Temporal Variability in Routine Mobility Documentation as an Early Signal of Functional Decline in Skilled Nursing Facilities: A Feasibility and Methods Study.

OBJECTIVE: To establish the feasibility of a temporal variability metric derived from routine mobility documentation in skilled nursing facilities and to examine its exploratory association with functional outcomes. DESIGN: Retrospective cohort study with methods and feasibility focus. SETTING: Single 120-bed skilled nursing facility in California. PARTICIPANTS: Inclusion required a length of stay (LOS) of ≥7 days, a minimum of 2 mobility documentation entries, and no hospice or comfort-focused care designation. Among 214 screened admissions, 182 residents met inclusion criteria (mean age 79.6±8.4y; 61% female; median LOS 24d). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: The primary outcome was the feasibility of calculating documentation variability using the coefficient of variation (CV) of weekly mobility entry counts. Secondary outcomes included functional trajectory classification (improvement, stability, and decline) and discharge disposition. RESULTS: Documentation variability was successfully calculated for all 182 eligible residents (100% calculability), exceeding the 90% feasibility threshold. Mean CV was 0.62±0.24 with no floor or ceiling effects. Functional trajectories were classified as improvement (32.4%), stability (52.7%), and decline (14.8%). Residents in the highest variability quartile showed elevated decline rates (30% vs 9%; P=.03), longer median LOS (31 vs 21d; P=.01), and lower community discharge rates (39% vs 67%; P=.02). A supplementary analysis of mean documentation frequency alone yielded no outcome gradient (P=.31), supporting temporal dispersion as the operative signal. All findings are exploratory and hypothesis-generating. CONCLUSIONS: Temporal variability in routine mobility documentation can be operationalized from existing electronic health record data and shows exploratory associations with functional outcomes. The approach requires no additional assessment burden. Multisite prospective validation with multivariable adjustment is needed to establish predictive validity and actionable thresholds.