JMIR Publications

ABSTRACT

Background:

With the growing burden of non-communicable diseases (NCDs) in low- and middle- income countries, the World Health Organisation recommended a stepwise approach of surveillance for NCDs. This is expensive to conduct on frequent basis and using interactive voice response mobile phone surveys (IVR) has been fronted as an alternative. However, there is limited evidence on how to design and deliver IVRs that are robust and acceptable to respondents.

Objective:

This study aimed to explore user perceptions and experiences of receiving and responding to an IVR in Uganda in order to adapt and refine the instrument prior to national deployment.

Methods:

A mixed methods design was used, comprised of a locally translated audio recorded IVR survey delivered in four languages to 59 purposively selected participants mobile phones in five survey rounds guided by data saturation. The IVR had modules on socio-demographic characteristics, physical activity, fruit and vegetable consumption, diabetes and hypertension. After the IVR survey, study staff called back participants and used a semi-structured interview; to collect information on the participant’s perceptions of IVR call audibility, instruction clarity, interview pace, language courtesy and appropriateness, the validity of questions, and the lottery incentive. Descriptive statistics were used for the IVR survey, while a framework analysis was used to analyse qualitative data.

Results:

The key findings that favoured IVR survey participation or completion included preference for brief surveys of 10 minutes or shorter, preference for evening calls between 6pm and 10pm, preference for courteous language, and favourable perceptions of the lottery type incentive. While key findings curtailing participation were suspicion about the caller’s identity, if the voice was unclear, skip patterns were confusing, difficulties with phone interface for the IVR interface , such as for selecting inappropriate digits for both ordinary and smart phones, and poor network connectivity for remote and rural participants.

Conclusions:

IVR surveys should be brief as possible and considerate of local preferences to increase completion rates. Caller credibility needs to be enhanced through either masking the caller, or prior community mobilisation. There is need to evaluate the preferred timing of IVR calls, as the finding of evening call preference is inconclusive, and might be contextual.