Informatics 99sep4

Journal of Informatics in Primary Care 1999 (September):14-16


A computerised clinical reference resource in primary care:  How is it being used?

James McMorran, PhD, MRCGP1, Jeremy Dale, PhD, MRCGP2

1 Visiting Senior Clinical Lecturer, Primary Care Unit, School of Postgraduate Medicine, University of Warwick
2 Professor, Primary Care Unit, School of Postgraduate Medicine, University of Warwick


Despite increasing interest in computerised decision support in general practice, there is little evidence about how such systems would be used. This study examined the use made by 128 GPs of a new computerised medical reference resource designed to be used as decision support tool. It found that, on average, GPs used the resource 1.35 times a day (range 0–12). The most common information used related to clinical problem solving. Some parts of the system, such as those relating to epidemiological information, appeared to be used infrequently. The reasons for the wide variation in the reported use of the system require further investigation.


          Decision support systems; general practitioners; computer use in general practice.


The availability of computers in primary care has increased more than tenfold since the 1970s, and there is now some degree of computerisation in more than 80% of general practice surgeries1. However, it appears that in most practices computers are still being used in a relatively limited way within the consultation2,3,4.

There is growing interest in computerised decision support in general practice. Computerised reference resources related to prescribing are being investigated5,6, as are more generalist computerised reference resources which allow the provision of a knowledge base relating to areas such as epidemiology, aetiology, investigation, diagnosis and prognosis of conditions7. These provide the potential for becoming an NHS standardised clinical reference that can be updated and modified by the users8.

The aim of this study was to explore how a computerised medical reference resource (Medilinks) was being used in general practice. This resource contains over 1.5 million words, has over 20,000 index terms, and was designed to be a 'look-up' tool within the primary care consultation.


Based on the results of a pilot study regarding areas of the system accessed by the user, a brief questionnaire was designed. This consisted mainly of closed questions relating to the frequency with which the system had been used and the parts that had been found most useful.

The study sample included all 133 GPs in the UK who had purchased a trial copy of the system (Medilinks) in July and August 1997.

Medilinks is a computerised reference resource that has been developed over the last five years. The information base contained in Medilinks was initially constructed based on the Oxford University Clinical School Curriculum9. Information is accessible via a pathological classification of disease entities, for example, information relating to ‘malignant melanoma’ is accessible via the link that ‘malignant melanoma’ is a subtype of ‘skin cancer’. Also information is accessible via a symptomatological classification, for instance, information relating to ‘red eye’ links in to the various causes, such as ‘iritis’. For each disease entity there is information relating to sub-elements such as epidemiology, aetiology, investigations, management and prognosis. The information base is continually updated and modified by a team of authors. Information contained within the Medilinks database is peer reviewed and, if available, evidence-based.

The design of the reference resource combines ideas from relational databases and semantic networks. Information is written in logically independent ‘packets’ and linked to other items in the resource by the hypertext semantic links, such as ‘type of’, ‘may be caused by’. The design of the system maximises ‘referential integrity’; that is, information is only updated in one ‘packet’ but referenced by many different semantic links and therefore information only needs to be updated once. This system design feature makes it possible to have many different authors concurrently updating and modifying the same resource.

The questionnaires relating to Medilinks were sent out in November and December, and a reminder was sent out six weeks later. All subjects potentially had use of the system for at least three months prior to receiving the questionnaire.


Five questionnaires were returned with explanations that the software had not yet been used because of lack of access to appropriate computer resources, and these subjects were excluded from the analysis of questionnaire data. Of the remaining 128 subjects, 93 (72.6%) returned completed questionnaires.

Analysis of the data relating to frequency of use was made with the assumption that a working week consisted of 5 days, and a month was composed of four working weeks. The average frequency of use during the month prior to completing the questionnaire was reported as 1.35 times per day (range 0–12; quartiles 0.3, 1.0, 2.0). Most (28/51) respondents said that their use of the system was approximately the same as one month previously, while 15/51 said they were using it more frequently.

The respondents were sub-divided according to the population median for frequency of usage. ‘More frequent’ users (44/93 subjects using the system more than once per working day) reported using the resource on average 2.28 times per day (quartiles1.5, 2.0, 2.5). The 49 ‘less frequent’ users reported using it on average 0.51 times per day (quartiles 0.1, 0.5,1.00).

The use made of the different types of information held within the resource were classified by respondents as ‘never’, ‘occasional’ and ‘frequent’ use (Table 1). ‘More frequent’ users appeared to use the resource for clinical problem solving, such as information relating to the diagnosis and investigation of diseases. ‘Less frequent’ users appear to make less focused use of the system, with no area of the system being described by more than 40% of respondents as being used frequently. ‘Less frequent’ users seem to be more negative about the information relating to clinical examination, symptoms and signs, basic knowledge, such as laboratory values, and other information than ‘more frequent’ users.

In addition, respondents were asked to state what part(s) of the system they found most useful. Only 49 respondents answered this question. The most useful parts of the system were listed as:

  • information relating to diagnosis/investigation (27)
  • clinical management (26)
  • evidence-based medicine (21)
  • prognosis (19)
  • clinical presentation (18)
  • epidemiological information (15)
  • pathological/aetiological information (12)
  never used occasional use frequent use never used occasional use frequent use
Epidemiological Information 2 (4.5%) 40 (90.1%) 2 (4.5%) 27 (55.1%) 17 (34.7%) 4 (8.2%)
Path/aetiological Information 2 (4.5%) 24 (54.5%) 18 (41%) 9 (18.4%) 31(63.2%) 9 (18.4%)
Clinical Presentation 6 (13.6%) 18 (41%) 20 (45.5%) 8 (16.3%) 25 (51.5%) 16 (32.7%)
Diagnosis/Investigation 1 (2.3%) 16 (36.4%) 27 (61.4%) 7 (14.3%) 26 (53.1%) 19 (38.8%)
Clinical Management 1 (2.3%) 19 (43.2%) 24 (54.5%) 3 (6.1%) 30 (61.2%) 16 (32.7%)
Prognosis Information 6 (13.6%) 28 (63.6%) 10 (22.7%) 12 (24.5%) 26 (53.1%) 11 (22.4%)
Evidence based Medicine 8 (18.2%) 30 (68.2%) 6 (13.6%) 13 (26.5%) 30 (61.2%) 6 (12.2%)
Causes of Symptoms & Signs 3 (6.8%) 29 (65.9%) 12 (27.3%) 16 (32.7%) 27 (55.1%) 6 (12.2%)
Relating to Clinical Examinations 21 (47.7%) 22 (50%) 1 (2.3%) 25 (51.0%) 24 (49.0%) 0 (0%)
Basic knowledge e.g. Laboratory Values 15 (34.1%) 20 (45.5%) 9 (20.5%) 18 (36.7%) 26 (53.1%) 5 (10.2%)
Information relating to Rules/ Regulations 23 (52.3%) 20 (45.5%) 1 (2.3%) 22 (44.9%) 25 (51.0%) 2 (4.1%)
Other information e.g. local guidelines 37 (84.1%) 6 (13.6%) 1 (2.3%) 41 (83.7%) 8 (16.3%) 0 (0%)

NB  If a questionnaire was returned with no response for a particular entry then it was assumed that these data were never used by the respondent.

Table 1:  Utilisation of computerised reference resource


This survey obviously has limitations. Those surveyed were a self-selected group of GPs who had purchased a computerised medical reference system. They are likely to have been a group with an existing interest in GP computing. Data comprised users’ reported use of the system, and lacking a record of actual use the validity of their perceptions about use is unclear.

In general, users are positive about the use of such a resource and are using the system at least as much as in the month previously. There was wide variation in the reported use of the resource, with most using the system over once a day. While some parts of the system were reported as being used regularly, particularly those relating to clinical problem solving (such as diagnosis/investigation, clinical management), others (such as epidemiological information) appeared to be perceived as less useful. The extent to which this reflects attributes of the decision support tool or the attributes and needs of GPs needs further study.

At present, the clinical reference is not directly accessible from GP computerised medical records systems. It is probable that frequency of use may have been greater had it been more readily available to the GP during the consultation.

More detailed studies relating to the use of this resource and comparison of use by different users (for instance, GP Registrars and GP Principals) are currently being undertaken. Also the information base has now been made accessible via the internet10 and research as to the information requirements of an internet primary care resource is being undertaken.



Social Surveys (Gallop Poll) Limited. Computerisation in GP Practices 1993 Survey. Department of Health (NHS Management Executive), London, 1993:40


Daniels A. GP survey on electronic communication. Oxford IM and T Consulting, 1991


Ahmed MA, Berlin A. Information technology in general practice: current use and views on future development. J Informatics in Primary Care 1997 (November):5–8


Richards H, Sullivan FM, Mitchell ED, Ross R. Computer use by general practitioners in Scotland. Br J Gen Pract 1998; 48:1473–1476


Purves IN. PRODIGY: implementing clinical guidance using computers. Br J Gen Pract 1998; 48:1552–1553


Walton R. An evaluation of CAPSULE, a computer system giving advice to general practitioners about drug prescribing. J Informatics in Primary Care 1996 (March):2–7


McMorran J, McMorran S, Wacogne I, Rowbotham C, Young-Min S, Crowther D, Pleat J. Medilinks – a computerised medical reference. Butterworth-Heinemann, 1995


Fox J, Glowinski A et al. Logic engineering for knowledge engineering: design and implementation of the Oxford System of Medicine. Artificial Intelligence in Medicine 1990; 2:323


McMorran J et al. Personalised Medical Reference. Biomedical Informatics Today 1993 (Autumn):17


ePulse Clinical Reference Guide.). Miller Freeman UK Ltd, 1999

On to News

Back to September 1999 informatics Contents Page

Back to informatics Index


  • There are currently no refbacks.

This is an open access journal, which means that all content is freely available without charge to the user or their institution. Users are allowed to read, download, copy, distribute, print, search, or link to the full texts of the articles in this journal starting from Volume 21 without asking prior permission from the publisher or the author. This is in accordance with the BOAI definition of open accessFor permission regarding papers published in previous volumes, please contact us.

Privacy statement: The names and email addresses entered in this journal site will be used exclusively for the stated purposes of this journal and will not be made available for any other purpose or to any other party.

Online ISSN 2058-4563 - Print ISSN 2058-4555. Published by BCS, The Chartered Institute for IT