The Contribution of Teleconsultation and Videoconferencing to Diabetes Care: A Systematic Literature Review

Background A systematic literature review was carried out to study the benefits of teleconsultation and videoconferencing on the multifaceted process of diabetes care. Previous reviews focused primarily on usability of technology and considered mainly one-sided interventions. Objective The objective was to determine the benefits and deficiencies of teleconsultation and videoconferencing regarding clinical, behavioral, and care coordination outcomes of diabetes care. Methods Electronic databases (Medline, PiCarta, PsycINFO, ScienceDirect, Telemedicine Information Exchange, ISI Web of Science, Google Scholar) were searched for relevant publications. The contribution to diabetes care was examined for clinical outcomes (eg, HbA1c, blood pressure, quality of life), behavioral outcomes (patient-caregiver interaction, self-care), and care coordination outcomes (usability of technology, cost-effectiveness, transparency of guidelines, equity of care access). Randomized controlled trials (RCTs) with HbA1c as an outcome were pooled using standard meta-analytical methods. Results Of 852 publications identified, 39 met the inclusion criteria for electronic communication between (groups of) caregivers and patients with type 1, type 2, or gestational diabetes. Studies that evaluated teleconsultation or videoconferencing not particularly aimed at diabetes were excluded, as were those that described interventions aimed solely at clinical improvements (eg, HbA1c). There were 22 interventions related to teleconsultation, 13 to videoconferencing, and 4 to combined teleconsultation and videoconferencing. The heterogeneous nature of the identified videoconferencing studies did not permit a formal meta-analysis. Pooled results from the six RCTs of the identified teleconsultation studies did not show a significant reduction in HbA1c (0.03%, 95% CI = - 0.31% to 0.24%) compared to usual care. There was no significant statistical heterogeneity among the pooled RCTs (χ 2 7= 7.99, P = .33). It can be concluded that in the period under review (1994-2006) 39 studies had a scope broader than clinical outcomes and involved interventions allowing patient-caregiver interaction. Most of the reported improvements concerned satisfaction with technology (26/39 studies), improved metabolic control (21/39), and cost reductions (16/39). Improvements in quality of life (6/39 studies), transparency (5/39), and better access to care (4/39) were hardly observed. Teleconsultation programs involving daily monitoring of clinical data, education, and personal feedback proved to be most successful in realizing behavioral change and reducing costs. The benefits of videoconferencing were mainly related to its effects on socioeconomic factors such as education and cost reduction, but also on monitoring disease. Additionally, videoconferencing seemed to maintain quality of care while producing cost savings. Conclusions The selected studies suggest that both teleconsultation and videoconferencing are practical, cost-effective, and reliable ways of delivering a worthwhile health care service to diabetics. However, the diversity in study design and reported findings makes a strong conclusion premature. To further the contribution of technology to diabetes care, interactive systems should be developed that integrate monitoring and personalized feedback functions.

[13] -Italy/Spain/Norway -2002 -18 months, follow-up planned (duration unknown) Secondary care. Bloodglucosemeter to send clinical data and lifestyle data (every seven to ten days) via telecommunication system (Internet / telephone line). Daily computer-generated feedback is provided, and if necessary, messages from physician (specialist in hospital) to advice patients. No details provided about feedback system and frequency of feedback.
-4. Observational studies without control group: n=32 -Four conditions: -Verification phase: clinical evaluation (n=3) -pilot clinical validation (n=12) -Demonstration phase: intranet (n=6) and internet (n=11) -DM type 1 -Adolescents and adults -Being well-compliant with therapy -Logfiles -Non-standardized questionnaire a) Decreased HbA1c (mean reduction of 1.23%, ns), one condition (intranet), NSD. c) Patient-caregiver communication increased (messages sent from caregiver to patients and vice versa). e) Reliable and helpful (caregiver) system, acceptable for patients. [14] -Italy/Spain/Germany -2003 -12 months, follow-up unknown Integrated care. Reflectometer and palmtop to transmit clinical data via multi access system (web access, telephone, interactive voice) to each agent involved in the care process; nurses, case managers, and specialists. Computer-generated feedback is provided via SMS or e-mail to patient and caregiver and educational messages are automatically sent to patients. Frequency of feedback not specified. Questionnaire) -Logfiles a) Decreased HbA1c in I (from 8.31 to 7.59, p<0.05), in C (from 8.86 to 7.95, p<0.05), after 6 months, NSD between groups. Patients Randomized, decreased HbA1c in I (from 8.24 to 7.44, p<0.05), in C (from 8.83 to 7.78, p<0.05), after 6 months, NSD. [15] - Germany -2002 -4-8 months, follow-up unknown Secondary care. Bloodglucosemeter to send clinical data via modem and telephone line to physician in diabetes centre. Personal feedback for proper dose adjustment by diabetes specialist via telephone advice. Frequency of feedback not specified.
-1. Experimental studies (RCT): n=43 -Two conditions: -Intervention: n=27 -Control (usual care): n=16 -Randomization by lots (2:1 in favour for telecare) -Fairly good matching of groups -DM type 1 -Having followed a structured diabetes education program -Taking part in the intervention for at least 50% -≥ 4 doses of insulin/day -Non-standardized questionnaire a) Decreased HbA1c in I (from 8.3 to 6.9 after 4 months, n=27, to 7.1 after 8 months, n=11), in C (from 8.0 to 7.0 after 4 months, n=16, to 6.8, n=10 after 8 months, NSD between groups. e) System appeared easy to use, patients' feeling of security increased through availability of BG-data and the possibility of consulting a caregiver within minimal time, without the need to travel to the diabetes center. f) Cost and time saving in I (saving consultation time although intensified contacts with caregiver), in I on caregiver's side patients time significantly increased.
Legend: 1 experimental study, 2 quasi-experimental study, 3 controlled observation, 3a cohort study, 3b case control study, 4 observational study without control group, 5 expert opinion. a) clinical values; b) quality of life, c) interaction, d) self-care, e) usability of technology, f) cost reduction, g)transparency of guidelines, h) equity (availability of health care to everyone) NSD; not statistically significant difference, ns, statistically not significant; I=intervention; C=control group Primary/ secondary care. Electronic communication network, linking the physicians' computer-based patient records (GPs and interns in hospital) to enable electronic data interchange. System provides computer-generated prompts for physicians to deliver feedback (messages). Frequency of feedback not specified.
-2. Quasi-experimental studies: n=275 -Two conditions: -Intervention: n=215 -Control (usual care): n=60 -Intervention group consisted of patients from GPs with highest number of referred patients. Average age in intervention group higher, less DM1 patients than control group -Having a high frequency of referrals to specialist -Non-standardized questionnaire a) Decreased HbA1c in I (from 7.0 to 6.8, p<0.05) in C (from 6.6 to 6.5, p=0.52), NSD between groups. c) Increased frequency of patient-caregiver communication (p<0.01), more complete information about patient care in I than in C. [17] -  -Interviews a) Decreased HbA1c in I (from 9.2 to 8.6, p<0.001, after 9 months), in C (from 9.3 to 8.9, after 9 months, p<0.05), NSD between groups. e) Difference in proportion of transmitted blood glucose results (40% more in I than in C), p<0.0001. [20] -  -Both groups were comparable (intervention-time, inclusion criteria (inadequate metabolic control, DM duration of over 5 year) -DM type 1 -Diagnosed with DM > 5 years -Poor metabolic control -Non-standardized questionnaire -Logfiles -Interviews a) Decreased HbA1c in I (from 8.4 to 7.9, p=0.053), increased in C (from 8.10 to 8.15, p=0.58), NSD between groups. c) Patients transmitted 3524 blood glucose readings, 1649 daily insulin adjustments, 24 exercise reports and ten diet modifications. Electronic communication with caregivers was limited, a total of 63 text messages were sent by all patients. Caregivers sent 118 text messages to patients (feedback and therapy modifications). Caregivers performed more therapy changes in I than in C due to the possibility to assess patient's condition on a frequent basis. d) Increased confidence in daily self management. e) Patients found the system has high utility, despite several technical problems. a) Decreased HbA1c in I (from 8.4 to 7.6), in C (from 8.9 to 7.6), after 12 months, NSD between groups b) General health status did not change in groups (SF-12) , quality of life improved in I (ns) and C (p<0.05); significant increase in knowledge in I (p<0.05) and in C (p<0.05). d) 80% Of patients reported that appointments in I did not interfere with daily life; in C 100% of patients reported daily interference with outpatients appointments. f) Time and costs saved by patients. Costs were lower (length of appointment 0.25h in I versus 0.5h in C). But 30% of the diabetes team and patient appointments were longer than expected due to technical problems (0.25h versus 1h). [26] -South Korea -2006 -12 weeks, follow-up unknown Tertiary care. Clinical data is entered daily in system via website or cellular phone (SMS). Automatic feedback (reminder) is generated in case patient has not forwarded data for more than a week. Personal feedback provided weekly by nurse in tertiary care hospital via SMS, telephone or Internet.    -Patients treated in project: n=336 -Patients treated by GP: n=225 -Patients treated in outpatient clinic: n=33 -DM type 2 -Already having a caregiver taking part in the intervention -Non-standardized questionnaire -Interviews a) Decreased HbA1c in UDP (from 7.8 to 6.8, p<0.0001), mean inclusion duration 3.2 years. Lipid profiles improved in I (Plasma cholesterol decreased (from 6.1 to 5.9, p<0.0001), plasma triglyceride decreased (from 1.9 to 1.7), p<0.0001) and diastolic blood pressure 86 to 83 (p<0.001). d) Data records of UDP cohort were most complete compared to other groups. e) GPs intend to continue participating in UDP despite shared care took more time. g) Standardized data transfer (protocol driven) between GP, diabetologist, laboratory established an effective infrastructure for shared diabetes care. [34] -China -2001 -12 weeks, follow-up unknown Secondary care. Dietary and clinical data is recorded in hand-held computer and sent twice a week via a modem to the diabetes team of a hospital diabetes clinic (composition of diabetes team not specified). System generates automatic feedback about content of food.
-2. Quasi-experimental studies: n=19 -Two conditions: -Intervention: n=10 -Control: n=9 -Each group used the DSM three months, served as the control group for another three months (cross over design); comparable groups -DM type not specified -Already having a caregiver taking part in the intervention -Non-standardized questionnaire -Interviews a) Decreased HbA1c in I (8.56 to 7.55 after treatment and 7.84 end of project 12 weeks) , in C (8.81 to 8.76 after treatment, and 8.40 after end project12 weeks). Mean difference was 0.825, p<0.019, n=19) e) The DSM was acceptable; 95% found it easy to use, 63% found it useful.   -Interviews a) Decrease of HbA1c (for each child, from 9.7 to 8.5; from 8.7 to 7.1; from 13 to 6.1, from 10.2 to 9.4; from 10.9 to 7.9), after 3 months. d) Better self-control (managing the sending of blood sugar), no hospitalizations, no school absences. [40] - -DM as a primary or secondary diagnosis -Adult home care clients -Already having a caregiver taking part in the intervention -Not cognitively impaired -Focus group -Interviews e) Patients and managers identified a higher degree of readiness for videoconferencing, patients because of the potential to support independence in their homes, managers because of efficiency of the system. Patients wanted to maintain their level of health, but with minimum intrusiveness, caregivers were more interested in measurable clinical outcomes (blood pressure, glucose), managers focused on cost-effectiveness. a) Decrease of HbA1c in I (from 8.7 to 7.8, p<0.001) in C (from 8.6 to 7.6, p<0.001), after 3 months,NSD between groups. b) Reduced diabetes related stress was observed in I and C, p<0.007, NSD between groups. d) More positive appraisal of their diabetes (p<0.05) in both groups. e) Most patients who received videoconferencing felt comfortable with videoconferencing and found it very convenient; overall satisfaction was high (score 4.3 of 5). Satisfaction with treatment increased in both groups (p<0.001), NSD between groups.
[42]  a) Decreased HbA1c in I (from 9.5 to 8.2, p<0.05), in C (from 9.5 to 8.6), after 3 months, NSD between groups. Mean weight reduction (4%). [45] -United States -2004 -12 months, follow-up unknown Secondary care: Self-management therapy video consultation (on nutrition) between patient at home and specialist in hospital. The telediabetes program had been in operation for 10 years. Used equipment not specified. Registered nurse conducts educational session with patient by videoconferencing.
-4. Observational studies (case series): n=60 -One condition -DM type not specified -Already having a caregiver taking part in the intervention -Non-standardized questionnaire -Interviews c) Sustainability of the telediabetes program depends on a feedback system; the effectiveness of the process depends on an interactive, ongoing collaboration between patient and caregiver. f) Reduced travel time for patients and caregivers. g) Administration took a long term view of the value of telemedicine service, service delivery followed national diabetes standards and a well defined cycle of care within a long-term quality improvement program and consistent education program resulted in sustainable diabetes care. h) The system provided access to specialized health care to remote areas. [46] -United States -2004 -12 weeks, follow-up unknown Secondary care: Physician and physical therapist in hospital connected with patient and nurse in medical centre for the treatment of diabetic foot ulcers. Both equipped with a video conferencing unit and a television monitor. The hospital has a handheld camera for real-time transmission of close-up images of the foot and a document camera for real-time transmission of foot X-ray images. Personal feedback by nurse during weekly session.
-2. Quasi-experimental studies: n=140 -Two conditions:   a) Decreased HbA1c in I (weekly monitoring) (from 8.3 to 8.1, p=0.22) and in C (daily monitoring), (from 8.7 to 8.8, p=0.78), after 24 months, NSD between groups. Adjusted mean values HbA1c in I (weekly monitoring) (from 8.1 to 7.8, p=0.20) and in C (daily monitoring), (from 8.6 to 8.7, p=0.79), after 24 months, NSD between groups f) Proportion of one or more hospital admissions decreased in daily monitoring group (77 to 43, p<0.01), and increased in the weekly monitoring group (73 to 106, p<0.01). The change in the average number of hospital bed days was eight days lower in daily monitored group than in the weekly monitored group (p<0.0001). Unscheduled primary care clinic visits were lower in the daily monitoring group (67 to 16) than in the weekly monitoring group (108 to 116), (significant difference between the two groups, p<0.01). [50] -United States -2005 -12 months, follow-up unknown Primary care: Patient-centered care coordination/home telehealth program, based on Wagner's Chronic Care Model, provides selfmanagement and decision support, via electronic reminders and care coordinator. The system used an in home dialogue box (via patients' cell phone) to answer questions about health status, Answers were daily sent to internet to care coordinator who response in case of alarming values. Besides a twoway audio video connectivity and videophone were used (see [48]). analysis of costs f) Significant difference between I and C in needbased primary care visits, increasing in I (7.6%) and decreasing in C (12%), (p<0.01). The likelihood of 1 or more emergency department visits decreased in I and C (significant differences between groups, p<0.0001). I-group had a relative lower likelihood of having 1 or more hospitalizations than patients in the control group (control for HbA1c, ns difference between I and C). h) Increase in access to care (I) -4. Observational studies (case series): n=5 -One condition -DM type not specified -Logfiles -Visit logs -Interviews e) Technology related problems (telecommunication; connectivity) were the primary cause of installation difficulties (at patients' home). Patient education and training are the most critical success factors. Patient education and training accounted two third of the in-home time for installation of equipment. Nurse installers are patient centric rather than technology centric patient centric rather than technology centric