Article Text
PDF
Statistics from Altmetric.com
Introduction
Every day, I write (or, more correctly, dictate) numerous letters. I also read letters from other doctors and healthcare professionals. Despite this, I do not recall any formal teaching about how to write a letter, nor can I recall reading any publications on the subject. This perhaps explains why so many clinical letters are hopelessly inadequate—too short (mainly doctors), too long (mainly nurses), too impenetrable due to jargon, some even too offensive.1 My own letters are of course perfect (!). So, what is the secret?
Why write a letter at all?
Although letters documenting clinical interactions between a patient and their hospital doctor are commonplace in the UK National Health Service (NHS), this does not hold universally. There are many countries where letters are seldom written. Perhaps these countries have otherwise immaculate record keeping, but I cannot comprehend how they operate successfully without good communication. I write a letter following almost every clinical encounter—not only after outpatient consultations and discharge summaries (to the patient's general practitioner (GP) and copied to other involved health professionals) but also after inpatient consultations, and telephone discussions with patients, doctors and others. This is a lot of work (especially for my secretary), so why bother?
I believe the reasons for writing are twofold. First, to impart information to other professionals, most importantly the GP, but also other professionals involved in the patient's care, and often the patient too. Second, to provide me with an essential summary of what I thought at the time on the next occasion I see the patient. Without my letters, I am lost in the follow-up clinic, and spend valuable time going over old ground, rather than catching up with new information—this explains my frequent rage at lost medical records although the evolution of electronic storage has helped. Some suggest that clinical letters may also serve as educational tools for patients and GPs—I am wary of this, as so often I am struggling to educate myself, let alone someone else. But if a letter is clear and precise, then any educational lessons should shine through. Finally, a typed clinical letter also represents an important medicolegal document, and hopefully might survive long after the hospital notes have been destroyed, or your illegible scrawls have long since faded with time.
I have for some time copied most of my letters to patients
To whom should you write?
Simply, anyone that is relevant. Most importantly, a hospital doctor should write to the person who referred the patient, most commonly the GP. This might seem obvious but it is a sad fact that some doctors omit the referring doctor. In my experience this most often occurs when I refer a patient to a specialist clinic. Here I do wish to be educated, and I am interested to hear how the patient with a brain tumour gets on in the long term, not just after the first visit. If you run a specialist clinic, remember to include the referrer, each time you see the patient. I am equally astonished to review patients who have suffered complications of their neurological disease which no one has thought I might be interested in. Orthopaedic surgeons pinning hips of Parkinson's disease patients are common culprits here, and one memorable epilepsy patient was surprised to discover that no one had told me of his near drowning experience, and subsequent 3 months in an intensive care unit, after falling into a canal after a seizure. And all this had occurred in my own hospital (each department has its own medical records rather than one universal set, which means it is impossible to know without the patient telling you, and not unreasonably, many patients assume that someone would have told us). I routinely ask patients at clinic visits which other clinics they are attending, and copy my correspondence to those clinics.
The evolution of electronic patient management systems is improving communication but few of us have the time in busy clinics to scroll through pages of computer script searching for any other medical interactions that might have taken place since we last saw the patient. The golden rule is to be over inclusive in to whom you copy your letters—after all, the recipients need only (confidentially) bin them if they are not interested.
Should you include the patient in your correspondence? In England and Wales, it is a legal requirement, albeit one that seems to be routinely ignored. While not a legal requirement in Scotland, I have for some time copied most of my letters to patients. There are some difficult areas; the patient with symptoms suggestive of early Parkinson's disease or motor neuron disease who you decide to watch, but without sufficient certainty to make a diagnosis for example. I used to think that copying letters to patients suffering from functional symptoms was difficult but now realise that this is the perfect situation to include them in correspondence (even if they do continue to insist no one has taken their case seriously or explained to them the diagnosis!). Copying letters to patients certainly concentrates one's mind about what should, and should not, be included in a letter.1 There may be specific issues in areas such as psychiatry2 but increased openness between doctors and patients can in general only be a good thing.
Do patients want to see their letters? This is simply too broad a question to be answered with a simple yes or no. Some do, a few do not, and many are quite indifferent. It is clear to me that many patients do not read the letters I send them and often forget that I have sent copies to them. I have had very few complaints, occasional corrections of fact (usually, but not always, trivial) and many more positive comments about copying letters. In my own limited personal interactions with health services, including spinal surgery, I have never been copied into correspondence, but I should like to have been. For patients in whom you are proposing potentially complex changes to their treatment (eg, epilepsy or Parkinson's), copying them in is only likely to enhance the chances of the changes occurring correctly.
Some doctors address their clinic letters primarily to patients, copying in the GP and other professionals.3 I have not adopted this approach. In the UK, the system we operate is that GPs refer to consultants for a consultation, and thus it seems appropriate that the results of our consultation are addressed back to the referrer, although I think it right to include patients in the distribution. Proponents of addressing clinical letters to patients note that it discourages medical jargon which is in some ways laudable, but also can potentially lead to rather ‘dumbed down’ letters. Medicine is technical at times, and while in favour of plain English, I would prefer to see ‘left L4/5 microdiscectomy’ listed in my letter rather than ‘slipped disc operation’. Other professions use their own technical language in communication with clients, as anyone who has had interaction with lawyers or architects will know, and it's educational to find out what ‘soffit’ or ‘inter alia’ actually mean.
What should a clinical letter contain?
Letterhead
The letterhead needs to be simple, uncluttered and contain the details of who you are, where you have seen the patient (especially vital if you work in more than one hospital) and how to contact you, including postal address, telephone and fax number (usually secretary) and ideally email address. Including the names of the other 24 neurologists who work (to a greater or lesser extent) in your centre, along with their areas of expertise, qualifications and secretaries, together with a colourful hospital logo is of no interest whatsoever to the reader, and simply takes up space, but this practice is remarkably common. The figure indicates my own letterhead.
Computer generated letter templates associated with electronic patient management systems are becoming increasingly common. My own experience of these is poor. In my area, they produce almost identical looking letters from any of three hospitals, which means GPs frequently address queries to the wrong hospital. Almost half a page is taken up with demographic details and we are forbidden (and technically unable) to customise them (which surely should be an advantage of such a system), so my colleagues' names clutter up the right hand margin. Remarkably, the system cannot cope with even simple formatting such as underlining, and a weird gremlin turns quotation marks into open boxes. Nor did the system include a spell checker until very recently.
Structure
Most clinical letters I have to read have no structure of any kind. Many are so short that structure is almost superfluous (ear, nose and throat doctors, this means you). Others are a prolonged stream of consciousness which rambles on interminably before suddenly halting at a seemingly random point (neurologists can be guilty here). Neither version makes for an engaging or informative read. At the other extreme however, some letters contain simply too many headings derived from a template, making it very difficult to build up a coherent picture. The figure provides the simple three layer structure that I use (history, examination, opinion). This might be customised for some specialties (ie, surgical letters might include operative procedures which are as important to surgeons as drugs are to us). My structure also provides an easy audit tool—if your examination section is routinely longer than your history section (surgeons take note), something is amiss. If your opinion section is routinely a sentence (or less), or simply a précis of what you have already written, then reconsider.
Diagnostic summary list
I started to do this after working for a consultant who used this system; it seemed so utterly obvious that I could not understand why so few others used it. I still do not, and even within my own department, I am in a minority, although heartened to note that many of my younger colleagues have seen the light. GPs will appreciate it (they are even busier than you, and receive numerous letters everyday, and do not have time to read every word). It also makes life much easier when you review the patient several months later. But the most important aspect is that it forces you to commit to paper what you actually think (see below). The diagnostic list is sometimes the most difficult aspect of the entire letter writing process, and it must be kept up to date on subsequent visits. If you take only one piece of advice from this article, let it be this one, and I shall be satisfied. If you use it, then the first item on the list should be the reason why you are involved, even if it might not be the patient's biggest problem. It should also include relevant (or potentially relevant) other medical aspects but should stop short of listing every single interaction with medical services over a patient's lifetime.
The diagnostic list is sometimes the most difficult aspect of the entire letter writing process
An example of how I write a clinical letter.
If you decide to ask (and do ask politely, rather than demand) the GP to do something, such as prescribe a drug or do a blood test, then put it clearly in bold at the beginning of the letter, and make it clear what you have told the patient to do; I usually tell the patient to contact the surgery when they have received a copy of my letter but not before. Anything that needs to be done urgently is best addressed by a quick hand written letter which the patient can deliver directly to their GP.
Content and length
It might seem obvious, but the content should be relevant to the problem in hand, and all my letters contain social details on first contact as well as the clinical features. Most of my letters are a page of A4 or less. Complex patients (common in neurology) of necessity lead to longer letters, but if you are frequently writing letters longer than two pages, then not many people will be reading them (psychiatrists and nurse specialists, this may be particularly pertinent to you).
Opinion
The most important section of the letter is the ‘Opinion’. Yet astonishingly, I frequently read letters which seem to avoid providing any kind of opinion at all (at least you can never be wrong with this tactic). You are being consulted for your opinion; if you do not provide one, you are missing the point of your existence. In this section, you should be explicit; what is your differential (and why), what tests, if any, have you ordered, what are your plans for follow-up and what have you told the patient.
Sign off
This is easy if you are a consultant, but a little more complicated if you are in training. If the latter, you must make it clear who you are working for either in the body of the letter, “I saw this man today in Dr Davenport's clinic …,” or by signing yourself off as Dr Junior, ST4 to Dr Davenport. The GP needs to know who the organ grinder is, although no doubt appreciates the monkey's effort.
Editing
I ask my secretary to prepare complex letters, or cases where I am contemplating (or even reading up about), in draft form. I always end up editing these letters to make them shorter, removing numerous unnecessary words and phrases, which emphasises the difference between the spoken and written word. There is therefore a case for every letter to be edited, honed to perfection, but my own life is too short and I suspect yours is too. If you do edit, do it yourself—get an electronic copy, fiddle away and send it back for printing. Nothing distresses secretaries more (apart from mumblers) than to receive a hard copy covered in ink, most of which they cannot read, and will spend an hour trying to decipher.
Also, if you are like me, and from time to time allow your agitation with those less perfect around you show in print, then having a human spam filter in the shape of your secretary is wonderful. My own secretaries are remarkably good at diplomatically suggesting maybe I might like to reword a certain letter, although after a decade as a consultant I am pleased to report this happens less often than it used to, a sign of my delayed but eventual maturation process.
Dictating
Most of us still dictate letters which are then typed by a secretary. As an aside, it alarms me that NHS hospitals increasingly think that typing pools (sometimes not even located in the UK) might be the way forward (ie, cheaper) and miss the fundamental point that one's secretary is so much more than a typist. But I digress. Dictation is another skill no one ever taught me, but is fundamental. Whether you dictate between patients or at the end is irrelevant, whatever suits your style, but you must learn to translate your thoughts into a coherent letter. I still cannot bring myself round to the idea of dictating in front of the patient4; like defecation, it is something to be done alone with just your thoughts for company. And you must learn how to speak coherently at the same time as thinking. Mumbling, coughing, pausing pregnantly, talking with a mouthful and not spelling out unusual words are all ideal ways in which to ensure alienation from your secretary (and you need her more than she needs you, never ever forget that). This is more of a challenge if English is not your first language, but many of the worst culprits in this area have no such excuse.
I frequently read letters which seem to avoid providing any kind of opinion at all
Grammar and English
The ability to write good English is a skill most of us struggle with. Such niceties no longer seem to feature highly on school curricula, and mastering spoken English as a second language is hard enough, let alone the written word. To read a well written letter is a rare pleasure in my experience, and to read a well written letter which is also clinically correct even more so. I cannot teach you this, and by the time you have qualified in medicine, further education in the use of words and grammar is unlikely if you are a native English speaker, which means I shall continue to feel the prick of irritation when you tell me that you are disinterested in cricket, your patient's illness course has been fluctuant or they are nauseous (unless you really do mean they make you sick).
Read your own letters
Ideally read them before they are posted, although I appreciate this is not always possible, and rapid communication is important. In the hospital that I attend only once a week, my letters go out unsigned, but I check all of them the following week, correcting errors and issuing new letters for the more serious mistakes. And do not just look for errors, think about how the letter reads; have you actually provided an opinion, is it clear what happens next and have you copied in all the appropriate participants?
Conclusions
You will by now have realised that good letter writing is one of the larger bees in my rather crowded bonnet, but I think this one is justifiable. It is an area where little attention has been paid, which is why there are so many awful letters out there. I hope that this article highlights some of the issues, and maybe I might have persuaded some of you to alter, or at least reconsider, your style of letter writing?
Acknowledgments
This article is dedicated to my two secretaries, Blossom Martis and Linda Milne, who cope heroically with all that I throw at them, as well as those who have gone before them and had to deal with my enthusiasm for writing letters. I acknowledge the helpful feedback I have received regarding letter writing over the years from all sources, and in particular the suggestions made by Drs Meryl Peat and Keith Hopcroft regarding this article.
Footnotes
Competing interests None.
Provenance and peer review Commissioned; not externally peer reviewed.
Other content recommended for you
- Prospective assessment of patient directed outpatient communication from a patient and general practitioner perspective
- Write2me: using patient feedback to improve postconsultation urology clinic letters
- How to set up a clinical database
- Plasma α-synuclein predicts cognitive decline in Parkinson’s disease
- Letters to patients: sending the right message
- Transitions in Parkinson's disease in primary care: protocol of a longitudinal mixed methods study
- The management of motor neurone disease
- Open questions on the nature of Parkinson’s disease: from triggers to spreading pathology
- Comparison of activities of daily living impairments in Parkinson's disease patients as defined by the Pill Questionnaire and assessments by neurologists
- Female gender doubles executive dysfunction risk in ALS: a case-control study in 165 patients