Table 7 Summary of the results from ‘ intervention fidelity ’ studies included in the review (n = 33 studies)

Concerns regarding: Randomisation, i.e., not getting treatment Loss of control (re: allocation) Mistrust

• Ensuring minimal standard of care for control group or minimal intervention for controls [29, 40] or alternatives to randomised controlled trials [27].

• Use trusted sources of information (e.g. doctor or family) [87].

• Participant education regarding the benefits of randomisation [73].

• Including peers and locals in intervention delivery [20, 55, 58, 82, 105, 108, 124].

Health intervention not culturally appropriate (with community values and beliefs)

• Community involvement in development, action research method or ‘cultural immersion’ [27, 29, 31, 35, 56, 67, 70, 71],[74, 77, 80, 85, 107, 108, 113, 114],[118, 122].

• Culturally tailored programs [19, 29, 52, 58, 68–70, 85, 88, 118].

Distance for delivery of intervention (rural groups)

• Group-delivered (one-on-one) telephone intervention [88].

Negative framing of health information, emphasis on disparities

• Positive and progress emphasised in health information [37]*.