individuals from the PITCH (Protective Immunity from T cells in Healthcare workers) study. Overall, in comparison to PITCH where 100% of tested individuals (n=93) generated anti-Spike antibodies after vaccine doses, 89% of patients within OCTAVE seroconverted 4 weeks after second vaccine dose. By corollary, approximately 11% of patients across all disease cohorts fail to generate antibodies that react to SARS-CoV-2 spike 4 weeks after two vaccines. Failure to generate spike reactive antibodies was found at a higher proportion in some specific patient subgroups, particularly AAV (72.4%), HD-IS (16.7%) and HepD (16.7%). Importantly, all recruited AAV patients had received Rituximab; a targeted B cell depletion therapy. Furthermore, even in those who seroconverted, 40% of patients across disease cohorts generate lower levels of SARS-CoV-2 antibody reactivity compared to healthy subjects after two SARS-CoV-2 vaccines; the functional significance of these findings in providing protection from subsequent SARS-CoV-2 exposure is not currently known. In contrast to the observed serological response, evaluation of the Spike-specific T cell response revealed that across all patient sub-groups (including AAV) a response similar to healthy individuals was generated. Our data argue strongly for further vaccination strategies to optimise humoral immune responses against SARS-CoV-2 in patients with chronic diseases and/or patients on immune suppressive therapies. Trial Registration: The trial is registered on ISRCTN 12821688.Funding: This work was supported by the Medical Research Council COVID-19 Immunity – National Core Study (IMM-NCS) [grant number MC-PC-20031]. Staff at the Cancer Research UK Clinical Trials Unit (CRCTU) are supported by a core funding grant from Cancer Research UK (C22436/A25354). PK and EB are supported by the NIHR Birmingham Biomedical Research Centres at the University Hospitals Birmingham NHS Foundation Trust and the University of Birmingham Biomedical Research Centres. EB and PK are supported by an NIHR Senior Investigator award. PK is funded by WT109965MA. SJD is funded by an NIHR Global Research Professorship (NIHR300791). TdS is funded by a Wellcome Trust Intermediate Clinical Fellowship (110058/Z/15/Z). DS is supported by the NIHR Academic Clinical Lecturer programme in Oxford. LT is supported by the Wellcome Trust (grant number 205228/Z/16/Z), the U.S. Food and Drug Administration Medical Countermeasures Initiative contract 75F40120C00085. and the National Institute for Health Research Health Protection Research Unit (NIHR HPRU) in Emerging and Zoonotic Infections (NIHR200907) at University of Liverpool in partnership with Public Health England (PHE), in collaboration with Liverpool School of Tropical Medicine and the University of Oxford. The PITCH (Protective Immunity from T cells to Covid-19 in Health workers) Consortium, is funded by the UK Department of Health and Social Care with contributions from UKRI/NIHR through the UK Coronavirus Immunology Consortium (UKCIC), the Huo Family Foundation and The National Institute for Health Research (UKRIDHSC COVID-19 Rapid Response Rolling Call, Grant Reference Number COV19-RECPLAS).Declaration of Interest: None to declare. Ethical Approval: This study was approved by the UK Medicines and Healthcare Products Regulatory Agency on the 5th February 2021 and the London and Chelsea Research Ethics Committee (REC Ref:21/HRA/0489) on 12th February 2021, with subsequent amendments approved on 3rd March 2021, 19th April 2021 and 26th April 2021).

Objectives To evaluate 6-month effectiveness of ustekinumab versus tumour necrosis factor inhibitor (TNFi), analysing predictors of low disease activity (LDA)/remission. Methods PsABio is a prospective, observational cohort study of patients with psoriatic arthritis (PsA) at 92 sites in eight European countries, who received first-line to third-line ustekinumab or a TNFi. Comparative achievement at 6 months of clinical Disease Activity Index for Psoriatic Arthritis (cDAPSA) LDA/remission, and minimal disease activity (MDA)/very LDA using propensity score (PS)-adjusted multivariate logistic regression was assessed. Results In the final analysis set of 868 participants with 6-month follow-up data (ustekinumab, n=426; TNFi, n=442), with long-standing disease and a high mean cDAPSA score (31.0 vs 29.8, respectively), proportions of patients in ustekinumab/TNFi treatment groups achieving cDAPSA LDA at 6 months were 45.7%/50.7%. cDAPSA remission was achieved in 14.9%/19.2%, and MDA in 26.4%/30.8% of patients. PS-adjusted odds ratios (OR; 95% confidence interval (CI)) of reaching cDAPSA LDA and MDA were 0.73 (0.46 to 1.15) and 0.87 (0.61 to 1.25) with ustekinumab versus TNFi, indicating no significant difference. High baseline body mass index or high cDAPSA were associated with a lower chance (OR (95% CI)) of reaching cDAPSA LDA with TNFi (0.94 (0.89 to 0.99) and 0.64 (0.52 to 0.79), respectively). Predictive factors were similar to previously published evidence, with cDAPSA and 12-item Psoriatic Arthritis Impact of Disease scores and chronic widespread pain at baseline appearing as new risk factors for unfavourable outcome. Safety data were similar between groups. Conclusion Treatment targets were reached similarly after 6 months of treatment with ustekinumab and TNFi. © Author(s) (or their employer(s)) 2021.