New chronic disease prevalence estimates: summary and comparison with previous values and QOF
2017-04-25
We recently published new estimates of disease prevalence for a range of chronic diseases as part of a wider disease and risk factor prevalence profile which draws together all the estimates from Fingertips.
The new estimates are for:
- Hypertension
- Coronary heart disease
- Stroke
- Heart failure
- COPD
- Depression
- Peripheral arterial disease (PAD)
These estimates were developed by Imperial College and are available for General Practices and lower tier Local Authorities.
The estimates for hypertension, COPD, stroke and coronary heart disease are ‘updates’ of existing values originally created by the Association of Public Health Observatories. The new estimates are not directly comparable with previous values for a number of reasons:
- Different data sources (see table)
- Different age bands
- More model predictors
- Improved estimation methods
| Disease | Data source (new estimate) | Data source (old estimate) |
|---|---|---|
| Hypertenion | Health Survey for England | Health Survey for England (Mindell et al. 2012) |
| CHD | Whitehall II study (Marmot and Brunner 2005) | Health Survey for England |
| Stroke | Whitehall II study | Health Survey for England |
| PAD | Whitehall II study | NA |
| COPD | CPRD1(Herrett et al. 2015) | Health Survey for England |
| Depression | Health Survey for Engand | NA |
| Heart failure | CPRD | NA |
Methods
We used the fingertipsR package to identify and download the prevalence data via the Fingertips API. The full code is available at here.
We extracted relevant profileIDs and areatype codes using the profiles and area_types functions. The profileID is 37, and the relevant domainIDs are 1938132820 and 1938133099. The area type IDs are 7 for GPs and 101 for lower tier local authorities respectively.
Using this information we can download the data from Fingertips using the fingertips_data function (the downloads may take a few minutes). The first few records are shown below.
| IndicatorID | IndicatorName | ParentCode | ParentName | AreaCode | AreaName | AreaType | Sex | Age | CategoryType | Category | Timeperiod | Value | LowerCIlimit | UpperCIlimit | Count | Denominator | Valuenote | RecentTrend | ComparedtoEnglandvalueorpercentiles | Comparedtosubnationalparentvalueorpercentiles |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 92659 | Estimated prevalence of hypertension (18+) | E92000001 | England | Country | Persons | 18+ yrs | 2015 | 20.77941 | NA | NA | NA | NA | Cannot be calculated | Not compared | Not compared | |||||
| 92659 | Estimated prevalence of hypertension (18+) | E12000001 | North East region | E06000001 | Hartlepool | District & UA | Persons | 18+ yrs | 2015 | 23.33363 | 21.44856 | 25.33096 | NA | NA | Cannot be calculated | Not compared | Not compared | |||
| 92659 | Estimated prevalence of hypertension (18+) | E12000001 | North East region | E06000002 | Middlesbrough | District & UA | Persons | 18+ yrs | 2015 | 19.93244 | 17.96331 | 22.05939 | NA | NA | Cannot be calculated | Not compared | Not compared | |||
| 92659 | Estimated prevalence of hypertension (18+) | E12000001 | North East region | E06000003 | Redcar and Cleveland | District & UA | Persons | 18+ yrs | 2015 | 24.20986 | 22.38044 | 26.13846 | NA | NA | Cannot be calculated | Not compared | Not compared | |||
| 92659 | Estimated prevalence of hypertension (18+) | E12000001 | North East region | E06000004 | Stockton-on-Tees | District & UA | Persons | 18+ yrs | 2015 | 20.09999 | 18.15705 | 22.19445 | NA | NA | Cannot be calculated | Not compared | Not compared | |||
| 92659 | Estimated prevalence of hypertension (18+) | E12000001 | North East region | E06000005 | Darlington | District & UA | Persons | 18+ yrs | 2015 | 21.18244 | 19.07864 | 23.45100 | NA | NA | Cannot be calculated | Not compared | Not compared |
The downloaded dataset needs further preparation:
- selecting relevant indicators
- recoding data and extracting disease categories
- categorising estimate types into “New”, “QOF” and “Retired”
Analysis
Overall summary
Summary statistics (note there are no local authority values for Isles of Scilly or City of London, due the small numbers and instablity of modelled estimates) are show in the table.
| AreaType | Disease | Estimate type | Count | Min | 10th centile | 25th centile | Median | Mean | 75th centile | 90th cemtile | Max | SD | Missing (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| District & UA | Chd | New | 324 | 6.66 | 7.15 | 7.43 | 7.88 | 7.92 | 8.26 | 8.7 | 10.5 | 0.65 | 0.00 |
| District & UA | COPD | New | 324 | 1.49 | 2.18 | 2.53 | 2.99 | 3.01 | 3.47 | 3.9 | 4.9 | 0.66 | 0.00 |
| District & UA | Depresssion | New | 324 | 8.54 | 12.96 | 14.07 | 15.05 | 15.03 | 16.21 | 17.2 | 19.2 | 1.72 | 0.00 |
| District & UA | Heart failure | New | 324 | 0.58 | 1.11 | 1.37 | 1.52 | 1.53 | 1.71 | 1.9 | 4.6 | 0.36 | 0.00 |
| District & UA | Hypertension | New | 324 | 12.79 | 17.11 | 19.12 | 21.07 | 20.79 | 22.80 | 23.8 | 28.2 | 2.72 | 0.00 |
| District & UA | PAD | New | 324 | 0.95 | 1.03 | 1.08 | 1.15 | 1.15 | 1.21 | 1.3 | 1.7 | 0.10 | 0.00 |
| District & UA | Stroke | New | 324 | 3.04 | 3.43 | 3.56 | 3.71 | 3.72 | 3.87 | 4.0 | 4.3 | 0.23 | 0.00 |
| District & UA | Undiagnosed hypertension | New | 324 | 9.16 | 10.89 | 11.61 | 12.22 | 12.15 | 12.75 | 13.3 | 14.4 | 0.93 | 0.00 |
| GP | Chd | New | 7528 | 3.81 | 6.61 | 6.94 | 7.29 | 7.37 | 7.71 | 8.3 | 12.2 | 0.70 | 0.73 |
| GP | Chd | QOF | 7586 | 0.00 | 1.72 | 2.47 | 3.23 | 3.22 | 3.98 | 4.6 | 20.6 | 1.16 | 0.00 |
| GP | Chd | Retired | 7499 | 0.03 | 2.99 | 3.78 | 4.65 | 4.65 | 5.49 | 6.3 | 12.9 | 1.35 | 0.00 |
| GP | COPD | New | 7528 | 0.13 | 1.43 | 1.88 | 2.41 | 2.42 | 2.94 | 3.4 | 13.6 | 0.78 | 0.50 |
| GP | COPD | QOF | 7586 | 0.00 | 0.82 | 1.25 | 1.80 | 1.91 | 2.45 | 3.1 | 10.8 | 0.93 | 0.00 |
| GP | COPD | Retired | 7499 | 0.62 | 2.02 | 2.39 | 2.91 | 3.00 | 3.56 | 4.1 | 6.5 | 0.81 | 0.00 |
| GP | Depresssion | New | 7528 | 5.16 | 10.86 | 13.70 | 15.78 | 15.34 | 17.35 | 18.8 | 23.5 | 2.95 | 0.00 |
| GP | Depresssion | QOF | 7586 | 0.00 | 3.98 | 5.72 | 7.79 | 8.16 | 10.12 | 12.7 | 34.1 | 3.55 | 0.00 |
| GP | Heart failure | New | 7520 | 0.09 | 0.77 | 1.07 | 1.38 | 1.36 | 1.64 | 1.9 | 11.9 | 0.46 | 0.00 |
| GP | Heart failure | QOF | 7586 | 0.00 | 0.35 | 0.51 | 0.71 | 0.76 | 0.94 | 1.2 | 7.6 | 0.37 | 0.00 |
| GP | Hypertension | New | 7528 | 2.00 | 16.47 | 19.00 | 21.24 | 20.86 | 23.11 | 24.7 | 49.3 | 3.55 | 0.00 |
| GP | Hypertension | QOF | 7586 | 0.00 | 9.27 | 11.88 | 14.22 | 14.01 | 16.33 | 18.3 | 54.8 | 3.74 | 0.00 |
| GP | Hypertension | Retired | 7502 | 4.22 | 18.91 | 22.07 | 24.95 | 24.60 | 27.42 | 29.6 | 44.7 | 4.42 | 0.00 |
| GP | PAD | New | 7528 | 0.58 | 0.87 | 0.94 | 1.00 | 1.00 | 1.06 | 1.1 | 1.8 | 0.09 | 0.64 |
| GP | Stroke | New | 7528 | 2.10 | 3.27 | 3.51 | 3.72 | 3.70 | 3.91 | 4.1 | 6.2 | 0.32 | 0.73 |
| GP | Stroke | QOF | 7586 | 0.00 | 0.83 | 1.23 | 1.71 | 1.71 | 2.14 | 2.5 | 22.6 | 0.73 | 0.00 |
| GP | Stroke | Retired | 7503 | 0.16 | 1.31 | 1.67 | 2.04 | 2.03 | 2.39 | 2.7 | 5.4 | 0.57 | 0.00 |
| GP | Undiagnosed hypertension | New | 7528 | 3.76 | 10.46 | 11.33 | 12.12 | 12.00 | 12.77 | 13.4 | 20.1 | 1.24 | 0.00 |
Regional summaries
Summary variation plot
Prevalence distribution
Correlations
Correlation matrix of prevalence estimates
Highest correlations
| row | column | cor | p |
|---|---|---|---|
| Estimated prevalence of COPD (all ages) | Estimated prevalence of Heart failure (16+) | 0.759 | 0 |
| Estimated prevalence of COPD (all ages) | Estimated prevalence of hypertension (18+) | 0.790 | 0 |
| Estimated prevalence of COPD (all ages) | Estimated prevalence of hypertension (all ages) - retired | 0.826 | 0 |
| Estimated prevalence of COPD (all ages) | Estimated prevalence of peripheral arterial disease (PAD) (55-79 yrs), GP based | 0.761 | 0 |
| Estimated prevalence of COPD (all ages) | Estimated prevalence of stroke (55-79 yrs) | 0.712 | 0 |
| Estimated prevalence of COPD (all ages) | Estimated prevalence of stroke (all ages) - retired | 0.839 | 0 |
| Estimated prevalence of COPD (all ages) | Estimated prevalence of undiagnosed hypertension (18+) | 0.816 | 0 |
| Estimated prevalence of COPD (all ages) | Hypertension: QOF prevalence (all ages) | 0.730 | 0 |
| Estimated prevalence of COPD (all ages) | Stroke: QOF prevalence (all ages) | 0.759 | 0 |
| Estimated prevalence of Heart failure (16+) | Estimated prevalence of hypertension (18+) | 0.899 | 0 |
| Estimated prevalence of Heart failure (16+) | Estimated prevalence of hypertension (all ages) - retired | 0.930 | 0 |
| Estimated prevalence of Heart failure (16+) | Estimated prevalence of stroke (55-79 yrs) | 0.703 | 0 |
| Estimated prevalence of Heart failure (16+) | Estimated prevalence of stroke (all ages) - retired | 0.798 | 0 |
| Estimated prevalence of Heart failure (16+) | Estimated prevalence of undiagnosed hypertension (18+) | 0.893 | 0 |
| Estimated prevalence of Heart failure (16+) | Hypertension: QOF prevalence (all ages) | 0.806 | 0 |
| Estimated prevalence of Heart failure (16+) | Stroke: QOF prevalence (all ages) | 0.838 | 0 |
| Estimated prevalence of hypertension (18+) | Estimated prevalence of hypertension (all ages) - retired | 0.896 | 0 |
| Estimated prevalence of hypertension (18+) | Estimated prevalence of peripheral arterial disease (PAD) (55-79 yrs), GP based | 0.744 | 0 |
| Estimated prevalence of hypertension (18+) | Estimated prevalence of stroke (55-79 yrs) | 0.728 | 0 |
| Estimated prevalence of hypertension (18+) | Estimated prevalence of stroke (all ages) - retired | 0.790 | 0 |
| Estimated prevalence of hypertension (18+) | Estimated prevalence of undiagnosed hypertension (18+) | 0.952 | 0 |
| Estimated prevalence of hypertension (18+) | Hypertension: QOF prevalence (all ages) | 0.850 | 0 |
| Estimated prevalence of hypertension (18+) | Stroke: QOF prevalence (all ages) | 0.780 | 0 |
| Estimated prevalence of peripheral arterial disease (PAD) (55-79 yrs), GP based | Estimated prevalence of stroke (55-79 yrs) | 0.864 | 0 |
| Estimated prevalence of peripheral arterial disease (PAD) (55-79 yrs), GP based | Hypertension: QOF prevalence (all ages) | 0.703 | 0 |
| Estimated prevalence of stroke (55-79 yrs) | Estimated prevalence of stroke (all ages) - retired | 0.745 | 0 |
| Estimated prevalence of stroke (55-79 yrs) | Estimated prevalence of undiagnosed hypertension (18+) | 0.718 | 0 |
| Estimated prevalence of undiagnosed hypertension (18+) | Hypertension: QOF prevalence (all ages) | 0.820 | 0 |
| Estimated prevalence of undiagnosed hypertension (18+) | Stroke: QOF prevalence (all ages) | 0.783 | 0 |
Mapping local authority estimates
Using the ggmap package. (Kahle and Wickham 2013)
Highest and lowest prevalences
Cluster analysis
Group together local authorities and general practices on the basis similarity across prevalence estimates. Using k-means analysis.
Dimensions
References
Herrett, Emily, Arlene M. Gallagher, Krishnan Bhaskaran, Harriet Forbes, Rohini Mathur, Tjeerd van Staa, and Liam Smeeth. 2015. “Data Resource Profile: Clinical Practice Research Datalink (CPRD).” International Journal of Epidemiology 44 (3): 827–36. doi:10.1093/ije/dyv098.
Kahle, David, and Hadley Wickham. 2013. “ggmap: Spatial Visualization with ggplot2.” The R Journal 5 (1): 144–61. doi:10.1023/A:1009843930701.
Marmot, Michael, and Eric Brunner. 2005. “Cohort profile: The Whitehall II study.” International Journal of Epidemiology 34 (2): 251–56. doi:10.1093/ije/dyh372.
Mindell, Jennifer, Jane P. Biddulph, Vasant Hirani, Emanuel Stamatakis, Rachel Craig, Susan Nunn, and Nicola Shelton. 2012. “Cohort profile: The health survey for england.” International Journal of Epidemiology 41 (6): 1585–93. doi:10.1093/ije/dyr199.
Clinical Practice Research Datalink - a GP dataset↩