IAU statistical analysis protocols
A statistical analysis protocol (SAP) is a document intended to guide analytical processes. It includes the proposed evaluation design, statistical methods, and the limitations of the analysis, including how these should be considered when interpreting the results.
The Improvement Analytics Unit agrees the protocol with local teams at the evaluated site before the analysis begins, to ensure that all design and methods choices are made objectively and are not influenced by what is found in the data. In rare instances, it may be necessary to make changes to the design of the study at a later stage; when this happens, the protocol is appended accordingly.
The IAU welcomes comments and questions on these documents. Please get in contact with Arne Wolters.
This statistical analysis protocol describes a population-level evaluation of the COVID Oximetry @home programme rolled out by CCGs in England. This is a synthetic control analysis creating a counterfactual group to quantify the impact of the programme on hospital activity and mortality during the second wave of the pandemic.
This report documents the findings of the Improvement Analytics Unit’s feasibility analysis
to determine whether a regression discontinuity design would be an appropriate method for
evaluating the causal impact of the Covid Oxymetry @home programme.
This statistical analysis protocol sets out the methods used in a patient-level evaluation of the COVID Oximetry @home programme in England. This analysis takes advantage of the 65-year age criterion for eligibility to estimate the effect of the programme on hospital use and mortality using a regression discontinuity design.
This statistical analysis protocol sets out the methods used to do pooled (meta) and subgroup analyses of three multidisciplinary team (MDT) models for integrated care in the community. Using entropy balancing to create a counterfactual (comparison group), we will explore the effectiveness of community MDTs for specific patient groups, and also examine potential risk factors from household social context.
This statistical analysis protocol describes a population-level evaluation of the impact on hospital activity of two NHS New Care Model Vanguard programmes of integrated care separately implemented in three CCGs in England. The evaluation adopts a Generalised Synthetic Control approach to identify a set of control GPs as counterfactual comparator.
This statistical analysis protocol describes the evaluation of the impact of the Integrated Care Transformation Programme on secondary care activity in the Mid-Nottinghamshire Better Together vanguard region. In this analysis, we used a synthetic control method in order to identify a counterfactual group from a donor pool of comparable GP practices in other parts of England.
This statistical analysis protocol describes the evaluation of Extensive Care Service (ECS) and Enhanced Primary Care (EPC), two intervention programmes of the ‘Your Care, Our Priority’ Fylde Coast NHS Multispecialty Community Provider Vanguard New Model of Care programme. This is a matched control group study using genetic matching methods.
This statistical analysis protocol sets out the methods used in an evaluation of the enhanced health in care homes vanguard in Wakefield. We used genetic matching methods to compare the secondary outcomes of the residents in the vanguard care homes to a ‘local matched control group’ comprised of residents living in similar care homes in Wakefield.
This statistical analysis protocol sets out the methods used in an evaluation of multidisciplinary integrated care teams (ICTs) on hospital use. In this matched control study, we used genetic matching to choose a comparison group of individuals from the same area who were similar on a range of observable characteristics to patients referred to ICTS.
This statistical analysis protocol sets out the methods used in an evaluation of a set of care home initiatives on hospital use. In this matched control study, we used genetic matching methods to retrospectively select a comparison group of similar individuals living in care homes in similar areas of England.
This statistical analysis protocol sets out the methods used in an evaluation of the impact of a set of care home initiatives on their residents’ hospital use. In this matched control study, we used Genetic Matching methods to retrospectively select a comparison group of similar individuals living in care homes outside of Sutton CCG in similar areas of England.
Further reading
Work with us
We look for talented and passionate individuals as everyone at the Health Foundation has an important role to play.
View current vacanciesThe Q community
Q is an initiative connecting people with improvement expertise across the UK.
Find out more
