|Year : 2021 | Volume
| Issue : 3 | Page : 338-343
Effect of personalized human-centered dietary decision support system (PHCDDSS) on dietary knowledge, attitude, practice (KAP), and mean fasting blood sugar (FBS) among participants with type 2 diabetes mellitus (T2DM) in community-based settings of northern state of India
Dinesh Kumar1, Ashish Joshi2, Ashoo Grover3, Sunil Raina1, Ashok Kumar Bhardwaj4, Bhavya Malhotra2, Shruti Sharma2
1 Department of Community Medicine, Dr Rajendra Prasad Government Medical College, Kangra at Tanda, Himachal Pradesh, India
2 Foundation of Healthcare Technologies Society, New Delhi, India
3 Indian Council of Medical Research, New Delhi, India
4 Department of Community Medicine, Dr Radha Krishnan Government Medical College, Hamirpur, Himachal Pradesh, India
|Date of Submission||18-Feb-2021|
|Date of Decision||05-Apr-2021|
|Date of Acceptance||30-Apr-2021|
|Date of Web Publication||30-Sep-2021|
Dr. Dinesh Kumar
Department of Community Medicine, Dr Rajendra Prasad Government Medical College, Kangra at Tanda, Himachal Pradesh.
Source of Support: None, Conflict of Interest: None
Introduction: Acknowledging the promising role of information and technology, a study was planned to determine the effect of personalized human-centered dietary decision support system (PHCDDSS) on dietary knowledge, attitude, practice (KAP), and mean fasting blood sugar (FBS) among participants with type 2 diabetes mellitus (T2DM). Materials and Methods: A community-based randomized control trial was conducted among 400 individuals with T2DM randomized into the 12-month intervention group (PHCDDSS) and 400 to the control (usual care) group. Results: In the control and intervention groups, 84.7% and 87.0% participants completed the follow-up at the end of 12 months. Mean knowledge score showed a significant (P = 0.00) declining trend (from 28.3 to 22.2) in the control group, but increasing (from 28.9 to 35.4) in the intervention group. Unlike knowledge, mean attitude score observed a significant declining trend in both the groups but less in the intervention group. The mean FBS (in mg/dL) trend was found to be insignificantly declining in control (199.2–195.4) and intervention (194.8–183.1) groups but the decline was relatively less in control when compared with the intervention group. Conclusion: In study participants, PHCDDSS proved to be effective in improving knowledge and attitude toward role of diet in managing T2DM. The intervention showed promising effect in reduction of mean FBS and proportion of individuals with sugar control.
Keywords: Decision-support systems, diabetes mellitus, mobile health
|How to cite this article:|
Kumar D, Joshi A, Grover A, Raina S, Bhardwaj AK, Malhotra B, Sharma S. Effect of personalized human-centered dietary decision support system (PHCDDSS) on dietary knowledge, attitude, practice (KAP), and mean fasting blood sugar (FBS) among participants with type 2 diabetes mellitus (T2DM) in community-based settings of northern state of India. J Diabetol 2021;12:338-43
|How to cite this URL:|
Kumar D, Joshi A, Grover A, Raina S, Bhardwaj AK, Malhotra B, Sharma S. Effect of personalized human-centered dietary decision support system (PHCDDSS) on dietary knowledge, attitude, practice (KAP), and mean fasting blood sugar (FBS) among participants with type 2 diabetes mellitus (T2DM) in community-based settings of northern state of India. J Diabetol [serial online] 2021 [cited 2021 Nov 30];12:338-43. Available from: https://www.journalofdiabetology.org/text.asp?2021/12/3/338/327304
| Introduction|| |
Type 2 diabetes mellitus (T2DM) is a lifestyle-related disease which has insidiously grown into a global epidemic with China and India as its epicenters. In addition to glucose-lowering drugs, adherence to diet management is considered imperative to manage the disease., Self-management practices are crucial in managing chronic diseases due to their life-long nature; therefore, they now have been facilitated by exploring the extent of application of information and technology (IT) like mobile phone short message service (SMS) in disease management., Meta-analysis and systematic review observed reduction in change in mean body weight with usage of regular mobile phone SMS over 1–12 months focussed to promote healthy diet and physical activity.,, Extent of web-based applications is observed with significant dietary changes such as eating habits and intake of fat among participants along with improvement in knowledge about DM., Personalized feedback on and self-monitoring of diet proposed to be relevant and acceptable intervention strategy for managing the disease.
Maintaining a written record, either on paper or internet-based diary, of daily consumed food items is considered as an efficient method of self-monitoring and behavior change technique. Effectiveness of written record depends on adherence to record-keeping practice, i.e., how frequently diary is updated? Application of self-monitoring methods proved to be beneficial in adherence to dietary advice and weight management. On the contrary, record-keeping has also been observed as an additional burden, especially among young subjects, therefore often considered as a barrier in its sustainability. This has resulted in application of mobile phone technology for maintaining dietary record and tailored feedback to participants for observing their dietary changes. In young adults, personalized SMS based on dietary analysis proved to be contemplating and found assistive in dietary behavior change. Dietary messages on mobile phone also proved to be beneficial in improving knowledge about diet and nutrition. Significant improvement in knowledge and practice was also observed among participants with T2DM by mobile phone SMS. Role of mobile phone messaging as a reminder for healthy lifestyle has been further extended and found with beneficial effects among prediabetics in reducing incident diabetes. Messages simple and easy to understand and received at convenient timings give a chance to effective reading and increase potential for behavior change. Meta-analysis observed effectiveness in disease control of T2DM after longest SMS intervention duration of 12 months., After analysis of literature mentioning randomized control trials (RCTs), it has been observed that very few studies had long duration of intervention, i.e., 12 months, and also had limited study participants in the control and intervention groups. The current study was carried out as an RCT to assess the effect of 12-month personalized human-centered dietary decision support system (PHCDDSS) on improvement of knowledge, attitude, and practice (KAP) for diabetic diet and on reduction of mean fasting blood sugar (FBS) among physician-diagnosed participants with T2DM in rural settings of Himachal Pradesh.
| Materials and Methods|| |
The current study was carried out as an RCT in the state of Himachal Pradesh, India from March 2017 to February 2019. Study participants, i.e., individuals with physician-diagnosed T2DM, were randomized in control and intervention groups. Himachal Pradesh is a northern state of India with a population of about 7 million and have 12 district administrative units. As per 2011 census, about 90.0% of its population resides in rural, 4.0% in tribal, and rest conglomerate in urban areas. Study participants were recruited against a calculated sample size of 786 (intervention: 393; control: 393) cases to detect difference of 10.0 g/dL reduction in average FBG in the intervention group when compare with the control group at 1:1 with 80.0% study power and at 5.0% level of significance. Participants with age more than 30 years, with controlled or uncontrolled T2DM, and having android-based mobile phone with self or family member were recruited. Participants or any family member willing to record web-based online dietary information while using a mobile/computer or paper-based diary were also recruited. Participants with T2DM with mental and physical challenge resulting in difficulty to inform about record-keeping, who are temporary residents, and involved in other trials of dietary assessment were excluded from the study. Project staff, two laboratory technicians and two field assistants, were trained to collect information, anthropometry, biochemical assessment using a pretested structured questionnaire. Consecutively, eligible participants were enrolled from two district administrative units of state by project staff and then randomized into control and intervention groups with one-to-one ratio using a computer-generated random table method independently by a statistician. Baseline (BL) assessment was carried out at the time of recruitment, and follow-ups (FUs) were done after 6 (FU-1) and 12 months (FU-2) of BL assessment. Project staff, on all three visits (BL, FU-1, and FU-2), collected information on structured questionnaire about dietary assessment for KAP and FBS. Baseline characteristics such as age, gender, disease history, and anthropometry (height in meters and weight in kilograms) were assessed at the time of BL assessment. In both the groups, after obtaining informed consent and following BL data collection, project staff gave knowledge about natural history of T2DM, dietary requirements, and importance of medication adherence to participants of both the groups using standardized booklets. It was given in the form of education session for 20 min/person in home-based settings. Afterwards, participants of the control group were given “paper diary” to record daily dietary consumption such as recoding time, amount, and type of food item(s) consumed. Alternatively, participants in the intervention group were asked to provide information in “electronic diary” to the project staff. It was done as participants were unable to record their diet in “electronic dairy,” so daily diet information was collected over telephone and entered in web-based online portal. The diet details that are to be filled were similar in both paper and electronic diary. The difficulty in managing dairy for diet was enquired by the project staff and recorded. Content analysis of participants’ verbatim was done, and results are reported (Supplementary file). The administrator of online platform had right to monitor the dashboard and add participants, viewing and scheduling of SMS, and editing of information. Based on the web-based online registered information, participants of the intervention group received personalized tailored feedback in the form of SMS over their mobile phone as dietary alerts and reminders. The message system was developed online at project-related domain at a website. Based on the consumed diet in the last 24 h, SMS were auto-selected algorithmically from a bank of 175 messages and sent over to mobile phone. In addition, generic messages were also communicated to the participants of the intervention group daily in a weekly cycle like Monday for home remedies, Tuesday for knowledge about diabetes, Wednesday for physical activity, Thursday for healthy diet, Friday for nutrient role, Saturday for dietary advice for nephropathy, and Sunday for motivation. Dietary alerts about consumption of food item with high glycemic index (GI) were also given to participants, in case participants consumed food item with high GI than messages communicated while listing all food items with high and low GI. Through messages only, participants were encouraged to eat low GI food items and consult their treating physician routinely. Messages were communicated to convenient time as suggested by the participants during BL assessment. Monitoring for delivery of messages was done for individual online profile randomly selected participants where records mentioned number and type of delivered messages. In addition, four dummy participants were also included to observe delivery of messages for timing and completeness. The dummy participants were not directly or indirectly related to study planning, implementation, and analysis.
Knowledge assessment was done with 12 questions and attitude with 03 questions, with maximum scores of 12 and 3, respectively. Therefore, score was converted into a standard scale of 100 by dividing observed with maximum score along with multiplication factor of 100 and then mean score was calculated. Unlike knowledge and attitude, practice was assessed using a Likert scale from 0 (not at all) to 7 (very frequently) and was classified into “negative” with 04 and “positive” with 03 questions. There was no cut-off chosen to categorize positive and negative practices. Low score on negative and high score on positive questions for practice was considered favorable and mean score was calculated. Data entry operator entered collected information into Microsoft Excel 2007 and analyzed using EpiinfoTM (7.2) for Windows developed by Centres for Disease Control and Prevention, USA. Test of linearity and χ2 test for linear trend were applied to assess trend (from BL to FU-2) for means and proportions, respectively. Baseline characteristics were compared between control and intervention groups using unpaired Student’s “t”-test for means and χ2 test for proportions with statistically significant difference at 5% and exact P-values up to two decimal points are reported. Prior ethical approval was sought from the Institutional Ethics Committee (IEC).
| Results|| |
After enlisting, a list of 800 study participants was prepared and 400 were randomized each in control and intervention groups, of which 350 (87.5%) completed FU-1 and 348 (87.0%) completed FU-2 visit in the control group, and 339 (84.7%) participants were assessed in both FU-1 and 2 visits in the intervention group [Figure 1]. Refusal and migration out from the study area were main reasons for censoring. Interview of patients reported barriers for intervention which were related to non-availability of internet connection, financial constraints for recharge, and lack of interest in uploading information. Likewise, participants of the control group reported lack of adherence due to busy schedule due to household chores, inability to find time to record diet, relying on treating physician for routine management (supplementary file). Therefore, desired information was collected by project staff over telephone from participants. It was then entered on an individual “electronic diary” for tailored SMS.
|Figure 1: Flow diagram of recruited and randomized individuals with T2DM in urban and rural area for PHDDSS in Himachal Pradesh, India, 2017–19|
Click here to view
Mean age of censored and not censored participants was about 57 years and was statistically similar (P > 0.05). Distribution of BL characteristics in both control and intervention groups does not vary significantly (P > 0.05) in terms of age, gender, marital status, type of family, and educational status. In both groups, participants have mean age of about 56 years, mostly comprising women (61.5% and 65.5%; P = 0.23), nearly 80.0% married, and about half belonged to nuclear family. Majority of the participants were educated up to high, primary, and middle school standards in both groups [Table 1]. Assessment at the time of recruitment for disease characteristics was also observed to be statistically similar (P > 0.05) in both groups. Majority of the participants were having disease for less than 5 years (55.5% and 59.7%) and had blood sugar assessment at least once in last 6 months (~90.0%). Self-assessment of blood glucose was reported to be carried out by majority of the participants in both groups (71.0% and 66.7%; P = 0.19). Nearly half of the participants of both the groups reported monthly visit to their treating physicians for follow-up of disease control. Both groups were statistically similar for their mean body mass index (BMI) (25.6 and 26.4 kg/m2; P = 0.22), waist circumference (98.1 and 97.8 cm; P = 0.83), and FBS (199.2 and 194.8 mg/dL; P = 0.48) [Table 2].
|Table 1: Basic characteristics of recruited and randomized individuals with type 2 DM in urban and rural area for PHDDSS in Himachal Pradesh, India, 2017–19|
Click here to view
|Table 2: Disease characteristics of recruited and randomized individuals with type-2 DM in urban and rural area for PHDDSS in Himachal Pradesh, India, 2017–19|
Click here to view
From BL to FU-2, mean knowledge score showed a significant (P = 0.00) declining trend (from 28.3 to 22.2) in the control group but increasing one (from 28.9 to 35.4) in the intervention group. Mean knowledge score was similar in BL visit in both the groups but for FU-1 and 2 visits, knowledge score was significantly high in the intervention group. Unlike knowledge, mean attitude score showed a significant (P = 0.00) declining trend in both the groups but its score was significantly more in the intervention group for both follow-up visits. Unlike knowledge and attitude, practice was observed on a Likert scale, where mean practice score found toward “0” i.e., “not at all” in both the groups indicating favorable outcome for negative but unfavorable one for positive practices. Negative practices are found with a significant downward trend in both the groups; less in control (1.1–1.2) but more in the intervention (1.1–0.8) group. Though declining trend was significant in both the groups, but across all three visits, the mean score for negative practices was significantly lower in the intervention compared with the control group. Like negative, a significant declining trend was also observed for positive practices in both groups (control: 1.4 to 1.0; intervention: 1.4 to 1.1), but score was relatively high in the intervention group for FU-1 and 2 visits. Although mean FBS trend found to be insignificantly declining in control (199.0–195.0 mg/dL) and intervention (195.0 to 183.0; P = 0.06) groups, decline was relatively less (4.0 points) in the control compared with the intervention (8 points) group. The difference between mean FBS was more in FU-3 (195.0 vs. 183.0 mg/dL) when compared with BL and FU-1 visits. Though the trend found to be insignificant in both the groups, like mean FBS, decline for proportion of participants with unsatisfactory FBS level (>125 mg/dL) was more in intervention (from 78.9% to 72.2%) when compared with the control group (from 82.7% to 79.8%). Generally, proportion of participants with unsatisfactory FBS level was low in the intervention group but it was significantly low in FU-1 and 2 visits when compared with the control group [Table 3].
|Table 3: Change on KAP and FBS of recruited and randomized individuals with type-2 DM in urban and rural area for PHDDSS in Himachal Pradesh, India, 2017–19|
Click here to view
| Discussion|| |
The present study showed a promising effect of said intervention PHCDDSS, which is a mobile phone SMS feedback system based on daily dietary intake recordings in an electronic dairy in comparison to control-paper-based dairy (without feedback) method. Its effect was observed on mean KAP scores and FBS level of individuals with T2DM already on management by their treating physician. Overall, trends were changing and statistically significant for mean KAP score in both the control and intervention groups across all three visits. In the beginning, BL KAP and FBS scores were statistically similar (except for negative practices) and deviation in scores explored to be significantly different during follow-up visits indicating an intervention effect. Compared with the control group, improvement in mean knowledge score and decline in mean score for negative dietary practices stayed significant in the intervention group during FU visits. Trend was found to be unfavorable for attitude and positive dietary practices as mean attitude score declined and mean score for positive practice stayed low (unfavorable) even in the intervention group. However, significantly high mean attitude and positive practice score during FU visits indicate relatively less decay in attitude and positive practice among participants of the intervention group. Though insignificant, but an expected fall in mean FBS was found in the intervention group which is more in comparison to the control group. Intervention showed positive effect on participants with T2DM, though its trend remained the same in both the groups but its proportion was significantly lower in the intervention group during FU visits.
So, the analysis confirms the beneficial effect of current intervention on dietary knowledge and reduction of mean FBS. Various RCTs have shown positive effect of tailored messaging in disease management in terms of providing dietary advice, medication adherence, and diabetes control, focussing on utilization of messaging to inform participants about following guideline-based dietary advices, what to and not to eat, observed with beneficial result in the management of T2DM., Mobile phone-based application proved to be useful which was enabled with provision of real-time-based information about consumed food items and based on which participants received messages focussed on fruits, vegetables, and junk foods. At the end of 6 months study, participants in the intervention group observed a significant reduction in eating practices of energy-dense nutrient-poor food (EDNP) items. Moreover, messages made participants think about EDNP food items and were twice likely to reduce their intake by over half a serve., Apart from changing practice, SMS also observed improvement in dietary knowledge over 7 weeks period along with increased consumption of fruits and vegetables. SMS effect on attitude was seemed indifferent but significant improvement in knowledge and practice was observed among participants with T2DM. Even, retrospective analysis among people with T2DM showed promising effect of text messaging on raising awareness and engaging participants. Apart from the usefulness of SMS in improving knowledge, practice, and behavior, it was also found to be effective in reducing mean FBS and improving glycemic control.,,
The current study needs to be viewed with limitations; first, lack of adherence to intervention due to lack of formal education, non-interest, non-availability of internet connection possibly impacted outcomes in the intervention arm. Additional inclusion criteria like willingness and motivation to use internet and feeding information certainly would have improved internal validity of the study. Secondly, FU by project staff could have impacted the results and limit external validity of the study. We focussed on internal validity and moreover, participants appreciated the messages but they find lack of time to feed information for messages. Thirdly, as proportion of participants with unsatisfactory FBS indicated, due to financial limitations, glycated hemoglobin could not be assessed for further validation of disease control. Lastly, assessment of 24-h urinary nitrogen for protein intake was proposed but could not be undertaken as collection of 24-h urine was considered a cumbersome process by a large majority of participants.
The current study supports available evidence wherein application of PHCDDS in community-based settings proves to be feasible and effective in improving disease knowledge and reducing mean FBS. After reviewing available literature, this study adds value in terms of limited experience which was carried out in community-based settings with considerable large sample with long duration intervention of 12 months. It also gives an opportunity to build IT platform for participants with chronic diseases under programmatic settings to improve self-monitoring for disease management. As anticipated, such efforts pose problem of record keeping as a laborious exercise, which can be overcome by intermittent feeding of desired information. Application of IT has beneficial potential under National Program for Prevention and Control of Cancer, Diabetes, Cardiovascular diseases, and Stroke (NPCDCS) to augment strategies. Thereafter, operational research studies need to be carried out to further improve its applicability and effectiveness.
We thank the Department of Health Research (DHR), Government of India, for providing financial assistance.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Zheng Y, Ley SH, Hu FB. Global aetiology and epidemiology of type 2 diabetes mellitus and its complications. Nat Rev Endocrinol 2018;14:88-98.
Krishan P, Bedi O, Rani M. Impact of diet restriction in the management of diabetes: Evidences from preclinical studies. Naunyn Schmiedebergs Arch Pharmacol 2018;391:235-45.
Kumar D, Raina S, Sharma SB, Raina SK, Bhardwaj AK. Effectiveness of randomized control trial of mobile phone messages on control of fasting blood glucose in patients with type-2 diabetes mellitus in a northern state of India. Indian J Public Health 2018;62:224-6.
] [Full text]
de Jongh T, Gurol-Urganci I, Vodopivec-Jamsek V, Car J, Atun R. Mobile phone messaging for facilitating self-management of long-term illnesses. Cochrane Database Syst Rev 2012;12:CD007459.
Liu F, Kong X, Cao J, Chen S, Li C, Huang J, et al
. Mobile phone intervention and weight loss among overweight and obese adults: A meta-analysis of randomized controlled trials. Am J Epidemiol 2015;181:337-48.
Palmer M, Sutherland J, Barnard S, Wynne A, Rezel E, Doel A, et al
. The effectiveness of smoking cessation, physical activity/diet and alcohol reduction interventions delivered by mobile phones for the prevention of non-communicable diseases: A systematic review of randomised controlled trials. PLoS One 2018;13:e0189801.
Afshin A, Babalola D, Mclean M, Yu Z, Ma W, Chen CY, et al
. Information technology and lifestyle: A systematic evaluation of internet and mobile interventions for improving diet, physical activity, obesity, tobacco, and alcohol use. J Am Heart Assoc2016;5:e003058.
DiFilippo KN, Huang WH, Andrade JE, Chapman-Novakofski KM. The use of mobile apps to improve nutrition outcomes: A systematic literature review. J Telemed Telecare 2015;21:243-53.
Cotter AP, Durant N, Agne AA, Cherrington AL. Internet interventions to support lifestyle modification for diabetes management: A systematic review of the evidence. J Diabetes Complications 2014;28:243-51.
Hebden L, Chey T, Allman-Farinelli M. Lifestyle intervention for preventing weight gain in young adults: A systematic review and meta-analysis of RCTs. Obes Rev 2012;13:692-710.
Burke LE, Wang J, Sevick MA. Self-monitoring in weight loss: A systematic review of the literature. J Am Diet Assoc 2011;111:92-102.
Kerr DA, Harray AJ, Pollard CM, Dhaliwal SS, Delp EJ, Howat PA, et al
. The connecting health and technology study: A 6-month randomized controlled trial to improve nutrition behaviours using a mobile food record and text messaging support in young adults. Int J Behav Nutr Phys Act 2016;13:52.
Shoneye CL, Dhaliwal SS, Pollard CM, Boushey CJ, Delp EJ, Harray AJ, et al
. Image-based dietary assessment and tailored feedback using mobile technology: Mediating behavior change in young adults. Nutrients 2019;11(2):435.
Brown ON, O’Connor LE, Savaiano D. Mobile myplate: A pilot study using text messaging to provide nutrition education and promote better dietary choices in college students. J Am Coll Health 2014;62:320-7.
Goodarzi M, Ebrahimzadeh I, Rabi A, Saedipoor B, Jafarabadi MA. Impact of distance education via mobile phone text messaging on knowledge, attitude, practice and self efficacy of patients with type 2 diabetes mellitus in Iran. J Diabetes Metab Disord 2012;11:10.
Ram J, Selvam S, Snehalatha C, Nanditha A, Simon M, Shetty AS, et al
. Improvement in diet habits, independent of physical activity helps to reduce incident diabetes among prediabetic Asian Indian men. Diabetes Res Clin Pract 2014;106:491-5.
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 II): Assessment of participants’ perceptions on efficacy. J Med Internet Res 2013;15:e282.
Huang L, Yan Z, Huang H. The effect of short message service intervention on glycemic control in diabetes: A systematic review and meta-analysis. Postgrad Med 2019;131:566-71.
Zhuang Q, Chen F, Wang T. Effectiveness of short message service intervention to improve glycated hemoglobin control and medication adherence in type-2 diabetes: A meta-analysis of prospective studies. Prim Care Diabetes 2020;14:356-63.
Registrar General & Census Commissioner of India. Ministry of Home Affairs. Government of India. Available from: http://www.censusindia.gov.in/2011-prov-results/prov_data_products_himachal.html (accessed on 21 March 2020).
MedCalc. Medical Calculator Software: bvba Acacialaan 22, 8400Ref (StataCorp. Stata Statistical Software: Release 11). College Station, TX: StataCorp LP; 2009.
[Table 1], [Table 2], [Table 3]