Association between secure patient–clinician email and clinical services utilisation in a US integrated health system: a retrospective cohort study

OBJECTIVE: To assess associations between secure patient–clinician email use and clinical services utilisation over time. DESIGN: Retrospective cohort study between July 2010 and December 2013. Controlling for a utilisation surge around first secure email use, we analysed difference of differences between propensity score-matched groups of secure patient–clinician email users and non-users for utilisation 1–12 months before and 7–18 months after first email (users) or a randomly assigned index date (non-users). SETTING: US integrated healthcare delivery system. PARTICIPANTS: 9345 adults with first secure email use between July 2011 and July 2012 and continuous enrolment for ≥30 months and 9345 adults without secure email use between July 2010 and July 2012 matched to users on demographics, health status, and baseline utilisation. PRIMARY OUTCOME MEASURES: Rates of office visits, patient-initiated phone calls, scheduled telephone visits, after-hours clinic visits, emergency department visits, and hospitalisations. RESULTS: After controlling for multiple factors, no statistically significant differences in utilisation between secure email users and non-users occurred. Utilisation transiently increased by 88–237% around first email use. Annual rates of patient-initiated phone calls decreased among secure email users, 0.2 fewer calls per person (95% CI −0.3 to −0.1), from a mean of 4.1 calls per person 1–12 months before first use to a mean of 3.8 calls per person 7–18 months after first use. Rates of patient-initiated phone calls also decreased among non-users, 0.1 fewer calls per person (95% CI −0.2 to 0.0), from a mean of 4.2 calls per person 1–12 months before the index date to mean of 4.1 calls per person 7–18 months after the index date. CONCLUSIONS: Compared with non-users, patient use of secure email with clinicians was not associated with statistically significant differences in clinical services utilisation 7–18 months after first use.