Introduction. The purpose of this study was to characterize the recovery of physical activity following knee arthroplasty by means of step counts and flight counts (flights of stairs) measured using a smartphone-based care platform. Methods. This is a secondary data analysis on the treatment cohort of a multicenter prospective trial evaluating the use of a smartphone-based care platform for primary total and unicondylar joint arthroplasty. Participants in the treatment arm that underwent primary total or unicondylar knee arthroplasty and had at least 3 months of follow-up were included (n=367). Participants were provided the app with an associated smart watch for measuring several different health measures including daily step and flight counts. These measures were monitored preoperatively, and the following postoperative intervals were selected for review: 2–4 days, 1 month, 1.5 month, 3 months and 6 months. The data are presented as mean, standard deviation, median, and interquartile range (IQR). Signed rank tests were used to assess the difference in average of daily step counts over time. As not all patients reported having multiple stairs at home, a separate analysis was also performed on average flights of stairs (n=214). A sub-study was performed to evaluate patients who returned to preoperative levels at 1.5 months (step count) and 3 months (flight count) using an independent samples T test or Fisher's Exact test was to compare demographics between patients that returned to preoperative levels and those that did not. Results. The mean age of the step count population was 63.1 ± 8.3 years and 64.31% were female, 35.69% were male. The mean body mass index was 31.1 ± 5.9 kg/m. 2. For those who reported multiple stairs at home, the mean age was 62.6 ± 8.3 years and 62.3% were female. The mean body mass index was 30.7 ± 5.4 kg/m. 2. . As expected, the immediate post-op (2–4 days) step count (median 1257.5 steps, IQR 523 – 2267) was significantly lower than preop (median 4160 steps, IQR 2669 – 6034, p < 0.001). Approximately 50% of patients returned to preoperative step counts by 1.5 months postoperatively with a median 4,504 steps (IQR, 2711, 6121, p=0.8230, Figure 1). Improvements in step count continued throughout the remainder of follow-up with the 6-month follow-up visit (median 5517 steps, IQR 3888 – 7279) showing the greatest magnitude (p<0.001). In patients who reported stairs in their homes, approximately 64% of subjects returned to pre-op flight counts by 3 months (p=0.085), followed similar trends with significant improvements at 6 months (p=0.003). Finally, there was no difference in age, sex, BMI, or operative knee between those that returned to mean preoperative step or flight counts by 1.5 months and 3 months, respectively. Discussion and Conclusion. These data demonstrated a recovery curve similar to previously reported curves for patient reported outcome measures in the arthroplasty arena. Patients and surgeons may use this information to help set goals for recovery following total and unicondylar knee arthroplasty using objective activity measures. For any figures or tables, please contact the authors directly

Introduction. Total knee arthroplasty (TKA) is highly successful due to pain reduction, patient satisfaction, and increased range of motion (ROM) during activities of daily living (ADL). ROM recovery is critical for successful outcomes, however ROM values are typically captured during routine physical therapy (PT) appointments via simplified measures (e.g. goniometric maximum passive ROM). These measures are imprecise, potentially neglecting patients’ home experiences. Accordingly, improved measurement methods are necessary to realistically represent ROM recovery. A validated inertial measurement unit (IMU) method continuously capturing knee ROM was deployed assessing knee ROM recovery during PT appointments and during patients’ routine daily experiences. Our objectives were to 1) continuously capture knee ROM pre-/post-TKA via IMUs and 2) divide each day's data to PT/non-PT segments comparing ‘gold standard’ ROM measurements (PT periods) with non-invasive home measurements (non-PT periods). Given patients are verbally/physically encouraged during PT, we hypothesized PT and non-PT metrics would be significantly different including 1) greater kinematics, 2) shorter times, and 3) greater activity level during PT compared to non-PT. Methods. Following IRB approval, IMUs captured long duration, continuous (8–12 hours/day, ∼50 days) knee ROM pre-/post-TKA. Post-TKA metrics were subdivided to PT/non-PT time periods including maximum ROM, gait phase ROMs (stance/swing), gait times (stride/stance/swing), and activity level. Clinical ROM and patient reported outcome measures (PROMs) were also captured before/after TKA. Statistical comparisons were completed between pre-TKA, post-TKA PT, and post-TKA non-PT metrics. Correlation analyses were completed between IMU, clinical ROM, and PROMs. Results. 10 TKA patients (3M, 69±13 years) enrolled. Patient compliance with sensor use was high (pre-TKA: 9.7±1.8 hours/day, 6±1 days; post-TKA: 8.5±2.3 hours/day, 37±4 days). Patients received 1.5±0.5 standard PT sessions per postop week (1.1±0.2 hours/session). All patients were well-healed at 6-weeks post-TKA with no additional surgical interventions required. Dividing ROM data showed distinct qualitative differences between PT and non-PT periods. Specifically, maximum knee ROM was significantly less during PT than outside PT during late rehabilitation weeks. Additionally, PT stance and swing phase ROM were significantly greater during PT throughout recovery. No differences in stride/stance/swing time were noted between PT/non-PT periods during recovery. However, significant activity level differences were noted throughout recovery. Discussion. This study highlights limitations utilizing clinic captured ROM establishing recovery. Notably, IMU ROM measurements allow capturing richer information than discrete simplified clinical measures. Maximum flexion during PT was likely less than non-PT due to exercises completed (i.e. high passive ROM vs. low ROM gait performance). PT gait flexion was likely greater than non-PT because of ‘white coat effects’ wherein patients are closely monitored clinically, whereas gait is completed as desired at home. Interestingly, temporal metrics were equal between PT/non-PT implying clinician's presence encourages improved kinematics but has no impact on ambulation timing. Activity level was significantly greater during PT than non-PT likely resultant from clinicians enforcing high activity levels throughout PT sessions. In total, these results imply data captured clinically represents optimum patient performance whereas data captured non-clinically represents realistic patient performance. For any figures or tables, please contact authors directly