NICE health technology evaluations: the manual

Evidence on health inequalities

Health inequalities can occur throughout the course of disease and treatment and can be measured in many ways. In economic evaluations, health inequalities can be defined by the impact of health technologies on the quality-adjusted life years (QALYs) experienced by groups in the general population with different social characteristics. Social characteristics related to health inequalities include socioeconomic status, level of deprivation and ethnicity. Distributional cost-effectiveness analysis (DCEA) is a method for estimating this effect.

The overall impact of a health technology on health inequalities in QALYs is determined by a range of factors. These include inequalities in disease prevalence, treatment uptake and treatment success. These factors can vary in importance depending on the disease area and the relevant social characteristics. It is important that DCEAs reflect the sources of variation that have the biggest impact overall on health inequality. Not doing so could result in bias. If variation in potentially influential parameters has not been accounted for, this should be justified.

The preferred source of robust health inequalities data will depend on the condition and whether it is mainly treated in primary or secondary care. Datasets such as the Clinical Practice Research Datalink and Hospital Episode Statistics can give an estimate of diagnosed prevalence by social characteristics in primary and secondary care, respectively. Other data sources may be appropriate for specific disease areas. For guidance on using datasets and registries see NICE's real-world evidence framework (PDF).

Evidence on health inequalities should stratify populations using the appropriate social characteristics, supported by a strong rationale (see the section on stratification of social groups). Using an inappropriate stratification technique to analyse health inequalities can lower the validity and relevance of a DCEA.

NHS England's CORE20PLUS5 approach, which defines a target population and identifies 5 clinical areas of national priority, could also be used as a guide to identify relevant health inequalities.

High-quality evidence on health inequalities should be gathered for the eligible population in the scope. When data is taken from a different population, for example a more broadly defined disease group, it should be accompanied by a description of the differences. Justification of its use, supported by evidence and expert opinion, should also be provided.  

Differences in treatment benefit across social groups should be shown with the best available evidence. This could include data from trials, real-world evidence or simulation modelling, depending on the population and technology being considered.

There may be social and structural barriers, as well as exclusion criteria, that prevent people from engaging in research. This can potentially lead to bias in evidence on health inequalities and should be documented so the committee can take it into account in its deliberations.