System Alignment for VaccinE Delivery (SAVED): Improving Rates of Influenza and Pneumococcal Vaccination Through Patient Outreach, Improved Medical Record Accuracy and Targeted Physician Alerts

The study's multifaceted approach will address the gap between current and optimal vaccination rates via patient-, provider- and system-level interventions intended to target local barriers. Patient-level outreach consisting of education and access information about influenza and pneumococcal vaccines will be delivered through Interactive Voice Recognition (IVR) calls and electronic patient portal messages. The study will have two patient-level outreach cycles. These outreach cycles will address the 2014-2015 flu season (Cycle 1) and the 2015-2016 flu season (Cycle 2). A non-blinded randomized controlled trial design will be utilized during Cycle 1. A total of 20,000 e-portal users and 10,000 non e-portal users who are identified in the Reliant Medical Group (RMG) Electronic Health Record (EHR) as not being up to date on their influenza vaccines will be randomized. E-portal users will be randomized to receive: (1) E-portal message with Interactive Voice Recognition (IVR) call; (2) E-portal message with no IVR call; (3) No e-portal message with IVR call; or (4) No e-portal message with no IVR call. Non e-portal users will be randomized to receive either (1) IVR call or (2) no IVR call. In Outreach Cycle 2, investigators will implement a broader outreach effort in which vaccination messages are sent to all e-portal users and IVR messages are sent to 15,000 patients who did not receive vaccinations in the 2014-2015 influenza season. The 15,000 patients receiving IVR calls will be randomly selected. In an attempt to improve the effectiveness of the intervention in Cycle 2, we will incorporate more detailed information (dates/times/locations on all RMG flu clinics) into both the IVR and e-portal messages. The rate at which e-portal messages can be distributed by RMG will limit the number that can be sent out before the scheduled flu clinics. Patients will therefore be randomly assigned to one of two groups: (A) half of the e-portal users will be sent early messages (approx. 19,000 patients anticipated), while (B) the other half of the e-portal users will not receive early messages. Those in the early message group will receive a message containing flu clinic location, dates, and times. Those in the non-early intervention group will be sent a simplified e-portal message in November 2015 (if they have not already been vaccinated for the 2015-2016 influenza season). The study team plans to conduct qualitative interviews with 20 patients as well as 10 providers and staff in order to tailor intervention components during Cycle 1. Provider-level components will include targeted and recurring educational materials for providers and their staff regarding the indications, risks and benefits of influenza and pneumococcal vaccination, and suggesting strategies for overcoming local barriers to vaccination. Beginning in year 2, providers will also receive individualized report cards providing feedback on their immunization rates. System-level components will include both the capability to capture patient reported vaccinations and automatic updating of EHR vaccination records with pneumococcal and influenza immunizations given at local hospitals and skilled nursing facilities. These system-level components will reduce inappropriate alerts and improve provider confidence in patient records. The key objectives are: 1. To improve rates of influenza and pneumococcal vaccination in eligible patient populations via: 1. Patient-level outreach targeted at unvaccinated patients; 2. Provider- and staff-level educational interventions and system support. 2. To improve the capture of vaccinations administered to Reliant Medical Group (RMG) patients in the community, hospitals and nursing facilities via system-level electronic Health Information Exchange (HIE)