Supplementary written evidence submitted by Dr Timothy Ferris, National Director of Transformation, NHS England and NHS Improvement

(DDA0067)

 

Thank you for giving me the opportunity to present evidence to your committee on the right to privacy and digital data as part of the oral evidence session on 20 April.

 

As part of my testimony, I committed to providing further information on whether analyses of the demographic distribution of National Data opt-outs have taken place.

 

I can confirm that NHS Digital monitors demographic factors and publishes National Data Opt-out Statistics. These cover the number of people who have a national data opt-out, broken down by age, gender, and a variety of geographical breakdowns.

Individuals can also choose to opt-out locally, wherein none of their data is shared outside a GP practice for purposes other than direct care. The information on national data opt-outs is published annually or when the national opt-out rate changes by more than 0.1 per cent. The latest publication was released on April 28 2022 and covers 1 May 2021 to 1 April 2022. This can be explored through the National Data Opt-Out open data dashboard.

 

The graphs and tables below show regional and age-based opt-out data and the impact on bias in health data. For example, as people in their 30s have the highest rate of opt-out this group will be most impacted age group in datasets where opt-outs are applied:


 


April 2022

 

 

Table 1: Number of National Data Opt-outs, by Age and Gender, April 2022

Age

Gender

Opt-out

List size (population)

Opt-out Rate

0 to 9

Female

53,602

3,207,742

1.67

10 to 19

Female

108,265

3,467,646

3.12

20 to 29

Female

283,689

4,034,267

7.03

30 to 39

Female

361,982

4,490,352

8.06

40 to 49

Female

291,394

3,875,507

7.52

50 to 59

Female

291,166

4,052,355

7.19

60 to 69

Female

225,445

3,247,431

6.94

70 to 79

Female

162,631

2,652,903

6.13

80 to 89

Female

60,268

1,349,524

4.47

90 +

Female

10,138

350,409

2.89

0 to 9

Male

57,031

3,370,947

1.69

10 to 19

Male

108,457

3,632,404

2.99

20 to 29

Male

190,413

4,038,989

4.71

30 to 39

Male

271,161

4,685,181

5.79

40 to 49

Male

248,067

4,177,728

5.94

50 to 59

Male

238,266

4,211,726

5.66

60 to 69

Male

179,678

3,205,100

5.61

70 to 79

Male

134,472

2,388,840

5.63

80 to 89

Male

46,489

1,014,491

4.58

90 +

Male

5,563

172,203

3.23


 

 

 

There were 3,328,566 national data opt-outs as at 1 April 2022. 5.4 per cent of the population registered with a GP practice have a national data opt-out. The National Data Opt-out Statistics are used to inform operational delivery, policy development and the management of organisational performance.

 

Further standalone analysis has been published previously, including an analysis of the effect of opt-outs on the Hospital Episode Statistics (HES) Admitted Patient Care (APC), Outpatients (OP) and Accident & Emergency (A&E). Although there was a reduction of 575,722 patients recorded attending A&E in 2017 when national data opt-outs were applied to the data (national data opt-outs taken as at November 13 2018), table 2 and chart 3 below show that the largest change in the distribution of patients attending A&E by ethnicity is less than 0.1 percentage points, indicating there was minimal change in the distribution of ethnicities with the removal of national data opt-outs.

 

Table 2: Distribution by ethnicity of patients recorded attending A&E before and after national data opt-outs were applied, 2017-18

 

 

Ethnicity

Distribution of attendances by ethnic category before opt-outs

were applied

Distribution of attendances by ethnic category after opt-outs

were applied

 

Percentage Point Difference

White

70.68%

70.68%

-0.00%

Mixed

1.78%

1.78%

-0.01%

Asian or Asian British

6.73%

6.79%

0.06%

Black or Black British

3.71%

3.66%

-0.05%

Other / Not Known / Not Stated

17.10%

17.09%

-0.01%


 

National Data Opt-outs were applied, by Ethnicity, 2017-18

 

However, your challenge has initiated further focus on understanding the impact of opt-outs and the opt-out population as compared to the national population. We are actively considering how to further explore this topic in greater depth.

 

May 2022