Background: The Patient-Report Outcomes Measurement Information System (PROMIS) is increasingly used as a general-purpose tool for measuring orthopaedic surgery outcomes. This set of questionnaires is efficient, precise, and correlates well with specialty-specific measures, but impactful implementation of patient-specific data, especially at the point of care, remains a challenge. Although clinicians may have substantial experience with established patient-reported outcome measures in their fields, PROMIS is relatively new, and the real-life meaning of PROMIS numerical summary scores may be unknown to many orthopaedic surgeons. Questions/purposes: We aimed to (1) identify a small subset of important items in the PROMIS Physical Function (PF) item bank that are answered by many patients with orthopaedic conditions and (2) graphically display characteristic responses to these items across the physical function spectrum in order to translate PROMIS numerical scores into physical ability levels using clinically relevant, familiar terms. Methods: In a cross-sectional study, 97,852 PROMIS PF assessments completed by 37,517 patients with orthopaedic conditions presenting to a tertiary-care academic institution were pooled and descriptively analyzed. Between 2017 and 2020, we evaluated 75,354 patients for outpatient orthopaedic care. Of these, 67% (50,578) were eligible for inclusion because they completed a PROMIS version 2.0 physical function assessment; 17% (12,720) were excluded because they lacked information in the database on individual item responses, and another < 1% (341) were excluded because the assessment standard error was greater than 0.32, leaving 50% of the patients (37,517) for analysis. The PROMIS PF is scored on a 0-point to 100-point scale, with a population mean of 50 and SD of 10. Anchor-based minimum clinically important differences have been found to be 8 to 10 points in a foot and ankle population, 7 to 8 points in a spine population, and approximately 4 points in a hand surgery population. The most efficient and precise means of administering the PROMIS PF is as a computerized adaptive test (CAT), whereby an algorithm intelligently tailors each follow-up question based on responses to previous questions, requiring only a few targeted questions to generate an accurate result. In this study, the mean PROMIS PF score was 41 ± 9. The questions most frequently used by the PROMIS CAT software were identified (defined in this study as any question administered to > 0.1% of the cohort). To understand the ability levels of patients based on their individual scores, patients were grouped into score categories: < 18, 20 ± 2, 25 ± 2, 30 ± 2, 35 ± 2, 40 ± 2, 45 ± 2, 50 ± 2, 55 ± 2, 60 ± 2, and > 62. For each score category, the relative frequency of each possible response (ranging from “cannot do” to “without any difficulty”) was determined for each question. The distribution of responses given by each score group for each question was graphically displayed to generate an intuitive map linking PROMIS scores to patient ability levels (with ability levels represented by how patients responded to the PROMIS items). Results: Twenty-eight items from the 165-question item bank were used frequently (that is, administered to more than 0.1% of the cohort) by the PROMIS CAT software. The top four items constituted 63% of all items. These top four items asked about the patient’s ability to perform 2 hours of physical labor, yard work, household chores, and walking more than 1 mile. Graphical displays of responses to the top 28 and top four items revealed how PROMIS scores correspond to patient ability levels. Patients with a score of 40 most frequently responded that they experienced “some difficulty” with physical labor, yard work, household chores, and walking more than 1 mile, compared with “little” or “no” difficulty for patients with a score of 50 and “cannot do” for patients with a score of 30. Conclusion: We provided a visual key linking PROMIS numerical scores to physical ability levels using clinically relevant, familiar terms. Future studies might investigate whether using similar graphical displays as a patient education tool enhances patient-provider communication and improves the patient experience. Clinical Relevance: The visual explanation of PROMIS scores provided by this study may help new users of the PROMIS understand the instrument, feel empowered to incorporate it into their practices, and use it as a tool for counseling patients about their scores.

Lower extremity fractures pose challenges due to prolonged healing times and limited assessment methods. Integrating wearable sensors with machine learning can help overcome these challenges by providing objective assessment and predicting fracture healing. In this retrospective study, data from a gait monitoring insole on 25 patients with closed lower extremity fractures were analyzed. Continuous underfoot loading data were processed to isolate steps, extract metrics, and feed them into three white-box machine learning models. Decision tree and Lasso regression aided feature selection, while a logistic regression classifier predicted days until fracture healing within a 30-day range. Evaluations via 10-fold cross-validation and leave-one-out validation yielded stable metrics, with the model achieving a mean accuracy, precision, recall, and F1-score of approximately 76%. Feature selection revealed the importance of underfoot loading distribution patterns, particularly on the medial surface. Our research facilitates data-driven decisions, enabling early complication detection, potentially shortening recovery times, and offering accurate rehabilitation timeline predictions.

INTRODUCTION: End‐of‐life (EOL) caregiving is associated with stress and burden, especially for persons with dementia. Little is known, however, about how dementia diagnosis, family structure, and co‐residence influence the prevalence of antidepressants and anxiolytics (psychiatric prescriptions) among spouses and adult children during EOL caregiving.METHODS: This was a retrospective cohort study of spouses (n = 82,321) and adult children (n = 367,888) linked to decedents with (n = 35,482) and without dementia (n = 121,548) between 1998 and 2016. Multivariable logistic regression analyses examined differences in prescription rates by decedent dementia status and family characteristics.RESULTS: Decedents’ dementia status was associated with increased odds of psychiatric prescription for all family members (wives: odds ratio [OR] = 1.22 [1.13–1.31]; husbands: OR = 1.15 [1.00–1.32]; sons: OR = 1.01 [1.01–1.14]; daughters: OR = 1.05 [1.01–1.10]). Co‐residence with the decedent reduced odds of prescriptions for daughters (OR = 0.78 [0.72–0.84]) and sons (OR = 0.73 [0.66–0.81]).DISCUSSION: Dementia status, family structure, and co‐residence may serve as useful indicators for assessing psychosocial needs, allowing health‐care professionals to better identify and support families during EOL caregiving.Highlights: Antidepressant and anxiolytic prescriptions may be influenced by the availability, proximity, and structure of family associated with persons with dementia; therefore, assessing family dynamics is essential in holistic dementia care. Family caregivers of individuals with dementia were more likely than non‐dementia caregivers to have a psychiatric prescription in the year leading up to the decedent's death. Findings underscore the added stress of dementia end‐of‐life (EOL) care and the need for greater support during this critical pre‐death period. Families in which children co‐resided with the decedent and those with more available members generally had lower odds of receiving psychiatric prescriptions. This suggests that family structure, living arrangements, and availability may serve as useful indicators for assessing psychosocial needs, allowing health‐care professionals to better identify and support families requiring assistance during EOL dementia caregiving.