Fill This Form To Receive Instant Help
Fill This Form To Receive Instant Help
Homework answers / question archive / PRISMA 2020 flow diagram for new systematic reviews which included searches of databases and registers only Screening Identification Identification of studies via databases and registers Records identified from*: Databases (n = ) Registers (n = ) Records removed before screening: Duplicate records removed (n = ) Records marked as ineligible by automation tools (n = ) Records removed for other reasons (n = ) Records screened (n = ) Records excluded** (n = ) Reports sought for retrieval (n = ) Reports not retrieved (n = ) Included Reports assessed for eligibility (n = ) Reports excluded: Reason 1 (n = ) Reason 2 (n = ) Reason 3 (n = ) etc
PRISMA 2020 flow diagram for new systematic reviews which included searches of databases and registers only Screening Identification Identification of studies via databases and registers Records identified from*: Databases (n = ) Registers (n = ) Records removed before screening: Duplicate records removed (n = ) Records marked as ineligible by automation tools (n = ) Records removed for other reasons (n = ) Records screened (n = ) Records excluded** (n = ) Reports sought for retrieval (n = ) Reports not retrieved (n = ) Included Reports assessed for eligibility (n = ) Reports excluded: Reason 1 (n = ) Reason 2 (n = ) Reason 3 (n = ) etc. Studies included in review (n = ) Reports of included studies (n = ) *Consider, if feasible to do so, reporting the number of records identified from each database or register searched (rather than the total number across all databases/registers). **If automation tools were used, indicate how many records were excluded by a human and how many were excluded by automation tools. From: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi: 10.1136/bmj.n71 For more information, visit: http://www.prisma-statement.org/ PRISMA 2020 Checklist Section and Topic Item # Checklist item Location where item is reported TITLE Title 1 Identify the report as a systematic review. 1 ABSTRACT Abstract 2 See the PRISMA 2020 for Abstracts checklist. 2 INTRODUCTION Rationale 3 Describe the rationale for the review in the context of existing knowledge. 4-5 Objectives 4 Provide an explicit statement of the objective(s) or question(s) the review addresses. 5-6 METHODS Eligibility criteria 5 Specify the inclusion and exclusion criteria for the review and how studies were grouped for the syntheses. 6 Information sources 6 Specify all databases, registers, websites, organisations, reference lists and other sources searched or consulted to identify studies. Specify the date when each source was last searched or consulted. 8-9 Table 1 Search strategy 7 Present the full search strategies for all databases, registers and websites, including any filters and limits used. 6-7 Selection process 8 Specify the methods used to decide whether a study met the inclusion criteria of the review, including how many reviewers screened each record and each report retrieved, whether they worked independently, and if applicable, details of automation tools used in the process. 6-7 Data collection process 9 Specify the methods used to collect data from reports, including how many reviewers collected data from each report, whether they worked independently, any processes for obtaining or confirming data from study investigators, and if applicable, details of automation tools used in the process. 6-7 Data items 10a List and define all outcomes for which data were sought. Specify whether all results that were compatible with each outcome domain in each study were sought (e.g. for all measures, time points, analyses), and if not, the methods used to decide which results to collect. 9 and 13 10b List and define all other variables for which data were sought (e.g. participant and intervention characteristics, funding sources). Describe any assumptions made about any missing or unclear information. 9 and 13 Study risk of bias assessment 11 Specify the methods used to assess risk of bias in the included studies, including details of the tool(s) used, how many reviewers assessed each study and whether they worked independently, and if applicable, details of automation tools used in the process. 6 Effect measures 12 Specify for each outcome the effect measure(s) (e.g. risk ratio, mean difference) used in the synthesis or presentation of results. 6-7 Synthesis methods 13a Describe the processes used to decide which studies were eligible for each synthesis (e.g. tabulating the study intervention characteristics and comparing against the planned groups for each synthesis (item #5)). 6-7 PRISMA 2020 Checklist Section and Topic Item # Checklist item Location where item is reported 5 13b Describe any methods required to prepare the data for presentation or synthesis, such as handling of missing summary statistics, or data conversions. 13c Describe any methods used to tabulate or visually display results of individual studies and syntheses. 7-8 13d Describe any methods used to synthesize results and provide a rationale for the choice(s). If meta-analysis was performed, describe the model(s), method(s) to identify the presence and extent of statistical heterogeneity, and software package(s) used. 7-8 13e Describe any methods used to explore possible causes of heterogeneity among study results (e.g. subgroup analysis, meta-regression). n/a 13f Describe any sensitivity analyses conducted to assess robustness of the synthesized results. 5 Reporting bias assessment 14 Describe any methods used to assess risk of bias due to missing results in a synthesis (arising from reporting biases). 6-7 Certainty assessment 15 Describe any methods used to assess certainty (or confidence) in the body of evidence for an outcome. 6 RESULTS Study selection 16a Describe the results of the search and selection process, from the number of records identified in the search to the number of studies included in the review, ideally using a flow diagram. 8-9 16b Cite studies that might appear to meet the inclusion criteria, but which were excluded, and explain why they were excluded. n/a Study characteristics 17 Cite each included study and present its characteristics. 7-8 Risk of bias in studies 18 Present assessments of risk of bias for each included study. 9-10 Results of individual studies 19 For all outcomes, present, for each study: (a) summary statistics for each group (where appropriate) and (b) an effect estimate and its precision (e.g. confidence/credible interval), ideally using structured tables or plots. Results of syntheses 20a For each synthesis, briefly summarise the characteristics and risk of bias among contributing studies. 20b Present results of all statistical syntheses conducted. If meta-analysis was done, present for each the summary estimate and its precision (e.g. confidence/credible interval) and measures of statistical heterogeneity. If comparing groups, describe the direction of the effect. 14 20c Present results of all investigations of possible causes of heterogeneity among study results. 11 20d Present results of all sensitivity analyses conducted to assess the robustness 10 10 13-14 PRISMA 2020 Checklist Section and Topic Item # Location where item is reported Checklist item of the synthesized results. Reporting biases 21 Present assessments of risk of bias due to missing results (arising from reporting biases) for each synthesis assessed. 9-10 Certainty of evidence 22 Present assessments of certainty (or confidence) in the body of evidence for each outcome assessed. 11-12 DISCUSSION Discussion 23a Provide a general interpretation of the results in the context of other evidence. 16 23b Discuss any limitations of the evidence included in the review. 18 23c Discuss any limitations of the review processes used. 18 23d Discuss implications of the results for practice, policy, and future research. 17 OTHER INFORMATION 24a Indicate where the review protocol can be accessed, or state that a protocol was not prepared. 15 24b Describe and explain any amendments to information provided at registration or in the protocol. 16 Support 25 Describe sources of financial or non-financial support for the review, and the role of the funders or sponsors in the review. 17 Competing interests 26 Declare any competing interests of review authors. 18 From: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi: 10.1136/bmj.n71 For more information, visit: http://www.prisma-statement.org/ Abstract Diabetes Mellitus (DM) exists as a serious public health issue that affects more than 400 million individuals globally. The chronic condition is caused by insulin deficiency or insulin resistance in the body. The absence of adequate supporting evidence has weakened the DM's intervention program. This study systematically reviews the literature and identifies the research gap in the big data uses in the diabetes management process. This study gathers information through the scientific literature on the usefulness of data in DM and organize them in the context of analytical modeling, quantitative empirical, and qualitative studies. The selected literature consistently indicated that big data plays a vital role in care gaps identification in communities and nations. The studies also evidenced that big data technology can support prediabetes intervention strategies by identifying risk factors and adopting preventive strategies. The review further evidenced that big data technology offers tremendous support to value-based methods of diabetes management. However, the limited research on big data implementation in diabetes management created challenges in obtaining objective evidence, hence, the application of big data in DM need to be further explored. Table of Contents Abstract............................................................................................................................................................................ 1 1. Introduction ............................................................................................................................................................... 3 1.1 Background ................................................................................................................................................................. 3 1.2 Objectives ................................................................................................................................................................... 4 2. Methodology ............................................................................................................................................................. 5 3. Qualitative Studies .................................................................................................................................................... 8 3.1 Population Health Analytics ....................................................................................................................................... 8 3.2 Machine Learning and Prediabetes Intervention......................................................................................................... 9 3.3 MHealth Wearable Devices ...................................................................................................................................... 11 4. Quantitative Empirical Studies .................................................................................................................................. 12 4.1 Population Health Analytics ..................................................................................................................................... 13 4.2 Prediabetes Intervention: Risk Stratification ............................................................................................................ 13 5. Analytical Modeling Studies ...................................................................................................................................... 14 5.1 Machine Learning Models and Prediabetes Screening ............................................................................................. 15 5.2 MHealth Assessment Model ..................................................................................................................................... 16 6. Health Policies ............................................................................................................................................................ 17 7. Economic Implications ............................................................................................................................................... 17 8. Conclusion .................................................................................................................................................................. 17 9. Limitations .................................................................................................................................................................. 18 References ....................................................................................................................................................................... 19 2 1. Introduction 1.1 Background Diabetes Mellitus (DM) is public health issue that affected more than 400 million individuals globally (Watson, 2020). DM is a disease that inhibits the proper breakdown of consumed food into energy. The condition occurs when the pancreas fails to produce insulin or produces an inadequate insulin amount (Watson, 2020). The condition also occurs when the pancreas generates adequate but highly inefficient insulin. Notably, insulin refers to a peptide hormone generated by the pancreas's beta cells that supports glucose breakdown for energy in the body. Inadequate insulin or insulin resistance elevates blood glucose levels, thus damaging various body organs. High blood glucose (hyperglycemia) creates failure or reduced functionality of blood vessels (large and small), kidneys, eye, nerve, feet, and heart (Deshpande, Harris-Hayes & Schootman, 2008). The complex non-communicable condition is prevalent among obese and elderly individuals in society. The main forms of DM affecting people in modern society include prediabetes, Type I diabetes, Type II diabetes, and gestational diabetes. Currently, cases of DM have increased at high rates in different societies around the world. Notably, the universal surge cases of obese individuals have stimulated a rapid increase in Type II diabetes conditions. Type II DM currently accounts for approximately 91% of diabetes cases in developed nations globally (Bullard et al., 2018). Diabetes Mellitus requires effective management and preventive strategies to reduce financial costs and maximize patients' welfare. Effective intervention strategies facilitate the reduction of complications associated with the condition, including heart attacks, lower limb amputation, kidney failure, adult-blindness, heart attack, and mortality. Diabetes management is also a costly operation that demands long-term commitment among healthcare providers, patients, family members, governments, and communities. The management process of over 400 million cases of diabetes around the globe occurs at a total cost of around $2.2-$2.3 trillion (Wang & Alexander, 2016). Big data technology can support management and preventive strategies for Diabetes Mellitus. 3 DM has a complex pathogenesis that features several processes like the pancreas's β-cells autoimmune destruction, such as insulin resistance and insulin deficiencies. Insulin deficiency or resistance usually creates abnormal protein metabolism, fat, and carbohydrates in the human body (Irwin, 2004). The anomaly creates dysfunctions in various body organs in a manner that reduces patients' welfare and performance. Today, poor integration strategies of big data technology with DM preventive and management strategies have increased suffering and mortalities among individuals with this condition. The weakness in the DM's intervention program is associated with the absence of supportive evidence. The low diagnostic rate of diabetes cases in the world indicates that intervention programs lack adequate evidence-based, which leads to their weakness (Ellahham, 2020). For instance, around 30% of DM cases occurring in the global population remain untreated and undiagnosed (Deshpande, Harris-Hayes & Schootman, 2008). Therefore, studies should be conducted to address the evidence-gap existing in the management and preventive programs for Diabetes Mellitus. The systematic review methodology aims to present the research gap in big data application in the DM management process. The study is based on the observation that the application of big data in DM management significantly lacks in the scientific literature. Therefore, the study generates information that can increase the efficiency of big data integration with Diabetes Mellitus management programs. The study generates comprehensive information about big data-based population health analytics, big data implementation in prediabetes intervention, diabetes prevention models implementation, and value-based diabetes management. Notably, by reviewing the current applications of big data in the established care for Diabetes Mellitus, it is feasible to consider the future potential by conducting more research on the influence of big data in diabetes care. Through the technological systems in place, huge amounts of healthcare data are being produced, and the solution is harnessing the data to generate actionable insights. 1.2 Objectives The study's objective is based on the identification of the efficiency and effectiveness of big data in the diabetes management process. In essence, the research study seeks to determine how big data technology can support prediabetes intervention strategies by identifying risk factors and adopting preventive strategies. This 4 objective is explored through two sub-objectives: (1) to determine how big data can generate predictive analyses for the rate of developing Diabetes Mellitus and help to develop effective preventative care; and (2) to identify how big data can support the management of diabetes mellitus. The study's specific focus is on how population health analytics generated from big data can be utilized in the identification of individuals with care gaps. Still, the study queries the potential for machine learning as a prediabetes intervention to generate predictive models and risk factors for the occurrence of diabetes mellitus. The potential for mHealth wearable devices in facilitating real-time data collection and communication to care teams is also considered in the study. 2. Methodology The research methodology was designed to facilitate the determination of the impact of big data on the management of Diabetes Mellitus. The research methodology utilized in this research was the secondary data collection using a systemic literature review approach. The method was selected for this study due to its ability to generate high-quality and reliable evidence for future research. Therefore, the systematic literature review approach was expected to generate high-quality and reliable information regarding the significance of big data on the diabetes management process (Xiao & Watson, 2019). Generally, the systematic review methodology encompassed three diversified and typical studies, including analytical modeling, quantitative empirical, and qualitative studies. The systematic review methodology facilitated the generation of comprehensive information about big data and DM management strategies. Thus, this systematic literature review aimed at providing a thorough summary of every accessible primary research tool in response to research questions. This method utilized secondary data; the information obtained was sourced from second-hand data and explanations from articles published in peer-reviewed journals and reviews. The PubMed, CINAHL, PsycINFO, Scopus, and MEDLINE sites were utilized to search for articles published between February 20th, 2000, and August 29th, 2020. The study was limited to peer-reviewed articles published in the 21st Century considering the period witnessed significant growth in big data technology and its integration with healthcare intervention programs. Big data integration with DM intervention programs also occurred within the selected period. In this study, these secondary sources 5 were expected to describe, synthesize and interpret the information provided in primary sources. The literature suggests similar studies in Consumer research (2-3 references), blockchain (2-3 references), Study 3(2-3 references), etc. Moreover, since the study was based on secondary data collection, it did not involve physical human respondents. However, the secondary sources of information were analyzed to determine the level of their quality maximization purposes. In the course of the study, the researchers considered several issues, including the initial aim of the study and the reliability of the information source in evaluating the credibility of the process. As noted in Noble and Smith (2015), the reliability of the information in a research study significantly influences the results' credibility and validity. Similarly, sticking to the study's initial purpose reduces the probability of not meeting the study's objectives. Besides, the study also considered the relevance of gathered information to the context of the study before admitting such information. According to Staron and Meding (2019), research processes risk being redundant when they do not sieve through collected data to determine the appropriateness of such information. Still, the research study was based on considerations of the intended audience to determine the scope of the study. Inherently, the target audience would determine the language used, including jargon's relevance in defining the primary terms. Through an accurate definition of the target audience, research studies have the opportunity to appropriately tuning their purpose to suit the specific demographic of the target audience. Still, this study also considered the accuracy and relevance of data and information by limiting the literature to recent publications. Indeed, recent publications effectively ensure the relevance of information collected in research processes through the admission of appropriate publications (Noble and Smith, 2015). 6 Table 1: Literature Table Author Riihimaa (2020) Type Qualitative Bai, Nqalini, and Majumdar (2019) Qualitative Contreras and Vehi (2018) Qualitative Klonoff (2007) Qualitative Brooke and Rege (2015) Qualitative Bellazzi et al. (2015) Qualitative Wang et al. (2017) Qualitative Klonoff (2013) Qualitative Krosel et al. (2016) Qualitative Hamine et al. (2015) Qualitative Title Source Journal of Medical Impact of machine Artificial Intelligence learning and feature selection on type 2 diabetes risk prediction Emerging Research in Analysis and detection Computing, of diabetes using data Information, mining techniques—a Communication and big data application in Applications. health care Journal of medical Artificial intelligence Internet research for diabetes management and decision support: literature review The artificial pancreas: Journal of diabetes how sweet engineering science and technology will solve bitter problems. mHealth Technologies Glucose Intake and in Prediabetes and Utilization in PreDiabetes Care Diabetes and Diabetes Journal of diabetes Big data technologies: science and technology new opportunities for diabetes management Advances in Nutrition A systematic review of application and effectiveness of mHealth interventions for obesity and diabetes treatment and selfmanagement The current status of Journal of diabetes mHealth for diabetes: science and technology will it be the next big thing? Mobile Health Empowering diabetes Technologies-Theories patient with mobile and Applications health technologies Journal of medical Impact of Internet research mHealth chronic disease management on treatment adherence 7 Rahman (2017) Quantitative Razaviia et al. (2015) Quantitative Musacchio et al. (2020) Quantitative and patient outcomes: a systematic review The advantages and disadvantages of using qualitative and quantitative approaches and methods in language "testing and assessment" research: A literature review Population-level prediction of type 2 diabetes from claims data and analysis of risk factors Journal of Education and Learning Big Data Artificial Intelligence and Big Data in Journal of medical Diabetes Care: A Internet research Position Statement of the Italian Association of Medical Diabetologists 3. Qualitative Studies In general, PubMed, CINAHL, PsycINFO, and MEDLINE searches yielded a total of 300 relevant peerreviewed articles on the topic of big data implementation in DM management. However, quality assessment facilitated the selection of five high-quality sources for the study topic. Qualitative studies selected for this study were expected to provide in-depth information about the research topic (Rahman, 2017). The argument is based on the knowledge that qualitative studies utilize an inductive approach to evaluating phenomena in a particular area. The inductive approach allows them to generate comprehensive information using a limited number of secondary studies. The approach suits the topic of big data implementation in DM management, considering that it is a new area with a limited number of supportive research evidence. 3.1 Population Health Analytics Qualitative studies consistently indicate that diabetes management is complicated by the fact that its symptoms in the early stages are not apparent. In most cases, the symptoms develop severity once the disease reaches critical levels of advancement. Notably, newly-diagnosed individuals with Type II DM feel perfectly normal. 8 At this point, managing the condition using big data technology becomes highly important (Riihimaa, 2020). At this stage, big data regulates glucose control to prevent the emergence of complications related to diabetes. The efficiency of big data in assessing cases of diabetes at an early stage makes it highly effective in generating population health analytics. Therefore, qualitative studies confirm that population health analytics produced by big data can facilitate the identification of diabetic individuals with care gaps. Population health analytics features tests conducted using big data technology to determine individuals' susceptibility to diabetes. The strategy also facilitates the determination of the status of their diabetes conditions. The efficiency associated with big data makes it an effective tool for population health management. In the studies, researchers observe that managing a population with diabetes is extremely difficult in the absence of reliable and accurate data. Researchers also observe that unreliable information about population health creates challenges in determining patients' compliance with various intervention strategies for diabetes (Bai, Nalini & Majumdar, 2019). In this regard, scientists have introduced the Enterprise Data Warehouse (EDW), which serves as an evidence-based for diabetes population management. EDW allows healthcare professionals to establish and maintain proper diabetes registries about their population with diabetes. EDW is based on data obtained using advanced tests like A1c, microalbumin, blood pressure, and fasting lipids tests. Moreover, EDW offers healthcare providers reliable evidence-based in their assessment of diabetic individuals' susceptibility to conditions like heart diseases, hypertension, and stroke. 3.2 Machine Learning and Prediabetes Intervention Diabetes Mellitus is diagnoses manually by healthcare professionals or automatically through automatic devices. However, the automated diagnostic strategy is superior to the manual process due to its high level of efficiency and sensitivity. Studies indicate that symptoms of diabetes are undetectable through manual diagnostic techniques in its early stages (Contreras & Vehi, 2018). In this case, advancement in Machine Learning (ML) has enhanced diagnosis and detection of diabetes in its early stage. ML technologies also reduce medical risks and minimize workloads for healthcare professions. The performance level makes ML an effective approach to creating automated diabetes management systems. The statement proves that machine 9 learning can be used to generate predictive models for Diabetes Mellitus, which guide intervention-relevant programs. Machine learning serves as a crucial decision-support tool for healthcare professions during the delivery of care to their diabetic patients. ML approaches support these professionals in customizing medications for diabetes to optimize compliance and welfare. Studies indicate that DM management in patients is improved and simplified by smartphone applications, pumps, and new sensors (Contreras & Vehi, 2018). ML approach in diabetes management enhances the assessment and improvement of glycemic controls, improves A1c levels, and lowers hypoglycemic episodes. Recently, published qualitative research featuring an "artificial pancreas" has indicated that combining algorithm-based insulin pumps with continuous glucose measurement reduces hypoglycemic episodes and improves an individuals' welfare in diabetes self-care programs (Klonoff, 2007). New generation sensors have demonstrated high sensitivity and accuracy regarding the hypoglycemia assessment during prediabetes intervention. Furthermore, scientists have created new algorithms that facilitate the suspension of insulin infusion during glucagon administration and hypoglycemia (Brooke & Rege, 2015). Newly developed algorithms offer an effective and safe system for individuals with a high susceptibility to hypoglycemia. Machine learning advancement has led to the development of applications that easily enable users to analyze and track their health data. The applications also provide insight to diabetic patients under self-care programs regarding the improvement of their health conditions. Presently, high-quality applications offer a comprehensive nutrition database to inform individuals about nutritional content by scanning barcodes. Applications enable users to search for healthy menus in restaurants, thus improving their conditions (Bellazzi et al., 2015). Currently, smartphone sensors that utilize symbolic reasoning and machine learning have been unveiled to facilitate quantification and recognition of high-level lifestyles that enable diabetic patients to informed decisions about their welfare. For instance, a smartphone camera system maximizes caregivers' and patients' engagement in daily wound care activities that hasten the healing process. Advanced technologies in diabetes management low healthcare expenses, eliminate travel costs and increase the patients' comfort. New 10 technologies imply that patients can receive healthcare services from qualified providers at the comfort of their offices or homes. 3.3 MHealth Wearable Devices Advanced healthcare models facilitate the sharing of care responsibility between diabetes patients and their caregivers. The healthcare models utilize an empowerment concept to maximize patients' engagement in managing their diabetic conditions. Notably, patient engagement is enhanced through sharing decision-making responsibilities, self-management, and education. Currently, mobile health (mHealth) occurs as an important technology for maximizing patient engagement in DM management processes (Wang et al., 2017). In this regard, mHealth represents mobile communication devices' utilization in delivering health information and care services. For this purpose, important mobile communication devices include patient monitoring devices, personal digital assistants, tablets, and mobile phones. Particularly, mHealth has a broad role in the healthcare sector, including managing disease outbreaks, tracking diseases, diagnostic support, treatment, communication with providers, remote monitoring, remote data collection, and public education. Mobile health applications address different challenges associated with diabetes management, including psychosocial care, diabetic foot screening, and high blood pressure. Other diabetes management challenges that mHealth addresses include glycemic control, education, diabetic retinopathy screening, obesity, medication adherence, physical activity regulation, and nutritional control (Klonoff, 2013). Glycemic control occurs as a key function of mHealth applications that facilitates DM management. Mobile Health applications support blood glucose self-monitoring processes that create informed diabetes management strategies like therapy optimization, behavioral adjustment, and nutritional adjustment. In nutritional control programs, mHealth enables patients to record food intake to determine daily calorie intake concerning the targeted amount (Krošel et al., 2016). The new mHealth devices generation named wearables allows individuals to assess food intake based on bite frequency and amount. Wearable mHealth devices play an essential role in monitoring physical activity, weight control, blood pressure control in diabetes patients. The development of this application is based on the knowledge that physical activity monitoring poses a colossal challenge to the diabetes management process. For example, 11 most diabetes patients confess that they are generally inactive or maintain physical activities below the recommended level. In this case, wearable mHealth devices enable patients to monitor the level of their daily physical activities. The devices combine gyroscopes with accelerometers to determine an individual's body movement. The weight management capacity of mHealth devices makes them important prediabetes intervention tools (Hamine et al., 2015). BMI level regulation serves as an important improves glycemic control in diabetic patients. Mobile health programs also consist of integrated automatized blood pressure assessment features. Blood pressure readings obtained using the mHealth program are immediately stored and sent to caregivers. Alerts and daily remind features in mHealth applications enhance patients' adherence to medication in diabetes intervention programs. The program has been unveiled at a period where pharmacological intervention program is a challenge in diabetes management. Mobile health applications address medication adherence problems through features like text-message reminders. Studies consistently indicate that mHealth programs use increases medication adherence among diabetes patients (Hamine et al., 2015). Mobile health programs are also utilized in offering personalized education regarding self-management practices. The commonly used mHealth methods in offering diabetes self-management education are short message services (SMS) and emails (Hamine et al., 2015). Moreover, the programs can be utilized in performing diabetes management activities like foot screening and retinopathy screening. Diabetic retinopathy screening promotes early retinal disease detection and treatment to prevent its advancement to critical stages that cause blindness in individuals. Diabetic foot screening facilitates the detection of peripheral vascular disease and peripheral neuropathy. Early peripheral vascular disease detection and management ensures that it does not create complications that necessitate lower limb amputation. 4. Quantitative Empirical Studies MEDLINE, PubMed, Scopus, CINAHL, and PsycINFO searches yielded a total of 50 quantitative peerreviewed articles that were relevant to the topic of big data implementation in DM management. A quality assessment procedure that followed the search process yielded only five reliable quantitative peer-reviewed to the study topic. Quantitative studies are expected to generate generalizable and reliable results for the topic. 12 However, few quantitative studies were obtained for this study, considering that big data implementation in diabetes mellitus management is a new topic. Notably, quantitative studies utilize deductive logic, which implies using a huge research population or a wide range of reference materials (Rahman, 2017). Still, a high research rate in big data and healthcare intervention means that the available quantitative studies are sufficient to test hypotheses for this study. 4.1 Population Health Analytics Quantitative studies indicate that big data analysis can facilitate assessing the care gap in a population of individuals with diabetes. Big data facilitate the creation of predictive analysis frameworks that allow providers to accurately determine the population of an individual with diabetes. Notably, it plays a vital role in the generation of population health analytics. Massive heterogeneous databases have been utilized in the utilization of Diabetes Mellitus cases in societies around the world. The databases also support healthcare professionals in determining the population that is highly susceptible to diabetes (Razavian et al., 2015). In quantitative studies, researchers in the health sector collect data of a large population in the country, which is compared with other collected through heterogeneous sources. For instance, research performed by researchers from Philadelphia and the University of Philadelphia offered a description of the emerging datadriven population health assessment strategy based on ML techniques. The team of researchers aimed at determining risk factors for Type II diabetes and its predictive models. The Type II diabetes assessment model was based on administrative data regarding the individuals' insurance data, pharmaceutical records, laboratory results, and health services access. The study featured 4.1 million people and occurred within four years (Musacchio et al., 2020). The study identified new risk factors for Type II DM thus improving the screening operation. The study indicated that ML models have over 50% higher predictive probability than conventional risk factors (Musacchio et al., 2020). The study further revealed that ML models reduce diabetes assessment costs, thus improving patients' satisfaction. 4.2 Prediabetes Intervention: Risk Stratification Risk stratification through newly developed machine learning models supports prediabetes intervention programs. A study involving 65,000 patients with Type II diabetes indicated that current assessment models 13 could identify 10% of individuals with the highest susceptibility to diabetes (Musacchio et al., 2020). The study suggested that the proportion of individuals in society who require prediabetes intervention strategies lower their chances of developing the condition. Notably, identifying individuals with high chances of living with undiagnosed diabetes cannot be obtained using different databases. Therefore, machine learning algorithms like neural networks should be utilized in diagnosing diabetes in individuals. Created surveillance algorithms have the capacity to differentiate between Type I and Type II diabetes through electronic medical records. For instance, wearable MHealth is an important technological tool that supports value-based methods of diabetes management. MHealth interventions have demonstrated efficiency in addressing blood glucose and obesity risk factors for Diabetes Mellitus. The technology promotes weight loss in individuals by regulating food intake and maximizing an individual's participation in physical exercises. The applications can lower people's sedentary lifestyles at a rate of 47.2 minutes per day (Musacchio et al., 2020). MHealth interventions further contribute to the reduction of blood glucose levels in individuals. 5. Analytical Modeling Studies Scopus, MEDLINE, PsycINFO CINAHL, and PubMed searches generated 20 peer-reviewed analytical modeling studies related to the topic. A quality assessment procedure yielded only three quality articles for the study. Notably, analytical modeling studies were intended to provide multiple perspectives on big data implementation in DM management topics (Sainfort et al., 2013). Analysis of these models enhanced understanding of the structural components of diabetes management tools. Big data analytics encompasses merging and aggregating heterogeneous and massive datasets. The studies were used based on the consideration that big data analytics support clinical modeling processes that facilitate the detection of diseases before they advance to critical stages. Notably, the modeling strategy serves as indirect confirmations for possible results in clinical assessments. Scientists have developed different diabetes prediction models using the data mining strategy. Models used in predicting the best treatment strategies for diabetes typically utilize regression-based data mining strategies to 14 available diabetes data. The models usually depend on an SVM algorithm in determining the best treatment strategy for diabetes in individuals of different ages (He, Shu & Zhang, 2019). The treatment models indicate that pharmacological intervention for young patients should be delayed while old patients should be offered prescription drugs immediately after diagnosis. Additionally, soft computing-based prediction models enable caregivers to assess the types and level of risks in diabetic patients. The model is tested using real-time medical data through the Genetic Algorithm. This model's obtained results enable healthcare professionals to determine their patients' susceptibility to complications like kidney disease, stroke, eye diseases, and heart attack. Furthermore, a superior hybrid diabetes prediction model is created through the combination of Classification and Regression Trees (CART), Self-Organizing Feature Maps (SOFM), and Generic Algorithm. Predictive analysis models support healthcare professionals in making an accurate response to patients' treatment needs and expectations. The models support healthcare professionals in making informed clinical and financial decisions when offering healthcare services to their diabetic patients (Cichosz, Johansen & Hejlesen, 2016). For example, Hadoop's predictive analysis algorithm can determine the diabetes prevalence, associated complications, and available treatment strategies. Notably, Hadoop refers to a data processing software developed by the Apache company. In the diabetes management process, Hadoop serves as an analytic tool and a data organizer (Wang & Alexander, 2016). The data processing platform is utilized in diabetes management due to its capacity to process large quantities of health data by allocating data sets to multiple servers (Eswari, Sampath & Lavanya, 2015). In this case, every server solves specific components of a huge problem and combines the solutions to generate the final result. In diabetic treatment, the data processing platform evaluates patterns like BMI, serum insulin, plasma glucose concentration, diabetes pedigree, and diastolic blood pressure. Hadoop's implementation in big data analytics creates a systematic strategy of improving the affordability, quality, and availability of diabetes care services. 5.1 Machine Learning Models and Prediabetes Screening Analytical modeling studies indicate that machine learning models facilitate prediabetes screening. The screening operation revolves around detecting various risk factors for diabetes, including BMI, hypertension, diabetes, age, and glucose level. The artificial Neural Network (ANN) model utilizes a classification strategy 15 in determining different risk factors for diabetes. The ANN model functions based on the fact that neural networks serve as accurate predictive factors for various diseases (Tigga & Garg, 2020). The model facilitates a simultaneous assessment of different neural network categories, including support vector machine, multilayer perceptron, feedforward neural network, generalized regression neural network, radial basis neural network, and probabilistic neural network. Different predictors for diabetes in this model include alcohol intake, gender, hypertension, family history, BMI, waist circumference, and age. The SVM model performs well in prediabetes assessment programs due to its ability to create a unique and effective solution to a problem. Notably, the model integrates different forms of data in nonlinear and flexible patterns. SVM model assessment using six risk factors, namely hypertension, waist circumference, body mass index, daily alcohol intake, gender, and age, proved that it has an accuracy of 69.9-70.2% in prediabetes screening (Choi et al., 2014). A similar performance assessment process indicated that the ANN model has an accuracy of around 73.23 percent in prediabetes screening (Choi et al., 2014). Generally, the ANN and SVM ML models predict diabetes at the undiagnosed stage. Utilizing the models in screening operations enables healthcare professionals to implement intervention strategies for diabetes in its early stages. 5.2 MHealth Assessment Model MHealth can be viewed as a health assessment model that saves time and money and improves patients' welfare. The model combines hardware (mobile phones and tablets) and software features (patient monitoring applications). The model facilitates the automatic gathering and transmission of data through smartphone technology. The model encompasses different features, including carbohydrate intake monitoring, food intake timing, hypoglycemic assessment, and medication intake assessment features. The model consists of glucose monitors that enable individuals to transmit big data to providers on the secure and individualized website. The transmitted big-data using mHealth supports a value-based method of managing their diabetic conditions. Sharing big data also facilitates the regulation of exercises, food intake, and medication intake among diabetic patients. 16 6. Health Policies The growing enthusiasm for big data in diabetes management warrants increased consideration of data, privacy, and regulatory issues. It is crucial to consider the current health policies pertaining to the implementation of data analytics in healthcare. Besides, there is an inherent need to develop suitable health policies to guide the implementations and operationalization of such interventions in the healthcare sector. One of the most significant aspects entails access and benefit-sharing of data in achieving the full benefits of data analytics (Vayena et al., 2018). An accurate definition of appropriate safeguards and data users' roles is essential in ensuring that such data is used for the public good. Besides considering policies on data use and access, privacy issues take precedent in the implementation of big data in diabetes management. According to Gostin, Halabi, and Wilson (2018), the integration of effective privacy safeguards positively influences data access. Indeed, the health policies should prioritize data protection and privacy to increase the data subjects' privacy options. Similarly, health policies should consider the regulation of potential benefits accrued from the exploitation of personal data. Still, data protection enhances accountability and transparency in the sharing and utilizing data to build trust among users. In essence, these policies should focus on the promotion of transparency through iterative processes encompassing all stakeholders. 7. Economic Implications The implementation of big data in diabetes management processes is dependent on the successful integration of innovation and business strategies. In particular, the integration of innovative approaches such as imagebased diagnostics can increase access and equity in healthcare services. The innovation strategy is based on providing updated programs and interventions to facilitate communication between care teams. In turn, the approach has implications on business development as it encourages investments in wearable devices and mHealth apps. Besides, data handling across different platforms provide positive business implications through hosting serves and database management services. The interventions have a vast market potential based on the increasingly growing enthusiasm for data analytics in the management of healthcare. 8. Conclusion The study indicates that Diabetes Mellitus exists as a serious public health problem in the 21st Century. The chronic condition is caused by insulin deficiency or insulin resistance in the body. The systematic review study aimed to reduce the research gap existing in the big data uses in the diabetes management process. Notably, the systematic review methodology was utilized in gathering data and information for the study from analytical modeling, quantitative empirical, and qualitative studies. The selected reliable studies consistently indicated that big data plays an important role in care gaps identification in communities and nations. The 17 studies also indicate that big data technology can support prediabetes intervention processes based on identified risk factors for the condition. The studies further indicate that big data offers tremendous support to value-based methods of diabetes management. However, the limited nature of the research on big data implementation diabetes management created challenges in obtaining comprehensive and accurate evidence. Therefore, further studies should be conducted to minimize the evidence gap in the big data-based diabetes management process. 9. Limitations The quantitative and qualitative studies indicate that big data implementation in diabetes management is a new healthcare sector concept. The situation means that few studies have been conducted on this topic. Therefore, more studies should be conducted in the future to enrich the diabetes management practice. In particular, a study should be undertaken to assess big data's effectiveness in determining care gaps in nations and societies. Similarly, more studies should be conducted to evaluate future machine learning approaches in determining risk factors for diabetes. The study is based on the fact that machine learning approaches are expected to improve significantly with time in the 21st Century. Consequently, further research should be conducted to determine big data technology's effectiveness in preventing and managing diabetes in patients. Moreover, future studies should address the efficiency of future versions of mHealth wearable technologies in diabetes management. The studies should also explore other technologies developed to improve diabetes management. 18 References Bai, B. M., Nalini, B. M., & Majumdar, J. (2019). Analysis and detection of diabetes using data mining techniques—a big data application in health care. In Emerging Research in Computing, Information, Communication and Applications. Bellazzi, R., Dagliati, A., Sacchi, L., & Segagni, D. (2015). Big data technologies: new opportunities for diabetes management. Journal of diabetes science and technology, 9(5), 1119-1125. Brooke, M. J., & Rege, A. (2015). mHealth Technologies in Prediabetes and Diabetes Care. In Glucose Intake and Utilization in Pre-Diabetes and Diabetes. Bullard, K. M., Cowie, C. C., Lessem, S. E., Saydah, S. H., Menke, A., Geiss, L. S., ... & Imperatore, G. (2018). Prevalence of diagnosed diabetes in adults by diabetes type—United States, 2016. Morbidity and Mortality Weekly Report, 67(12), 359. Deshpande, A. D., Harris-Hayes, M., & Schootman, M. (2008). Epidemiology of diabetes and diabetes-related complications. Physical therapy, 88(11), 1254-1264. Choi, S. B., Kim, W. J., Yoo, T. K., Park, J. S., Chung, J. W., Lee, Y. H., ... & Kim, D. W. (2014). Screening for prediabetes using machine learning models. Computational and mathematical methods in medicine, 2014. Cichosz, S. L., Johansen, M. D., & Hejlesen, O. (2016). Toward big data analytics: review of predictive models in management of diabetes and its complications. Journal of diabetes science and technology, 10(1), 27-34. Contreras, I., & Vehi, J. (2018). Artificial intelligence for diabetes management and decision support: literature review. Journal of medical Internet research, 20(5), e10775. Conway, C. M., & Kelechi, T. J. (2017). Digital health for medication adherence in adult diabetes or hypertension: an integrative review. JMIR diabetes, 2(2), e8030. Ellahham, S. (2020). Artificial intelligence in diabetes care. The American journal of medicine. Eswari, T., Sampath, P., & Lavanya, S. (2015). Predictive methodology for diabetic data analysis in big data. Procedia Computer Science, 50, 203-208. Hamine, S., Gerth-Guyette, E., Faulx, D., Green, B. B., & Ginsburg, A. S. (2015). Impact of 19 mHealth chronic disease management on treatment adherence and patient outcomes: a systematic review. Journal of medical Internet research, 17(2), e52. He, B., Shu, K. I., & Zhang, H. (2019, April). Machine Learning and Data Mining in Diabetes Diagnosis and Treatment. In IOP Conference Series: Materials Science and Engineering (Vol. 490, No. 4, p. 042049). IOP Publishing. Irwin, S. (2004). Primary Prevention and Risk Reduction for Cardiovascular/Pulmonary Disorders—Preferred Practice Pattern 6A. In Cardiopulmonary Physical Therapy. Klonoff, D. C. (2007). The artificial pancreas: how sweet engineering will solve bitter problems. Journal of diabetes science and technology, 1(1), 72-81. Klonoff, D. C. (2013). The current status of mHealth for diabetes: will it be the next big thing?. Journal of diabetes science and technology, 7(3), 749-758. Krošel, M., Švegl, L., Vidmar, L., Dinevski, D., & Bonney, W. (2016). Empowering diabetes patient with mobile health technologies. In Mobile Health Technologies-Theories and Applications. Musacchio, N., Giancaterini, A., Guaita, G., Ozzello, A., Pellegrini, M. A., Ponzani, P., ... & de Micheli, A. (2020). Artificial Intelligence and Big Data in Diabetes Care: A Position Statement of the Italian Association of Medical Diabetologists. Journal of medical Internet research, 22(6), e16922. Rahman, M. (2017). The advantages and disadvantages of using qualitative and quantitative approaches and methods in language "testing and assessment" research: A literature review. Journal of Education and Learning, 6(1). Razavian, N., Blecker, S., Schmidt, A. M., Smith-McLallen, A., Nigam, S., & Sontag, D. (2015). Population-level prediction of type 2 diabetes from claims data and analysis of risk factors. Big Data, 3(4), 277-287. Riihimaa, P. (2020). Impact of machine learning and feature selection on type 2 diabetes risk prediction. Journal of Medical Artificial Intelligence, 3(10). Sainfort, F., Kuntz, K. M., Gregory, S., Butler, M., Taylor, B. C., Kulasingam, S., & Kane, R. L. (2013). Adding decision models to systematic reviews: informing a framework for deciding when and how to do so. Value in health, 16(1), 133-139. Tigga, N. P., & Garg, S. (2020). Prediction of type 2 diabetes using machine learning classification methods. Procedia Computer Science, 167, 706-716. Wang, L., & Alexander, C. A. (2016). Big data analytics for medication management in diabetes mellitus. International Journal of Studies in Nursing, 1(1), 42. Wang, L., & Alexander, C. A. (2016). Big data analytics as applied to diabetes management. 20 European Journal of Clinical and Biomedical Sciences, 2(5), 29-38. Wang, Y., Xue, H., Huang, Y., Huang, L., & Zhang, D. (2017). A systematic review of application and effectiveness of mHealth interventions for obesity and diabetes treatment and selfmanagement. Advances in Nutrition, 8(3), 449-462. Watson , S. (2020). Everything You Need to Know About Diabetes. Healthline. Retrieved from https://www.healthline.com/health/diabetes Xiao, Y., & Watson, M. (2019). Guidance on conducting a systematic literature review. Journal of Planning Education and Research, 39(1), 93-112. Noble, H., & Smith, J. (2015). Issues of validity and reliability in qualitative research. Evidence-based nursing, 18(2), 34-35. Staron, M., & Meding, W. (2019, November). Ensuring reliability of information provided by measurement systems. In International Workshop on Software Measurement (pp. 1-16). Springer, Berlin, Heidelberg. Vayena, E., Dzenowagis, J., Brownstein, J. S., & Sheikh, A. (2018). Policy implications of big data in the health sector. Bulletin of the World Health Organization, 96(1), 66. Gostin, L. O., Halabi, S. F., & Wilson, K. (2018). Health data and privacy in the digital era. Jama, 320(3), 233-234. 21 22
