- M C Record,
- J A Spencer,
- R H Jones,
- K P Jones
- Department of Primary Health Care, Medical School, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH
- Correspondence to: Dr K P Jones.
- Accepted 24 August 1994
Objective: To ascertain the views of primary care professionals about the current purpose, uses, potential, and workload implications of the statutory general practice annual report.
Design: Postal questionnaire survey.
Setting: General practices in the Northern region.
Subjects: All practices in the region that were singlehanded, fundholding, non-fundholding and with more than five partners, and a one in three20random sample of all non-fundholding practices (n=318). Results - 263 practices responded (83%). The20report took a median of 12 hours to produce (95% confidence interval 11 to 15 hours; interquartile range 7-35). The main perceived purpose of the report was to monitor practice activity (165 respondents; 63% (95% confidence interval 57% to 69%)), but 44 respondents (17%; 13% to 22%) produced it only because it was contractually required. Practices included statutory and non- statutory data in these20reports and would have liked comparative practice activity information (155 respondents; 59%) and “good ideas” (165 respondents; 63%) fed back to them. Respondents would have liked the annual report used to improve practice development planning (122 respondents; 46% (40% to 52%)), to facilitate audit (115 respondents; 44% (38% to 50%)),20and to influence resource allocation (104 respondents; 40% (34% to 46%)). One hundred and eighteen practices (45%; 39% to 51%) would produce20an annual report even if not contractually required. Data collected were perceived to be already available20elsewhere.
Conclusions: Primary care professionals have20concerns about the current annual report. They20would prefer to collect relevant, standardised data which could lead to better audit, planning, and resource allocation.
The new contract for general practitioners in 1990 required each practice to produce an annual report
These reports have been criticised as being of little practical value either to health authorities or to practices
This study shows that 31% of those primary care professionals surveyed were concerned about the time taken to produce the report, although the reported median completion time was only 12 hours
Other concerns included problems with data handling, inconsistent recording, imprecise definitions, software difficulties, changing requirements, and poor central organisation and administration
However, there was a strong commitment to the concept and use of practice based data collection and a wish to collect relevant, standardised data
General practice reports have been produced by primary health care teams for years, the earliest recorded being in 1970, though 70% of a surveyed group of practices producing them had done so only from 1984. Data most commonly included were practice histories, consultation and visiting rates, immunisation rates, practice population profiles, personnel and staff changes, as well as written contributions from primary health care team members. These poineers thought that such retrospective practice reports could aid planning in the practice by collating relevant data, helped foster greater spirit and communication, and increased accountability to users of the service.2
The new contract for general practitioners in 1990 included a requirement for each practice to produce an annual report.3 The report was designed to increase the amount of information available to family health services authorities about primary care. The required data included details about premises; staff with reimbursable salaries; prescribing formularies and arrangements for repeat prescriptions; how patients' comments about the practice were obtained; general practitioners' commitments outside the practice; and secondary care referrals, including outpatient and inpatient referrals. The health departments saw general practice annual reports as part of the government's drive to increase accoutability within general practice and also as an input into the medical audit arrangements outlined in Working for Patients.4 However, annual reports have been criticised as being of little practical value either to health authorities or to the practices producing them.5, 6
This paper surveys primary care professionals' views about the workload implications of producing this mandatory document and its perceived usefulness.
In summer 1993 all singlehanded, fundholding, and “large” practices (defined as a a non-fundholding practice with six or more partners) plus a one in three random sample of all other non-fundholding practices in the Northern region were sent a postal questionnaire. This asked about the current purpose, uses, potential, and estimated workload implications of producing the statutory practice annual report. The questionnaire, containing closed multiple choice and open textual questions, was sent to the senior partner but requested the views of other primary health care team members. Respondents were also asked for details about their list size, number of partners, location, type of computer system, and whether they had a practice manager. Results were analysed by the SPSS PC+7 and confidence interval analysis software packages. Descriptive statistics with confidence intervals and categorical comparisons by means of the X2 test are reported. Continuous variables were compared by the Mann-Whitney and Kruskal-Wallis tests, as appropriate. Ninety five percent confidence intervals for percentages are calculated to within +/-6% unless stated otherwise.
Two hundred and sixty three out of a possible 318 responses were received (83%). Response rates varied with the status of the practices but the differences were not statistically significant. Tables I-III give the profiles of the responding practices in terms of location, numbers with a practice manager, average list size, and types of computer system used.
|Status||No (%) of practices responding||No (%) with practice manager||Mean list size (SD)|
|Singlehanded||98 (37)||54 (55)||2 111 (658)|
|Non-fundholding||87 (33)||60 (80)||5 638 (2 202)|
|Fundholding||58 (22)||57 (98)||10 927 (3 402)|
|Large||20 (8)||19 (95)||13 501 (2705)|
|Total||263 (100)||200 (76)||6 138 (4 488)|
|Location||No (%) of responding practices|
|Inner city||40 (15)|
|Urban or suburban||127 (48)|
|Did not specify||18 (7)|
|No (%) of responding practices|
|AAH Meditel||45 (17)|
Purpose of annual report
Roughly two thirds and half of respondents, respectively, thought that practice activity monitoring and assisting in audit were important functions of the annual report (table IV). Of the 103 practices (39%) indicating that they completed the annual report because of its contractual nature, 44 (17% of the total; 95% confidence interval 13% to 22%) indicated that the contractual obligation was the only reason they produced it. Over half of these (23; 52%) were singlehanded practices. This difference was significant (df=3, X2=8.5, P<0.05).
|Perceived purpose||No (%) of responding practices [95% confidence interval]|
|For monitoring practice activity||165(63) [57to69]|
|To assist inaudit||130(49) [43to56]|
|To review practice team activities overyear||108(41) [35to47]|
|To inform family health services authority about practice needs anddifficulties||106(40) [34to46]|
|To provide practice developmentplan||68(26) [21to31]|
|To inform practicepopulation||17(6) [4to10]|
Producing annual report
Two hundred and thirteen practices (81%) indicated that general practitioners were involved in producing the annual report but in only 117 practices (44%) did the practice nurse play a part. Clerical staff also contributed: receptionists in 142 practices (54%), computer operators in 96 (37%), and secretaries in 112 (43%). In practices that had a manager he or she almost invariably played a part (199/200; 95% confidence interval 97% to 100%).
The number of practices reporting that attached district health authority employed staff contributed to the annual report was less than that for practice employed staff, in 34 practices health visitors or district nurses, or both, contributing to the report (13%; 95% confidence interval 9% to 17%). Other participating staff included community psychiatric nurses, physiotherapists, chiropodists, midwives, and counsellors.
Table V indicates the estimated median and interquartile ranges of times for members of the primary health care team to complete the annual report (including only those questionnaires where a time was indicated). Table VI gives a breakdown of the median and interquartile ranges for the time taken to complete the annual report in the different types of practices. The differences were significant (Kruskal-Wallis test: df=3, X2=13.2, P=0.004). Those respondents who believed that the only purpose of the annual report was statutory did not spend significantly more or less time completing the task than all other respondents (“statutory only” group: median time 12.5 hours (interquartile range 5- 41); all other respondents: median time 12.0 hours (7-34); P=0.78).
|Estimated time taken (hours)|
|Team member||No (%) of responding practices||Median||95% Confidence interval||Interquartile range|
|Practice manager||175 (73)||6||5 to 8||3-12|
|General practitioner||180 (75)||3||3 to 4||2-8|
|Receptionist||111 (46)||4||3 to 6||2-12|
|Clerical staff||78 (33)||4||2 to 7||2-10|
|Computer operator||84 (35)||4||2 to 5||2-7|
|Practice nurse||79 (33)||2||2 to 3||1-4|
|District nurse||26 (11)||2||1 to 2||1-2|
|Health visitor||22 (9)||2||1 to 4||1-4|
|Other||11 (5)||3||1 to 20||2-18|
|Total||239 (100)*||12||11 to 15||7-35|
*Twenty four respondents did not complete this part of the questionnaire.
|Estimated time taken (hours)|
|Status||No(%) of responding practices||Median||95% Confidence interval||Interquartile range|
|Singlehanded||87 (36)||12||8 to 13||5-17|
|Non-fundholding||79 (33)||13||10 to 21||8-34|
|Fundholding||54 (23)||22||10 to 37||6-48|
|Large||19 (8)||32||12 to 58||12-58|
|Total||239 (100)*||12||11 to 15||7-35|
*Twenty four respondents did not complete this part of the questionnaire.
The median time taken for practices to develop their annual report if they included a business plan was 29 hours (95% confidence interval for the median 26 to 43; interquartile range 11-51 (n=68)) and 11 hours (10 to 13; 6-22 (n=168)) if they did not include one. This difference was significant (Mann-Whitney U test: z=-4.1, two tailed P<0.0001).
An open ended question asked what difficulties or problems, if any, respondents had had in completing the report. One hundred and eighty five questionnaires from respondents (70%) included comments. Of these, one fifth stated that they had encountered no problems. The main difficulty cited by those who had was the time taken to produce the report. This was mentioned by 82 (55%) of those commenting - 31% of all respondents (95% confidence interval 26% to 37%). Some of the other difficulties encountered were problems with data handling, including duplication of data (particularly referral data), inconsistent recording, imprecise definitions, and problems with computer software; changing requirements for particular data; lack of motivation; and poor organisation and administration of the reports by the family health services authority. Several respondents questioned the accuracy and quality of the data produced, in particular those relating to referrals to secondary care. Finally, several non-computerised practices reported difficulties in collating the data manually.
Data included in annual report by practices
Two hundred and fifty four practices (97%; 95% confidence interval 94% to 98%) included outpatient referral data, 228 (87%; 83% to 91%) inpatient referral data, 226 (86%; 82% to 90%) data on practice premises, 154 (59%; 53% to 65%) diagnostic data, 170 (65%) data on employed staff working hours, 153 (58%) information on staff duties, and 204 (78%; 73% to 83%) information on past staff training. A few practices also included prescribing data. These data are statutorily required in the annual report.
Practices also included other, non-statutory data in their annual report. Seventy seven practices (29%) included morbidity data, 76 (29%) details of staff training required, 162 (62%) the number of surgery consultations, 155 (59%) the number of home consultations, 141 (54%) the number of night consultations, and 221 (84%; 95% confidence interval 80% to 89%) details about the attached staff in the practice. Thirty two practices (12%; 8% to 16%) also included various additional data - from audit figures to other referral information (emergency, self, and private). Seventy two practices (27%) included a practice development plan in their annual report.
Future uses of annual report and feedback
From a practice perspective 122 respondents (46%) thought that improvement of practice development planning was possible, as well as using the document to facilitate audit (115; 44%) and clinical research (54; 21% (95% confidence interval 16% to 25%)). One hundred and four respondents (40%) also saw the annual report as a valuable tool in influencing the resource allocation process, helping in the targeting of practices with problems and developmental needs (110; 42%), and informing family health services authorities about primary care team activities (114; 43%). Respondents indicated that annual reports could be used to inform the processes of purchasing care (79%; 30%) and health needs assessment (114; 43%). Eight practices (3%; 95% confidence interval 1% to 6%) stated that the annual report should be abandoned altogether in the textual responses to this question.
Respondents stated that they would like referral information (154; 59%), morbidity information (118; 45%), and comparative practice staffing information (81; 31%) fed back to them. One hundred and fifty five practices (59%) indicated that they would like comparative practice activity information fed back to them and “good ideas” to be disseminated (165; 63%). Textual responses were often favourable towards the collection of data and feedback to practices, but some caution about standards was also expressed. For instance, one respondent commented that the current “standard of data [was] too poor to be meaningful.”
Finally, one hundred and eighteen practices (45%) said they would still produce a document similar to the annual report if it was not contractually required. Table VII shows the responses grouped by status of the practices. Differences were significant (df=3, X2=9.74, P<0.025).
|Status||No (%) of practices responding||No (%) that would produce similar document|
|Singlehanded||93 (37)||31 (33) [24-44]|
|Non-fundholding||83 (33)||37 (45) [34-56]|
|Fundholding||56 (22)||38 (68) [54-80]|
|Large||20 (8)||12 (60) [36-81]|
|Total||252 (100)||118 (47) [41-53]|
Differences were significant: df=3, X2=9.74, P<0.025).
This questionnaire survey has shown that primary health care professionals have some concerns about the statutory annual report. Production of the report is considered to be time consuming by many, and there are doubts about the quality and usefulness of the data. There seems to be no clear consensus about the purpose or purposes of the report, and a substantial minority of practices see its only purpose as a contractual obligation. There was a greater proportion of singlehanded practices in this second group.
Data collected for the annual report were often found to be duplicated. For instance, statutory data included in annual reports about reimbursed staff (hours worked, training in the past five years, and duties) and practice premises are recorded by the family health services authorities. Together with the cost of general practice computing, these account for more than 25% of general medical services expenditure (NHS Management Executive, personal communication 1994). Inpatient referral data are also required to be provided in annual reports, yet these data are recorded by hospitals for pricing contracts. That statutory data are not recorded by all practices may indicate either that these practices see little value in their collection or that family health services authorities have waived the need for practices to collect them as they are already held on management information systems. Arguably duplication of data has merit, as nominally comparable datasets may be valuable, especially when existing records are incomplete. Nevertheless, it was clear from many of the respondents' comments that they found this annoying and an unnecessary burden when their administrative workload had already been increased by imposition of the new contract.9, 10
Within the non-statutory data included in annual reports the variety is diverse and, as with statutory data, sometimes repetitive - for example, staff training details, night visiting figures, and attached staff details are all held by other agencies. One explanation for the inclusion of all these data is that individual practices may be using the document to review the year's activities and the data serve to reinforce any objectives or problems that the report highlights.
Time spent producing the annual report was highlighted by respondents, the interquartile ranges of the data showing that there is a wide variability in the estimated time spent compiling the document, especially in practices that have embraced a more business-like attitude (as implied by having a business plan). Though there are important implications in terms of staff workload, this investment of time was apparently considered worth while by some practices, as exemplified by one response from a fundholding practice: “We regard the report as a marvellous summary of what is going on each year.” Probably the figures for the time taken to produce the report are an underestimate, as respondents were not asked to include the time taken to log and enter the data into computerised and manual systems.
The level of participation of attached staff as opposed to practice staff seems to be low and might be explained by the contractual obligation for annual reports not extending to these practice staff groups. Extending this obligation to attached staff, who may already collect data for their own managers, might lead to the collection of data which could produce a workload redistribution, better manpower planning, and a more integrated primary care team. The actual proportions of employed practice clerical and nursing staff concerned in producing the report may be greater than reported in this survey, as the questionnaire did not identify the numbers of such staff in each practice.
Practices see a value in producing a document similar to the annual report. They would also like to see some data fed back to them to enable comparisons to be drawn with other practices in the area and to learn from the experiences of other practices in developing innovative solutions to problems. This would be dependent on data being collected in a standardised way and being fed to practices in an acceptable and effective format.11, 12
As Wilkinson has suggested,6 the current structure of the annual report and the value of the data collected via this mechanism should be re-examined in order to reach a compromise which fulfils the needs of both the primary care team and the health authorities and is not too time consuming. Other workers have shown that collection of standardised data in general practice is possible, is beneficial to the practice, and is actively used by them.13 We propose that a standardised minimum dataset should be collected via a mechanism similar to the annual report under agreed and appropriate definitions which focuses on relevant items not available elsewhere, such as primary care consultation rates and patterns, intrapractice referrals, and other non-reimbursable practice activity.
This dataset could then be brought together with other relevant data on aspects of primary care provision collected by non-practice organisations (such as family health services authorities, district health authorities,14 the Office of Population Censuses and Surveys, and provider units), including inpatient referral data, socioeconomic data, and cervical cytology screening data. In this way accurate, comprehensive, and useful comparative feedback could be provided to practices and managerial bodies to facilitate better planning, audit, and targeting of resources.
In conclusion, we have shown that primary care professionals at present are not wholly convinced that the effort expended in producing an annual report is worth while. However, they perceive a role for data collection and activity reporting if these could be made more relevant to the needs of general practice and the patients it serves.
We thank the practices in the Northern region that participated in this research and for their contributions, Dr T D van Zwanenberg and Ms F MacDonald. We are also grateful to David Newrick for help with the analysis. The grant to undertake this project on the development of general practice annual reports was provided by the Northern Regional Health Authority.