However, inconsistency in the association

was found when compared with the study reporting a correlation value (Paasche-Orlow and Roter 2003) (Table 3). The current study found that 38 of the communication factors investigated were associated with patient ratings of satisfaction with care and, for those factors for which correlation values were reported, most had a fair correlation. The number of potentially modifiable communication factors associated with satisfaction with care and the magnitude of their association partially support interventions of communication skills training valuing patient autonomy. Previous investigations of effectiveness of theory-based training of communication skills (eg, patient-centred care and

shared decision-making) have reported no effect on satisfaction with care (Brown et al 1999, Edwards et al 2004, Uitterhoeve et al 2010). It is possible that previous trials PFI-2 mw have tested interventions built on communication factors that are not evidence-based. Based on the results Proteasome inhibitor review of our review a small number of communication factors were found that could form the basis for intervention for communication skills training. However, those factors valuing patient autonomy were inconsistently associated with satisfaction with care (eg, verbal expressions valuing patient-centred care). Patientautonomy approaches involve a biopsychosocial perspective to understand patient’s experiences, share responsibility and develop relationships based on emotional support (Abdel-Tawab and Roter 2002). Our findings (eg, length of consultation, showing interest, and being caring) sustain the understanding of patients’ experiences and developing relationship based on emotional support rather than sharing responsibility. Interestingly Phosphatidylinositol diacylglycerol-lyase consistency found among verbal, nonverbal and interaction style for being caring shows that behaviours without speech content of emotional support should be also considered during the interaction. Over half of the identified factors in the current review (n = 75) were never associated

with satisfaction with care. We found fewer communication factors, and a weaker association with patient ratings of satisfaction with care, than reported in previous systematic reviews (Beck et al 2002, Hall et al 1988). The poor association seems unexpected for some communication factors used by clinicians, such as using psychosocial questions, using social niceties and smiling. Training protocols aimed at improving clinician communication skills proposed in the USA and recommended in health settings in other countries such as Honduras, Trinidad and Tobago, and Egypt emphasise the optimisation of these factors (Negri et al 1999). Based on the results of our study, training protocols and communication interventions should be checked for communication factors not likely to deserve attention.