DTEXT – text messaging intervention to improve outcomes of people with type 2 diabetes: protocol for randomised controlled trial and cost-effectiveness analysis

Type 2 diabetes is predicted to be the leading contributor to the burden of disease in Australia by 2020 [1]. In 2014–15 it was estimated that one million Australian adults (5%) had type 2 diabetes [2], and if trends continue, this figure could triple to three million by 2025, due to factors such as high rates of overweight and obesity, low physical activity and poor lifestyle behaviours [3]. The majority of Australians with type 2 diabetes do not meet the recommended guidelines for optimum diabetes management [4]. Research shows that diabetes self-management, which includes management of symptoms, treatment, physical and psychosocial consequences, and lifestyle changes [5], can improve glycaemic control and reduce complications in people with type 2 diabetes [6].

Type 2 diabetes, if not managed well, can progress to complications such as coronary artery disease, stroke, kidney failure, limb amputations and blindness [7]. These complications lead to an increased burden of the disease for the health system and the individual [8‐11]. By reducing type 2 diabetes related complications, there would be a significant reduction in hospitalisations and the economic burden posed by diabetes on the health care system and people with the disease [12]. Modifiable lifestyle factors such as a poor diet and physical inactivity leading to overweight and obesity are driving the development of type 2 diabetes in the majority of cases [8]. A recent study showed that the majority of Australian adults with type 2 diabetes performed little or no exercise (66%) and had inadequate vegetable (93%) and fruit (51%) consumption [13]. The Australian Institute of Health and Welfare (AIHW) reported that only 55% of Australians with type 2 diabetes effectively manage their condition, and this figure reduces to 38% for Indigenous adults [14]. The AIHW also reported that only 18% of Australian’s complete the Annual Cycle of Care [13], the recommended best practice for the management of type 2 diabetes [7].

International studies suggest that attendance at diabetes education sessions can be as low as 30% due to logistical, medical, or financial reasons; or a perceived lack of benefit [15]. In Australia, it has been reported that over 40% of people with type 2 diabetes have no access to diabetes education programs [16], and only 24% of people saw a diabetes educator in the last year [14]. A study of Australian adults aged 18–39 years with type 2 diabetes showed that 59% chose not to participate in a structured education program with the perception that programs do not cater to their needs or concerns [17]. People not attending diabetes education have a fourfold increased risk of complications [18], therefore self-management of type 2 diabetes is critical to delaying or preventing disease progression and complications [19‐21].

The use of mobile phone text message interventions to complement existing health care is a rapidly emerging field [22‐24]. Text message interventions have the potential for high population reach; require minimal resources; and can be easily translated into the routine practice and existing services. In Australia, 96% of adults use a mobile phone [25], and it has been shown that usage is high amongst people from lower socioeconomic areas, rural and remote regions, the aging population and those with higher body mass index, health comprising behaviours and lower levels of self-rated health [26‐29]. These socio-demographic characteristics align closely with those having a higher prevalence of type 2 diabetes [30], suggesting that a text message intervention could be an appropriate method to support self-management of their condition.

Research has shown that text message interventions can improve the short term health of cardiac patients [31], help promote smoking cessation [32] and assist in weight loss [29] by providing prompts, information, reminders and support. Studies have also shown that text messaging can aid in lowering blood pressure [33], and improve medication adherence [34], physical activity [35], weight loss [36] and glycaemic control [24, 37, 38], however, these studies have been limited due to factors such as a short study duration, small sample size and lack of theoretical basis [39‐42].

A recent systematic review and meta-analysis by Faruque et al. 2017 for people with type 1 and 2 diabetes showed that text messaging interventions resulted in some improvement in glycaemic control, but not other clinical outcomes [43]. A systematic review by Dobson et al. 2017 reported that improvements in glycaemic control from text message interventions were inconclusive for people with sub optimal (HbA1c 7.1–8%) or poorly controlled diabetes (HbA1c ≥8.1%) [22]. Currently no evidence exists for a text message intervention to improve the health and support self-management of adults with type 2 diabetes in Australia.

This study aims to determine the effectiveness and cost effectiveness of a text message intervention (DTEXT) on improvements in glycated haemoglobin (HbA1c) and self-management for people with type 2 diabetes.

The randomised controlled trial will involve two parallel arms, an intervention arm and a control arm. The intervention arm will receive the 6 month text message intervention (DTEXT). Both arms will receive usual care from their treating doctor and associated health professionals.

The text message intervention was developed by an expert panel of clinicians, academics and health promotion staff using evidence-based guidelines and recommendations [7], and the use of appropriate language for Australian adults with type 2 diabetes [47, 48]. The content of the messages has a readability level of grade 5–6 (10–11 years of age) [49]. Each text message was limited to 160 characters in length, and no message was repeated during the intervention. The messages were tested with 20 people with type 2 diabetes in a 5 week pilot program to determine message acceptability and ensure message content, tone and structure was appropriate. Adjustments to the messages were made based on pilot participant feedback.

The DTEXT text messages were designed using Michie et al’s COM-B system from the Behaviour Change Wheel [50]. This system increases the individual’s capability, opportunity and motivation to enhance their self-management of health and lifestyle related behaviours, with the aim to improving diabetes control. Individual behaviour change is a critical component to achieve self-management of conditions [20]. The behaviour change techniques used in the text messages were based on the Behaviour Change Technique Taxonomy [51] and include: providing information about the health consequences, prompts and cues, problem solving, self-monitoring of behaviour and outcome, goal setting of behaviour and outcome, instruction on how to perform a behaviour, graded tasks, social support, social comparison, self-talk, credible source and reducing negative emotions. Behaviour change techniques are recommended for disease management programs using mobile technology [52].

The text message intervention will consist of daily text messages for months 1–3; and four messages per week for months 4–6. The messages will be designed to be sent at random times between 8 am and 5 pm, with message themes coinciding with relevant times of the day e.g. breakfast messages sent in the early morning, healthy dinner options sent in the afternoon. They include modules on nutrition, physical activity and sedentary behaviour, diabetes care and compliance with the diabetes Annual Cycle of Care, weight management, medication adherence and smoking cessation. The messages will be semi-personalised with occasional use of the participant’s name receiving the messages, occasional use of the researcher’s name (Karen) sending the messages, and tailored so that only smokers receive smoking cessation messages and only those taking medication will receive medication adherence messages (see Table 1).

A technology platform to deliver text messages will be developed using the software company Message Media. The platform will allow for automation of the text message intervention. The intervention will be designed as a uni-directional program, however participants will be able to reply to text messages with responses addressed by the study team, if required. Participants can withdraw from the study at any time using an automated opt-out STOP function.

The study will examine if the text message intervention, DTEXT, leads to improved glycated haemoglobin and self-management of type 2 diabetes compared to usual care. If successful, DTEXT will provide evidence to support scalable text message interventions as an appropriate tool to complement existing health care services and assist in improving the health of Australians with type 2 diabetes.

The findings from DTEXT have the potential to inform the New South Wales Ministry of Health on the scalability and translation of the intervention across the health care system. If the intervention is effective it could provide benefits on a population, individual and health service delivery level. These benefits could include: improving the population reach of diabetes education and support through an intervention that is equitable across location, culture and socioeconomic status; enhances diabetes self-management; and reduces diabetes related complications and hospitalisations with a low-cost, low resource intensive intervention that can enhance existing models of care.