Embedding a text message recall system to increase the uptake of cardiovascular screening in Australian general practice. Results of the Text to Detect randomised controlled trial

Abstract Background Australian guidelines recommend cardiovascular disease (CVD) risk screening for adults aged 45-74 years, but less than one in three have blood pressure and cholesterol recorded within their general practitioner (GP) records. Medicare subsidised ‘Heart Health Checks’ were introduced in primary care in 2019, however uptake has been suboptimal. Purpose The aims of this study were to; 1) implement a text mediated Heart Health Check recall program within general practice at scale, 2) determine whether uptake can be increased with nudge reminder text messages and 3) evaluate the impact on risk factor measurement and medication prescription. Methods Australian general practices were invited via advertisement and mailing list. Medical record software identified patients eligible for a Heart Health Check. Practices were randomised to one of three arms: 1) standard SMS: 1 SMS inviting patients to see their GP for a Heart Health Check, 2) nudge reminder: standard SMS + nudge reminder two weeks later, 3) control: no SMS. Aggregate, deidentified, practice-level data was extracted at baseline and 3-months. The primary outcome was the proportion of Heart Health Checks delivered at a practice level. Secondary outcomes included updated CVD risk factor assessment and new medication prescriptions. Analyses used ANOVA and Tukey's HSD. Results Between 2022 and 2023, 206 general practices participated, each with 560 median eligible patients (IQR:297-909). A total of 70,679 recall text messages and 34,406 nudge reminder messages were sent. Heart Health Check uptake was 11 and 16 times higher in the standard and nudge arms (respectively) compared to control (0.4 vs 4.5 vs 6.3 per 100 patients, p<0.001), with a 40% statistically significant increase (p=0.02) for 2 SMS vs 1 SMS. Updated cholesterol values were 31.6 and 60.2% higher in standard and nudge arms (respectively) compared to control (13.7 vs 16.8 vs 17.4 per 100 patients, p=0.02). Both active arms were statistically significantly different from control (standard p=0.047, nudge p=<0.001). Updated blood pressure was 23.0 and 26.8% higher in the standard and nudge arms (respectively) compared to control (13.7 vs 16.8 vs 17.4 per 100 participants), however only the nudge arm was statistically significantly different (p=0.38). New medication prescription (lipid or blood pressure-lowering) was 36.8% higher in the nudge arm compared to control (32.1 vs 43.9 patients per practice, p=0.03). Conclusions This study demonstrates the effectiveness of text mediated recall of eligible patients to increase the uptake of cardiovascular risk screening in Australian general practice. The findings will inform the implementation of new Australian guidelines and policy reform.