Space-time pattern mining for recorded depression prevalence in England from 2011 to 2022. An interactive map application

Space-Time Pattern Mining for recorded depression prevalence from 2011 to 2022, based on the Anselin Local Moran’s I algorithm.

The unit of analysis of this geospatial analysis was the Lower Super Output Area (LSOA). There are 32,844 LSOAs across England, with an average population of 1500 people (Office for National Statistics, 2021). In all analyses, we used the LSOA boundaries published by the Office for National Statistics as at March 21, 2021 (Office for National Statistics, 2021).

The diagnosed depression prevalence was derived using the data published by NHS Digital. Figures showing the recorded prevalence of depression in England by general practitioner (GP) practice are published annually in the Quality and Outcomes Framework (QOF) administrative dataset, which also reports how the QOF-recorded prevalence has changed since the previous year (NHS Digital, 2020, pp. 2019–2020). For this study, we combined all available data on depression published by NHS Digital and created time-series recorded depression for each LSOA from 2011 to 2022. The annual aggregate data on diagnoses of depression per LSOA has been calculated based on the weighted averages of the number of patients diagnosed with depression per LSOA divided by the total number of registered patients in each LSOA. In terms of coverage, the data for Quality and Outcomes Framework (QOF) have been collected annually at an aggregate level for each of the 6470 (97.5%) GP practices in England, with approximately 61 million registered patients aged 18 years and above; thus, the dataset offers nationwide insights.

This online app depicts interactively the Cluster and Outlier Analysis, using the Anselin Local Moran’s I algorithm (Anselin, 1995), to identify local indicators of spatial association (LISA) and correct for spatial dependence. The conceptualisation of spatial relationships parameter value was set as the ‘Contiguity edges corners’, the standardisation option was set as ‘Row’, and the number of permutations was set as 999. The LISA refer to statistically significant spatial clusters of small areas with high values (high/high clusters) and low values (low/low clusters) of depression, as well as high and low spatial outliers in which a high value is surrounded by low values (high/low clusters), and outliers in which a low value is surrounded by high values (low/high clusters).