• 47,623 patients discharged from Chelsea & Westminster and West Mid Hospitals
• 5 new SMS messages with one line re-written
• Each patient received one SMS version only (between-subjects RCT)
• We achieved between 3% – 8% increase in response (+32% over standard SMS)
a. Original: We would like you to think about your recent experience in the [FFTDepartment] at Chelsea and Westminster Hospital. How likely are you to recommend our [FFTDepartment] to friends and family if they needed similar care or treatment?
b. Social Validation: Hello. Please join the 94% of people who think it’s a good idea to give feedback to hospitals by answering the following: how likely are you to recommend Chelsea and Westminster [FFTDepartment] to friends and family if they needed similar care or treatment?
d. Requests for Help: Hello. Will you help us improve services at Chelsea and Westminster Hospital for future patients? Please tell us how likely you are to recommend our [FFTDepartment] to friends and family if they needed similar care or treatment.
e. Perceptual Completeness: Hello. To conclude your recent experience at Chelsea and Westminster Hospital, we would like to hear your views. How likely are you to recommend our [FFTDepartment] to friends and family if they needed similar care or treatment?
• Simple between-subjects Randomised Control Trial (RCT)
• 8 wards across 2 hospitals in London, UK
• 7 weeks for SMS text line replacement
• 2 weeks for a pre-SMS message (data not shown)
• Total of 47,623 patients received one of the variants
• Sequential send of SMS variants (1,2,3,4,5,1,2,3,4,5, 1,2, etc.) to control for time of day and time of year
• Reciprocity & Liking SMS (‘The staff at Chelsea and Westminster Hospital who looked after you recently would benefit from your feedback…‘) consistently gave the highest response rates across all wards/hospitals compared to any other SMS tested – between a 3% – 8% increase over the original SMS
• Strength of result high (p<001 compared to original)
• Margin of error below 5% on larger wards, trend continued across smaller wards
• Even the ‘scores’ (mean rating) of hospital stay were found to be influenced by the message type (in some cases with significance p<05)) – it should be emphasised that these differences are marginal. In all, now 1 in 4 discharged patients responded, and alonside that the ‘score’ for the hospital has risen.