Dislocation of the tongue: the focus of medical IT should be innovation, not digitisation

James Clemence, University of Cambridge School of Clinical Medicine, Addenbrooke's Hospital, Hills Road, Cambridge, CB2 0SP

Medical information technology systems such as the electronic medical record (eMR) are becoming ubiquitous in clinical practice. Digitisation brings considerable advantages over traditional paper methods – ease of information dissemination, interoperability between primary and secondary care, and the ability for simultaneous access of records by multiple professionals. Electronic systems remove the requirement for miles of shelving for storage, allow for theoretically fast data retrieval for audit, and remove the inherent ambiguity of poor handwriting in paper records.

Such outcomes are without doubt laudable, but merely emphasising the portability, reduced space requirements, and increased accountability is missing the greatest strengths of computer systems. Computers are not simply dumb data-entry terminals, and to use them as such deprives systems of considerable capability. At their most simple, an eMR is a digitised medical note – familiar to all clinicians with a white space box and subtitled sections of history of presenting complaint, past medical history, etc. This is in line with historical tradition, and offers a simple and clear way of organising information for clinicians involved in, or new to, a particular case.

However, software implementations of this are frequently inelegant. Large white space boxes for information, a plethora of dropdown lists, and inflexible or unrealistic categories for diagnosis mar the user’s experience. Frequently, dropdown lists rely on awkward catch-all defaults: a particularly humorous example seen recently was the list of anatomical sites for a dislocation including ‘Tongue’.

For those unable to touchtype, eMR systems tend to be slower than handwriting notes, a common frustration vented on the wards. Additionally, whilst an approximate equivalent to the handwritten medical note, concrete assumptions in the system make this a frustrating method of inputting data. The data input side of eMR software could be drastically simplified and made more intuitive, embracing some of the improvements which have developed in other IT fields over the past few years, especially in web-based systems.

The biggest weakness in the current eMR approach is the subsequent retrieval of records, data manipulation, and conversion of information from a note into a clinic letter, other communique etc. and information dissemination. For example, in an audit situation, once provided with a list of patients meeting a diagnostic criteria (via a Read code), information has to be extracted manually from referral letters and clinic notes in order to assemble the dataset. An inherent weakness of this is the implicit assumption of a rigorously consistent Read code policy – something which in practice often has considerable inter-clinician variability. Patients without Read codes are missed, and copying data to the dataset, frequently manually, is a superfluous step that potentially allows aberrant data to enter results.

Modern systems should allow the entering of clinical information simply, but then develop retrieval systems allowing the clinical data stored in clinic notes to be accessed, indexed for search, and extracted without requiring the current degree of manual action. Clinic letters, for example, should not just sit as Microsoft Word documents and scanned images of hospital letters attached to the rest of the record. Instead, they should be indexed for search, and allow extraction of useful clinical data for research, which can improve both delivery of care for that patient in particular, and for the wider population in general.

Putting data into computer systems is the future of practice, rendering paper records obsolete. Part of the problem is that the solutions we have currently are a half-way house, bridging the gap between outmoded paper systems and modern electronic ones. However, as adoption of electronic systems begins apace, and makes increasing inroads into secondary care, we need to think carefully about all the roles these systems need to perform. We need to aim beyond the superficial objectives of merely replacing paper, and mitigating the accountability, legibility, and portability issues associated with paper notes.

Development of systems in the absence of end user input invariably leads to half-baked solutions. Clinicians, as end users, need to be involved throughout the design process, and not merely to clarify to how the current system works. Innovation needs to become the watchword here, stepping beyond the current focus of eMR systems, and instead looking at innovative solutions which embrace IT for some of its biggest strengths – data manipulation and retrieval, rather than purely digitising the system we have used for years.


Story image from Wikimedia Commons.