Intended for healthcare professionals

Editorials

The paperless general practice

BMJ 1996; 312 doi: https://doi.org/10.1136/bmj.312.7039.1112 (Published 04 May 1996) Cite this as: BMJ 1996;312:1112
  1. Ian N Purves
  1. Director Sowerby Unit for Primary Care Informatics, Department of Primary Health Care, Newcastle University, Newcastle upon Tyne NE2 4HH

    It is coming, but needs more professional input

    The contract between health commissioning authorities and general practitioners states that “a doctor shall keep adequate records of the illness and treatment of his patients on forms supplied to him for the purpose.” “Form” is clearly a paper form, yet in 1993 a large national survey found that 8% of general practices were already paperless.1 The regulations will probably soon be changed to remove the limitation on paperless records. Are general practitioners and their clinical information systems ready for this legitimisation?

    Certainly electronic medical records have been admissible in medical litigation and criminal cases for some years,2 though the Civil Evidence Act requires that the computer system should be created for the purpose that it is being used; there should be proper hardware and software maintenance; electronic records should be contemporaneous; and there should be a full audit trail of additions and deletions. The required audit trails are specified in version 3 of the requirements for accreditation for general practice computer systems,3 and most systems now conform to at least these sections. Furthermore, there is now case law of electronic medical records being used as evidence.4 Practices also need to register under the Data Protection Act 1984.5 The implication of all these changes is that it is (or soon will be) legal to be paperless.

    What about confidentiality? This is an important issue of the moment,6 7 and the debate about maintaining patient confidentiality needs to be concluded before paperless clinical records spread their wings across any wide area network. Nevertheless, in the “trusted base” of general practice the nine Anderson principles of data security6 should be achievable given consideration and some changes to systems.

    Of course, the paperless practice includes more than simply the medical record: it also encompasses administration and other issues relating to clinical information systems. Paperless records in their raw electronic text form add only availability and legibility to their paper form and lack paper's ability to carry figurative annotations.8 An electronic record in coded form, however, opens the door to many forms of added value. These include automated restructuring of records (such as for problem lists); queries on data (such as for disease registers or quality assurance); decision support systems (such as PRODIGY9); speeding, guiding, and validating data input (such as through templates); mailmerge functions (such as for standard semi-automated referral letters); and electronic messaging (such as for laboratory results).

    Nevertheless, an important part of the record will remain free text: the patient's story needs to be captured adequately to enable effective communication through medical records.10 This requirement for recording the patient's story needs free text narrative, which is perhaps not always recorded. Of course, some data currently need to be coded as free text searches but on free text are limited. The proposed “narrative model” of the medical record10 challenges current clinical and medical informatics views and perhaps will move us forward.

    In summary, what are the pros and cons of the electronic record and the paper record? The losses are not yet clear, but where they are visible, as outlined above, they are looking tolerable. The gains improve practice, perhaps significantly, as can be seen from the two recent systematic reviews that covered clinical information systems,11 12 which at their heart have the coded electronic medical record.

    In current clinical systems the gains derive from the data entered. With little data in a system there is little gain; this has been the major barrier to progress towards paperless practice. Once a reasonable amount of data have been entered, however, the data start to work for the clinician and the patient—and this provides the incentive to learn new skills. Most general practice systems support the basics of paperless practice, but a few still do not. Technical innovation is also still required in relation to computer interfaces, though one of the greatest difficulties is reaching a professional consensus so that interfaces can be engineered with enough “intelligence” to make them quick and intuitive tools.

    The major issues that need addressing are professional: the production of good practice guidelines for medical records; a review of the purpose, structure, and content of medical records; the authoring of knowledge bases to improve interfaces; and educating general practitioners about what constitutes a quality record and the best use of clinical information systems. The system suppliers also have work to do. The requirements for accreditation for general practice computer systems need a more effective set of user requirements and evaluation of developments, and these in turn need more input from the clinicians who will use these systems. The time of top down processes, led by management consultants, should come to an end. Given enthusiasm by the clinical professions, the NHS Executive should see a gain from funding collaboration between the professions and system suppliers to develop these future systems.

    So are we ready for paperless practice? The short answer has to be a cautious yes. In fact over 10% of general practices probably already are paperless. Furthermore, the quality of their records has been shown to be good.13 The slightly longer answer is still yes, though work is needed to ensure that the computer as a tool is integrated into the consultation, and that in turn needs efforts from the professions as well as from system developers.

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    View Abstract