RT Journal Article
SR Electronic
T1 Using digital notifications to improve attendance in clinic: systematic review and meta-analysis
JF BMJ Open
JO BMJ Open
FD British Medical Journal Publishing Group
SP e012116
DO 10.1136/bmjopen-2016-012116
VO 6
IS 10
A1 Dan Robotham
A1 Safarina Satkunanathan
A1 John Reynolds
A1 Daniel Stahl
A1 Til Wykes
YR 2016
UL http://bmjopen.bmj.com/content/6/10/e012116.abstract
AB Objectives Assess the impact of text-based electronic notifications on improving clinic attendance, in relation to study quality (according to risk of bias), and to assess simple ways in which notifications can be optimised (ie, impact of multiple notifications).Design Systematic review, study quality appraisal assessing risk of bias, data synthesised in meta-analyses.Data sources MEDLINE, EMBASE, PsycINFO, Web of Science and Cochrane Database of Systematic Reviews (01.01.05 until 25.4.15). A systematic search to discover all studies containing quantitative data for synthesis into meta-analyses.Eligibility criteria Studies examining the effect of text-based electronic notifications on prescheduled appointment attendance in healthcare settings. Primary analysis included experimental studies where randomisation was used to define allocation to intervention and where a control group consisting of ‘no reminders’ was used. Secondary meta-analysis included studies comparing text reminders with voice reminders. Studies lacking sufficient information for inclusion (after attempting to contact study authors) were excluded.Outcome measures Primary outcomes were rate of attendance/non-attendance at healthcare appointments. Secondary outcome was rate of rescheduled and cancelled appointments.Results 26 articles were included. 21 included in the primary meta-analysis (8345 patients receiving electronic text notifications, 7731 patients receiving no notifications). Studies were included from Europe (9), Asia (7), Africa (2), Australia (2) and America (1). Patients who received notifications were 23% more likely to attend clinic than those who received no notification (risk ratio=1.23, 67% vs 54%). Those receiving notifications were 25% less likely to ‘no show’ for appointments (risk ratio=.75, 15% vs 21%). Results were similar when accounting for risk of bias, region and publication year. Multiple notifications were significantly more effective at improving attendance than single notifications. Voice notifications appeared more effective than text notifications at improving attendance.Conclusions Electronic text notifications improve attendance and reduce no shows across healthcare settings. Sending multiple notifications could improve attendance further.