Department of Health

Website of the Department of Health

Please note that this website has a UK government access keys system.

General practice survey 1998 - Health authority tables

  • Last modified date:
    9 February 2007

The sample for the survey was drawn from the Electoral Registers in September 1998. It comprised 1,000 residents of each of the 100 English Health Authorities (HAs) making a total of 100,000 persons who were sent the questionnaire. Within each HA the selection of residents was random so as to make the selected sample statistically representative.

In total 61,426 completed questionnaires were received back. In general higher levels of returns were received from the rural HAs. The greatest number received was from North & East Devon (684); the lowest from Camden & Islington (447). Only 11 HAs returned fewer than 550 questionnaires. The numbers returned by the individual HAs are shown on some tables (eg Table 27).

The Tables

There is a total of 29 tables, each containing several questions. The questions are not always in the order in which they appear on the questionnaire. Necessarily they have to be grouped according to the 'base' - ie the group or sub-group of respondents that was asked a particular question. Also appended to this note (Annex 1) is an index of the tables identifying the questions each table contains.

Each table is presented on a standard set of five separate sheets. On these sheets the 100 HAs are listed alphabetically within Health Authority Clusters. These clusters are the product of a system, designed by the Office of National Statistics, which classifies HAs into similar types using socio-demographic data from the 1991 Census. There are eleven such clusters, but they can also be organised, by grouping 10 of them into pairs, into an alternative structure of six clusters.

The identification of these clusters is set out below:

No. of HAs in Cluster

Eleven ClustersSix Clusters

4

Inner LondonInner London

12
6

Coal Fields )
Ports & Industry )		Mining & Industry

9
7

Manufacturing )
Mixed Economies )		Urban Centres

9
7

Resort & Retirement )
Services & Education )		Mature Areas

4
20

Most Prosperous)
Growth Areas)		Prosperous Areas

15
7

Mixed Urban & Rural )
Coast & Country )		Rural Areas

The cluster totals are presented on Sheet 1 of each table. The remaining sheets (2-5) comprise rows of data relating to the individual HAs.

Weighting the Data

The proportions of men and women of different age groups among those responding in each Authority were not identical. The data presented in these tables therefore have been weighted to standardise the age and sex profiles of every HA to the latest available national estimates for England (mid-1997).

For each row in each table however the actual number of respondents is shown as well as the weighted (total) base from which the percentages are derived.

The weights for the totals shown on Sheet 1 of each table have been adjusted to take account of the varying sample fractions necessarily employed in each HA to generate a selected sample of 1000 in each one. Each of 6 and each of 11 clusters has also been age and sex standardised to mid-1997 national estimates.

Confidence Intervals and Margins of Error

All survey data are estimates of the 'true' proportion in the population sampled. With random sampling methods it is possible to estimate the margins of error either side of each percentage indicating a range within which the 'true' percentage will fall.

These margins of error vary according to both the percentage estimates from the sample, and the number of people included in the sample. The table below indicates the"95% confidence intervals" that users of the tables must allow, taking both of these criteria into account. That is, the table shows the range in which we would expect the "true" percentage to fall 95 times out of 100. For example, if the estimated value is 50% and the sample size is 500, the true value is likely to be between 46% and 54%.

CONFIDENCE INTERVALS FOR HAs

Sample size10%25%50%75%90%

75

3% 17%

15% 35%

39% 61%

65% 85%

83% 97%

100

4% 16%

17% 33%

40% 60%

67% 83%

84% 96%

150

5% 15%

18% 32%

42% 58%

68% 82%

85% 95%

200

6% 14%

19% 31%

43% 57%

69% 81%

86% 94%

250

6% 14%

20% 30%

44% 56%

70% 80%

86% 94%

300

7% 13%

20% 30%

44% 56%

70% 80%

87% 93%

400

7% 13%

21% 29%

45% 55%

71% 79%

87% 93%

500

7% 13%

21% 29%

46% 54%

71% 79%

97% 93%

600

8% 12%

22% 28%

46% 54%

72% 78%

88% 92%


CONFIDENCE INTERVALS FOR HA CLUSTERS

Sample Size

10%

25%

50%

75%

90%

500

7% 13%

21% 29%

46% 54%

71% 79%

87% 93%

750

8% 12%

22% 28%

46% 54%

72% 78%

88% 92%

1000

8% 12%

22% 28%

47% 53%

72% 78%

88% 92%

2000

9% 11%

23% 27%

48% 52%

73% 77%

89% 91%

3000

9% 11%

23% 27%

48% 52%

73% 77%

89% 91%

4000

9% 11%

24% 26%

48% 52%

74% 76%

89% 91%

5000

9% 11%

24% 26%

49% 51%

74% 76%

89% 91%

10000

9% 11%

24% 26%

49% 51%

74% 76%

89% 91%

15000

10% 10%

24% 26%

49% 51%

74% 76%

90% 90%

Tables 14 and 15 present the answers to questions which applied to relatively few respondents. In these tables the bases are so small (and therefore the margins of error so wide) that the data have not been presented at the level of individual HAs. Only Sheet 1 (ie the cluster totals) of these tables are included in the set.

Notes on the Tables
* = less than 0.5%

- = NIL.

Download tables

as Excel spreadsheets:

Additional links

GP patient surveys

From 2007, a series of surveys was developed to collect information to reward GP practices according to performance on elements of access and choice.

Ordering publications

Contact details for obtaining hard-copy DH publications.

Help viewing PDFs

Help viewing PDFs

Useful tips to ensure you get the most from PDFs and the free Reader program. Topics covered include accessibility, troubleshooting and searching files.

Access keys