Assessment and Measurements
We assessed provider knowledge and comfort in their fall risk evaluation and management skills before and after the educational intervention using an 11-item multiple-choice questionnaire and a 4-item confidence questionnaire. The confidence questions used a 7-point Likert scale, with 0 indicating “no confidence” and 7 indicating ”lots of confidence.” The questions were administered via a paper survey. Qualitative comments were derived from evaluations completed at the end of the week.
The fellows’ practice of fall risk screening and management was studied from May 2018, at the completion of Mobility week, to May 2019 for the post-intervention period. A 1-year timeframe before May 2018 was used as the pre-intervention period. Eligible visit types, during which we assumed fall risk was discussed, were any office visits for patients 65+ completed by the patients’ PCPs that used fall risk as a reason for the visit or had a fall-related diagnosis code. Fall risk visits performed by other clinic providers were not counted.
Of those patients who had fall risk screenings completed and were determined to be high risk (STEADI score ≥ 4), data were analyzed to determine whether these patients had any fall-related follow-up visits to their PCP within 60 days of the STEADI screening. For these high-risk patients, data were studied to understand whether orthostatic blood pressure measurements were performed (as documented in a flowsheet) and whether a PT referral was placed. These data were compared with those from providers who practiced in clinics within the same system but who did not participate in the mini-fellowship. Data were obtained from the organization’s EHR. Additional data were measured to evaluate patterns of deprescribing of select high-risk medications, but these data are not included in this analysis.
Analysis
A paired-samples t test was used to measure changes in provider confidence levels. Data were aggregated across fellows, resulting in a mean. A chi-square test of independence was performed to examine the relationship between rates of FRMP adoption by select provider groups. Analysis included a pre- and post-intervention assessment of the fellows’ adoption of FRMP practices, as well as a comparison between the fellows’ practice patterns and those of a control group of PCPs in the organization’s other clinics who did not participate in the mini-fellowship (nontrained control group). Excluded from the control group were providers from the same clinic as the fellows; providers in clinics with a geriatric-trained provider on staff; and clinics outside of the Portland metro and Medford service areas. We used an alpha level of 0.05 for all statistical tests.
Data from 5 providers were included in the analysis of the FRMP adoption. The sixth provider changed practice settings from the clinic to the ED after completing the fellowship; her patient data were not included in the FRMP part of the analysis. EHR data included data on all visits of patients 65+, as well as data for just those 65+ patients who had been identified as being at high risk to fall based on a STEADI score of 4 or higher.