WV Osteopathic Medical Association . The study was acknowledged as exempt by the West Virginia University Institutional Review Board ( Protocol # 1903504512 ).
PROCEDURES Data Collection
The survey instrument was built and refined following examination of peer-reviewed recruitment and retention research literature , including existing instruments , and a gray literature search of recruitment and retention initiatives in WV and similar rural contexts in the United States ( US ) using Google as the search engine . The survey was pilot tested by three physicians who were currently licensed and practicing across WV to gain feedback on accessibility , structure , content , and wording . A final REDCap ® electronic survey link 17 , 18 was approved and distributed via email to the study population . The study population included all physicians with an e-mail address on file with the WV Board of Medicine or WV Osteopathic Medicine Association ( N = 9,108 ). The WV Board of Medicine directly emailed this survey request and link to its licensees that included a statement of support for the project from the WV Department of Health and Human Resources Bureau for Public Health . As the WV Board of Osteopathic Medicine does not allow direct email of surveys to their licensees , the WV Osteopathic Medical Association distributed the survey by email to its listserv , which includes more than 95 % of osteopathic physicians licensed to practice in the state .
All responses were voluntary , and all participants were presented questions in the demographics , finances , education , and training background sections . Branching logic applied the question “ Please list the zip code of any additional site of practice four + hours per week ,” with years of practice at this site and hours per week at this site as follow-up questions for every zip code entered . For the Pipeline Programs section , additional questions were only presented if a respondent selected one or more options provided in the question matrix : “ Throughout your training , did you participate in any pipeline programs such as those below ? If so , when ?” Respondents had to indicate “ Yes ” to serving as a preceptor in the last 12 months for the three additional questions to be presented . Branching logic opened additional question items as respondents opted in dependent on applicability to respondent ’ s experience . Respondents were asked to check all factors that applied in response to the question “ Which of the following has influenced your selection of a practice site in WV ?” The factors included financial considerations , quality of life , community attributes , work environment , and other . Remaining forced choice questions related to these factors were presented to all respondents .
The survey concluded with two open-ended items presented to all participants that specifically addressed retention through exploration of factors promoting and challenging respondents ’ continued practice in WV . Specific questions were as follows : “ What are the main reasons you continue to practice at your current site ( s ) in WV ?” and “ Are there any factors that may cause you to consider leaving your current site ( s ) in WV ?”
Operational Definitions
Practice type was categorized as primary or specialty care . Primary care was defined as family medicine , internal medicine , obstetrics / gynecology , pediatrics , and psychiatry with all other disciplines categorized as specialty care . Psychiatry was included in the definition of primary care to match the definition of primary care of the state ’ s legislatively mandated Rural Health Initiative ( RHI ) led by the WV Higher Education Policy Commission ( HEPC ). The HEPC partners with the three academic health centers in WV to allocate state RHI funding to address pipeline , recruitment , and retention strategies . 19 Respondents self-reported across study measures . High school was used as a proxy for rural background 10 and “ rural ” was defined using the Federal Office of Rural Health Policy Rural Health Grants Eligibility Analyzer Tool . 20 “ Disadvantaged background ” was defined as having graduated from a high school listed in the Teacher Cancellation Low Income Directory , established in 1997 , that lists low-income schools and educational service agencies that are eligible for federal loan forgiveness programs . 21
DATA ANALYSIS
Survey responses were collated via RED- Cap ® and exported into SPSS ( v . 26 ) for further analysis . Demographic variables were used to describe the sample . Qualitative responses to the two open-ended questions exploring factors related to retention underwent an inductive analysis technique to form themes and patterns throughout ; major messages and ideas expressed by respondents were compiled and reviewed with consensus by the research team . Data presented here represent a subset of data collected from the larger study .
RESULTS
The final sample size was 428 after removing duplicates and incomplete records . All 428 respondents indicated having practiced in WV at some point after receiving their license , 424 respondents provided a response for the year they received their license to practice in WV , 416 reported they practiced at least four hours per week in WV at the time of the survey , and 41 respondents indicated that although they had practiced in WV , their primary practice location was out of state .
Approximately 62 % of respondents were male ( n = 263 ), and 29 % female ( n = 126 ) with 9 % preferring to self-identify gender or not answer . Eighty-nine percent of respondents identified as white ( n = 381 ) and 4 % identified as Hispanic or Latino ( n = 17 ). Respondents ’ age ranged from 25 to 80 years old or older , with the highest percentage of respondents ( 17.8 %) between the ages of 60 and 64 ( n = 76 ).
Forty-four percent of respondents selfidentified as a native of WV ( n = 189 ; n = 239 not a native of WV ). Seven respondents who identified as a native of WV indicated they went to high school out of state . Self-identity as a native of WV was used in analyses . Of respondents who graduated from high school in WV , 17 % attended a high school in rural WV and 26 % attended high school in an urban area . 20 Thirty-nine percent of all respondents were from a disadvantaged background . 21 High school graduation year ranged from 1953 to 2009 ( n = 411 ) with the highest number of respondents graduating in 1973 ( n = 17 ) and median graduation year of 1991 .
West Virginia Medical Journal • March 2022 • 33