BACKGROUND/RATIONALE:
Secure Messaging (SM) has been demonstrated to be a patient-desired, efficient form of expanding services that is specifically mentioned by the Institute of Medicine as a part of patient-centered continuous care. The VA is aggressively adopting SM to increase patient access to care, and to enhance the efficiency of care delivery. In August, 2011, Secretary Shinseki announced a goal of having 3 million veterans using SM by 2014. Effective use of SM is a high priority for our operational partners. However, most SM use is reactive, not proactive.

OBJECTIVE(S):
The study will be conducted in two phases. Phase 1 is formative and will consist of expanding our understanding of the current implementation of pre-visit Secure Messaging at the Worcester Community-Based Outpatient Clinic (CBOC). Phase 2 will consist of a formal proactive Secure Messaging implementation program at the Providence VAMC.

METHODS:
The settings are the Worcester CBOC and the Providence VAMC. We will examine both patient and provider experiences with pre-visit Secure Messaging in primary care settings using the PACT model. Semi-structured interviews will be used. Selection procedures include chart abstraction to identify those using SM prior to appointments.

For Phase 1, we will collect administrative data of secure messages that will be coded for content. We will also conduct one-on-one, semi-structured, audio-recorded interviews with both PACT teamlet members and Veteran patients.

Phase 2 will be based on our Phase 1 findings. We are planning an implementation program targeting Primary Care teams at the Providence VAMC.


FINDINGS/RESULTS:
SDP Phase 1 Coding: 331 messages. Content of pre-visit secure messages were coded using the following coding schemes: 1a&b) Requests for information or action (Torp), 2) Physician responses to patient requests (Sittig), 3) Socio-emotional exchange (RIAS), 4) Task-focused exchange (RIAS), 5) Message complexity( Tang), 6) Message content (Torp), 7) Global affect ratings (RIAS), 8) Formality and chattyness (White). All percent values are approximates.
1a) Requests for information (Torp)-Medical treatment (6%), symptoms (5%), and administrative (4.5%) were the most mentioned topics.
1b) Request for action (Torp)-Medical treatment (6%), other administrative action (3.9%), and tests (3.8%) were the top action topics mentioned.
5) Message complexity (Tang)-Simple messages (35%), medication requests (17.5%), and symptoms (14%) ranked highest regarding message complexity.
6) Message Content (Torp)-Logistics/other (34%), medication (21.5%), and FYI (13.5%) ranked highest regarding message content.
7&8) Global Affect Rating (Rias, White)-(slide also includes Formality and Chattiness)-Messages were coded as informal/noncourteous (49%) (vs. formal/courteous=19%), respectful (33%), concise (32%), chatty (29.8%), showing friendliness/interest (25%), and hurried (18%).
Also included (slides mismarked)-are:
3) Socio-emotional exchange (RIAS)-Concern/worry (11.2%), shared decision making (5.8%), and build partnership (5.2%) ranked highest in this category. The physician initiated the message about 5.2% of the time.
4) Task-Focused Exchange/Request for Information (RIAS)-Ranked highest were giving information (36%), Transition (24%), check understanding (20%), and directing behavior (18%). There were more closed ended questions asked (8%) vs. open ended questions (6%) .

SDP Phase 2: We conducted trainings with 13 teams at two VA facilities. All teams found the training useful, and several teams adapted the pre-visit secure message to meet their specific needs. After completion of the trial, teams completed an evaluation assessing their experience with pre-visit secure messaging. Analysis of this data is ongoing.

In total, 1967 messages were sent to patients. Patient responses were content coded using the Taxonomy of Requests by Patients (TORP) classification system. Of the 1967 messages sent, 756 (38%) were read by patients, and 201 (10%) patients replied with topics for discussion at their upcoming appointment. Many patients who replied listed their 3 topics, however some simply replied generally. All but one message was sent by the patient themselves. Most mentioned topics were regarding symptoms (other than pain) (47.96% of messages), pain symptoms (31.63%), Prescriptions/medications (42.86%), treatment (39.8%), tests (34.69%). Note, percent of messages with these topics add up to more than 100% as there were multiple topics per message. After visits, we completed follow-up phone calls to assess patient experience of shared decision making during the visit, comparing patients who responded to messages with those who did not read or respond, and analysis is ongoing. We are now completing a chart review to evaluate what percentage of patient topics were addressed in visits.

IMPACT:
The findings of this study will help provide evidence towards pre-visit SMs to better provide patient-centered care. This may meet the priority of patient-centered care by improving communication between providers and Veterans in an easy-to-use format and optimize use of EHRs.

---

External Links for this Project

NIH Reporter

Grant Number: I01HX001002-01
Link: https://reporter.nih.gov/project-details/8413325

---

PUBLICATIONS:

Journal Articles

1. Luger TM, Hogan TP, Richardson LM, Cioffari-Bailiff L, Harvey K, Houston TK. Older Veteran Digital Disparities: Examining the Potential for Solutions Within Social Networks. Journal of medical Internet research. 2016 Nov 23; 18(11):e296. [view]

Conference Presentations

1. Atkins D, Ho M, Hogan TP, Kirsh S, Shimada SL. Technology-assisted new models of care: Implementation challenges and evaluation in the VA. Presented at: AcademyHealth Annual Research Meeting; 2014 Jun 9; San Diego, CA. [view]

DRA: Other Conditions
DRE: Treatment - Efficacy/Effectiveness Clinical Trial
Keywords: none
MeSH Terms: none