A web-based clinical decision support system for gestational diabetes: Automatic diet prescription and detection of insulin needs

https://doi.org/10.1016/j.ijmedinf.2017.02.014Get rights and content

Highlights

  • Clinical assessment time decreased by 27.389% per Gestational Diabetes patient.

  • Face-to-face visits reduced by 88.556% with telemedicine enhanced by decision support.

  • Automatic detection of 100% patients who needed insulin therapy.

  • Safe and effective automatic diet prescriptions without clinicians’ intervention.

  • High satisfaction and trustworthiness of patients with the proposed system.

Abstract

Background

The growth of diabetes prevalence is causing an increasing demand in health care services which affects the clinicians’ workload as medical resources do not grow at the same rate as the diabetic population. Decision support tools can help clinicians with the inspection of monitoring data, providing a preliminary analysis to ease their interpretation and reduce the evaluation time per patient. This paper presents Sinedie, a clinical decision support system designed to manage the treatment of patients with gestational diabetes. Sinedie aims to improve access to specialized healthcare assistance, to prevent patients from unnecessary displacements, to reduce the evaluation time per patient and to avoid gestational diabetes adverse outcomes.

Methods

A web-based telemedicine platform was designed to remotely evaluate patients allowing them to upload their glycaemia data at home directly from their glucose meter, as well as report other monitoring variables like ketonuria and compliance to dietary treatment. Glycaemia values, not tagged by patients, are automatically labelled with their associated meal by a classifier based on the Expectation Maximization clustering algorithm and a C4.5 decision tree learning algorithm. Two finite automata are combined to determine the patient’s metabolic condition, which is analysed by a rule-based knowledge base to generate therapy adjustment recommendations. Diet recommendations are automatically prescribed and notified to the patients, whereas recommendations about insulin requirements are notified also to the physicians, who will decide if insulin needs to be prescribed. The system provides clinicians with a view where patients are prioritized according to their metabolic condition. A randomized controlled clinical trial was designed to evaluate the effectiveness and safety of Sinedie interventions versus standard care and its impact in the professionals’ workload in terms of the clinician’s time required per patient; number of face-to-face visits; frequency and duration of telematics reviews; patients’ compliance to self-monitoring; and patients’ satisfaction.

Results

Sinedie was clinically evaluated at “Parc Tauli University Hospital” in Spain during 17 months with the participation of 90 patients with gestational diabetes. Sinedie detected all situations that required a therapy adjustment and all the generated recommendations were safe. The time devoted by clinicians to patients’ evaluation was reduced by 27.389% and face-to-face visits per patient were reduced by 88.556%. Patients reported to be highly satisfied with the system, considering it useful and trusting in being well controlled. There was no monitoring loss and, in average, patients measured their glycaemia 3.890 times per day and sent their monitoring data every 3.477 days.

Conclusions

Sinedie generates safe advice about therapy adjustments, reduces the clinicians’ workload and helps physicians to identify which patients need a more urgent or more exhaustive examination and those who present good metabolic control. Additionally, Sinedie saves patients unnecessary displacements which contributes to medical centres’ waiting list reduction

Introduction

Diabetes prevalence is increasing all over the world due to sociocultural changes. It is becoming a public health problem which is originating an increment in the demand for health services [1]. The most common type of diabetes is type 2 (T2DM), whose risk factors are a poor diet, physical inactivity, advanced age or high blood glucose (BG) during pregnancy, affecting both mother and child [2]. The incidence of Gestational diabetes (GDM) is also increasing, as women tend to delay childbearing to older ages, and if new diagnosis criteria, that has recently proven to be cost-effective [3] are adopted, the prevalence could be doubled. Besides causing perinatal complications such as foetal macrosomia, shoulder dystocia or caesarean section [4], women affected by GDM have a 7-fold increased risk of developing T2DM compared with those who had a normoglycaemic pregnancy [5]. Clinical guidelines for GDM [6], [7], [8] establish that pregnant women with GDM should register self-monitoring data (glycaemia, diet, ketonuria) and attend periodic visits to the medical centre every 1 or 2 weeks. It is recommended that patients measure their BG levels at least four times a day: before breakfast (breakfast preprandial) and after the three main meals (breakfast postprandial, lunch postprandial and dinner postprandial). Patients also have to monitor their fasting urine ketones every day using urine strips. During medical visits, clinicians assess if a treatment adjustment is needed based on the patients’ monitoring data.

Telemedicine systems, widely applied to diabetes care, have shown that they are able to improve clinical outcomes and self-care by registering electronically diabetes monitoring data [9], [10], [11], and are especially useful in rural areas where, due to a lack of health care professionals, patients have to travel tens of kilometres to reach the specialized assistance [12]. In GDM, telemedicine has proven to reduce face-to-face visits as it allows remote patient monitoring so visits are only scheduled if a change in the therapy is required [13], [14]. A recent study [15], reported that telematics reviews require less time than face-to-face visits (4.6 min vs. 11.7 min). Some studies conclude that telemedicine, through a more accurate and frequent monitoring, improves glycaemic control and reduces GDM complications [16], [17]. But there is still some controversy in the area as other studies [13], [18], [19], [20] obtained that, although not inferior, telemedicine does not improve glycaemia values or neonatal outcomes compared to traditional care.

Limitations and research gaps identified in telemedicine systems for diabetes include usability, real-time feedback, and decision support capabilities [21]. Although telemedicine saves the patients unnecessary displacements and improves the access to specialized assistance, it might raise the clinicians’ workload to unacceptable levels, as it generates a greater amount of patient data that needs to be evaluated. The use of decision support tools (DSTs) integrated in telemedicine systems, are pointed out as a feasible solution to limited human resources in order to analyse the information generated [22]. Expert systems are computer systems designed to imitate the decision-making ability of a person who has expert knowledge and experience and they usually incorporate a knowledge base containing accumulated experience and an inference or rules engine [23]. An increasing number of expert systems has been proposed to manage diabetes care and to build DSTs, including diabetes diagnosis, therapy adjustment and support to patient self-management [24], [25], [26], [27]. The complexity of diabetes care which is affected by multiple variables and the lack of a gold standard affect the formalization of knowledge when building expert systems in diabetes. DSTs are able to prevent clinicians from undertaking the time consuming task of row data assessment helping to automatically inspect self-monitoring data. Clinicians can interpret the information obtained by the automatic analysis of DSTs, such as recognized glycaemic patterns or weekly statistics, to make a more accurate high level evaluation. DSTs based on expert systems can also suggest treatment changes using protocol-based reasoning [28], automating some of them and asking for medical verification only when a more thorough decision needs to be taken. The validation of advice generated by expert systems and DSTs is a time-consuming task which requires clinical co-operation.

Clinical decision support systems (CDSS) contain DSTs to assist in decision-making tasks. CDSS have been developed in different clinical domains to generate several modes of decision support, including alerts when critical values are detected, reminders and advice for therapy prescription [29]. CDSS can help to optimize the clinicians’ time indicating those patients which are more critical and therefore need a deeper examination. CDSS have shown to improve the practitioners’ performance, but the effects on patient outcomes in acute care have been inconsistent [30]. In GDM, the use of CDSS has been very limited. The DIABNET system was designed to be used by physicians during face-to-face visits and proposed qualitative diet modifications and quantitative changes in insulin therapy based on the offline analysis of self-monitoring data [31] and therefore did not aim to reduce them. The MobiGuide system, which is based on computerized clinical guidelines, proposes personalized decision support for GDM patients and their caregivers and is adapted to a mobile environment [32].

This paper presents Sinedie (Smart and educational system for gestational diabetes), a telemedicine and educational health platform for GDM care enhanced by decision support capabilities. The following sections describe the system functionalities and design and present the evaluation results related to the usage of the system during a clinical trial at “Parc Tauli University Hospital” (PTUH) in Spain.

Section snippets

Materials and methods

Sinedie is a web based CDSS that aims to provide a safe and effective platform to manage GDM patients’ care. The main objective is to enhance healthcare processes, by improving the access to specialized healthcare assistance, reducing face-to-face visits to prevent patients from unnecessary displacements and optimizing the clinicians’ time by reducing the clinical evaluation time required per patient.

Validation study

We uploaded to Sinedie the monitoring data from 25 patients who were followed up at the hospital (since diagnostic until either insulinization or delivery) during an average period of 63.761 ± 37.073 (median = 69.000; percentile5 = 17.600; percentile25 = 29.000; percentile75 = 87.000; percentile95 = 125.000) days per patient. The data set contains a total of 6025 BG measurements, 124 ketonuria values, and 226 diet non-compliance. Patients attended a total of 75 face-to-face visits where physicians

Discussion

Our first design goal, of face-to-face visits reduction (88.556%), was successfully achieved as patients treated with Sinedie only have to go to consultation when they are required by the physician to evaluate a therapy adjustment (mainly due to the need to start insulin). The average duration per visit did not increase with the use of the system (6.752 ± 3.319 min in conventional care vs. 15.000 ± 0.000 with Sinedie), even though patients in the intervention group attended to less visits than

Conclusion

The use of Sinedie reduced face-to-face visits as well as the time clinicians devote to patients’ evaluation, which enhances clinicians’ efficiency to overcome their growing workload. The system detected all cases that required a therapy adjustment, including all patients that needed an insulin therapy, obtaining no false negatives.

The CDSS presented (Sinedie) improves the access to specialized assistance, allowing patients to send their monitoring data from home, which prevents unnecessary

Conflicts of interest

None of the authors of this article have competing financial interests or personal relationships with other people or organizations that could inappropriately influence their work.

Author contributions

E. C-R., G. G-S., M.E. H. and M. R contributed to the conception and design of the study. E. C-R., G. G-S., B. P., M. V. and M. R. acquired the data. E. C-R., G. G-S., M.E. H., M. V. and M. R analysed and interpreted the data. E. C-R., G. G-S. and M.E. H. drafted the article. All authors critically revised the manuscript for important intellectual content and approved the final version to be submitted.

Summary table

What was already known on the topic?

  • Little research on the effects of a clinical

Acknowledgments

This work has been funded by the Spanish grant Sinedie (PII0/01125 and PI10/01150), co-funded by FEDER. We would like to thank the patients and professionals from the “Parc Tauli University Hospital” for their collaboration and support in this research.

References (41)

  • International Diabetes Federation, Chapter 1. What is diabetes? in: IDF Diabetes Atlas, 6th ed., Brussels, Belgium,...
  • A. Duran et al.

    Introduction of IADPSG criteria for the screening and diagnosis of gestational diabetes mellitus results in improved pregnancy outcomes at a lower cost in a large cohort of pregnant women: the St. carlos gestational diabetes study

    Diabetes Care

    (2014)
  • B.E. Metzger et al.

    Hyperglycemia and adverse pregnancy outcomes

    New Engl. J. Med.

    (2008)
  • D.R. Coustan et al.

    Screening and diagnosis of diabetes mellitus during pregnancy. Treatment and course of gestational diabetes mellitus

    UpToDate

    (2014)
  • Grupo Español de Diabetes y Embarazo

    Asistencia a la gestante con diabetes. Guña príctica actualizada en 2014

    Avances en Diabetologña

    (2015)
  • Federación Internacional de Diabetes, IDF GDM Model of Care. Implementation protocol. Guidelines for healthcare...
  • R.L. Bashshur et al.

    The empirical evidence for the telemedicine intervention in diabetes management

    Telemed. E-Health

    (2015)
  • R. Bellazi et al.

    A telemedicine support for diabetes management. the T-IDDM project

    Comput. Methods Progr. Biomed.

    (2002)
  • V. Mohan et al.

    Telemedicine in Diabetes Care: in rural India, a new prevention project seeks to fill in the screening gap

    IEEE Pulse

    (2014)
  • F. Carral et al.

    Web-based telemedicine system is useful for monitoring glucose control in pregnant women with diabetes

    Diabetes Technol. Ther.

    (2015)
  • Cited by (97)

    • Quality of patient-reported outcome reporting in trials of diabetes in pregnancy: A systematic review

      2022, Diabetes Research and Clinical Practice
      Citation Excerpt :

      After excluding titles and abstracts not meeting the inclusion criteria, we reviewed the full text of 206 studies. Of these, 35 articles with 5112 patients identified a PRO as a primary or secondary outcome and were included in the study [31–65] (shown in PRISMA flowchart Figure S1). One hundred and seventy-one studies did not include a PRO in their primary or secondary outcomes, however 27 studies included PROs as exploratory outcome (Table S1).

    • Digital health and telehealth for pregnancy

      2022, Diabetes Digital Health and Telehealth
    View all citing articles on Scopus
    View full text