Attachment C: Pilot Methods and Results

Attachment C: Pilot Methods and Results

Purpose of the Pilot Project

From June to August 2016, John Snow, Inc. (JSI) conducted a pilot project for AHRQ REV with Cambridge Health Alliance (CHA; Cambridge, MA) and Altamed (Los Angeles, CA). The purpose of the pilot was to determine the feasibility of the data collection methods and instruments, burden placed on respondents for the different data collection activities, and value of the data collected. These findings will inform the data collection activities to be implemented in the full project with nine primary care sites across the country.

Pilot Project Methods

All data collection instruments and interview guides were tested in the pilot. The table below offers an overview of the data collection activities, the research settings, and research participants from the pilot.

Form	Research setting	Research participants
Primary Care Site Organizational Characteristics Form (Attachment G)	CHA	1 medical director
Primary Care Site Patient Characteristics Form (Attachment H)	CHA	1 medical director
Work Flow Mapping Preliminary Interview Guide (Attachment I)	CHA	1 nurse practitioner, 3 nurses, 2 medical assistants, 3 front desk staff
Work Flow Mapping Group Interview Guide (Attachment J)	CHA	1 primary care physician, 1 nurse practitioner, 3 nurses, 2 medical assistants, 2 front desk staff
Work Flow Mapping Follow-Up Interview Guide (Attachment K)	CHA	1 nurse practitioner, 3 nurses, 3 medical assistants, 2 front desk staff
Patient Interview Guide (Attachment L)	Altamed	7 recently discharged patients
Community Agency Interview Guide (Attachment M)	CHA	1 assistant director and 1 social worker from an aging services center

The workflow mapping activities at CHA occurred over several weeks during the one-hour time slot typically used for biweekly CHA staff meetings. In the time between data collection activities at CHA, the research team analyzed and synthesized the findings from previous data collections to prepare for the next ones. The community agency interview took place separately outside of the staff meeting, as no CHA were involved in the data collection. The data collection and analysis process are described for the pilot activities at CHA first, followed by the patient interviews at Altamed.

Primary Care Site Organizational and Patient Characteristics Forms

Implementation Description and Feasibility: The medical director at CHA completed the Primary Care Site Organizational Characteristics and Patient Characteristics Forms. The director was able to respond to all questions on these forms, and was able to look up the information required and record it.

Use and Value of Data Collected: The Primary Care Site Organizational Characteristics Form provided helpful context for later data collection activities, such as the number and type of different staff at CHA, electronic medical record management, and care coordination with nearby hospitals. Some of the questions asked are expected to prove useful later when comparing and contrasting the nine primary care sites, such as patient-centered medical home status and extent of administrative support from beyond the specific primary care site. The Primary Care Site Patient Characteristics Form also provided useful context for the later data collection activities. For instance, it was helpful to understand the demographics of this primary care site’s patients when speaking with primary care staff about the transitional care services delivered. This information is also expected to be useful when comparing and contrasting the nine primary care sites.

Work Flow Mapping Preliminary Interview Guide

Implementation Description and Feasibility: Two pairs of research team members, one interviewer and one note-taker each, interviewed an assortment of four different role types was interviewed using the preliminary interview guide. Some staff were interviewed individually, while others were interviewed in pairs if they had the same role type.

Use and Value of Data Collected: The notes from the interviews were aggregated and analyzed to understand the transitional care workflow at CHA. Transitional care processes were distilled into specific steps by role type and by time period (during hospitalization, between discharge and the primary care visit, during the visit, and after the visit). These processes were also represented on an initial workflow map that noted which transitional care processes applied to which patients, as well as which processes had inconsistencies with regard to what services were delivered, how, and their reliability. This information allowed the investigators to understand the primary care site’s infrastructure (e.g. use of electronic medical records), staff roles and responsibilities, workflow sequence and timing, and challenges to delivering transitional care services. This data collection activity was crucial to gathering this information at a specific level, and informed a much more efficient group workflow mapping session. These individual and paired interviews also allowed research participants to speak freely without more high-ranking staff members in the room (e.g. front desk staff or medical assistants may not speak as openly with physicians or other senior clinical staff in the room).

Work Flow Mapping Group Interview Guide

Implementation Description and Feasibility: Dr. Amy Boutwell, co-principal investigator of this project, facilitated a group workflow mapping session with CHA staff. Dr. Boutwell reviewed the notes from the preliminary interview guide, which allowed her to bring the session by describing the research team’s understanding of the transitional care activities at the site. Dr. Boutwell then clarified and aligned this description with how the primary care staff viewed the transitional care activities at their clinic. A workflow map based on this discussion was drawn on a flipchart in real-time during this discussion.

Use and Value of Data Collected: The group workflow mapping session was useful to bring the staff together to discuss the transitional care workflow. Individual staff members’ experiences discussed in group context helped to highlight differences in transitional care processes across different groups of patients and how their responsibilities related to those of other staff members. The group workflow mapping session was essential to creating a whole picture of the transitional care activities at CHA.

Work Flow Mapping Follow-Up Interview Guide

Implementation Description and Feasibility: As with the preliminary interviews, two pairs of research team members, one interviewer and one note taker each, interviewed an assortment of four different role types for the follow-up interviews. The staff was interviewed in pairs, with the exception of one triplet. Previously collected information on their roles and responsibilities was brought in for their review, clarification, and confirmation. This reference material made it easy to notice any gaps in information, expound on details, and consider improvements to the transitional care workflow.

Use and Value of Data Collected: The follow-up interviews were helpful to gain final clarity on roles, responsibilities, and the workflow after the group workflow mapping session. They were also helpful to identify and discuss the impact of potential failures in the workflow, and consider potential solutions. These interviews offered valuable insight for the REV model and also had direct application to CHIA’s efforts to improve their transitional care processes.

Community Agency Interview Guide

Implementation Description and Feasibility: A telephone interview was conducted with the director of an elder care agency in Cambridge MA, which worked closely with CHA on its Hospital to Home program, its Center for Medicare and Medicaid Services (CMS) sponsored program for Medicare beneficiaries. This interview was conducted on the phone by a JSI staff member using an interview guide. A second JSI staff member took notes on the call. A frontline social worker from the elder care agency also sent written comments on the interview guide.

Use and Value of Data Collected: The interview and communication with the Elder Service Agency was useful in describing this community agency’s role in the hospital to home program. The interview allowed JSI staff to document the role that elder agency staff played during the care transition process (during hospitalization, between discharge and the primary care visit, during the visit, and after the visit). The agency provided health education and referrals to important non-medical services such as meals on wheels, personal care assistance, and transportation at various points in the care transition process. The interview and communication also provided valuable information on the socio-economic challenges that occur in the care transitions process for elderly patients.

Patient Interview Guide

Implementation Description and Feasibility: Two research team members conducted interviews via telephone with patients who had been discharged. The target time to reach the patients was between two to six days post-discharge. Several lessons regarding feasibility were learned in this process. First, it was important that the interviewers were equipped to handle concerns brought up by patients during these interviews, or were able to correctly route the concerns to the appropriate party who would be able to address these concerns. Second, the original introductory script was too wordy and has been slightly condensed for ease of understanding. Finally, it was also necessary to establish rapport and trust with caregivers, who sometimes posed a potential barrier to interviewing patients.

Use and Value of Data Collected: The patient interviews were enlightening with regard to: what patients experienced with regard to transitional care services delivered, in relation to their expectations; current transitional care policies from both the hospital and primary care practice; and best practices. Because of the patient interviews, the research team was able to identify a clear gap between what transitional care should be delivered versus what was actually received by the patients. These interviews offered valuable insight for the REV model.

Key Pilot Project Findings

• Transitional Care Workflow from Primary Care Perspective (Primary Care Site Forms, Workflow Mapping Interviews)

• Significant variation in transitional care workflow dependent on patient type and individual staff:

• Patients can be in or out of network, or belong to specific high-risk groups (complex care management, hospital to home program, skilled nursing facility). Each of these patient groups requires different transitional care services; and

• Staff in the same role may not do all the same things, or may do them differently.

• Many challenges in delivering transitional care, or their root causes, occur before the primary care visit.

• Key reported challenges from staff included:

• Not knowing that a patient was discharged;

• Not reaching patients between discharge and the primary care visit;

• Scheduling and assisting patients with a timely follow-up visit;

• Unavailable/incomplete/inaccurate discharge summary;

• Patient not engaged in discussing the hospitalization during visit; and

• Medication reconciliation inconsistently delivered across the care transition process.

• Transitional Care Workflow from Patient Perspective (Patient Interviews)

• Patients greatly value open lines of communication between the patient and their primary care physician (PCP):

• Some patients expressed that they would have liked to have received a call from their PCP during their hospitalization;

• Most patients expect that the PCP will know everything about their hospital stay by reviewing the hospital records and medication list before the follow-up visit; and

• The primary care follow-up visit is highly valuable to the patient as a means for them to bring their primary care provider up to speed with their current health status post-discharge.

• Most patients interviewed desired a follow-up visit with their PCP and assistance with making this appointment prior to discharge.

• Nearly half of the patients interviewed had specific medical and non-medical needs at the time of discharge that had not been addressed, and reported feeling lost on how to obtain an in-home care provider, transportation, stable housing, disability placard, diabetes medication, and durable medical equipment.

• Transitional Care Workflow from Community Agency Perspective (Community Agency Interviews)

• Many patients are not routinely assessed for non-medical needs during the care transition process.

• Community Agency staff believe that referrals to important community services such as nutrition support, personal care assistance, transportation and behavioral health often occurs too late in the care transition, which can have adverse consequences on patients’ health.

Implications

The AHRQ REV pilot project demonstrated that these data collection instruments and activities were both feasible and valuable. For both logistical planning and estimating time burden, the pilot also allowed a test of the duration of these data collections. The primary care site forms gathered background information on the primary care site efficiently to serve as background for workflow mapping activities and for future cross-site comparisons. The phased approach to the workflow mapping activities allowed for time-constrained providers to participate at their convenience, for researchers to analyze, build on, and refine high-quality information, and to ultimately produce insights for the REV model and immediate process improvements in the field. The community agency interviews provided useful information on the role an elder service agency can play in the care transition process and how the referral process can be improved. Finally, the patient interviews yielded unique information about primary care-based transitional care perceptions and activities that can only be obtained from this population, and does not currently exist in the research literature.

11/2016 The Re-engineered Visit for Primary Care (AHRQ REV) – Attachment C 6