Mariken Stegmann

Discussion  In the two‐round Delphi process, GPs, medical specialists, and patient representatives  reached consensus on lists for the most relevant items in correspondence between primary  and secondary care.   It was agreed that referral letters should contain medical facts and information about the  current problem, but that a detailed description of the physical examination and  investigation findings was not necessary. Information about social context, diagnosis and  treatment to date was considered unnecessary. It was considered that the specialist letters  should contain the same medical facts as the referral letters. Also, it was agreed that  diagnosis and treatment should be described in detail, and that items such as “what the  patient has been told” and “whom the patient should contact” should be mentioned in the  conclusion of the letter.  Interestingly, the panels considered that it was important to share information about  resuscitation plans in both the referral and specialist letters. Recent studies have indicated  that timely discussion of preferences for end‐of‐life care is important to improve quality of  life and care in this phase [16]. Our findings are in line with earlier research that doctors take  these discussions seriously and need to have clear information to do this effectively [17].  Although outside the scope of this study, it would be interesting to elaborate on the  responsibilities in this matter as perceived by GPs, oncology specialists and patients.  Concerning the content of specialists’ letters, the most striking requirement was for the  inclusion of detailed information, including the rationale, about diagnosis and treatment.  Earlier research by our group showed that detailed information about diagnosis is often  available in correspondence, but that key information about possible treatment options and  the justification is often lacking [14]. Similarly, it is typically the case that no explicit  information is given about whether the treatment is being done with palliative or curative  intent [14]. This is important because GPs are not only involved in providing care for patients  during and after cancer treatment [6,7] but also because they are expected to be formally  involved in that care [9]. After a patient has visited an oncology specialist, the GP is able to  answer questions that arise and to discuss the diagnosis and the treatment options after the  patient has digested the initial information. In this way, GPs can bridge the gap between the  patient and specialist by delivering care close to home, potentially allaying patient distress  and possibly even reducing the burden on secondary care.  In this study, we also asked the panels to indicate their preferences regarding the use of  undefined abbreviations in correspondence. We previously reported that abbreviations are  common in medical practice (e.g., "BP" for "blood pressure" or "abd." for "abdomen") [14].  However, some medical abbreviations can have different meanings depending on the  context and specialism (e.g., "OAC" for both "oral anti‐conceptive" and "oral anti‐ coagulant") and may cause confusion [18]. To ensure mutual understanding, abbreviations  should be avoided, even if their meaning is customary and obvious to the writer.  Although the study was performed in the Netherlands, we think that the results may also be  of interest to doctors in other healthcare systems. The care for cancer patients often  involves several doctors and it is important for them to share all important information.  Besides, our results have the potential to inform medical training, they can be used to  develop guidelines and they can help develop formats for letters generated from electronic  patient files.  7 89 Delphi consensus study of correspondence