1 Introduction

Unplanned acute care use, including emergency department (ED) visits and hospital admissions, is a major driver of healthcare costs following total knee arthroplasty (TKA).¹ Despite the overall success of TKA as a durable treatment for activity-limiting osteoarthritis—with national registries reporting under 5 % revision rates at 10 years^2–5—postoperative recovery is increasingly challenged by a more complex patient population along with shifts in surgical practice and recovery protocols. Patients undergoing TKA today often have more comorbidities, a factor associated with greater risk of postoperative complications and unplanned visits in the early postoperative period.^6,7

Close

Fig. 1 Kaplan-Meier estimation of the acute care (survival) function for the smart implantable device (SID) and control groups. The dotted lines show the median post-operative day of the acute care visits for each group.

Export to PPT

Close

Fig. 2 Frequency (%) of primary diagnosis associated with acute care visit (emergency department of hospital) for SID (orange) and controls (blue). The SID group had fewer acute visits associated with knee pain, acute postprocedural pain, and chest pain.

Export to PPT

The 90-day postoperative window has become the standard benchmark for monitoring admissions and acute care utilization, as most complications occur within this period and bundled payment models, such as CMS Comprehensive Care for Joint Replacement, define the 90-day episode as the period for which providers are held financially responsible.¹ ED visits in this timeframe are commonly prompted by pain, wound concerns, limb swelling, and medical complications including cardiopulmonary and neurological events.^6,7

The concept that patient monitoring with smart implantable devices (SIDs) could impact acute care resource use has been well explored in other domains including remote monitoring of pacemakers and implantable cardioverter defibrillator devices in cardiology, continuous glucose monitoring in diabetes, and pulse oximetry monitoring in COVID patients. These have been shown to reduce unplanned healthcare encounters and enhance outcomes.⁸ Patients with access to relevant health data experience increased confidence in their health status, reducing acute care usage.⁹

Current advances in TKA design allow for sensors to be implanted within a tibial extension of the knee prosthesis to create smart knee implants,¹⁰ thereby raising similar questions about post-TKA resource allocation. Frequent, objective activity and gait quality data following TKA with SIDs are available to the patient and clinician during the postoperative recovery period. Beyond musculoskeletal recovery, activity metrics such as step count, walking speed, and cadence may also provide indirect insight into cardiopulmonary status—the leading cause of readmission after TKA.^11,12 Gait data facilitate satisfactory gait rehabilitation outcomes and may support earlier recognition of complications such as stiffness or instability. Monitoring may also help reassure patients who might otherwise present to the ED for pain or swelling and allow providers to direct care through targeted outpatient interventions, including physiotherapy or medication adjustments.

Here we present the first administrative claims analysis directly comparing traditional to SID TKAs, focusing on 90-day resource utilization (ED visits and hospital admissions) and cumulative all-cause costs. We hypothesized that SID TKA patients would have lower 90-day acute care utilization and associated costs compared with patients who received traditional implants.

2 Methods

3 Results

The analysis included 1621 patients with SID knees (1709 TKAs) and 3041 patients (3418 TKAs) implanted with traditional knees. Due to the large sample sizes from this claims data analysis, the p-values indicated significant differences in age, yet ranges and median values were similar. SID patients were more likely to be women, have commercial insurance and BMI over 30, but less likely to have diabetes, hypertension or hyperlipidemia. The Charlson Comorbidity Index (CCI) was similar between groups (Table 1). Primary diagnoses associated with acute care events differed for relative frequency of other acute postprocedural pain (G8918) and chest pain (R0789 & R 079).

Cumulative incidence of ED visit (SID: 11.4 %; control: 15.6 %), hospital admission (SID: 6.0 %; control: 8.8 %), or any acute care (ED or hospital admission) (SID: 14.1 %; control: 19.2 %) was statistically significantly lower in the SID group than the control group (p < 0.05) (Table 2). A Kaplan-Meier curve shows the time dependent acute care usage, illustrating about half the acute care occurs in the first two weeks after surgery, and thereafter a slower, but steady use of acute care continues through 90 days (Figure 1).

Median total cost of all healthcare services from the day after surgery (day 1) to day 90 was significantly lower in the SID group ($3747) than the controls ($5629) (Table 3). The mean (median) cost of caring for patients with SID was 61 % of the cost of control patients in the immediate 90-day post-operative period. Patients that utilized acute care within 90 days of TKA cost 207 % (SIDs median) and 213 % (controls median) more than their counterparts who did not use acute care. SID patients also had lower costs during day 91–365. The SID acute care group had significantly lower median acute care cost ($7774) than the control group ($12,012) in the day 1 to day 90 timeframe, and in the day 91–365 period ($5322 and $8,454, respectively) (Table 3).

The top 10 primary diagnoses associated with acute care usage are related to pain, hypertension, syncope, and the diagnoses of osteoarthritis or presence of an artificial knee (Figure 2). Primary diagnoses were grouped with the first 5 digits of the codes. Pain in the Knee (M2556∗) was the most common primary diagnosis, followed by Other acute postprocedural pain (G8918∗) and hypertension (I10∗). Differences between groups were not statistically significant.

4 Discussion

In patients undergoing primary TKA, SID use was associated with fewer ED visits, fewer hospital admissions, and lower all-cause cost compared with control (traditional TKA) patients. These findings suggest that remote monitoring via the SID platform can deliver tangible benefits for both patients and the healthcare system by reducing unplanned care utilization and associated expenditures.

Our results are consistent with prior work in orthopedics demonstrating the value of remote patient monitoring (RPM) and patient facing electronic health apps. Chahal et al. reported significantly fewer ED visits (13 % vs 25.8 %) and hospital admissions (3.5 % vs 8.3 %) among joint arthroplasty patients enrolled in an RPM program compared with non-monitored patients.¹³ Similarly, a randomized clinical trial involving hip and knee arthroplasty patients showed a marked reduction in rehospitalization rates in the RPM group (3.4 % vs 12.2 %, p = 0.01).¹⁴ Mobile health apps have been shown to reduce opiate use for knee pain after TKA based on app feedback that advised pain medication use, exercise or rest and when to call the clinic.⁹ These data support the current study's observation that providing patients and clinicians with real-time recovery information can reduce escalation to acute care.

The value of RPM has also been firmly established in other fields, particularly cardiology. In conditions such as hypertension, heart failure, and atrial fibrillation, RPM has been shown to lower acute care use by enabling clinicians to identify complications early and intervene before patient decline.^15–18 Our findings extend this evidence base to orthopedics, supporting the growing role of implantable monitoring technologies in postoperative care.

Our findings are also consistent with prior claims-based analyses; the Humana dataset reported by Burnett et al. demonstrated substantial post-TKA 90-day episode costs, with a mean cost, defined as allowed amount, of $17,331.¹⁹ Burnett's work highlights the value of claims analyses for understanding post-TKA resource use, and the present study follows the same line of inquiry using a different data source. Where Humana provides a well-characterized single-payer view, the multi-payer structure of the Komodo dataset captures encounters across both commercial and public payers, offering an expanded perspective on post-TKA utilization in a broader patient population. In the current study, cost data were calculated using Komodo Health's validated imputation methodology, which estimates allowed amounts based on CMS pricing rules and benchmarks from over 4.4 billion claims. This approach has been shown to mirror external claims database benchmarks such as Health Care Cost Institute (HCCI) and Merative MarketScan, ensuring consistency and reliability of cost estimates across health plan groups (or “insurance markets”).[Komodo Health White Paper, 2023]

Beyond clinical outcomes, the cost implications are substantial. Unplanned admissions within 90 days of TKA have been shown to cost between $7000 and $12,000 per event, with aggregate costs adding roughly $300–$700 per case.^20,21 These expenses tend to cluster in the early postoperative period and account for a large portion of the bundled 90-day episode of care.^22,23 Even small reductions in ED visits or hospital admissions—as seen in this study—could therefore result in meaningful cost savings.

Several factors may explain the lower post-acute care use observed among SID patients in this study. Although the SID does not directly assess the wound, continuous, objective recovery data can alert clinicians to early deviations in mobility or activity patterns that often accompany wound-related issues (e.g., unexpected reductions in step count or gait quality). These signals may prompt earlier outpatient evaluation of potential wound concerns or stiffness, enabling timely interventions that can prevent ED visits or admissions. Real-time feedback may also improve patient confidence and reduce acute care visits driven by patient anxiety or uncertainty.

Although research has shown that patient expectations influence TKA outcomes,^24–26 preoperative education alone has not consistently improved outcomes.^27,28 Even with consistent communication from surgeons, health literacy and preoperative anxiety may preclude information retention.²⁹ Providing patients with a clear, concrete reference schema—such as age- and gender-adjusted recovery curves—helps them understand expected outcomes and offers an objective basis for efficient and effective patient-provider communication.

Early feedback from SID clinician users also suggests that patients call the office less frequently during recovery, though this observation requires formal evaluation. Incorporating SID data into routine care may further support individualized modification of rehabilitation plans, allowing providers to optimize physical therapy. Knowing that their activity and gait are being monitored may reassure patients and motivate them to adhere more closely to rehabilitation recommendations. Although this study was not designed to test the above mechanisms directly, they offer reasonable explanations for the observed reduction in post-acute care use and warrant further investigation in prospective research.

The current study has several limitations. It was retrospective and relied on claims data, which do not capture outcomes such as patient-reported pain, function, or satisfaction. Claims are collected for billing rather than research, so they are subject to coding errors and missing information, though these issues would likely affect both SID and traditional groups equally. In addition, SID-derived activity data in the first postoperative week are often incomplete as patients set up and synchronize their devices, which limits the reliability of early postoperative comparisons; therefore, any difference in acute care use during days 0–7 cannot be attributed with confidence to monitoring. The dataset included only insured individuals, which may limit generalizability to uninsured or underinsured populations. SID patients may represent a different socioeconomic or engagement group, e.g., more tech savvy, higher compliance or treated at higher resource centers. The costs available in the database are estimated using the proprietary expected payment, which may over or underestimate actual payments. Despite these limitations, the findings provide early evidence that SIDs in TKA may help reduce downstream healthcare use and costs within the critical 90-day postoperative period.

5 Conclusions

Use of SIDs in primary TKA was associated with fewer ED visits and hospital admissions, as well as lower 90-day total healthcare costs. These cost reductions reflect both a lower incidence of unplanned care and lower costs per patient. These findings support the potential of implant-derived recovery data to enhance care coordination, align postoperative management with patient expectations, increase patient confidence in their health status, and reduce the burden of post-TKA acute care within value-based healthcare models. Future prospective studies are warranted to investigate the causal mechanisms of the observed lower acute care utilization in the SID group and to explore how smart implant technology can be effectively integrated into bundled payment and value-based care frameworks.

Guardian/patient's consent

IRB approval/consent was not necessary as the data from the claim database were deidentified in this source file.

Credit author statement

KH: writing, original draft preparation, review, and editing; JY and SL: conceptualization, supervision, and project administration; PA and ES: data collection, analysis, and interpretation; MMEO and SL: writing, review and critical revision. All authors have read and approved the final manuscript and agree to be accountable for all aspects of the work.

Ethics

Not applicable. This article is based on deidentified database data and does not involve human participants or animal studies.

Funding statement

This work was supported by Canary Medical Inc (CMI). The CMI purchased access to the claims database.