20, 95% CI 0.06 to 0.33, n = 661) were poorly and positively correlated. Partnership building is the use of partnership statements, paraphrasing, and requests for patient’s opinion (Hall et al 1994). Interestingly, giving information to educate patients had a fair, positive correlation with satisfaction with consultation (pooled r = 0.28, 95% CI 0.04 to 0.48, n = 281), however, findings from individual studies were inconsistent for similar constructs, with r values ranging from –0.02 to 0.20 (Table 3). Individual studies

found fair to moderate correlations between verbal communication factors and satisfaction. The strongest associations were observed for use of negative questions (r = 0.30) to gather information; language reciprocity (r = 0.48) and expressions of uncertainty (r = 0.40) as facilitators; expressions of support and sympathy (r ranging from 0.19 to 0.58); listening (r = 0.27) and engaging (r = 0.22) to involve patients. LY2835219 in vitro They were reported to have a positive correlation with satisfaction with consultation (Table 3). Language reciprocity is the use of similar words by both the BMS 354825 patient and the clinician (Rowland-Morin and Carroll 1990), and expression of uncertainty is the direct and unambiguous expression of uncertainty (eg, use of the expression ‘I don’t know’) (Gordon et al

2000). Use of psychosocial questions (r = –0.15, 95% CI –0.29 to 0.00) and use of social niceties such as the expression ‘Thank you’ (r = 0.15, 95% CI –0.07 to 0.36) were not correlated with satisfaction with the consultation. Nonverbal factors: Pooled analysis was possible for four nonverbal factors employed by clinicians reported in seven studies (Bensing 1991, Comstock et al 1982, Greene et al 1994, Hunfeld et al 1999, Mead et al 2002, Smith et al 1981, Street and Buller 1987) (Figure 3). The nonverbal factors of length of consultation (pooled r = 0.30, 95% CI 0.08 to 0.49, n = 260) and nonverbal caring expressions of support (pooled r = 0.24, 95% CI 0.10 to 0.36, n = 197) had a fair, positive correlation with satisfaction with consultation. Showing interest as a facilitator

had a fair, positive correlation (pooled r = 0.23, 95% CI 0.05 to 0.39, crotamiton n = 127). Individual studies showed that the strongest associations were reported for discussing prevention (r = 0.53) (Smith et al 1981) and ability to decode body language, defined as the ability to understand patients’ nonverbal body language expressions except facial expression (r = 0.36) (DiMatteo et al 1979, Dimatteo and Taranta 1979, DiMatteo et al 1980). Positive associations were also found for ability to decode (r = 0.16) and encode (r = 0.30) tone of voice (DiMatteo et al 1979, Dimatteo and Taranta 1979, DiMatteo et al 1980) and shared laughter (r = 0.34) (Greene et al 1994) to facilitate and involve patients (Table 4). Use of nonverbal factors that appeared to avoid negative communication (r =-0.