Type of data	Themes
	Use of template	Establishing patient agenda	A biopsychosocial perspective	Agreeing health plans
	3D nurse review observations	Template intrusive, was followed closely and determined questions and structure. Nurses often explicitly referred to it to explain content of review. Data entry interrupted flow. Unfamiliarity with template slowed them down	Elicited wide-ranging patient concerns, explored in detail. Musculoskeletal concerns less likely to be explored. Often categorised simply as pain and mobility problems. Validation and prioritisation of patient’s agenda in most cases	Covered formally in every review because of template. Template questions about quality of life and PHQ-9 questionnaire could prompt exploration of psychosocial issues	Management plans removed from nurse responsibility, but some nurses negotiated actions concerning long-term conditions within their own expertise
Usual-care nurse review observations	Template structured the reviews but not intrusive as patients and nurses were familiar with it. Usually completed during review	Restricted to reason for review. Other unrelated problems occasionally raised but not explored in depth	Evident in many observations but mainly taking the form of social enquiry	Usual conclusion to review was to summarise actions agreed or confirm no change to management
3D GP review observations	Template mostly followed but in a more free-form way than by 3D nurses. Some overtly referred to template when checking review was complete and printing health plan. Data entry less intrusive than in nurse reviews. Template use sometimes not obvious	Varied in extent to which previously compiled agenda was used. Not all GPs explored problems on patient agenda because: they lacked expertise; old problem; nothing new to add; or considered not relevant. Some new problems were identified	In two-thirds of reviews there was evidence of in-depth understanding of psychosocial issues. In others, often where not obviously relevant to problems to be addressed, there was no evidence	Health plans agreed in almost all cases. Occasional patient suggestions but mainly GP suggestions agreed to by patient and all formulated by GP rather than patient. Written plan not always provided
Usual-care GP review observation	Had to rely on computer to look up information in patient record, which was time consuming	Patient wanted to talk about other problems, not LTCs. GP reviewed LTCs and medications at length, then addressed other problems	Not evident	Prescriptions given and stated when to review

LTC = long-term condition. PHQ-9 = nine-item Patient Health Questionnaire.