The communication of important clinical information in the management of our patients continues to occur in the large part through clinical letters, electronic or otherwise. Poor letter writing which misses important information and does not highlight key considerations leads to suboptimal care. This is certainly true in the crisis and home treatment team environment, where patients present with increased levels of risk and are often discharged solely to the care of their general practitioner (GP). Unfortunately, omission of information often occurs. A study of letters from mental health services in London found a mismatch between what GPs felt they needed to know and what was being included; for example, only 17% of the letters in this study included a diagnosis.1 Evidence suggests that structured letter formats take no longer to read, are preferred by GPs, improve comprehension of letters and reduce the risk of omission of information by the author and reader.2
Anecdotally, we observed variation in our medical correspondence to our GP colleagues and thus undertook a quality improvement project looking at improving the quality and consistency of letters produced by our team. This project was used as a developmental opportunity for one of our trainees under consultant supervision to learn more about the quality improvement process and to solidify trainees’ understanding of the psychiatric history, mental state examination and risk assessment as core psychiatric competencies. We determined clinical letter standardisation by including widely recognised components of the psychiatric history and the requirements of our service operational policy. We included diagnosis with ICD-10 code, formulation, risk assessment plan and capacity to consent to the care plan.
In phase 1, using our standardised template, we examined all existing letters from a randomly selected period (January and February 2018) within the service. Of these, 44% included past psychiatric history, 38% included past medical history, 19% included personal history and none included forensic history. Mental state examinations also contained omissions, with 62% documenting speech, 56% documenting thoughts and 63% documenting perception. From the perspective of a crisis and home treatment service, it was noteworthy that history of self-harm was included in 38% of letters, risk of suicide in 81%, harm to others in 50%, risk to dependants in none, and risk of self-neglect and vulnerability documented in 25% each.
The standardised letter template was developed collaboratively, communicated to and agreed with all medical practitioners of all grades in the service. It was universally implemented and allowed to be used for 6 months. After this we again assessed the content of medical letters, including all those produced for a second randomly selected period (September and October 2018). A significant improvement was seen. Following implementation of the template all aspects improved: past psychiatric history (96%), past medical history (92%) and forensic history (67%). It was reassuring that documentation of risk of suicide increased to 100%, self-harm to 96%, harm to others to 96%, harm to dependants to 80% and vulnerability to 84%. Regarding the mental state examinations, speech was documented in 92%, thoughts in 100% and perception in 96%.
Overall our results demonstrate that standardisation of doctor's letters continues to be an area for improvement within psychiatric services, but that relatively simple methods and collaborative efforts can lead to significant quality improvements. We hope our approach warrants consideration from the wider pool of colleagues as we meet our duties to improve the quality of services3 and especially commend engaging trainees in the process as a learning opportunity in an attempt to add as much value as possible.