Intended for healthcare professionals

General Practice

Collecting data in general practice: need for standardisation

BMJ 1996; 312 doi: (Published 06 January 1996) Cite this as: BMJ 1996;312:33
  1. D C Newrick, research associatea,
  2. J A Spencer, senior lecturera,
  3. K P Jones, senior lecturera
  1. a Department of Primary Health Care, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH
  1. Correspondence to: Mr Newrick.
  • Accepted 28 September 1995

The variety of practice annual reports that have appeared since 19701 2 demonstrates that decisions about which data are best collected to assess the quality of patient care and for forward planning are usually left to individual practices to determine.

Subjects, methods, and results

To examine existing data collection and views about feedback we sent a postal questionnaire for self completion to all singlehanded practices, all fundholding practices, and all non-fundholding practices with over five partners, plus a one in three random sample of all other non-fundholding practices in the Northern region in September 1993 (n=211). The response rate was 79%.

Respondents spent substantial amounts of time entering data into computer systems or manual records each month (see table). In computerised practices, deciding on a standardised coding system and ensuring that team members consistently entered data caused problems.

Mean, standard deviation, and 95% confidence intervals for the estimated amount of time spent in hours each month recording data by respondents, grouped by practice status and practice's views on methods of collecting/reporting data

View this table:

Levels of recording of outpatient referrals and attendances at accident and emergency departments were high though low on diagnostic details. Chiropody, physiotherapy, and alcohol-drug counselling referrals were recorded by fewer respondents, as were social service referrals. Collection of data about surgery consultations, domiciliary visits, and night visits by general practitioners were high (all over 90%), but telephone consultations were recorded at lower levels. Recording of practice nurse surgeries was high, but nurse telephone consultations and home visits were recorded at under half the rate for general practitioners. Immunisations, vaccinations, cervical cytology, births, and deaths were recorded at extremely high levels. Least recorded data were for socioeconomic details of patients.

Satisfaction with the current feedback from family health services authorities was high (172, 70%) and 154 (62%) practices were prepared to standardise their data collection; the main concern about standardisation related to the time and cost. The main preferences for feedback of data were by individual practice (137) and by whole family health services authority (111). Over half the respondents (143) were in favour of using “spotter” practices to collect data from primary health care teams.


Extrapolating our data to all England and Wales suggests that about 1230000 hours are being spent every month in collecting data in general practice. Considerable duplication of data collection by other bodies exists—for example, for referrals, accident and emergency attendances, and item of service statistics. Historically, general practices have answered to the rest of the health service for what they do, recently through the ill perceived annual reports.3 Data collection has been focused more on meeting administrative demands from outside. Despite high cost computerisation, there is little vision about which data should be recorded, how to use it, how it can best inform practices in their activities, and how it should be standardised.

Telecommunications could lead to fast and effective transfer of large amounts of data over wide area networks on an unprecedented scale. For this to be successful, however, data collection needs to have more focus, agreed standards, and consistent cross mapping so that it can eventually provide data for health resource planning and public health and epidemiological uses. This would also reduce general practice workloads—at present questionnaires are sent to practices because there is no other way of getting this information. Data collection in primary health care is clearly crucial to use fully the personal knowledge that teams derive from extended day to day contact with patients.4 The potential of large computer databases in general practice has yet to be realised.5 Standardised software systems, such as Scotland's GPASS, are one way forward, but until standards of data integrity and format are agreed nationally there seems little likelihood of realising the full potential of data collection in primary care and ending costly duplication.


  • Funding None.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.