Image Distribution and other Aspects of Radiologist – Clinician Communication

Günther Gell
Dept. of Medical Informatics, LKH-Univ.-Klinikum Graz
A-8010 Graz, Austria

Objective

This is a position/discussion paper to raise awareness that PACS and image distribution may have indirect but important and possibly also negative effects on medical practice. Such effects should not just happen as accidental side effects but should be anticipated and planned in a conscientious and responsible way.

Introduction

In the traditional interaction between clinicians and radiologists the clinician sends the patient to the radiologist with a specific request for information, e.g. to verify or exclude a diagnosis, or at least to narrow down the range of possible diagnoses. The radiologist performs the examination and interprets the images. The results are communicated to the clinician, primarily in the form of a written report, occasionally also orally (in emergency cases, or during a clinical conference) and, if needed, with the images or a selected subset of images.

This traditional pattern is based on the assumption that radiology examinations need a highly specialized knowledge to perform the procedures and interpret the results efficiently and effectively.

In practice, there exist many variations of the basic pattern: if new examinations like ultrasound (seem to) need less knowledge for interpretation and carry no particular risks (no potentially dangerous radiation exposure as with X-rays), then they tend to be established also outside of imaging departments. Also, if a highly specialized clinical speciality depends heavily on images, there is a tendency to develop and incorporate the skills and the equipment for imaging within this speciality (neurosurgery etc.).

There are few objective criteria to determine the optimal distribution of imaging tasks between radiologists and clinicians – optimal in the sense of optimal outcome for efficient and effective patient care. Different actors like adminstrators, radiologists, clinicians (of course in itself a very inhomogeneous group), industry etc. have different views and different interests.

Of course communication plays a key part in this complicated structure and changing one element may change the pattern of cooperation between radiologists and clinicians substantially. Making for example all images available to clinicians, even prior to the availability of a radiology report, may induce clinicians to interpret images and start therapeutic actions without waiting for the report. This depends also from the delay till the report becomes available to the clinicians. Free access to images may also reduce the interest in joint clinical conferences thus reducing direct communication and mutual feedback. Such a change, which may have advantages and disadvantages should not be an accidental byproduct of a technical measure but the result a consensual analysis and strategy to use the new possibilities of digital imaging, image management and image communication in a coherent way to serve the overall purposes of patient care.

Of course the main factors are not technical systems but the (lack of) confidence of the clinicians in the quality of radiological reports (in comparison with his own interpretation).

Radiological Image Distribution

As we have seen, image distribution is an example, where the change of one apparently technical parameter of a system may have quite far reaching effects and therefore is an interesting object of analysis. What are the factors that favor electronic image distribution?

·       Technical feasibility and technical logic. Just as digital imaging methods lead to PACS, i.e. digital image management and archiving, the existence of PACS leads to image distribution – an elegant and technically interesting solution to avoid the cost, delay etc. of conventional film lending.

·       The obvious need of clinicians to see the images and not only the reports in many circumstances, particularly if anatomical and morphological details and not only pathology is needed for therapy (e.g. surgery).

·       The increasing clinical use of image processing, e.g. in computer aided surgical navigation, modelling in orthopaedic or dental surgery, radiotherapy planning etc. – the list is growing very fast.

·       The growing interest for the electronic patient record, which will include all kind of medical images.

The reasons above, in particular the second and third, are strong enough to justify the technical development and implementation of image distribution. Given the present status of PACS there are a number of possible solutions: a viewer under Windows that accesses the PACS archive or an intermediate server via DICOM query, Java applets or WWW-browsers with appropriate servers. Some apparently technical decisions like push or pull, have important consequences and should therefore not be made only because of technical considerations but as an implementation of an organisational policy.

For a project for image distribution therefore the following questions should be answered before starting a technical implementation:

·       What are the goals of image distribution (or remote image access)

·       What are the rules for image distribution/image access for different users and different uses

·       What level of data protection and, data security, speed and reliability is required

·       What functionality of the viewer is needed – dependent on the different user groups

·       How do we evaluate, if the goals have been reached

·       Are there any milestones, where the deployment of the system will be stopped or even reversed if important goals cannot be reached.

A generalised Approach to Clinician – Radiologist Communication

The PACS System builder will usually not have the official competence to set or alter the rules that govern clinician – radiologist communication. He may however do so effectively by introducing apparently purely technical measures, blurring the responsibility for possible negative effects. Therefore we suggest the following course:

·       Insist on getting competent answers for the questions of the preceding chapter, in particular concerning image distribution/access policies, preferably from joint working groups of clinicians and radiologists

·       If consensus cannot be reached, the safe approach – primum non nocere (firstly do not harm) – would be to stick to the established rules of image distribution and copy them with the electronic system

·       Establish a strategy to streamline and improve the whole system of clinician – radiologist communication, the ordering of exams, the access of radiologists to clinical information, reporting and report distribution, clinical feedback for radiologists, video conferencing and ”asynchronous” interactions and image distribution.

Results

In our own system in the Graz University Hospital we try to follow the strategy outlined above. For each clinic that wants access to radiologic images we set up a task group consisting of radiologists and clinicians under the coordination of medical informatics, that sets up a consensual policy for image access, which is then implemented by the image distribution software with a licensing process (1). Radiologists and clinicians have mutual access to patient data (in the RIS and HIS respectively).

This year we are implementing a pilot project to reduce the delay between reporting and report writing by digital dictation (not speech recognition), where the audio report will be available immediately together with the images for the clinicians as a preliminary information. Some experiments with commercial speech recognition systems have not been successful (in the sense that they were not accepted for routine use).

Discussion and Conclusion

Image distribution is an example that shows that the introduction of apparently purely technical improvements may change the practice of medicine in a way where the final consequences are difficult to assess. We therefore propose to start a discussion between the concerned parties to analyse possible consequences and to define a consensual policy how to integrate new technical possibilities and order to get the optimal solution for efficient and effective patient care.

References

1.  Gell G., Hecke P., Zeilinger G. Radiological Image Distribution: Problems and Solutions. 15. International EuroPACS Meeting, Pisa, (1997) Tipografia Editrice Pisana, 43-46

Corresponding Author:

Prof. Günther Gell

Department of Medical Informatics

LKH-Univ.-Klinikum Graz

Engelgasse 13

A-8010 Graz (Austria)

Fax : + 43 316 385-3590

e-mail: gell(at)email.kfunigraz.ac.at

URL: http://www.kfunigraz.ac.at/imiwww/

Oral presentation at EuroPACS'98, Barcelona, Spain