Mariken Stegmann

A major strength of this study is that it provides important information that can be used by  all doctors to improve communication. Another strength is the quality of the item list, which  had large content validity because it combined data from a previous study and an extensive  literature search, and because participants were asked to suggest additional items. To  ensure comprehensibility of the item list, all items were discussed in several meetings, the  list was assessed among GPs and oncology specialists, and explanations of all items were  provided in writing to participants from the patient panel.  The panel quality was key to producing valid outcomes from this study, so we composed  each panel with great care by purposive sampling. This was both a strength and a limitation.  On the one hand, participants were recruited from across the Netherlands and sampled  based on a list of desired characteristics, with the experts showing a high level of consensus,  indicating a high probability that the study results could be generalised to cancer care in the  Netherlands and in countries with comparable healthcare systems. On the other hand, it is  possible that our cohort was not representative of all healthcare providers, given the  response rates of 42% and 41% respectively for GPs and medical specialists.  This study is the first to use an iterative approach to identify what information is perceived  as relevant and what is perceived as redundant in communication between primary and  secondary care. Given that administration, including writing letters, is a time‐consuming task  for doctors [19], it is important that the focus is on providing only that information  considered relevant to the recipient. Such an approach can enhance the effectiveness of  communication, lessen administrative burdens and potentially result in cost‐saving.  Conclusion  Our research focused on communication between primary and secondary care before and  during the initial treatment of cancer in the Netherlands. Using a consensus approach with  panels of GPs, oncology specialists and patients, we obtained a set of items that were  deemed relevant for inclusion in referral letters and specialist letters by all groups. The  recommendations of this study can be used to develop guidelines for the correct  composition of these letters and could be incorporated in computerised systems in both  primary and secondary care that generate letters from electronic patient records. Additional information  Funding information  This study was supported by a grant from the Stoffels‐Hornstra Foundation, Coevorden, the  Netherlands.  Role of the funding source  This funding source had no role in the design of this study nor had it any role during its  execution, analyses, interpretation of the data, or decision to submit results.  Conflict of interest  The authors declare that they have no conflict of interest.  90 Chapter 7