Of the included studies, 24 used cross-sectional and 3 used longitudinal designs AZD8055 (Table 1). The most commonly investigated clinicians were physicians (n = 24 studies) and included inhibitors studies used videotape, audiotape, observation and surveys to collect information on

verbal, nonverbal and/or interaction style factors (Table 1). The studies also used a variety of tools to code both communication factors and satisfaction. The most frequently used tool was the Roter Interactional Analysis System used in 8 studies (Gilbert and Hayes 2009, Gordon et al 2000, Graugaard et al 2005, Hall et al 1994 studies I and II, Hall et al 1981, Mead et al 2002, Paasche-Orlow and Roter 2003). Quality: The most common methodological flaw of included studies was lack of appropriate statistical adjustment for confounding factors. In general, included studies also failed to report whether the coder was aware of prognostic factors at the time of outcome assessment ( Table Luminespib in vivo 2). No longitudinal analysis investigated the association between communication factors

and satisfaction with care such as symptom relief. Therefore all the data obtained by the review were from cross-sectional analyses. In total, 129 communication factors were identified in the review, 75 (58%) of which were not associated with satisfaction with care. Correlation values were reported for 108 of the 129 identified communication factors. Association between communication factors and satisfaction with the consultation was investigated for 106 factors of those 108 reporting correlation values. They have

been categorised into Digestive enzyme verbal factors, nonverbal factors, or interaction style. Verbal factors: Pooled analysis was possible for seven verbal factors employed by clinicians reported in nine studies (Bensing 1991, Comstock et al 1982, Hall et al 1994 studies I and II, Paasche-Orlow and Roter 2003, Putnam et al 1985, Smith et al 1981, Stiles et al 1979, Street and Buller 1987) (Figure 2). Use of closed questions to gather information as a facilitator of communication was poorly and negatively correlated with satisfaction with consultation (pooled r = –0.10, 95% CI –0.18 to –0.01, n = 574). Verbal expressions of empathy had a fair, positive correlation (pooled r = 0.21, 95% CI 0.09 to 0.33, n = 253) and psychosocial talk (pooled r = 0.15, 95% CI 0.05 to 0.