METHODS: We selected individuals with a positive HCV antibody test, but no RNA test result, newly reported to DOHMH between February 28th and October 2nd, 2015. We piloted three text scripts with varying degrees of personal health information among 105 individuals to assess recipient reaction and participation. We selected a final script based on these results and with input from DOHMH Communications, IRB, and Legal staff. The text script included the following initial consent message: “The NYC Health Dept received info about your health from your doctor. Reply YES for this info. Reply STOP to quit, HELP to contact us.” We randomized 385 individuals with valid mobile numbers and home addresses to receive either texts messages (n=191) or the standard letter (n=194). The intervention was disseminated in two waves as eligible individuals were reported to DOHMH. Outcomes included the proportion of delivered messages, and consent and opt-out rates. We compared the proportion of subsequent RNA tests reported among those randomized to receive the text messages or the letter.
RESULTS: Pilot results indicated no difference in recipient reaction to different text scripts; therefore, we selected the script with the most specific health information. Of 296 initial texts sent (pilot and main study combined), 79.7% were successfully delivered. Of initial text recipients, 21.6% consented to further messaging and 8.5% opted out. To date, among those randomized to receive the texts or the letter, 12.0% of those notified by text had a subsequent RNA tests, compared to 12.9% of those notified by letter (p=0.80).
CONCLUSIONS: Direct-to-patient text messaging requires substantial initial resource investment but provides additional means for health departments to communicate with patients at a lower on-going cost than mailings. While text messaging may be an effective way to reach some individuals, multiple approaches are likely needed to increase linkage to care among HCV-infected individuals.