Background
Methods
Eligibility
-
Participants: Aged between 2 and 18 years or described as ‘paediatric’, with a clinical diagnosis of T1DM (with or without comorbidities). For the purpose of this review this is also the definition of children and young people [8].
-
Intervention: Interventions that supplement usual care with a health care strategy primarily delivered through a technology-based medium for the engagement of patients in self-management behaviours. Studies using continuous glucose monitoring (CGM) were included when participants were still required to conduct blood sugar checks and so they could be deemed to be purely an informational tool used to prompt self-management. Interventions targeting only parents/primary caregivers or other health stakeholders and not the child directly were excluded. Interventions which included a technology component as a secondary medium for delivery were excluded.
-
Comparison: A comparison condition was not required for inclusion.
-
Outcomes: Key behaviours (SMBG, insulin administration, physical activity and diet), prerequisites to behaviour (psychological supports) and indicators of behaviour (HbA1c) as highlighted by NICE guidance relating to diabetes self-management [2].
-
Study design: Randomised or non-randomised studies and cohort studies. Single-case studies were excluded.
Search methods
Data extraction and study quality
Measures of treatment effect
Results
Included studies
Study | Design | Medium | Setting | Age range | N | % Male | No. of patients lost to analysis | Intervention | Control group | Relevant outcomes | Adherence |
---|---|---|---|---|---|---|---|---|---|---|---|
Berndt et al. (2014) [30] | 2 group RCT | Mobile phone app | Germany | 8–18 | 68 | 39.7 | 0 | Encouraged to use 3/day for 4 weeks | Usual care | HbA1c, self-efficacy | Below recommended amount |
Boogerd et al. (2013) [31] | 2 group RCT | Interactive online tool | Netherlands | 11–21 | 62 | 35.5 | 12 | Access for 9 months | Usual care | HbA1c, self-efficacy, knowledge | 35–65% engagement with components |
Clements & Staggs (2017) [21] | 2 group non-RCT | Mobile phone app | USA | 10–16 | 81 | 49.0 | 11 | Access for at least 60 days | Usual care | SMBG | Synced 0.22 times a week |
Dyal et al. (2017) [22] | Cohort | Interactive online tool | Canada | 8–12 | 13 | 53.0 | 2 | Access for 3 months, encouraged by avatars to log 3 times/day | None | SMBG | 69% logged in the recommended amount |
Franklin et al. (2006) [34] | 3 group RCT | Text messaging support system | Scotland | 8–18 | 126 | 53.8 | 1 | Received 1–2 messages a day for 1 year | Usual care | HbA1c, self-efficacy, knowledge | Not relevant |
Freeman et al. (2013) [41] | 2 group RCT | Internet-based videoconferencing | USA | 12–19 | 92 | 59.2 | 21 | Up to 10 1–1.5 h sessions in a 12 week period | Usual care | Working alliance | Completed 7 sessions on average |
Frøisland et al. (2012) [38] | Cohort | Mobile phone app | Norway | 13–19 | 12 | 41.7 | 0 | Access for 3 months, encouraged to use at least two 3 day periods | None | HbA1c | Not reported |
Giani et al. (2016) [23] | Cohort | CGM | USA | 8–17 | 61 | 52.0 | 0 | Provided for 6 months | None | SMBG, insulin administration | Used average of 5.4 days a week at baseline, 3.4 days a week at 6 months |
Goyal et al. (2017) [12] | 2 group RCT | Mobile phone app | Canada | 11–16 | 92 | 45.7 | 1 | 1 h tutorial at start to enable independent usage over 6 months | Usual care | SMBG, insulin administration | 65% had low or very low engagement |
Han et al. (2015) [29] | 3 group RCT | Text messaging | USA | 10–17 | 30 | 43.0 | 0 | Received 1 text message a day for 26 weeks | Usual care | HbA1c | 2/4 educators accessed the program |
Henkemans et al. (2017) [42] | 3 group RCT | Interactive robot | Netherlands | 7–12 | 28 | 48.1 | 1 | Played a quiz with robot during 3 consecutive clinic appointments | Usual care | Knowledge, need satisfaction | Only 100% completers included |
Harris et al. (2015) [16] | 2 group RCT | Internet-based videoconferencing | USA | 12–19 | 90c | 55.0 | 24 | Up to 10 1–1.5 h sessions in a 12 week period | Face-to-face skills sessions | SMBG | Completed 5.8 sessions on average |
Herbert et al. (2016) [24] | Cohort | 2-way text messaging software | USA | 13–17 | 23 | 39.0 | 0 | Daily interactive prompts and educational text messages for 6 weeks | None | SMBG | 78% of text messages responded to |
Kowalska et al. (2017) [28] | 2 group RCT | Computer software with automatised food and insulin calculation | Poland | < 18 | 106 | 39.6 | 2 | Encouraged to use for at least 50% of meals for 26 weeks | Usual care | HbA1c | 41.5% used the recommended amount |
Landau et al. (2012) [17] | 2 group RCT | Internet-based glucose monitoring system | Israel | 11–20 | 70 | 15.0 | 3 | Maximum of weekly calls for 6 months | Usual care | SMBG | 66.6% accessed the recommended amount |
Lehmkuhl et al. (2010) [19] | 2 group RCT | Telehealth behavioural therapy | USA | 9–17 | 323 | 28.1 | 0 | 3 × 15 min sessions a week for 12 weeks | Usual care | SMBG | All sessions completed |
Mulvaney et al. (2010) [32] | 2 group RCT | Internet-based program | USA | 13–17 | 72 | 55.6 | 0 | 6 multimedia stories over 11 weeks | Usual care | HbA1c, problem solving | 63–76% engaged with various components |
Mulvaney et al. (2012) [26] | Cohort | Mobile phone-based ecological momentary assessment | USA | 12–17 | 50 | 50.1 | 0 | 2 calls a day for 10 days | None | SMBG, insulin administration | 59.4% total call records with complete data |
Mulvaney et al. (2012b) [39] | 2 group non-RCT | Personalised automated text messaging | USA | 13–17 | 28 | 57.0 | 5 | 10 text messages a week for 3 months | Usual care | HbA1c | 2.9 messages a week responded to |
Newton et al. (2009) [27] | 2 group RCT | Text messaging | New Zealand | 11–18 | 78 | 47.0 | 0 | Weekly text messages/ pedometer reminders over 12 weeks | Usual care | Physical activity | All messages sent |
3 group RCT | Personalised video tapes | Sweden | 2–18 | 332 | 83.0 | 0 | 2 mailed videotapes | Usual care | SMBG, HbA1c | Use range between 1 and 20 times | |
Nunn et al. (2006) [35] | 2 group RCT | Telephone support and educational program | Australia | 3–16 | 139 | 56.0 | 16 | Bimonthly 15–30 min telephone calls for 7 months | Usual care | HbA1c, knowledge | Not stated |
Pinsker et al. (2011) [40] | Cohort | Website | USA | Omitteda | 52 | 51.9 | 20 | Given access for 6 months | Non-users | HbA1c | Logged in 4 or more times during the study |
Rachmiel et al. (2015) [25] | 2 group non-RCT | CGM | Israel | 1–17 | 149 | 47.7 | 0 | Provided with for 1 year | Usual care | SMBG | 38% used 75% of the time, 50% stopped using by 6 months and 66% by 1 year |
Raiff et al. (2016) [13] | 2 group RCT | Internet-based program | USA | 13–18 | 52 | 58.5 | 11 | Given access for 20 days | Non-users | SMBG | All participants used on at least 10 days |
Rami et al. (2006) [36] | 2 group crossover RCT | Mobile phone-based support program | Austria | 10–19 | 36 | 55.6 | 0 | Received 1 text a week for 3 months and access for 6 months | Usual care | HbA1c | 25% engaged < 50% of the recommended amount |
Schiaffini et al. (2016) [14] | 2 group RCT | Website | Italy | Omittedb | 29 | 37.9 | 2 | Access for 5 years, monthly reminders to access | Usual care | SMBG, insulin administration | 2 patients disengaged in the fifth year |
Whittemore et al. (2010) [33] | 2 group RCT | Interactive internet sessions | USA | 13–16 | 12 | 42.0 | 0 | 1 session a week for 5 weeks | Internet education intervention | HbA1c, self-efficacy | 83% completed all sessions |
Whittemore et al. (2012) [18] | 2 group RCT | Interactive internet sessions | USA | 11–14 | 320 | 45.0 | 0 | 1 session a week for 5 weeks | Internet education intervention | SMBG, self-efficacy | 78% completed at least 4 sessions |
Whittemore et al. (2016) [15] | 2 group RCT | Internet psychoeducational program | USA | 11–14 | 124 | 37.4 | 0 | Prompted to login 2 times a week for 4 weeks | Open access website | SMBG, self-efficacy | 85% logged in at least once overall |
Interventions reporting on frequency of SMBG
Study | Clinical significance of findings | Clinical relevance |
---|---|---|
Self-monitoring of blood glucose (SMBG) | ||
Clements & Staggs (2017) [21] | Positive link found but 1/3 did not use app so it is likely that individuals who often SMBG use the app rather than the app encouraging more SMBG. Factor of increase (2.3) reported but exact figures not given so difficult to interpret i.e. could be increase from 1 to 2.3 which would impact very little or 3 to 6.9 which would impact a great deal | Likely low |
Harris et al. (2015) [16] | 25-item measure but does not give any measure of frequency and so impossible to say | Unclear |
Mulvaney et al. (2012) [26] | 1/3 of the cohort sent through no data, percentage of missed blood glucose tests reported but actual frequency of SMBG not given and there is very high variability between the groups i.e. some participants missed very few and some missed almost all tests thus for some findings could be clinically meaningful but not for others | Unclear |
Rachmiel et al. (2015) [25] | Difference between continuous and intermittent users expected to have clinical significancethough would also lead to concomitant increase in discomfort from testing | Likely high |
Raiff et al. (2016) [13] | An increase of 2 to 4 tests per day would be expected to result in clinical improvements | Likely some |
Frequency of SMBG and insulin administration behaviour | ||
Giani et al. (2016) [23] | Improvements found amongst those who monitored regularly at the beginning of the study but not those who initially monitored poorly. This suggests the intervention was effective but only for those who did not need it | Likely low |
Schiaffini et al. (2016) [14] | Can be assumed that the intervention increased SMBG and insulin administration to a degree that would be clinically significant. Unclear whether this is due to the electronic platform or feedback from the clinical team | Likely some |
HbA1c | ||
Mulvaney et al. (2012b) [39] | Nature of control group (matched historical control) is inappropriate to enable estimation of clinical significance | Unclear |
Pinsker et al. (2011) [40] | No statistical significance found but demographic characteristics of the two groups are not provided and so it is not clear if groups were matched | None |
Rami et al. (2006) [36] | No statistical significance reported | None |
HbA1c and self-efficacy | ||
Franklin et al. (2006) [34] | Improvements only reported in the intensive insulin group suggesting that the technology-based intervention was not primarily responsible for differences | None |
Self-efficacy | ||
Berndt et al. (2014) [30] | Measure likely taken immediately after receiving information from the clinical team and so probably does not reflect real changes in self-efficacy | Likely low |
Whittemore et al. (2012) [18] | Intervention appears to have been no more effective than control | Likely low |
Need satisfaction (SDT) | ||
Henkemans et al. (2017) [42] | Unclear how findings relate to clinical outcomes | Unclear |