Clinical Negligence and the Electronic Patient Record

Lawyers will be affected by the changes in general practice and the electronic patient record (EPR).

• Computers are now used widely in all areas of primary care and in general practice, both for recording the notes of clinical consultations and for storing the results and correspondence. EPRs are also increasingly being used in hospitals and, as part of the NHS upgrade, there are plans for the system to become integrated across the NHS (from primary to tertiary care).¹
• Computers have changed the way in which doctors work and, in so doing, have also changed standards of practice. They have opened doctors up to new risks and yet have become the means of limiting other risks.
• Lawyers need to be aware of these changes and, in particular, of how to request all the information needed in cases of clinical negligence. They may benefit from a greater understanding of the EPR and its development and working.

I propose to describe the major changes taking place and to raise some of the main issues that computerisation of medical records will present in clinical negligence cases.

• For over fifty years, the majority of general practitioners kept their consultation entries in chronological order on the standard Lloyd George cards, with the correspondence and results sections tucked into the back of the envelope.
• Any doctor could walk into a surgery and work with the system since it was ubiquitous, simple, data-transferable and generally caused serious problems only when the entries were illegible, inadequate or where fire or loss of records took place.
• Lawyers involved in personal injury work or clinical negligence were also able to consider the records and have immediate understanding of the basic events that had taken place.
• The Lloyd George system raised issues such as the interpretation of the manuscript, whether post incident alterations or entries had been made and whether the doctor had viewed the previous records or the incoming results.

• As standards in general practice rose, GPs were encouraged to keep records in chronological order and to keep the consultation record separate from the correspondence and results section. Separate cards were added to record repeat prescribing, treatment summary and immunisations.
• The fear of litigation caused doctors to write ever-longer entries and, as the paperwork increased, the folders became thicker and more difficult to access.
• In short, today - for most practices - the paper notes have become unmanageable and, even patients with relatively uncomplicated problems, can have records of several hundred pages, thus making it almost impossible for any doctor to read the entire previous history.

• Until 1 October 2000, every doctor was required to keep a written record on the Lloyd George card. Many practices were already starting to keep "additional records" on computer and some had already abandoned the written notes entirely and gone "paper-free".
• What had started as a prescribing tool to save time and handwriting, quickly became a means of recording immunisations, cervical smear results and patient summaries. Recording a basic diagnosis was possible and then free-text became available for making a full consultation record. The EPR was born.
• Today, there is hardly a practice without a computer in use. Government offered a high proportion of reimbursement for both software support and hardware for "approved systems". So, by 1995, there were many different systems in operation, often with only small numbers of users and with each user operating markedly different levels of use even within the same practice.
• Some doctors would be using the computer simply to prescribe drugs and to operate an age-sex register; others were entering large amounts of data and clinical entries on which to manage a practice population and undertake research and audit. What was needed more than anything else was some standardisation and transferability of data between doctors. For this purpose Dr James Read suggested a system of coding.

Read/Standardised Nomenclature of Medicine (SNOMED) codes are an attempt to classify medical activity and offer to be the means by which medical records will, in future, be transferred from one practice to another.

Not dissimilar in principal to the Linnaean classification of species in the nineteenth century, James Read produced an international classification of medical activity to include disease names, operations and procedures. The aim was to allow easy transfer of information between primary, secondary and tertiary care and be easy to use by clinical staff, researchers, administrators and planners.

• The doctor selects the Read/SNOMED code and, consequently, coding is only as good as the knowledge of coding by that doctor. Every computer system in use allows, encourages, or insists on coding of every consultation.
• Where coding is obligatory, it tends to force the doctor into entering a diagnosis and, in this sense, the code can prejudice the record and a doctor uncertain of the diagnosis can be pressed into entering a code for which there is no firm evidence, but which may later be judged as supporting that diagnosis.
• Coding is easy for conditions such as "sore throat" or "ear pain." Such entries are likely to be accurate; however, the more complex codes are less reliable, and many are entered wrongly.
• Busy doctors have learnt shortcuts to avoid coding by entering codes such as "unwell generally" or "chat to patient". Some codes are well known as being ridiculous; for example "doctor walked out" or "accident caused by soap, stew or curries" and "late effect of foreign body entering orifice."
• To the best of my knowledge, the significance of the selecting of the code has yet to be tested in a case in court but, if a doctor has entered a code and not acted upon it, he or she is likely to be open to criticism.

From October 2000 it became lawful to make only a computer entry in the NHS record and to abandon the manuscript entry. This was agreed, providing the health authority approved. Approval is only given when the system in use conforms to Registration for Accreditation 99 (RFA 99) standards and the doctors sign up to the "good practice guidelines for general practice electronic patient records".
It may be helpful to be aware of some of the pros and cons of the EPR and the table below lists these:

There are several areas of clinical risk. The most serious is the inability to transfer data between practices; the second is the limitation of the computer screen for viewing data.

Transfer of notes between doctors

• The ability to transfer medical information between doctors is paramount in the reduction of risk and the safety of the patient.
• At present, the various computer systems are not intercompatible and information has to be entered anew at each registration or patient transfer.
• However "GP2GP", which is up and running for EMIS practices, allows transfer of medical information between primary care practices using the same system.
• Urban patients move practices frequently and currently all that follows the patient is a print-out from a previously used system. This can be illegible due to cost saving on printer ribbons and it is not uncommon to have two or three such print-outs enclosed with the notes. Information from each print-out would need to be entered on to the new system where a patient may remain registered for a short while only.
• Clinical negligence is much more likely to occur when the full history of the patient is not easily available and this particularly applies to the itinerant population at high risk of illness.
• Standardisation of data transfer, which allows transfer and immediate integration of computer data between practices is now essential and long overdue.

Viewing notes speedily

Information fatigue:

• Just as the old Lloyd George notes became unmanageable, so huge amounts of data are even more difficult to view and recall on a screen.
• We might imagine having to read files by turning pages on a computer and not being able to skip back or forward quickly, insert a sticky marker or make a margin note.
• It is still much quicker visually to scan a pile of hospital letters rather than to call each one up on a screen and consider it. Once the screen has flipped, the preceding record has gone and comparison is only possible when the records are printed out.
• Many systems produce displays that do not easily provide an overview or allow the reader to personalise the viewing. Those that do are often overloaded with information and require frequent use of the mouse to obtain important data.

Until this is solved, the correspondence sections of the records in the Lloyd George envelopes are probably a safer depository of information than scanned records. Scanned records are easy to recall for the most recent events and correspondence but not easy to visualise a complex patient history.

Computer usage in general practice poses several particular problems for those working in clinical negligence.
Knowing what information may be available, recognising that the standard of practice has been changed and finding new solutions to old issues, makes knowledge of the EPR vital for clinical negligence lawyers.

The whole record and nothing but the whole record

• Solicitors now need to request not just the entire record but to be told which software is being used and to what extent the system is being used by the practice.
• The record request will need to state that a copy of every scanned item is required as well as a complete report and all the paper records. Already cases have arisen where a paper record was not entered on a computer system and was negligently overlooked.
• One system in common use provides a comprehensive report to include repeat prescribed medication but does not print out when each repeat prescription was issued. An additional report has to be requested and the request should emphasise that full prescribing detail be provided.
• On receipt of the print-out, it is vital to check that every page is legible and that the margin dates are copied fully.

Bolam standards

• As primary care changes, so does the standard of reasonable practice by which general practitioners are judged. It is a judge who must decide what is negligent practice; however, with different general practitioners all using different systems, there is less standardisation.
• Experts will find it more difficult to give an opinion without knowledge of the system in use and what information was presented to the doctor by the computer system.
• Experts may be required to have some working knowledge of the systems and of how the data are entered - in particular, whether the system offered warnings to the doctor which were available and should have been heeded. For instance, a doctor working a system which constantly reminds about an overdue smear and then fails to act, has less defence than a doctor whose recall system is entirely practice nurse-based.

There are also some well-recognised issues in clinical negligence that will now present differently.

What notes would the responsible practitioner have seen?

• When reconstructing a series of consultations it has frequently been necessary to give opinion as to which of the previous consultations a practitioner would have read or considered.
• With the paper record, a doctor is more likely to have read the penultimate consultation if it was on the same page and less likely if a crucial consultation occurred some entries back in the record. In general, when an entry is made immediately below another, the preceding entry cannot be overlooked.
• With the EPR, some management screens will present the doctor with a problem list and checking prior consultation entries has to be done by a series of mouse clicks. In others, the previous free text is displayed automatically before any new entry can be made.
• For instance, a doctor using the EMIS system is presented with a page of the previous encounters before being able to add the next encounter.
• A doctor using INPS Vision system could enter a new consultation without ever having to view the previous text entry and could be relying entirely on the history given by the patient.
• In surgeries where some doctors use the paper records and some the computer records, doctors are obliged to try and check both. During a 10-minute consultation, checking and integrating events leaves less time for clinical assessment.
• Barristers rightly will be asking what information each system presented to the doctor, and would the ordinary and competent doctor check the previous record in these circumstances.
• We can assume no longer that an adequate past history is on the screen in front of the doctor.

Have the notes been altered?

• Suspicious handwriting and changes in ballpoint pen or ink were often the means by which fraudulent entries were detected in the paper record. The computer record can more easily be altered but it leaves an audit trail.
• Detection of the sort of changes made by Dr Shipman needs a high index of suspicion and also the technical skill required to identify the audit trail.

Different computer - different style

People driving a sports car will be tempted to speed, to cut corners and will have a high accident risk; hence, they pay higher insurance.
To some extent, using a computer in the surgery can be analogous in that a system that encourages a structured approach to making a record may take longer to use but will produce a better record and less risk of medical error.
The LV5 version of EMIS, for example, encourages a separate entry for History, Examination, Comment and Problem; whereas, some systems allow the doctor to enter the entire consultation under one free text area which, although quicker, is less likely to produce a safe structured entry. The former will be more likely to provide a sound defence.
I have no doubt that, within ten years, some systems will be identified as much more liable to medical accident than others.

Automated warnings

• Computers don't just keep consultation records, they also provide warnings about prescribing and, in certain circumstances, a doctor has to override a warning in order to issue a prescription.
• There are also programmes that support and add to the computer system in use - for example, the NHS's Clinical Knowledge Summaries, which will prompt drug choices in certain circumstances and will also provide warnings of errors, such as prescribing incorrect doses for the age of the patient.
• Systems frequently carry additional clinical advice, such as Mentor, the Oxford Handbook of Clinical Medicine, the British National Formulary and Electronic MIMS (EMIMS), which can all be installed, as well as other sources of information from Grays Anatomy to the local path lab list of normal values.

Accountability and length of consultations

With many systems it is possible to know how long the patient was waiting and how long each patient spent with the doctor. Most clinical negligence lawyers are already familiar with this information from the out-of-hours services and know how helpful this is for deciding whether a thorough assessment has been made.

The doctor as typist

• Typing skills are now essential for doctors. Data entry slows up those who do not have them and enables those who touch type to have time in the consultation to spend in patient contact or to use the computer more fully.
• Medical school training cannot be expected to take account of each system in use and some doctors will be working with systems with which they are unfamiliar.
• The question may be put as to whether the doctor has adequate skill to operate the system he is working and the training to find the appropriate information.
• Non-principals or locums cannot be expected to know precisely how each system has been used and where information is held on the system. Without full training the potential for locum doctors to err is greatly enhanced.

Out-of-hours

• Patients are now more likely to be seen by nurse practitioners at a walk-in centre and duty doctors at a primary care out-of-hours centre and, before doing so, to have been triaged by a message handler, triage doctor or nurse or to have been attended to by a paramedic.
• Almost all of these contacts are now subject to electronic records, many of which are then supplied in paper form to the patient's doctor. This represents a huge problem for scanning, storage and summarising, which is addressed differently in virtually every practice. The extent to which the GP record can be considered complete is dependent upon how much information is entered.

• The EPR has the potential for safer prescribing, clear legibility, easier presentation of information to the doctor and better-constructed consultations. Good systems will minimise clinical risk and encourage good practice.
• On the other hand, there is a serious problem with lack of ease of data transfer, information overload, the difficulty in using an electronic medium and the fact that patients could see doctors as more interested in their machines than in their patients.
• Poor systems make it difficult to view history and essential data, and discourage the construction of a good record.
• All systems could be safer than they are and only time will tell whether the EPR is to be yet another factor in the ever-rising tide of clinical negligence or a means by which clinical risk is minimised.

Dr Adrian Rogers LRCP MRCS MB BS D Obst RCOG MRCGP wrote this article and Medical Litigation kindly granted permissions to publish within Mentor. [First published in Medical Litigation April 2001] ©Medical Litigation and Dr A Rogers 2007.

1. House of Commons Health Comittee; The Electronic Patient Record; Sixth Report of Session 2006-2007.

• DH; NHS Executive - Electronic patient medical records in primary care. Changes to the GP Terms of Service. (2000)
• DH; Primary Care Computing - updated April 2006
• NHS Executive Letter 1/10/00 Electronic Patient Medical Records in Primary Care
• DH; Clear rules set out for patients' electronic records; May 2005
• DH; Electronic patient medical records in primary care changes to the GP terms of service. 2000