Skip to content

Ch 8d messaging

Chapter 8d -- Clinical Messaging

8d.1 Introduction

This chapter will deal with the clinical communications supporting day-to-day patient care, which involve the sending of data to and from GP systems.

An important part of the safe and effective use of clinical data received from elsewhere is to have an appreciation of the limitations such data may have, as a consequence of the methods of their capture in a remote system, and their subsequent transformation, transfer to and incorporation into the GP system. Mention will therefore be made of 'health warnings' practitioners should bear in mind, when using data originated elsewhere.

Where, as is nowadays the case, diverse mechanisms are emerging to support a particular business flow of information, guidance will be offered on the system-independent processes needed to handle such data and information transfers.

8d.2 Background

The messaging discussed here is the transfer of clinical data between computer systems, the data being structured (organised, coded in whole or in part, and assembled) to a standard observed by both systems. This structuring means that the receiver can 'understand' all or part of the data's meaning - as if they had been natively entered.

Every practitioner will be acquainted with the laboratory investigation reports, standards for which were developed in 1993-5 and which now have near-universal implementation after the Pathology Messaging Implementation Project (PMIP). GP2GP record transfer and then GP Summary Care Record uploads (see chapters 8b & 8e respectively) are the next most widely-used messaging applications. Other clinical messaging, though now on the threshold of a major expansion, has been slow to develop.

Currently, all new messaging travels via the NHS Spine, for which accreditation to the demographic and other services is needed -- but is not in place for many Trust systems. So the recent work to develop the NHS Interoperability Toolkit^1 (NHS ITK -- sets standards based on web services for (any) systems to transmit data securely between themselves) will make communication between GP and Trust systems easier to establish, and through lowering the barriers to entry, will involve more systems as participants to the communications.

This diversity in messaging, allied to the increasing number of intermediate solutions being offered by document transfer systems (vide infra) and to the possibilities of remote login to Trust systems, brings both opportunity and threat. With the opportunity to participate in a communication that has long been desired (e.g. inpatient discharge summary) comes the threat of confusion from a variety of means of implementing it.

The electronic transfer of documents (by NHSmail or by other means) and the transfer of data through web services has become sufficiently sophisticated that a clinician may not readily appreciate whether the data and text he or she is looking at is in the local or in a remote computer system, and is, or is not discoverable by a search of his/her system. This is likely not to matter for the immediate needs of the patient, but the clinician's responsibility for keeping adequate records does have implications for whether he/she: makes a local entry of data from a remote system that has been used for decision making; or makes a coded entry to allow later retrieval of data imported as text. It is particularly important to ensure that, when a patient record is transferred to a new general practice, information that is held on a remote system is not lost as a result.

8d.2.1 From this background, some guidance can be offered

  • Clinicians have a responsibility to be trained in and to understand in broad terms the limitations of the clinical communication tools they use.

  • For each new clinical messaging or communication facility offered to them, clinicians should ask some questions of the supplier, taking advice if necessary from their local support and through them from NHS CfH:

    • To what extent are data and information I am receiving to be incorporated into the local medical record and discoverable by searches on my system?

    • To what extent and in what form (coded or text) will data I am receiving be subsequently exported to the patient's next GP by a GP2GP record transfer?

    • For data which passes out of my clinical system, to whom will it go and what will it be used for?

    • Which are the standards governing the transfer, and are these all open or are any proprietary?

    • Does the transfer conform to NHS guidance on encryption & security^2?

    • Has the NHS CfH clinical safety process been used to assure this development, and if not, what stands in its place?

    • What training am I and any of my staff who are to use this software going to receive?

8d.3 Processes involved in handling clinical messaging data transfers

This section concentrates on incoming data handling. Outgoing data by messaging is less demanding, is covered by guidance above, and (apart from ETP, SCR and GP2GP) there are no applications in immediate prospect.

All the incoming messaging applications to be discussed specifically below (Pathology, Radiology, OOH encounter, A/E or OPD attendance and I/P discharge) have common characteristics in the way they need to be handled -- some preparatory work needs to be done to confirm validity and to deliver them to the right person and they then need to be processed by a clinician and filed or archived away. The main steps in this might be summarised;

  • An administrative process deals with corrupted or missing transmissions

  • An administrative process matches manually those patients or clinicians identified in the message for whom an adequate machine match is not possible

  • An administrative process assigns or reassigns messages to the work-stream for the clinician who is to attend to them

  • A clinical process of examining and acting on the contents on the message occurs; this may involve coding or re-coding some of the message content

  • An administrative process of filing the message content into the clinical record and clearing it from the work-stream occurs. This is often triggered by the clinician

A large number of diverse communications converge on every practising clinician and these are gradually moving to be presented through the medium of the clinical records system. There is clear potential for failure if the handling mechanisms do not offer maximum utility and efficiency, and this is a prospective systems design issue.

Based upon this, some guidance can be offered:

  • For every clinical communication stream, practices should appoint and train one or more members of staff (and at least one deputy to cover for sickness/leave) who is responsible for the administrative processes outlined above

  • Practices should have a protocol for handling laboratory results and other clinical messages which are intended for a clinician's attention but which arrive during his or her absence

8d.4 Pathology

The PMIP messages in widespread use were scoped for use for Haematology, Biochemistry and Microbiology reports, and because of their success in this and the lack of development of other clinical messaging vehicles, they have since been used for wider purposes, including for cytology and radiology reports.

In 2005-6 new pathology messages were developed by NHS CfH^3 to extend the scope of use to requesting and the full range of report disciplines and, supported by the Carter Report into Pathology^4, they await a political impetus for implementation. A current NHS CfH project^5 is using the new report message to pilot the transmission of neonatal Blood Spot testing results in to departments of Child Health.

PMIP messages usually identify the investigations conducted by Read codes but also permit un-coded investigations. Systems have offered facilities to allow practices to encode these tests but this is to be deprecated because of the risk of not doing this as the laboratory would, had it been faced with the same code-list choices. This would not only have the potential to cause a later misapprehension, but also, in the age of GP2GP transfers, could cause tests not to be recognised as equivalent in the destination GP system.

As a result of the historical independence of pathology laboratories, there are many labs which are conducting the same investigations, and reporting them using different units of measurement. Examples include the measurement of Haemoglobin concentration expressed in grammes per decilitre and grammes per litre. An initiative from within the Pathology^6 community is working to reduce these non-uniformities, but it will take many years before this is complete. Meantime, GP2GP record transfers and centrally-hosted GP systems with insufficient defences against these differences risk mixing these non-comparable data and leading to clinician confusion.

Based upon this, the following guidance may be offered;

  • Practices and laboratories using PMIP report messages should avail themselves of the extensive implementation advice^7 available

  • Practices should encourage their laboratories to use the National Message Assurance Service (NMAS^8) for periodic checking of their message output and where problems arise in use

  • Practices should be aware that where laboratory tests are not identified by Read codes, they will not be discovered by subsequent searches

  • Practices should encourage their laboratories to use valid Read codes for identifying investigations, and where such codes do not exist, to apply for their inclusion in the Bounded Codelist^9 that constrains use of the PMIP messages and in the new National Catalogue for Laboratory Medicine^10

  • Practices should exercise great caution in using facilities which may be provided to map uncoded investigations in laboratory reports to Read codes: it is in every way preferable for the laboratories themselves to do this mapping and to involve them in the problem

  • Practices should be aware of the limitations in aggregating report data (say into trend lines) caused by the changing of: codes used to identify investigations; of laboratory methods; and of laboratory reference ranges

  • Practices should be aware that the import of laboratory results from a patient's previous practice via GP2GP transfer may bring together instances of the same investigation carried out in different laboratories, using different methods and reported using different units of measurement

8d.5 Radiology

As has been mentioned earlier, the PMIP report messages are being used in a number of places to convey reports for radiology, a subset of Diagnostic Imaging (DI) modalities. New HL7v3 radiology request and report message standards have been developed in NHS CfH. As yet, there are no smartcard-enabled PACS/RIS (radiology) systems, and until this is the case, these messages will not be able to travel via the spine to GPs. However the NHS ITK mentioned earlier offers hope of an earlier step towards using these messages.

8d.6 Out Of Hours (OOH)

The dominant supplier for OOH centres (Adastra) offers an electronic document transfer of details of an OOH encounter and there is widespread use of this. It is text-based and so practices will need to make their own decisions and to personally encode them if they wish to store clinically-coded information relating to these.

Arising from this, guidance can be offered;

  • Practices receiving messages about OOH encounters should make decisions on clinical coding just as they would had the communication been received on paper

8d.7 A/E encounter, outpatients encounter, inpatients discharge.

These three communications are grouped together because they share similar characteristics. Messages developed by NHS CfH are not now planned to be implemented as part of a future development of the Summary Care Record. The path to developing the capability of GP systems to receive them is reasonably clear, the capacity of Trust systems to send them seems further off.

Meantime, a wide variety of initiatives is giving GPs access to some of this information in varying degrees of structure, ranging from structured messages based upon a project in 2006 in Kettering, to unstructured document transfer (e.g. into the Docman add-on). In this environment, the precautions advised of practices in section 8d.2 and 8d.6 are also applicable.

Back to top