Cost-Effectiveness of Routine Monitoringof Long-Term Conditions in Primary Care: Informing Decision Modelling with a Systematic Review in Hypertension, Type 2 Diabetes and Chronic Kidney Disease.

Abstract

Long-term conditions (LTCs) are major public health problems with a considerable health-related and economic burden. Modelling is key in assessing costs and benefits of different disease management strategies, including routine monitoring, in the conditions of hypertension, type 2 diabetes mellitus (T2DM) and chronic kidney disease (CKD) in primary care. This review aimed to identify published model-based cost-effectiveness studies of routine laboratory testing strategies in these LTCs to inform a model evaluating the cost effectiveness of testing strategies in the UK. We searched the Medline and Embase databases from inception to July 2023; the National Institute for Health and Care Institute (NICE) website was also searched. Studies were included if they were model-based economic evaluations, evaluated testing strategies, assessed regular testing, and considered adults aged >16years.Studies identified were summarised by testing strategies, model type, structure, inputs, assessment of uncertainty, and conclusions drawn. Five studies were included in the review, i.e. Markov (n=3) and microsimulation (n=2) models. Models were applied within T2DM (n=2), hypertension (n=1), T2DM/hypertension (n=1) and CKD (n=1). Comorbidity between all three LTCs was modelled to varying extents. All studies used a lifetime horizon, except for a 10-year horizon T2DM model, and all used quality-adjusted life-years as the effectiveness outcome, except a TD2M model that used glycaemic control. No studies explicitly provided a rationale for their selected modelling approach. UK models were available for diabetes and CKD, but these compared only a limited set of routine monitoring tests and frequencies. There were few studies comparing routine testing strategies in the UK, indicating a need to develop a novel model in all three LTCs. Justification for the modelling technique of the identified studies was lacking. Markov and microsimulation models, with and without comorbidities, were used; however, the findings of this review can provide data sources and inform modelling approaches for evaluating the cost effectiveness of testing strategies in all three LTCs.