Text message–based intervention to improve treatment adherence among rural patients with type 2 diabetes mellitus: a qualitative study

Text message–based intervention to improve treatment adherence among rural patients with type 2 diabetes mellitus: a qualitative study

p u b l i c h e a l t h 1 6 3 ( 2 0 1 8 ) 4 6 e5 3 Available online at www.sciencedirect.com Public Health journal homepage: www.elsevier.com/puhe ...

NAN Sizes 0 Downloads 0 Views

p u b l i c h e a l t h 1 6 3 ( 2 0 1 8 ) 4 6 e5 3

Available online at www.sciencedirect.com

Public Health journal homepage: www.elsevier.com/puhe

Original Research

Text messageebased intervention to improve treatment adherence among rural patients with type 2 diabetes mellitus: a qualitative study X.Z. Chen a, S.J. Yu a, C.Y. Li a, X.X. Zhan a,b, W.R. Yan a,* a

Department of Epidemiology and Biostatistics, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, 430030, PR China b School of Nursing and Rehabilitation, Xinyu University, Xinyu, 338004, PR China

article info

abstract

Article history:

Objectives: Adherence to treatment among most type 2 diabetes mellitus (T2DM) patients is

Received 1 February 2018

relatively poor in rural China. The present study aimed to explore the perspectives of rural

Received in revised form

T2DM patients and health workers on a text messageebased intervention (TMI) for

20 June 2018

increasing patients' adherence in rural China.

Accepted 22 June 2018

Study design: Qualitative study. Methods: Six focus group discussions with T2DM patients, six with village doctors, and three with public health physicians were conducted in Xianning city during 2015. Semi-

Keywords:

structured interview guides were employed to facilitate qualitative data collection. Audio

Type 2 diabetes mellitus

recordings of the sessions were transcribed verbatim, and theme analysis was performed.

Treatment adherence

Results: Based on the participants' reports, T2DM patients had insufficient knowledge about

Text messageebased intervention

diabetes and suboptimal adherence to treatment in rural China. Most of the participants

Qualitative study

had a positive attitude toward this novel TMI approach to improving patients' treatment adherence and knowledge. The perceived potential barriers to the utilization of TMI included poor eyesight and educational background and gradually losing interest during a long-term intervention. The suggestions for successfully implementing this strategy included family or social support, applicability of the text message content, adequate frequency and timing of sending the messages, and combining of messages with other educational formats. Conclusion: A TMI is a promising option for improving T2DM patients' adherence to treatment in rural China. The findings of the present study can contribute knowledge to the application of TMI in similar settings. © 2018 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.

* Corresponding author. Department of Epidemiology and Biostatistics, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, No. 13 Hangkong Road, Wuhan, 430030, PR China. E-mail address: [email protected] (W.R. Yan). https://doi.org/10.1016/j.puhe.2018.06.016 0033-3506/© 2018 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.

p u b l i c h e a l t h 1 6 3 ( 2 0 1 8 ) 4 6 e5 3

Introduction Diabetes mellitus (DM) is currently the leading cause of chief vascular complications such as heart disease, blindness, and lower extremity amputation.1,2 The International Diabetes Federation3 estimated that the number of patients (aged 20e79 years) with diabetes was 109.6 million in 2015 and was projected to reach 150.7 million by 2040 in China, and this high prevalence was accompanied by approximately $51 billion in diabetes-related health costs in 2015. Type 2 diabetes mellitus (T2DM) is the most common type of diabetes, accounting for 90e95% of DM patients. A meta-analysis over a wide year range (2000e2014) estimated that the prevalence of T2DM in rural and urban areas in China was 8.2% and 11.4%, respectively.4 Although its prevalence in rural areas in China is lower than that in urban areas, it is growing very fast. Between 2001 and 2006, there was a 4.5% increase in the number of residents with T2DM in rural areas whereas there was only a 3.7% increase in urban areas.5 In addition, the risks of all-cause mortality and complication attributed to DM were higher in rural areas than in cities (rate ratiorural 2.17 vs rate ratiourban 1.83, P < 0.05).6 The implementation of effective strategies for preventing and controlling the occurrence and progression of T2DM in China, especially in rural areas, is urgently needed. Previous studies7,8 have shown that poor adherence to treatment is the main hindrance to reducing the health costs and mortality due to T2DM, as poor adherence is strongly associated with an increased risk of diabetes-related hospitalizations and complications. Several studies9,10 reported that the rates of adherence to treatment among T2DM patients in China ranged from approximately 45.4%e52%, which are inadequate for producing good outcomes. Treatment adherence is positively associated with health knowledge and financial and education levels,7,11 suggesting that rural patients are more vulnerable to suboptimal adherence, as they have a lower level of health knowledge, lower income, and lower education level than their city counterparts. To improve patients' medication adherence, various health promotion strategies have been conducted in recent decades. Traditional intervention formats, such as in-person or face-toface counseling and group teaching, could improve adherence to treatment among T2DM patients. In Mainland China, these kinds of interventions on treatment adherence have also shown positive impacts among DM patients. For example, a pilot study12 conducted in the city of Wuhan in central China suggested that the levels of knowledge and fasting blood glucose were improved among T2DM patients who received group teaching education conducted by medical professionals for approximately about 6 weeks. In Zheng's randomized controlled trials,13 adherence was significantly improved among patients in the intervention group who received oneon-one individual education and a behavioral intervention conducted by healthcare experts compared with the control group who received usual care. However, these intervention formats have several limitations in their ability to be implemented in rural areas, such as the shortage of qualified trainer and funds, inconvenient transportation, and time constraints.14 Innovative interventions using modern information technology, such as web-based education and text

47

messageebased intervention (TMI),15,16 can circumvent some limitations of traditional formats because of their flexibility and cost-effectiveness, and they can also be serviced as not only a tool for delivering diabetes-related knowledge but also a reminder for taking medication or appointments. It has been reported that TMI could improve adherence to DM therapy and clinical outcomes in low-income areas.15,17 For example, findings from a randomized controlled trial17 demonstrated that TMI had better potential to improve treatment adherence than the control condition of usual care. Additionally, TMI to support diabetes education resulted in a more significant improvement in compliance among patients with T2DM than did face-to-face education programs.18 Two theoretical frameworks, the Health Belief Model (HBM)19 and the CommunicationePersuasion Model (CPM),20 could explain the theory of this TMI strategy. In the HBM, Janz stated that participants' perceptions of disease, individual modifying factors, and cues to action may facilitate the development of healthy behaviors. The model of improvement for treatment adherence via text messages employs this theory to increase patients' knowledge about DM, correct their misperceptions, generate healthy actions, and improve adherence to treatment through delivered health messages. Although the HBM could help patients develop healthy behaviors, in the CPM, McGuire explained that if a new behavior was not reinforced, it would gradually be abandoned by individuals. Hence, in the strategy of improving treatment adherence via text messages, text message communication could be used as reinforcement to consolidate health behaviors and then keep optimal compliance in the long term. In China, the high frequency of cell phones use (approximately 90% of rural residents using one)21 provides an excellent opportunity for TMI to easily reach T2DM patients. Furthermore, the expenditure of delivering a message is only ¥0.1 RMB, and receiving messages is free in China. Thus, using TMI to improve adherence is potentially feasible because of its convenience and low cost compared with other strategies.16 Before successfully designing and implementing a TMI strategy, attaining a better understanding of key stakeholders' attitudes toward TMI and exploring the potential barriers that hinder the acceptance of TMI are critical. Thus, the aim of the current qualitative study was to investigate patients' knowledge about T2DM and their medical adherence, and the perceptions toward TMI among rural T2DM patients, village doctors (VDs), and public health physicians (PHPs) who are responsible for the health management of T2DM patients in rural China.

Methods Study settings, sampling, and participants This qualitative study was conducted in three counties of Xianning city in the southeast of Hubei province of China, from MarcheApril 2015. By the end of 2015, the total population of Xianning city reached 2.5 million, and the gross domestic product per capital was approximately 41381.57 yuan ranking ninth among 17 cities in Hubei Province.

48

p u b l i c h e a l t h 1 6 3 ( 2 0 1 8 ) 4 6 e5 3

In our study, the study participants consisted of rural patients with T2DM, VDs from the local village clinics, and PHPs from the local township hospitals. In rural China, VDs and PHPs are both in charge of the health management of all residents with T2DM, such as the follow-up of patients, delivery of healthy lifestyle knowledge, prevention of complications, and monitoring of blood sugar. The PHPs also take responsibility for managing and analyzing patient data collected from VDs to learn about the status of the T2DM epidemic and then guide VDs to conduct health intervention strategies for T2DM patients across the entire town. In addition, VDs are responsible for providing medical treatment other than health management. Rural patients usually collect and refill medications from their VDs in the local village clinic, which suggests that the VDs are very familiar with the medication compliance situation among rural patients. Furthermore, there are two reasons that we selected VDs and PHPs as the study participants, in addition to the T2DM patients, in this study. First, their inclusion provides broader perspectives on the medication adherence among rural T2DM patients and perceptions toward TMIs given their different roles in the management of T2DM patients in rural areas. Second, as health administrators, VDs and PHPs are both involved in the healthcare management of patients with T2DM and are familiar with their adherence to treatment; therefore, including them in the study can help us explore and design a more effective and efficient TMI addressing the barriers to adherence. There is a growing awareness that focus group discussion (FGDs) can be applied to investigate participants' attitudes toward intervention strategies and assess underlying barriers to the implementation of an intervention, which are not easily measured by a simple questionnaire survey.22 Additionally, this method could facilitate the generation of a hypothesis in the initial stage of intervention and be helpful in designing or tailoring the formats and contents of the intervention text messages.22 Thus, we employed FGDs in the current study. A multistage sampling method23 was used in this study. First, a total of three counties (Chibi, Chongyang, and Jiayu) with similar distances in different directions to the center of Xianning city and different economic levels (low, medium, and high) were selected for this qualitative study. Second, a random number generator was used to select two towns from each county. Third, five FGDs, each with 4e6 participants, were held in each county, including one FGD that consisted of rural T2DM patients, one FGD with VDs in each selected town, and one group of PHPs in each county. Finally, a total of 15 FGDs were conducted with 69 individuals who were purposefully selected. The inclusion criteria for the FGD participants were as follow: (1) T2DM patients who were over the age of 18 years, had a history of taking antidiabetic medications for at least 2 years, and were familiar with text messages; (2) VDs who treated and followed up with T2DM patients for more than 2 years; and (3) PHPs who had a working history of managing and followed up with T2DM patients for more than 2 years. The exclusion criteria included the following: (1) patients who had mental disorders or other serious illnesses who were unable to join the interview; (2) participants who did not give their informed consent; and (3) patients who did not have a cell phone. If eligible, the participants were informed of the locations and time of interviews.

Data collection Two semi-structured interview guides were developed by the research team to facilitate all the FGDs. The interview guide for the nine FGDs with VDs and PHPs was the same and focused on eliciting the viewpoints of the VDs and PHPs regarding patients' adherence and knowledge of T2DM and attitudes toward and suggestions for the TMI. The other interview guide employed in the six FGDs with rural T2DM patients was essentially the same with a few minor differences from that of the VDs and PHPs; the patient interview guide included questions about their understanding of T2DM, status of adherence, and attitudes toward or suggestions for a TMI. All of the FGDs were conducted by one researcher with extensive experience in facilitating interviews, which minimized the discrepancies in the influence of participants' sociodemographic characteristics on the interviewee's responses. At the same time, another researcher kept field notes on verbal and non-verbal communication in the group interactions among the interviewees. Prior to each interview, participant characteristics (gender, age, and education level), other than their name and other identifiable information, were collected. Subsequently, participants were encouraged to engage with other participants and articulate their viewpoints rather than address them to a facilitator, and they were informed that there were no ‘right’ or ‘wrong’ answers to minimize the bias from participants adjusting their views on a subject to align with popular opinion or hesitating to express their ideas for fear of making a mistake. During the FGDs, when disagreements emerged, the facilitator encouraged participants to discuss the inconsistencies, elucidate their perceptions and clarify the reasons for their actions. For patients with a low educational background, the facilitator encouraged and allowed them to share their points of view in dialect and avoided using medical terms in the interviews. To obtain additional information about research goals, the application of the interview guide was adjusted gradually based on the interviewees' perspectives. Six FGDs with patients were held in the meeting room at each local village clinic; six FGDs with VDs were held in a meeting room at each local township hospital; and three FGDs with PHPs were held at each county-level Center for Disease Control and Prevention. Each group interview lasted approximately 1e1.5 h. After 15 interviews had been completed, there was no further information obtained, and theme saturation24 was thus reached at 15.

Data analysis Grounded theory,25 a methodology of generating inductive themes that are grounded in systematically gathered and analyzed data, and the constant comparison approach25 were the principal analysis methods in the current study. Nvivo 8 (QSR International Pty Ltd, Doncaster, Australia), a qualitative computer software program, was applied to manage and code the data. All of the audio-recordings were transcribed verbatim, and the field notes were employed to supplement the records. Initially, the researchers read and reread the transcripts and hand-written field notes in detail, and then they listed

49

p u b l i c h e a l t h 1 6 3 ( 2 0 1 8 ) 4 6 e5 3

recurrent themes or concepts relevant to the study objectives. Subsequently, transcripts were coded into related categories using the method of grounded theory and constant comparison, wherein the researchers coded units of information to develop initial codes (open coding), reorganized initial codes into categories based on their similarities and differences (axial coding), and finally identified the coding framework through further categorizing and organizing the axial codes (selective coding). To ensure credibility and reliability of the analysis, transcripts were independently coded and analyzed by two researchers. Next, a consensus-based coding framework was developed based on comparing the similarities and difference of the two coders' results. Meanwhile, any discrepancies or disagreements were resolved by discussion with other research members until a consensus-based code was reached. Finally, the agreed-upon coding framework was then imported into Nvivo 8, which was employed to link and analyze the data.

Results We analyzed all of the data with the help of qualitative computer software, Nvivo 8, and three themes developed: knowledge about T2DM, treatment adherence, and attitudes toward TMI to supplement T2DM treatment. Table 1 presents the detailed demographic characteristics of the participants.

Theme 1: knowledge about diabetes Rural T2DM patients expressed that they have insufficient knowledge about diabetes. A few patients did not know that T2DM was a common chronic disease and even believed that the disease was doomed to occur with increasing age: ‘I really don't know it is a chronic disease. The reason why I am suffering from this disease is that I reached the age of sickness.’ (patient in

group 1, female, 69 years old). The vast majority of patients reported that they knew only a few clinical symptoms of T2DM, such as frequent urination and increased thirst and hunger, and they usually misunderstood some common complications, such as diabetic foot and vision disorders, to be the sideeffects of taking hypoglycemic medications: ‘The kidney disease that I suffered is the side-effects of taking oral hypoglycemic agents.’ (patient in group 1, male, 52 years old). Some of the VDs reported that due to insufficient diabetes-related knowledge, some patients usually did not initiate timely diabetic therapy, waiting until the disease developed into a late stage. As one VD said: ‘I had a patient who had little knowledge about T2DM. Unfortunately, he was dead due to the diabetic foot, as he didn't realize the importance of taking long-term medicines and therefore often missed dosages.’ (VD in group 2, female, 41 years old). Most of the PHPs also expressed that if patients had more knowledge of diabetes and antidiabetic medications and realized the long-term benefits of taking medications, they could have better adherence to the treatment and improve their therapy outcomes: ‘Some patients did not realize that the serious health outcomes of diabetes are due to their insufficient diabetes-related knowledge, and they often forgot to take medicines, lost control on of their diet, and seldom do did physical exercise.’ (PHP in group 3, male, 42 years old).

Theme 2: treatment adherence Most of the rural patients had difficulties keeping good adherence to treatment. The VDs and PHPs reported that old patients generally had better adherence to medications than young patients: ‘I found young patients had poorer adherent behavior compared with older patients as young patients usually had better physical conditions or more social activities, which made them often refuse or forget to take medications.’ (VD in group 3, female, 42 years old). Additionally, patients with a long duration of T2DM showed better adherence than newly diagnosed

Table 1 e The characteristics of the participants. Characteristics Gender Male Female Age (years) <31 31e40 41e50 51e60 >61 Educational level Illiterate Primary education Junior high education High/vocational/technical education College education and above Duration of T2DM (years) <5 5e10 >10 T2DM, type 2 diabetes mellitus.

Public health physicians (N ¼ 19) [n (%)]

Village doctors (N ¼ 24) [n (%)]

T2DM patients (N ¼ 26) [n (%)]

6 (31.6) 13 (68.4)

18 (75.0) 6 (25.0)

8 (30.8) 18 (69.2)

e (25.0) (20.8) (37.5) (16.7)

e e e 12 (46.1) 14 (55.9)

7 (36.8) 5 (26.4) 7 (36.80) e

6 5 9 4

e e e 16 (84.2) 3 (15.8)

e e 4 (16.7) 19 (79.1) 1 (4.2)

e e e

e e e

7 6 10 3

(26.9) (23.1) (38.5) (11.5) e

11 (42.3) 12 (46.2) 3 (11.5)

50

p u b l i c h e a l t h 1 6 3 ( 2 0 1 8 ) 4 6 e5 3

ones. When discussing the reasons for interruptions in adherence over the course of treatment, forgetfulness in taking or refilling the medications was reported as the major reason by patients: ‘I think, for me, the biggest hurdle was forgetfulness, especially when lots of farm works needed to be done’ (patient in group 1, male, 55 years old). In addition, some patients adjusted the dose of their medications depending on how they felt without consulting with their doctors. Additionally, as mentioned before, some patients misunderstood the complications of T2DM as being the side-effects of medication, which made them intentionally reduce their adherence to the medication: ‘I often find that some patients reduce medication doses in private due to the fact that they are worried about the side-effects of medicines. However, the side-effects of medicine that they talked about were mostly complications.’ (VD in group 2, male, 38 years old). When they were asked about T2DM therapy, most of the interviewed patients knew only that other than taking antidiabetic medication, it is beneficial to have a healthy diet, and they did not realize the importance of other healthy behaviors such as physical activity, behaviors that reduce the risk of complications, and the self-monitoring of blood glucose. ‘I only knew that eating high-cereal fiber food and taking medication on time were good for me. I thought regular monitoring blood sugar was useless and could not cure my disease’ (patient in group 1, female, 54 years old). Some VDs also reported that: ‘To my knowledge, most of patients never do any physical activities or self-monitor blood sugar in their daily life in rural areas.’ (VD in group 2, male, 42 years old). Furthermore, some PHPs reported that most patients had incorrect conceptions about physical exercise. As one PHP had reported: ‘Most patients have low health literacy in rural China. For example, some patients often thought of farm work as exercise.’ (PHP in group 3, female, 39 years old). When discussing the use of text messages as an intervention strategy to improve treatment adherence, most of the participants thought that this method would work if the delivered messages included knowledge about T2DM (the fatality, life-long nature, and seriousness of this disease), hypoglycemic agents, and self-management (regular physical activity, healthy diet, etc.) because they thought these kinds of messages would help patients learn why, how, and what to do to keep optimal adherence, realize the benefits of adherence, and then facilitate the development of these healthy behaviors: ‘I hope this program will give patients more knowledge about T2DM and self-management and then help them realize the importance of good adherence.’ (VD, male, 41 years old). Some patients thought that delivering frequent messages would make them not only gradually learn the importance of adherence but also consolidate new healthy behaviors: ‘While I know the importance of adherence to treatment, I only can persist for a short time. So, I think this approach would be beneficial for me to keep healthy actions through delivering frequent messages’ (patient in group 1, male, 52 years old). Additionally, most of the patients suggested that text messages could function as reminders for taking medication and attending appointments to address the barrier of forgetfulness: ‘it would be better to send reminding messages, like ‘do you take your medication today’. I thought it would be useful to overcome forgetfulness’ (patient, male, 53 years old).

Theme 3: attitudes toward TMI to supplement D2TM treatment Rural areas are covered by a telecom network, and the participants reported that most patients owned cell phones and are able to read text messages once they receive them. Therefore, it is feasible to implement the TMI among rural T2DM patients.

The acceptability of TMI among patients The participants expressed positive perceptions toward improving adherence via TMI and believed that this novel strategy could be a useful, informative, educational, and remindful tool: ‘We are certainly willing to accept this way, because it will provide us lots of knowledge about this disease and be of great benefit to our adherence.’ (patient in group 1, female, 60 years old). The convenience of TMI was reported as the most important advantage for using it, as the messages can be read at any time and place. Similarly, some of the interviewed patients thought that the ease of storing messages is another advantage of using TMI, as they can ‘keep the useful information in their cell phones and read repeatedly in the future’ (patient in group 1, male, 55 years old). However, there still were a few participants who held the opposite opinion (the reasons are described below).

Barriers to implementing the TMI During the FGD, most of the participants were willing to participate in TMI, but some potential barriers were identified by the interviewers, as follows.

Low education level or health literacy Most of the patients commented that they understood only some basic diabetes-related terms, such as polydipsia, polyuria, polyphagia, and emaciation when they discussed T2DM: ‘It is difficult for me to understand medical terms. Some professional jargon is beyond my understanding.’ (patient in group 1, female, 54 years old). A few patients with a low education level (primary school and lower) expressed their concerns about participating in this intervention as they believed that it could be difficult for them to understand the delivered messages. Some VDs and PHPs also expressed similar concerns that ‘for most rural patients, poor education background could affect the effectiveness of the proposed intervention since they couldn't fully grasp the meaning of messages delivered’ (PHP in group 3, female, 36 years old).

Poor eyesight Reading messages require good eyesight. However, the eyesight of patients is usually poor due to their increasing age and diabetic retinopathy: ‘Most older rural patients are suffering from vision disorder. So, poor eyesight would hinder them reading messages.’ (PHP in group 3, male, 42 years old).

Gradual loss of interest Most of the patients reported that they may gradually lose enthusiasm about reading text messages and then dropout before the end of the intervention. Some of the VDs had

p u b l i c h e a l t h 1 6 3 ( 2 0 1 8 ) 4 6 e5 3

similar concerns and reported that ‘the situation would be improved, if participants’ clinical outcomes could be obviously improved or the contents of messages can meet individualized needs' (VD in group in group 2, male, 43 years old).’ A high frequency of delivering messages (once or twice a day) and messages with a lot of jargon and monotonous content were perceived as factors resulting in a loss of interest and message fatigue among patients: ‘If the contents of text include lots of professional terms, which will be hard for me to understand and then make me lose interest for participation.’ (patient in group 1, female, 53 years old).

Suggestions for TMI To successfully implement this strategy, some suggestions or facilitators to overcome potential barriers were discussed among the interviewed participants.

Social or family assistance While poor eyesight and education backgrounds were perceived as barriers to reading the messages, the patients provided some useful suggestions. For example, a patient with poor vision and education level expressed that ‘although it was difficult for me to understand or read delivered messages, my family members and/or neighbors could help me read and explain these messages.’ (patient in group 1, female, 72 years old). Additionally, they also suggested that text messages could be delivered to the mobile phones of their family members: ‘You can deliver messages about diabetic knowledge to the cell phone of my husband or son. So, they will tell and explain me about related-information.’ (patient in group 1, female, 54 years old).

The applicability of text message contents Participants suggested that the contents of messages should be attractive, straightforward, and simple to understand, avoiding lots of massive jargon: ‘I think text should not have too much jargon. Moreover, it must let patients easily understand what it means, and it shouldn't contain many contents in one message.’ (PHP in group 3, female, 29 years old). There was a general consensus among the patients that it would be best if the contents of the messages included basic knowledge about T2DM, medication information, reminders for taking medication and/or attending appointments, and knowledge about self-care and self-management: ‘I would be interested in this intervention if the program delivered knowledge about DM as well as how to keep a healthy lifestyle in routine life’ (patient in group 1, male, 59 years old). At the same time, most patients reported that getting information from authoritative professionals would increase the likelihood of their participation. In addition, a few patients articulated a strong desire for a tailored message content and a two-way message format, an approach through which receivers could keep in touch with providers: ‘if this intervention uses the format of two-way message, I won't dropout of this program halfway because I can have interactions with providers, which would help to solve my problems’ (patient in group 1, male, 65 years old).

Frequency and timing of sending messages Most of the patients preferred to receive messages two or three times a week, and a few patients mentioned that they would like to receive messages every day. Most VDs and PHPs

51

also expressed that the frequency of delivering messages could be twice a week: ‘it is appropriate for patients to receive at most two messages every week. If you send too many messages in a week, patients may be bored with it and then ignore these messages.’ (VDs in group 2, female, 35 years old). With respect to the best time for delivering the messages, most of the patients preferred receiving messages before dinner when they had more time to read messages and could easily obtain assistance from their family members: ‘I want to receive text messages before dinner, because I can let my family members help me read the text. I think receiving a text once every other day is best.’ (patient in group 1, male, 62 years old).

Other education formats Some education formats, such as voice messages, phone calls, and health education videos, were thought to be useful complements for addressing the potential limitations of implementing this innovative strategy. For example, some patients suggested that ‘researchers could deliver health information via text message service combined with voice messages to address the problem of low educational level and poor eyesight’ (patient in group 1, male, 65 years old). One PHP provided a suggestion that ‘researchers could build online platforms such as WeChat or QQ groups for participants, which can provide timely answers to participants' question about this strategy’ (PHP in group 3, male, 39 years old).

Discussion The purpose of this article was to describe the state of adherence to treatment among patients with T2DM, to explore the acceptability of TMI used as a tool to improve adherence and to provide suggestions for further conducting the TMI in rural China. The present study found that knowledge of DM and adherence to treatment were generally poor among rural patients with T2DM in China. Regarding attitudes toward TMI, although there were some potential barriers to implementing TMI, most of the participants expressed positive attitudes toward this innovative strategy. Our study found that the current status of adherence to treatment among rural T2DM patients was suboptimal in China. Forgetfulness in taking medication and inadequate knowledge of T2DM were perceived as the main barriers resulting in poor adherence, and these findings are consistent with a previous study.26 In addition, it is worthwhile to note that better adherence to medication is strongly associated with optimal glycemic control and a decreased risk of diabetes-related complications.8,27 Therefore, it is critical to explore a strategy to overcome the aforementioned barriers. Previous studies28,29 have shown that TMI has promise to be a good approach to address these obstacles, as it has the convenience of usage and message storage; these results were in accordance with our findings. Most of the participants in our study expressed that they would like to receive messages with knowledge about DM and reminders for taking medication to improve their adherence, and these findings were aligned with previous studies.28 Nonetheless, several potential barriers to implementing this strategy in rural areas were identified and described by the participants in the present study.

52

p u b l i c h e a l t h 1 6 3 ( 2 0 1 8 ) 4 6 e5 3

It is a great challenge to maintain participants' interests in participation and prevent high dropout rates during the implementation of the TMI. A previous study28 found that participants easily lost interest, and then dropped out before the end of the intervention. For example, in Buis's retrospective study,30 the dropouts increased over the course of the program, with only 43.74% of participants completing the program. To overcome this barrier, participants in our study suggested that health information providers adjust the content of the delivered messages based on the needs of participants and employ a two-way message (interactive communication) format. An interactive communication format in which public health providers can directly answer participants' question regarding T2DM via text messaging might meet the individual needs of patients, helping increase their interests in the program and decrease the number of unnecessary appointments, thus saving participants' time and costs. According to the theory of CPM20 used in the model of TMI for improving compliance, the format of interactive communication could be used as reinforcement for maintaining healthy actions in the long term. The findings from a longitudinal study also stated that the content of tailored messages was developed based on the participants' concerns for improving treatment adherence (e.g. messages providing knowledge about DM), as well as preferences (e.g. the timing of the delivery of the messages), which could help participants feel that someone cared about them and thus improve their engagement.31 In Nelson's pilot study,32 participants were provided with personalized messages based on their feedback, and this approach led to a higher engagement rate (approximately 84%). Participants in our study also recommended that the researchers deliver messages at an optimum frequency to avoid message fatigue, which could also increase engagement rates. This result was in agreement with a previous study as well.33 It is possible that delivering too many messages during intervention would gradually make participants feel bored, ignore the messages, and then dropout halfway through. In addition to the aforementioned suggestions, the interviewed patients in our study reported that if the delivered information was from experienced endocrinologists or doctors, it would increase their interest in participation. This suggested that patients would be more willing to trust and adopt health information from authoritative professionals. Participants in the present study also expressed concerns about poor education background and eyesight among T2DM patients in rural areas, which would be barriers for patients in reading and understanding the delivered information.33 Nevertheless, the study participants still expressed positive attitudes toward conducting TMI and offered lots of insightful suggestions for addressing these barriers. For example, they suggested that the contents of messages should be simple to understand and straightforward and that researchers could employ a mixed-method approach in which TMI could be combined with other education formats such as health education videos, phone calls, or voice messages; these findings are consistent with a previous study.34 Future studies in this filed should pay attention to these suggestions. Furthermore, researchers should carry out more prospective studies to

explore the feasibility and effectiveness of TMI to increase treatment adherence among T2DM patients in rural China.

Strengths and limitations First, three key stakeholders (patients, VDs, and PHs) were involved in the present study, which could provide a broad range of detailed information about the status of treatment adherence and more perceptions about TMIs in rural China. Thus, the researchers could gain an in-depth understanding of the conditions among T2DM rural patients and tailor efficient text messages for future intervention projects. Additionally, a qualitative study has the advantages of easily understanding patients' experiences of a disease and their attitudes toward topics; this knowledge can help researchers identify potential barriers and find solutions to these problems. There are several limitations in this research. First, this study was conducted in Xianning city in the Hubei province of China, where the situation of treatment adherence and usage of cell phones is closely related to the economic level and geographic context, which may restrict the generalization of our findings to a broader area given the disparities in geographic and economic backgrounds. Second, as with all focus groups, there is the limitation that the interviews may have caused some participants to reserve their comments or adjust their viewpoints to follow popular perceptions. Efforts were made to minimize this bias, such as encouraging interviewees to describe their views in detail and asking participants who usually reserved their comments and attitudes about others' viewpoints.

Conclusions In summary, this qualitative study demonstrates that TMI is feasible and that it has the potential to be accepted as a useful and novel strategy to improve adherence to medications among T2DM patients in rural areas in China. However, several potential barriers to its success are identified, such as poor eyesight, low health education level, and gradual lack of interest, and researchers conducting future studies on this topic should pay attention to these barriers when analyzing TMI in rural China.

Author statements Acknowledgments The authors would like to acknowledge the support of the Xianning Center for Disease Control and Prevention and the project ‘the Fundamental Research Funds for the Central Universities’ funded by the Ministry of Education and Ministry of Finance of China.

Ethical approval This study was approved by the Institutional Review Board of Tongji Medical College, Huazhong University of Science and Technology. Informed written consent was provided by all of

p u b l i c h e a l t h 1 6 3 ( 2 0 1 8 ) 4 6 e5 3

the participants prior to the study. All of the recorded information was kept confidential.

Funding This research received financial support from the Fundamental Research Funds for the Central Universities [2014ZZGH014] funded by the Ministry of Education and Ministry of Finance of China.

Competing interests The authors declare that there are no competing financial interests in this study.

references

1. Gao Q, He B, Zhu C, Xiao Y, Wei L, Jia W. Factors associated with lower extremity atherosclerotic disease in Chinese patients with type 2 diabetes mellitus: a case-control study. Medicine 2016:95. 2. Sun X, He J, Ji X-L, Zhao Y-M, Lou H-Y, Song X-X, et al. Association of chronic kidney disease with coronary heart disease and stroke risks in patients with type 2 diabetes mellitus: an observational cross-sectional study in Hangzhou, China. Chin Med J 2017;130:57. 3. IDF. IDF diabetes atlas. 7th ed. 2015. http://www.diabetesatlas. org/key-messages.html. 4. Yang L, Shao J, Bian Y, Wu H, Shi L, Zeng L, et al. Prevalence of type 2 diabetes mellitus among inland residents in China (2000e2014): a meta-analysis. J Diabetes Invest 2016;7:845e52. 5. Liu X, Li Y, Li L, Zhang L, Ren Y, Zhou H, et al. Prevalence, awareness, treatment, control of type 2 diabetes mellitus and risk factors in Chinese rural population: the RuralDiab study. Sci Rep 2016:6. 6. Bragg F, Holmes MV, Iona A, Guo Y, Du H, Chen Y, et al. Association between diabetes and cause-specific mortality in rural and urban areas of China. JAMA 2017;317:280e9. € ffski O. 7. Breitscheidel L, Stamenitis S, Dippel F-W, Scho Economic impact of compliance to treatment with anti diabetes medication in type 2 diabetes mellitus: a review paper. J Med Econ 2010;13:8e15. 8. Ho PM, Rumsfeld JS, Masoudi FA, McClure DL, Plomondon ME, Steiner JF, et al. Effect of medication nonadherence on hospitalization and mortality among patients with diabetes mellitus. Arch Intern Med 2006;166:1836e41. 9. Wu P, Liu N. Association between patients' beliefs and oral antidiabetic medication adherence in a Chinese type 2 diabetic population. Patient Prefer Adherence 2016;10: 1161e7. 10. Zhou Y, Liao L, Sun M, He G. Self-care practices of Chinese individuals with diabetes. Exp Therap Med 2013;5:1137e42. 11. Ke L, Zhang Y, Wang X, Li S, Yang W, Tong N. Assessment of diabetes care and the healthcare system in economically and transport underdeveloped rural mountain areas of western China: a cross-sectional survey. J Diabetes 2017;9: 475e81. 12. Fu M, Hu J, Cai X. Effectiveness of a community-based diabetes self-management intervention for Chinese adults with type 2 diabetes: a pilot study. Int J Nurs Pract 2015;21(Suppl. 2):132e40. 13. Zheng Y, Fan L. Effects of executing diabetes education intervention on the individual patients attitude, knowledge, behaviorial. Heilongjiang Nurs J 2005;11:353.

53

14. Lou Q, Wu L, Dai X, Cao M, Ruan Y. Diabetes education in mainland Chinada systematic review of the literature. Patient Educ Counsel 2011;85:336e47. 15. Mallow JA, Theeke LA, Barnes ER, Whetsel T, Mallow BK. Using mHealth tools to improve rural diabetes care guided by the chronic care model. Online J Rural Nurs Health Care 2014;14:43. 16. Fjeldsoe BS, Marshall AL, Miller YD. Behavior change interventions delivered by mobile telephone short-message service. Am J Prev Med 2009;36:165e73. 17. Arora S, Peters AL, Agy C, Menchine M. A mobile health intervention for inner city patients with poorly controlled diabetes: proof-of-concept of the TExT-MED program. Diabetes Technol Therap 2012;14:492e6. 18. Capozza K, Woolsey S, Georgsson M, Black J, Bello N, Lence C, et al. Going mobile with diabetes support: a randomized study of a text message-based personalized behavioral intervention for type 2 diabetes self-care. Diabetes Spectr 2015;28:83e91. 19. Janz NK, Becker MH. The health belief model: a decade later. Health Educ Q 1984;11:1e47. 20. Stiff JB, Mongeau PA. Persuasive communication. 3rd ed. 2016. 21. Ericsson. https://www.aliyun.com/zixun/content/2_6_906043. html. 2015. 22. Kitzinger J. Qualitative research. Introducing focus groups. Br Med J 1995;311:299e302. 23. Onwuegbuzie AJ, Leech NL. Sampling designs in qualitative research: making the sampling process more public. Qual Rep 2007;12:238e54. 24. Green J, Thorogood N. Qualitative methods for health research. Socio Res Online 2004;9:177e80. 25. Strauss A, Corbin J. Basics of qualitative research: techniques and procedures for developing grounded theory. 2nd ed. London: Sage; 1998. 26. Wong M, Kong AP, So WY, Jiang JY, Chan JC, Griffiths SM. Adherence to oral hypoglycemic agents in 26,782 Chinese patients: a cohort study. J Clin Pharmacol 2011;51:1474e82. 27. Lerman I. Adherence to treatment: the key for avoiding longterm complications of diabetes. Arch Med Res 2005;36:300e6. 28. Nelson LA, Coston TD, Cherrington AL, Osborn CY. Patterns of user engagement with mobile- and web-delivered self-care interventions for adults with T2DM: a review of the literature. Curr Diabetes Rep 2016;16:66. 29. Vervloet M, van Dijk L, Santen-Reestman J, van Vlijmen B, van Wingerden P, Bouvy ML, et al. SMS reminders improve adherence to oral medication in type 2 diabetes patients who are real time electronically monitored. Int J Med Inf 2012;81:594e604. 30. Buis LR, Hirzel L, Turske SA, Des Jardins TR, Yarandi H, Bondurant P. Use of a text message program to raise type 2 diabetes risk awareness and promote health behavior change (part I): assessment of participant reach and adoption. J Med Internet Res 2013;15:e281. 31. Osborn CY, Mulvaney SA. Development and feasibility of a text messaging and interactive voice response intervention for low-income, diverse adults with type 2 diabetes mellitus. J Diabetes Sci Technol 2013;7:612e22. 32. Nelson LA, Mulvaney SA, Gebretsadik T, Ho YX, Johnson KB, Osborn CY. Disparities in the use of a mHealth medication adherence promotion intervention for low-income adults with type 2 diabetes. J Am Med Inform Assoc 2016;23:12e8. 33. Baranoski AS, Meuser E, Hardy H, Closson EF, Mimiaga MJ, Safren SA, et al. Patient and provider perspectives on cellular phone-based technology to improve HIV treatment adherence. AIDS Care 2014;26:26e32. 34. Mohammed S, Siddiqi O, Ali O, Habib A, Haqqi F, Kausar M, et al. User engagement with and attitudes towards an interactive SMS reminder system for patients with tuberculosis. J Telemed Telecare 2012;18:404e8.