Capturing pharmacists’ impact in general practice: an e-Delphi study to attempt to reach consensus amongst experts about what activities to record

Background In the UK, there is ongoing integration of pharmacists into general practice as a new healthcare service in primary care. Evaluation of the service involves national measures that require pharmacists to record their work, on the general practice clinical computer systems, using electronic activity codes. No national agreement, however, has been established on what activities to record. The purpose of this study was to attempt to reach consensus on what activities general practice-based pharmacists should record. Methods The e-Delphi method was chosen as it is an excellent technique for achieving consensus. The study began with an initial stage in which screening of a general practice clinical computer system and discussion groups with pharmacists from two ‘pharmacists in general practice’ sites identified 81 codes potentially relevant to general practice-based pharmacists’ work. Twenty-nine experts (pharmacists and pharmacy technicians from the two sites along with experts recruited through national committees) were then invited by e-mail to participate as a panel in three e-Delphi questionnaire rounds. In each round, panellists were asked to grade or rank codes and justify their choices. In every round, panellists were provided with anonymised feedback from the previous round which included their individual choices along with their co-panellists’ views. Final consensus (in Round 3) was defined as at least 80% agreement. Commentaries on the codes from all e-Delphi rounds were pooled together and analysed thematically. Results Twenty-one individual panellists took part in the study (there were 12 responses in Round 1, 18 in Round 2 and 16 in Round 3). Commentaries on the codes included three themes: challenges and facilitators; level of detail; and activities related to funding. Consensus was achieved for ten codes, eight of which related to activities (general and disease specific medication reviews, monitoring of high-risk drugs and medicines reconciliation) and two to patient outcomes (presence of side effects and satisfactory understanding of medication). Conclusions A formal consensus method revealed general practice-based pharmacists’ preferences for activity coding. Findings will inform policy so that any future shaping of activity coding for general practice-based pharmacists takes account of pharmacists’ actual needs and preferences.