more_vert
WITH World Patient Safety Day having come and gone in 2020, we reflect on the role that written communication between doctors plays in safe health care.
The Australian Commission on Safety and Quality in Health Care has a Communicating for Safety Standard that “aims to ensure timely, purpose-driven and effective communication and documentation”. In 2002, Tattersall and colleagues called for improved standards in doctors’ letters. Yet almost two decades later, we feel that many doctors’ letters still fail the Communicating for Safety Standard’s aim.
Why are doctors’ letters still an issue? Can we learn ways of improving them from the human factors literature? Are there other barriers that prevent our letters meeting the standard?
Dual purpose
Doctors’ letters are often poorly written because they are trying to serve a dual purpose: as communication to the referring doctor, and as a record of the consultation (here and here). If letters are the rivers of communication between the specialist and the GP, then long and unfocused dual-purpose letters are rivers littered with unfiltered data (much of it already known to the referring doctor) that turn the water into a murky mess.
Instead, these letters should be distilled by the crystal-clear purpose of informing the reader (the referring doctor) of the salient issues relevant to the consultation.
Humans factors – cognitive load and reading speed
Cognitive load theory informs us that the mind has limited ability to process large amounts of disorganised information. Therefore, it is not surprising that the literature points to the need for clarity and structure in the letter to foster readability (here, here and here). From a safety perspective, Hannawa’s research outlines five core communication competencies – sufficiency, accuracy, clarity, contextualisation, and interpersonal adaptation – as essential for high quality patient care.
In addition to cognitive overload, psychologists have shown that we overestimate our reading speed. In particular, non-fiction writing, extremes of age, and English as a second language may reduce reading speed and adversely affect comprehension.
We therefore advocate for a “less is more” approach and encourage spending the time to write shorter letters. In our opinion, reading a doctor’s letter should not take more than 3 minutes (20% of a 15-minute consultation). This is consistent with Rash and colleagues’ conclusion that the letter should be no longer than 350 words or two pages. Effective letters control the flow; ineffective ones are like tropical thunderstorms that flood the reader’s mind with unwanted and unnecessary information, and potentially break the dams of patient safety.
Timeliness
Quality of the communication also relates to the timeliness of the letter’s reception (here, here and here). A high quality letter is both readable in a short amount of time and arrives in time. Some factors affecting the arrival time are beyond our control (eg, postal delivery service, efficiency and skill of the typist, and fears around privacy that keep the fax machine industry afloat by precluding emails), but some are. Imagine that your letter is the water your colleague needs while they’re wandering the desert of community practice, then do what you can to ensure that the water gets to them as quickly as possible.
Templates
Many authors have called for the use of standardised templates (here, here and here), as have some colleges in Australia through association with Medicare item numbers (here and here). Based on i) our experience (as a specialist pain medicine physician [AYH] and GP [FTA]); ii) our reading of the evidence; iii) existing guidelines; iv) safe communication principles; and v) Engel’s biopsychosocial model, we propose that the doctors’ letter should be broadly structured into three sections (see Box).
Box: Letter template
| Formulation (“a comprehensive synthesis of the multiple elements that led to a patient presenting at clinic or hospital with a problem”:
· A short summary of relevant history, examination, investigations, and previous treatments relating to the presenting complaint · A summary of other medical comorbidities (including previous major surgery) · A summary of mental health diagnoses or absence of such diagnoses · A statement on polypharmacy (if present) and/or a list of medications · A statement on metabolic health with preference to use of the terminology waist to height ratio · A summary of psychosocial factors that might contribute to the presentation or be protective |
| Expectation:
· A short summary of the patient’s goals or expectations to emphasise a patient-centred approach and to give the patient a “voice” in the letter |
| Management plan:
· A numbered and unambiguous summarised list of recommendations that specify who is responsible for follow-up of each recommendation and that outline contingency plans in the event of deterioration |
The formulation
We believe that of the three sections outlined above, the formulation requires the greatest skill and deserves the most attention because it is often non-existent or poorly executed. We therefore offer a formulation template (utilising numbered lists summarising pertinent information and grouping information according to topic) to assist writers in synthesising a formulation.
The role of the formulation is to synthesise and summarise why a patient presents to the clinic. The first issue in the formulation should summarise the history, examination, investigations and previous treatments relating to the “presenting complaint” in a maximum of three to five sentences. In the pain clinic, it should be the pain diagnosis; in endocrinology, it might be diabetes mellitus; in colorectal surgery it might be sigmoid carcinoma. A summary of all other biomedical comorbidities (including previous major surgery) such as hypertension, ischaemic heart disease, osteoporosis etc, should then follow for completeness. This summary of medical conditions can help paint a picture of how frail or complex the patient is. For example, a patient with a history of only hypertension is far less complex than the patient with congestive cardiac failure, chronic kidney disease and hypertension.
The above constitutes the bare minimum we expect in a clinical letter, but to meet the definition of a “comprehensive synthesis”, we have identified four other considerations that warrant inclusion: mental health, obesity, polypharmacy, and social determinants of health.
According to the Fifth National Mental Health and Suicide Prevention Plan, almost half of adult Australians will experience mental illness in their lifetime. Mental health issues therefore deserve prominence as a distinct category. We thus propose that there should be a summary of any formal mental health diagnoses (eg, depression, personality disorders, or post-traumatic stress disorder) or absence of such diagnoses.
In our experience, obesity and inappropriate polypharmacy are common but infrequently mentioned in the letter despite posing significant population health burdens (here and here). We therefore propose that a statement on polypharmacy (allowing the writer to summarise the current medication list and previous medication trials) and metabolic health be included as a cognitive aid for these issues to be included in the management plan. And on the subject of metabolic health, we prefer the “waist to height ratio” nomenclature over body mass index, and use it to avoid the term “obesity” because of its pejorative connotations.
“Social determinants of health” is defined as “the conditions in which people are born, grow, live, work and age”. We believe that a short summary of these conditions is in keeping with Engel’s biopsychosocial model and helps the reader better understand the patient’s context. In our formulation template, this final issue should summarise developmental and social factors that may contribute to the presentation. Both precipitating and maintaining negative factors (eg, adverse childhood experiences, long term unemployment, financial stress, social isolation) and positive protective factors (eg, family supports, high levels of education) should be described.
Medical education
We believe that a lack of prominence in medical curricula is partly to blame for suboptimal doctors’ letters. Few doctors receive formal training in formulation and writing letters, and letter writing education interventions can be expensive. However, in the context of competency-based medical education and patient safety, formulation and letter writing should be seen as essential skills – “entrustable professional activities”.
We propose evaluative judgement as a framework for improving the quality of doctors’ letters. This involves the use of self-assessment, peer assessment, feedback, rubrics and exemplars, and we have outlined above some suggestions that may contribute to these rubrics and exemplars. We also note that letter writing, by its nature, lends itself to teaching in an electronic environment such as a massive open online course (MOOC) or within a learning management system. Despite the role of MOOCs being unclear in medical education at this point, we believe that well designed, educationally sound, online learning activities may support evaluative judgement and improve letter writing practices by providing formative feedback.
Culture of autonomy
We also note that education may prove futile if culture does not change. Medicine has a culture of autonomy (here and here), and this culture may explain why some of us persist writing long dual-purpose letters that lack structure. We believe that patients and colleagues alike are the ones to suffer from this culture of autonomy – patients may experience harm because information may be missed, leading to clinical errors or misjudgement; and colleagues may feel disrespected by being forced to absorb large volumes of words that fail to make points precisely and accurately.
In an era of increasing patient complexity, a team-based approach, not autonomy, is required. This approach, to paraphrase Gawande, requires fewer “cowboys” writing long, poorly structured, dual-purpose letters that serve the cowboys’ needs, and more pit crew members writing short, well structured, single-purpose letters serving the patient’s and teams’ needs.
The future
In summary, the format, content and length of doctors’ letters should be evidence-based and tailored to a time-pressured audience. Good letter writing is an underappreciated skill that deserves greater prominence in the medical curricula. We believe that the introduction of electronic medical records may facilitate audit, big data research, and the use of artificial intelligence to provide feedback to the writer. Furthermore, investment in e-learning, medical education and translational research will play a role in developing methods for improving the standard of doctors’ letter writing in a cost-effective manner so that the next generation of writers, readers and patients may benefit.
Dr Andrew Yanqi Huang is an Anaesthetist and Specialist Pain Medicine Physician working at Austin Health and Eastern Health, and is a PhD candidate in the Department of Medical Education at the University of Melbourne.
Dr Ferghal Timothy Armstrong is an Addiction Medicine Specialist and GP who is about to commence Pain Medicine training.
Peter Mellow is a former University academic of 30 years, who is now Director of Learning and Teaching at Eastern Health.
The statements or opinions expressed in this article reflect the views of the authors and do not represent the official policy of the AMA, the MJA or InSight+ unless so stated.
Three comments from me:
1. As a group, we doctors have a good mastery of English but tend to use big words and long sentences when simpler plain English would work better. The US Military requires that all manuals are written to low readability score – ie, 8 or less years of formal education. That’s not because American soldiers are stupid: it is because the manual ,just be able to be understood instantly by a person under immense pressure. In medicine and nursing we call it the “3 am rule” — the procedure must be able to understood and acted upon promptly by a person in an emergency situation at 3 am.
2. Jargon and acronyms are used far too much in correspondence. We must’t assume that the language we use in our centre will be understood by the reader. Good example: PSA means prostate specific antigen to a GP and Patient Service Assistant to a hospital nurse or administrator.
3. Structured reports are very valuable, like the ISBAR oral handover routine. After receiving a few reports in uniform format, the recipient will now how to get to what they want to know quickly. It’s a pity Cerner is the dominant hospital electronic record system, because it makes the creation of templates and ‘boilerplate’ inserts very tedious.
I would be pleased if a lot of Specialists who I have referred patients to had bothered to write a letter at all. Professors seem to have been the worst offenders.
In my residency, the unit reviewed all discharge letters at the weekly audit.
Any letter longer than one page was sent back for editing before incorporation in the medical record.
Electronic medical records make it easy to copy-paste lots of information into a letter. However, as the saying goes, “rubbish in, rubbish out”. Although pre-set forms and structures can be useful for common presentations or referrals, nothing substitutes for up-front succint, timely and accurate information, pertinent to the purpose for the letter. For this, nothing is faster or more efficient than dictation. Even for this generation of native typers, navigating the multiple screens and security features of EMS inevitably slows it up.
My suggestion for hospital discharge letters is that the most senior team member who has cared for the patient dictates the “executive summary” of the discharge letter, then other staff – preferably clerical – append the details of test results etc.
For GP referrals to EDs or other specialists, I recommend dictating or typing an opening paragraph that summarises the reason and context for referral, with other information like medications appended. Just like we should do in verbal communication, the nub of the issue and what still needs doing should be clear and succinct, up-front.
It is my good fortune that my medical career escaped the worst ravages of the computer age. Now that medicine is universally computerised it is even more vital to have a concise, informative narrative of the clinical encounter.
The introduction to my edition of “The Economist Style Guide” states ” On only two scores can The Economist hope to outdo its rivals consistently. One is the quality of its analysis; the other is the quality of its writing” I can recommend no better tutor and exemplar in the art of writing informative letters than The Economist Style Guide or indeed, The Economist itself. In my own case, my secretary noted with appreciation that my letters consistently reduced to one page or less. I even received some un-solicited notes of approval from grateful referring doctors!
Terrific article, Andrew et al.
As a GP receiving hospital discharge letters, often written by someone who did not actually care for the patient and was not involved with decision making, who then uploaded everything from the patient’s electronic record into a discharge letter, it was frustrating to do the long read and still not find what was needed for a clinical handover back to the GP – but we feel lucky to get a letter and can’t complain. Often, the discharge medication list is even up to date!
Another nuisance is the use of acronyms throughout a discharge letter; as there is no standardised system of acronyms in medicine, these are confusing, and time consuming to google.
Finally, from an educational viewpoint, if writing hospital discharge letters is a learning activity for junior doctors, being able to synthesise the information and distil the clinical reasoning in a digestible discharge letter as you describe would increase the educational value.