Improving Patient Safety Systems for Patients With Limited English Proficiency
Appendix D: Methods
We used a robust mixed-methods approach to better understand the unique role of language barriers and cultural factors in patient safety events. Specifically, we aimed to answer two research questions:
- What are the roles of language barriers and cultural factors in patient safety events?
- How are hospitals addressing linguistic and cultural sources of error for LEP patients?
Data were drawn from the following four sources: (1) environmental scan; (2) interpreter pilot database; (3) qualitative interviews; and (4) a town hall meeting.
The research was conducted by a diverse multidisciplinary team of researchers from the Disparities Solutions Center, Mongan Institute for Health Policy at Massachusetts General Hospital and Abt Associates, a consulting firm in Cambridge, Massachusetts. General project oversight was provided by the Agency for Healthcare Research and Quality (AHRQ), which funded the project. All data were gathered and used in conformity with Institutional Review Board and Health Insurance Portability and Accountability Act requirements.
We conducted a systematic review of the peer-reviewed and grey literature on LEP and patient safety, using English articles in PubMed between 1966 and 2009. Search terms related to limited English proficiency (LEP) and cultural diversity were combined with terms related to patient safety and teams (adverse event, medical error, patient safety). To identify grey literature in this area, we reviewed the Web sites of several key organizations and examined existing and proposed standards issued by accrediting bodies that related to health care for LEP patients. We extracted, summarized, and tabulated the design, research methods, and findings of each article. We summarized the implications of each article for policy and practice, with a specific focus on two types of policy and practice recommendations: trainable team behaviors and hospital-level change.
Interpreter Pilot Project
We used a pilot project with interpreters to gather additional insight into the types of patient safety incidents commonly observed by interpreters but rarely recorded. We asked interpreters to document situations they thought negatively affected the health, health care, safety, and well-being of LEP patients. Prior to project start, interpreters received a brief training on patient safety terminology. Any significant quality and safety events were reported to hospital quality and safety leadership immediately; otherwise, incidents were reviewed by project staff and hospital leadership in Interpreter Services and Quality and Safety.
We collected information on 34 incidents between April 2009 and March 2010 (28 of which met the specific study criteria) and reviewed the cases for key themes and safety implications. For formal classification purposes, we applied the AHRQ patient safety classification Common Formats. We identified all incidents reported as Incidents since all events reached the patient. However, it was difficult to identify the emotional distress or other consequences the patient might have experienced, which made the application of the AHRQ Harm Scale inapplicable.
We conducted indepth interviews with frontline staff and clinical and administrative leaders to obtain input on conditions affecting safety for LEP and culturally diverse patients. The goal of the interviews was to gather both "on the ground" and institutional perspectives on the role of language and cultural factors in patient safety events. We interviewed frontline staff (interpreters, nurses, and other bilingual frontline staff) and hospital leadership (directors/managers of interpreter services; patient safety officers; and clinical/administrative nurse managers).
We conducted the interviews at three hospitals in the Boston area. All interviewees received a formal letter of invitation and a study fact sheet summarizing their rights as participants and background information about the project. Frontline staff each received a $50 cash incentive for their time.
Interviews lasted approximately 60 minutes. We conducted a total of 18 interviews both in person and by phone, with 9 interviews with frontline staff and 9 interviews with clinical and administrative leaders. All interviews were audio recorded and transcribed. Each interview was reviewed and coded for key themes, with a focus on the development of the two primary project deliverables, the Hospital Guide and TeamSTEPPS® Module. Quality assurance measures were taken to ensure that themes were identified in an unbiased and uniform process.
Town Hall Meeting
We conducted a "town hall" meeting to help inform our understanding of how hospitals are addressing linguistic and cultural sources of error. The purpose of the meeting was to gather information from a diverse range of hospitals on best practices for preventing, reporting, and documenting medical errors for LEP patients. This interactive phone discussion lasted 1 hour and 15 minutes and included 19 participants from hospitals, hospital associations, and one health plan.
Key leaders from several hospital associations were represented, including the National Association of Public Hospitals and Health Systems, the American Hospital Association, and the Joint Commission. The hospital associations and accrediting agencies helped to provide a broader perspective of common hospital practices and challenges.
Using the minutes from the town hall meeting, we developed a formal debriefing document that included key themes and implications for the content and development of the Module and Hospital Guide. This meeting was held as part of ongoing activities of the Disparities Solutions Center's Disparities Leadership Program.
Field Testing of Hospital Guide
In addition to consulting with our Executive Advisory Board, we field tested the Hospital Guide with a diverse range of hospitals across the country. We asked nine leaders in Quality and Safety at nine unique hospitals in the country to: (1) participate in an indepth 30-minute telephone interview to provide detailed feedback on the guide; and (2) share the guide with members of their implementation teams (e.g., representatives from Quality and Safety, Interpreter Services) to gather "real world" feedback on implementation. The goal of the field testing was to gather indepth and diverse perspectives from key decisionmakers in Quality and Safety and members of their implementation teams on the Hospital Guide's content, usability, ease of implementation, and overall design and structure.
Hospitals were specifically selected to represent diverse perspectives that included public, private, academic, and nonacademic hospitals in both rural and urban locations. We also selected two hospitals with experience implementing TeamSTEPPS. We developed a discussion guide that included 22 questions across three domains: (1) structure and design; (2) chapter-by-chapter content; and (3) usability, implementation, and impact. The interviews were transcribed and imported into Atlas.ti to identify key themes and a coding scheme.
To make sure the codes were applied correctly and systematically, a portion of the transcripts were double-coded by team members. Any coding inconsistencies were reviewed, discussed, and modified by consensus. Key themes and implications for the Hospital Guide were reviewed with the Executive Advisory Board and project staff, and revisions were made accordingly.