You are currently viewing a new version of our website. To view the old version click .
MDPIMDPI
Search all articles
Applied Sciences
Download PDF
  • Review
  • Open Access

11 November 2025

Predictive Factors of Inpatient Rehabilitation Stay After Elective Hip and Knee Replacement: A Scoping Review

and
1
IRCCS Ospedale Galeazzi-Sant’Ambrogio, Scientific Direction, 20157 Milan, Italy
2
Faculty of Medicine and Surgery, Vita-Salute San Raffaele University, 20132 Milan, Italy
*
Author to whom correspondence should be addressed.
Appl. Sci.2025, 15(22), 11957;https://doi.org/10.3390/app152211957
This article belongs to the Special Issue New Insights into Physical Therapy
Version Notes
Order Reprints

Abstract

Patient stratification strategies based on digital databases and advanced information technology can predict inpatient rehabilitation outcomes and support safe hospital discharge for patients who underwent joint replacement for hip and knee osteoarthritis. The degree of continuity between surgery and rehabilitation, the perioperative process integration, and the setting where rehabilitation is provided are crucial factors to improve care effectiveness, access, minimize readmissions, and cost increase. The primary aim of this scoping review of the literature is to identify perioperative variables that are predictive of inpatient rehabilitation stay after hip and knee arthroplasty for osteoarthritis. These factors are divided by time of assessment through the perioperative pathway and surgical procedure site. The secondary aim is to explore how different data sources and facilities are linked into a patient-centered perioperative pathway. An electronic search of the literature was performed on PubMed, Embase, and Scopus. No time restrictions were applied. All primary research studies investigating predictive factors of inpatient rehabilitation stay after hip and knee osteoarthritis were included. In total, 25 studies were included in the review. Age, caregiver presence, presence of comorbidities, sex, Body Mass Index, Risk Assessment and Prediction Tool composite score, pre-operative Clinician-Reported Outcome Measures, pre-operative Patient-Reported Outcome Measures, and post-operative Barthel Index of autonomy in the Activities of Daily Living were predictive of some degree of inpatient rehabilitation stay in more than one study. The studies were fairly distributed between retrospective and prospective, with multicentric databases more spread among the latter. Data collection occurred in acute hospitals more than in specialized rehabilitation facilities. Using comprehensive models supported by electronic health records and powerful information technologies, analyzing specific inpatient rehabilitation LOS as distinguished from surgical ward rehabilitation, using institutional registries, and including specific rehabilitation factors in these registries, and promoting vocabulary and federated data sharing can strongly enhance the predictivity of models investigating rehabilitation outcomes and support appropriate discharge from inpatient rehabilitation units.

1. Introduction

Patient stratification strategies to optimize inpatient length of stay (LOS) after hip and knee replacement have long been advocated for clinical, organizational, and allocative purposes, pushed by the growing burden of hip and knee osteoarthritis in aging societies [,,,,,,]. The elective nature of the procedure enables planning hospital admission, intraoperative arrangements, postoperative rehabilitation, and possibly also prehabilitation programs, promoting multidisciplinary, perioperative research to identify any possible measures to accelerate patient recovery and reduce waiting lists.
Medical progress, population aging, cost containment and sustainability issues have made the reduction in hospital LOS increasingly possible and necessary, resulting in different accelerated protocols like fast-track [], enhanced recovery after surgery [], short stay programs [], and same-day discharge protocols [], maintaining high standards of care, comparable safety [], and non-inferior to better outcomes [,,,]. At the same time, older patients being clinically fit for discharge may prolong inpatient stay because of disadvantaged living conditions and postoperative fragmentation of care, further delaying hospital admissions to the detriment of acute patients [,].
After the pandemic, promising information technologies are becoming more common in joint arthroplasty, not only supporting the reduction in hospital LOS, but making the prediction of hospital LOS easier, through the analysis of large, heterogeneous datasets collected from multiple units, facilities, and professionals [,,]. Clinicians, researchers, and administrators can track socio-behavioral data, patient visits, clinical observations, and procedures across the pathway, potentially distinguish rehabilitation-specific LOS from total hospital LOS, and link these data through the use of electronic patient identifiers. In this way, predictive factors can be more easily identified at different stages, supporting patient-centered, clinical, and organizational arrangements accordingly. Preoperative, intraoperative, and postoperative information on patient’s health, function, and social resources can support an optimal rehabilitation process in advance by making clinicians and administrators able to estimate the time of recovery, the potential complications, design patient-centered pathways (e.g., prehabilitation, nutrition, anesthesia, pain medications, iron diet, physical therapy intensity, Activities of Daily Living recovery), and organize facility shifts efficiently (e.g., from surgical ward to inpatient rehabilitation, and from inpatient rehabilitation to home or primary care facilities). Moreover, medical technology suppliers can identify, develop, and market useful devices to accurately measure clinically relevant, predictive conditions (e.g., wireless wearable accelerometers to measure physical activity levels before surgery) [,].
After the COVID-19 pandemic, substantial financial investments are being addressed to the digitalization of healthcare and the integration of acute and post-acute, intermediate and primary care, ideally merging data into patient-centered information systems able to track the patient’s pathway across different facilities and professionals [,]. Therefore, digital health records, artificial intelligence, and machine learning can improve the perioperative streamlining even further, provided adequate validation measures [,]. The goal is to preserve the healthcare system effectiveness, responsiveness, and sustainability building on proactive population health management across healthcare and community settings, both entering and leaving the hospital (e.g., allowing access to the patient’s history based on integrated medical records, activating post-acute care on time, facilitating safe home discharge, reducing complications and hospital readmissions due to care fragmentation, monitoring parameters from remote areas, preventing the exacerbation of chronic conditions outside hospitals).
In this context, the degree of continuity between surgery and rehabilitation, the perioperative process integration, and the setting where rehabilitation is provided are crucial factors to improve care effectiveness and access and avoid readmissions and financial losses, both in hip and knee arthroplasty [,,]. The more predictive variables are known to clinicians, researchers, and policy makers, the more room for patient-centered arrangements to improve postoperative outcomes and care continuity. In turn, sharing experiences about data collection and linking across different providers is key to a) efficiently map, integrate, and plan retrospective and prospective studies, and b) provide good quality digital sets to train and run artificial intelligence, machine learning, and statistical analyses on big data.
The primary aim of this scoping review of the literature is to identify perioperative variables that are predictive of inpatient rehabilitation stay after elective hip and knee arthroplasty. Then, among the studies identified by the review, we aim to describe how different databases (monocentric and multicentric) or registries (institutional, scientific societies) are linked among different professionals, units, and facilities, considering the importance of data integration across different levels.

2. Materials and Methods

The extended version of the Arksey and O’Malley framework for scoping reviews [] was adopted as a protocol for this research (Table 1). The protocol is also compliant with Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist (Supplementary Material S1).
Table 1. Scoping review protocol.
The Population, Intervention, Control/Comparison, Outcome, and Study (PICOS) framework was adopted to define the research question, inclusion, and exclusion criteria [] (Table 1). Put in the terms of the Population, Concept and Context (PCC) framework, the population is made up of patients undergoing joint replacement for hip and knee osteoarthritis, the concept is the investigation of potentially predictive factors of inpatient rehabilitation stay, and the context is all healthcare facilities providing inpatient rehabilitation stay (both as a part of general hospital LOS, as a part of postoperative surgical ward LOS, as a part of postoperative specialized unit LOS in the same facility where surgery was performed, and as a separated facility from the hospital where surgery was performed). Of the studies included in the research, the data collection methodology (prospective or retrospective) and links (professional/scientific registry, institutional registry, monocentric database, multicentric database) through any different sources and providers (general hospital, surgical ward, inpatient rehabilitation unit) were also considered.
A PRISMA flowchart was developed to describe study selection (Results, Figure 1). Two reviewers selected the studies. It was planned to use a third reviewer to resolve any discrepancies. However, no discrepancies needed to be handled, as population, intervention, outcome, and study design (or concept and context) were clear, and no intervention control or comparison was relevant to our scope. The protocol was not registered before initiating, as permitted by the Arksey and O’Malley extended framework for scoping reviews []. However, the full electronic research, along with data and results, is available in Supplementary Material S1 in order to ensure reproducibility and transparency.
Figure 1. PRISMA 2020 flow diagram (https://www.prisma-statement.org/prisma-2020-flow-diagram accessed on 2 November 2025) for database and register search.
When considering data collection and linking results, data management categories were grouped into professional/scientific society registry, institutional registry, multicentric database, and monocentric database. By data management, we mean the source of data repository and integration. The difference between an institutional registry and a database is the difference between institutional ongoing systems of surveillance and monitoring—see, for instance, European Reference Networks [], the Swedish Knee Arthroplasty Register [], or the Italian Registro Nazionale Artroprotesi []—and local, spontaneous initiatives from one or more hospitals and healthcare providers, no matter how extended in patient population and time, generally requiring approval from local ethical committees. The professional/scientific society registry category was also added to include data-driven, monitoring registries from the clinical records available to healthcare professionals (e.g., The American College of Surgeons National Surgical Quality Improvement Program) [].
Data sources were grouped into hospital (if data were collected from a hospital, either including surgical and/or subsequent inpatient rehabilitation stay) and inpatient rehabilitation facility (if data were collected from an external inpatient rehabilitation facility, like a physical medicine and rehabilitation center). For example, if a study is performed in a hospital providing both surgery and subsequent inpatient rehabilitation in the same facility, it is categorized as a hospital study. Otherwise, if a study is performed into a physical medicine and rehabilitation center that also retrieves a patient’s history from the previous surgical hospitalization, it is categorized as an inpatient rehabilitation facility study.

3. Results

In total, 66 studies were initially identified from 3 databases; 56 studies resulted after removing duplications, and 25 final studies resulted after applying PICOS criteria from abstract screening to full text reading (Figure 1). The results of the review are represented in Table 2.
Table 2. Final results.

3.1. Predictive Factors of Inpatient Rehabilitation Stay

The results of the literature review are grouped in Table 3, representing the number of factors investigated (“Analyzed”) and actually found to be predictive (“Predictive”). These factors are divided per time of assessment through the perioperative pathway (“Pre-existing patient characteristics; Pre-operative factors; Intra-operative factors; Post-operative factors”) and surgical procedure site (“Knee procedures; Hip procedures; Both procedures in the same study”).
Table 3. Predictive factors of inpatient rehabilitation stay data chart.
Pain was considered separately from other Patient-Reported Outcome Measures (PROMs), being a core symptom of osteoarthritis. BI was also considered separately from other Clinician-Reported Outcome Measures (CROMs), because of the regulatory importance it may have in allowing or not allowing hospital discharge []. The resulting PROMs employed for predictivity analysis were EQ-5D, GDS, HOOS-JR, IPAQ–Short Form, KOOS-JR, LEAS, Lequesne Index, MOSS-SS, OKS, PCS, PROMIS-Mental and Physical Health, SF-36, Veterans RAND-12 Items core–Mental and Physical Health, and WOMAC-function. The resulting CROMS employed for predictivity analysis were the Braden Scale, FIM, HHS, KSFS, NSQIP–Functional Status, MMSE, and Morse Fall Scale. IKSS combines both Patient-Reported and Clinician-Reported Measures; therefore, it was included in both categories.
Among the pre-existing patient characteristics analyzed, age was the most common predictive factor (7 times predictive up to 13 times investigated), followed by caregiver presence (4 up to 6), sex (4 up to 10), occupation, race, previous leg arthroplasty, use of drugs (NSAIDS, Aspirin), heart disease (1 up to 1), CCI (1 up to 2), anatomical region affected in case of same-study hip and knee population (1 up to 3), and general presence of comorbidities (2 up to 9). Area of residence, blood-borne transmissible infections, chronic steroid use, other specific comorbidities, educational level, marital status, payer status, Scottish Index of Multiple Deprivation, and smoking status demonstrated no predictive value in the studies included.
Among the pre-operative factors analyzed, PROMs (pain excluded) were 3 times predictive up to 15 times investigated, followed by composite scores (2 up to 3), CROMs (2 up to 8), BMI (2 up to 11), hospitalization year (1 up to 1), non-operated limb weight (1 up to 1), pain catastrophizing (1 up to 2), pain (1 up to 4), clinical chemistry (1 up to 6) and ASA score (1 up to 8). Expectations on pain and function, performance, physical activity measures, prehabilitation, and vital signs demonstrated no predictive value in the studies included.
Among the intra-operative factors analyzed, the operating surgeon was 1 time predictive up to 1 time investigated, followed by day of surgery in the week, duration of surgery/operational time, and blood loss/transfusion(s) occurred (1 up to 3). Surgical access technique and type of anesthesia demonstrated no predictive value in the studies included.
Among the post-operative factors analyzed, BI was 2 times predictive up to 3 times investigated, followed by physical activity measures (1 up to 1) and verticalization day (1 up to 2). Analgesia protocol, autonomy in ADL (BI excluded), autonomy in postural transitions, clinical chemistry, cognitive function, composite score, pain, performance measures, physiotherapy protocol, PROMs, surgical ward LOS, and walking autonomy and/or distance demonstrated no predictive value in the studies included.
In total, 17 studies included surgery and rehabilitation in a single hospital LOS. Eight studies considered inpatient rehabilitation LOS separately, either in the same hospital (e.g., patients moved from the surgical ward to an internal specialized rehabilitation unit) or in a separate facility (e.g., physical medicine and rehabilitation center). Among the latter, two studies considered surgical ward LOS as a potential predictor for inpatient rehabilitation LOS: one in which surgical stay and specialized rehabilitation stay were part of the same inpatient consecutive pathway, in the same orthopedic hospital [], and another in which a specialized rehabilitation center received patients from five different hospitals [].
Coudeyre et al. [] considered the place of discharge after surgical ward stay as a primary study outcome (home vs. physical medicine and rehabilitation center), and a composite RAPT score as a potential predictor. An RAPT score greater than 9 was found to be actually predictive of home discharge. Among the 25 patients with an RAPT score greater than 9, 5 patients were actually discharged to a physical medicine and rehabilitation center, as they lived in rural, remote areas. These patients had a significantly shorter inpatient rehabilitation LOS than other patients in the same cohort. Therefore, an RAPT score greater than 9, combined with living in an isolated location, was predictive of a shorter inpatient LOS compared with other patients who underwent rehabilitation in an external physical medicine and rehabilitation center. Then, the study was eligible for inclusion.

3.2. Data Collection and Linking

In total, 14 studies were conducted retrospectively and 11 studies were conducted prospectively. In total, 20 studies were completed, and 5 were protocols. Skov et al. [] and Skov [] are possibly duplicates of the same study protocol, the former in the version of a clinical trial registration and the latter in the version of a peer-reviewed publication, as the only slight difference is the expected population size (150 vs. 200). In total, 21 studies were observational and 4 were interventional (randomized controlled trials, RCT).
In terms of data management and source, prospective data collection was performed seven times through a monocentric database and three times through a multicentric database, in all cases based on a hospital database. Two of these [,] are possibly duplicates of the same protocol. No prospective data collection was performed through an inpatient rehabilitation facility. Retrospective data collection was performed nine times through a monocentric database and three times through a multicentric database. Two of the nine retrospective studies based on a monocentric database retrieved data from a specialist inpatient rehabilitation facility [,]. No studies retrieved data from institutional registries. One study retrieved data from a scientific society registry [].
More specifically, regarding the way these data were collected and linked, among the prospective studies, eight involve specific data collection that is not further detailed. One RCT protocol [] involves electronic data collection with RedCap software. Two observational, completed studies involved multiple sources of collection and dedicated data linking (internal clinical records, administrative database, and further specific data manually collected) [,]. Dedicated data linking means that no pre-established platform existed to allow for routine data collection (e.g., an electronic, integrated health record system), but the dataset is created specifically for the single study and progressively populated.
Among the retrospective studies, 12 used internal data; 5 collected data from clinical records, 5 collected data from an administrative database, registry, or both, and 2 used manually linked data. One drew data from the multicenter registry of a professional society44 and one used a single hospital’s internal data without providing further details []. Data collection and linking are represented in Table 4.
Table 4. Data collection and linking.

4. Discussion

Inpatient rehabilitation protocols are fairly standardized [,] and so can be the criteria for whether or not to authorize discharge depending on local regulation [,]. However, no literature reviews specifically investigate the predictive factors of inpatient rehabilitation LOS to our knowledge, despite the promising advancements of information technology. Moreover, the majority of studies resulting from this research did not distinguish between inpatient rehabilitation stay and acute or surgical stay, despite the necessity of developing specific model predictions of inpatient rehabilitation LOS [,].
Older age was the most frequent predictor of post-operative inpatient rehabilitation, both considered alone and associated with other perioperative factors, including pre-operative function (assessed with Clinician-Reported Outcome Measures) [,], performance (walking autonomy, distance, aids) [,], comorbidities (either in general or aggregated by CCI index) [,,,] and presence of a caregiver providing extra-hospital support [,,]. This is in line with the literature [,,,,,,] and supports the recommendation to consider age as part of a more comprehensive, pre-operative assessment, including potentially confounding factors like function and burden of care []. Interestingly, all the studies included here that found extra-hospital support as a predictor of postoperative LOS specifically considered inpatient rehabilitation unit LOS [,,], while the studies that did not find such a factor predictive generally used postoperative hospital LOS [,]. This finding suggests increasing awareness towards the key role played by social support in rehabilitation outcomes [].
The presence of extra-hospital, social, and household support was also recalled to hypothesize why male patients tend to have shorter LOS after hip and knee replacements in comparison to females [,,], as having someone who cares for them at home more frequently can increase the motivation to leave the hospital []. Females having longer postoperative LOS is consistent with the results of this review, according to four multivariate analyses, two focusing specifically on the inpatient rehabilitation unit [,] and two using general hospital LOS [,]. Regardless of the reasons that may explain this trend, it is prudent to consider the possibility of an extended LOS when patients undergoing rehabilitation are females. If the hypothesis about having a caregiver at home less frequently is true, taking into consideration the need for support as early as possible could improve safety and reduce inpatient stays.
Taking social factors into consideration could also have made the RAPT score more predictive [,] than other composite scores, including age, sex, comorbidities, and early post-operative performance []. The RAPT score was originally developed to predict the need for extended inpatient rehabilitation stay, combining age, sex, preoperative functional capacity with or without technical aids, social support at home, living alone, and patient’s preference towards discharge destination into a single, numerical output; the higher the score, the higher the possibility of prosecuting a rehabilitation outpatient, stratifying patients into three levels of risk (high, intermediate, and low) [].
In the two articles of this review where the RAPT score was investigated, however, it also demonstrated predictive value in terms of inpatient rehabilitation LOS. LeBrun et al. [] found RAPT to be a strong predictor of postoperative inpatient LOS even after controlling for other relevant covariates, testing the score on almost 18,000 patients, of whom 92% were discharged home. Coudeyre et al. [] found an RAPT score ≥9 to be predictor of home discharge; however, among the patients actually being clinically ready for home discharge, 20% were unable to comply with outpatient rehabilitation because of disadvantaged home arrangements, and were transferred to a physical medicine and rehabilitation unit, where they reported a shorter LOS in comparison to the inpatient rehabilitation LOS of other patients, presumably discharged here primarily because of a clinical need.
Interestingly, the possibility of using the score by modulating thresholds and patient characteristics makes the prediction adaptable to the local healthcare system needs [,]. The focus on key functional and social domains made the RAPT practical for routine preoperative assessment, with a good performance on predicting discharge destination in high-risk or low-risk patients; however, the predictive value seems to diminish in the case of intermediate-risk patients, losing granularity between the extremes [,,]. Taking advantage of big data analysis, machine learning, and electronic health records, possibly combined [], can provide more accurate predictions for more clusters of patients, intermediate risk included.
No singular PROM or CROM questionnaire resulted in predictive inpatient rehabilitation LOS in more than one study, either pre-operative or post-operative. Considering PROMs and CROMs on an aggregate level, three pre-operative PROMs (PCS; MOS-SS; SF-36-general health perception) demonstrated predictive value in two studies [,], and two pre-operative CROMS (HHS–function and activity; NSQIP–Functional Status) demonstrated predictive value in two studies [,]. All the results of this review investigated the predictive value of PROMs (either pre-operative and post-operative) along with other objective factors, including inpatient rehabilitation stay as a separate outcome from surgical ward stay [,]. Thanks to the possibility of integrating PROMs in routine clinical workflow [], the use of PROMs as predictors of post-operative LOS reported good-to-fair results in large populations. For instance, pre-operative PROMs (EQ-5D-5L, EQ-Visual Analog Scale, HOOS-PS, KOOS-PS, PROMIS-Fatigue and Depression Short Forms) collected on almost 5,000 hip and knee replacements demonstrated predictive value for post-operative LOS when analyzed with machine learning and regression models, providing cues to clinicians and healthcare administrators for capacity planning, patient expectation management, and patient risk management []. Also, the integration of ordinary clinical workflow records and a dedicated PROM database into a patient-centered, perioperative pathway (including post-discharge follow-up at 3 months) could predict the ideal rehabilitation protocol even before the assignment of surgery based on a machine-learning, second opinion []. On the contrary, pre-operative PROMs (KOOS-JR, PROMIS-Mental and Physical Health) collected on almost 3,000 knee replacements could predict extended postoperative stay, but not as effectively as the RAPT score did []. However, two of these studies [,] did not distinguish between surgical ward rehabilitation and specialist inpatient rehabilitation stay (in the same hospital or elsewhere).
In terms of predictive, pre-operative CROMs, both the questionnaires included in this review assessed patient function before surgery (HHS–function and activity; NSQIP–Functional status), both were collected from patients undergoing hip replacement, and both measured inpatient rehabilitation stay without distinguishing between the surgical ward and the specialist unit [,]. The introduction of prehabilitation in hip and knee elective arthroplasty is based on the hypothesis that functional improvement before surgery improves recovery outcomes [], post-operative LOS reduction included [,], although the beneficial effect of prehabilitation itself in reducing postoperative LOS remains controversial [,,,]. In order to better understand the extent to which prehabilitation can reduce postoperative LOS while also improving postoperative outcomes (e.g., pain and function) [,], it would be important to capture whether postoperative recovery measures are limited to surgical ward rehabilitation or extended rehabilitation units, specifying the facility where inpatient rehabilitation is provided, both in primary studies and in meta analyses or systematic reviews.
Pre-operative BMI also resulted predictive of inpatient rehabilitation stay in this review. The role played by BMI in the spread of osteoarthritis is generally accepted [] and used for planning operating lists, in combination with other factors []. Both the studies in which BMI demonstrated predictive value were conducted on knee arthroplasty, one considering general hospital stay [], and one distinguishing between surgical ward and specialist rehabilitation unit 38]. The former study found that BMI ≥ 25 predicted 2.3 more hospitalization days, while the latter study found that BMI ≥ 30 predicted one fewer rehabilitation unit and overall hospitalization day (surgical ward followed by specialist unit). The other nine studies that investigated BMI in this review did not find such a variable to be predictive. These findings not only confirm the controversial relation between overweight and postoperative LOS [,,], but possibly make it even more controversial, as being overweight had a protective value in terms of inpatient rehabilitation stay, not explainable by a concurrent extension of surgical ward stay, as this latter factor did not vary in relation to BMI [].
BI (and subsequent modifications) is a rank-order scale that measures disability impairments to complete ADL, and as such, it is used to measure rehabilitation improvements after surgical interventions []. The higher the score, the more independent the patient is in completing ADL, and in turn, the more likely the patient is to return home after a hospital stay. A total score ≥ 75 is commonly considered enough to discharge a patient at home, and along with a 30% improvement from inpatient rehabilitation to discharge, it is used as a threshold to authorize patient discharge in orthopedic hospitals of Italy. This is a reasonable explanation of why all the studies that adopted postoperative BI as a potential predictor of inpatient rehabilitation stay are all based in the same region, and all these studies (one concerning hip and knee procedures, two concerning knee procedures only) specifically measured inpatient rehabilitation unit stay [,,]. Two of them adopted inpatient rehabilitation stay in a specialist unit as a primary outcome, and found such a factor to be predictive (the higher the score, the shorter the LOS) [,].
Multicentric databases were more often employed in prospective studies (4 times up to 11) than in retrospective studies (2 times up to 14). Planning data collection from more than one facility is actually more common and easier than retrieving data from multiple facilities, which needs post hoc standardization procedures. Institutional registries sharing data from more facilities should facilitate data standardization and retrieval. However, no studies resulting from this review employed such registries, including prospective studies. Only Raad et al. [] retrieved data from the Registry of a Scientific Society, adding useful clinical information for research and quality improvement than insurance-based administrative databases []. Monocentric databases were fairly distributed between prospective studies (7 times up to 11) and retrospective studies (9 times up to 14). Monocentric databases should be easier to employ as long as data collection does not need the same standardization efforts that are needed among different facilities.
Only two studies, both retrospective, retrieved data from monocentric rehabilitation facilities. This may suggest an underdeveloped data infrastructure in specialist rehabilitation centers compared to acute hospitals. Shedding more light on these facilities could improve patient’s health monitoring after hospital discharge, possibly using integrated (and therefore multicentric) datasets accessible by different healthcare providers (e.g., hospital surgeon, inpatient rehabilitation specialist, outpatient rehabilitation specialist, and general practitioner), researchers, healthcare administrators, and policy makers (e.g., to capture gaps in care continuity and find locally fit solutions). The role of digitalization in promoting user-friendly, integrated health records here is clear, along with the role of AI and machine learning analyses in support of patient stratification, outcome prediction, and continuity of care planning and strengthening [], provided adequate protection measures.
This review has a number of limitations. First, the results of predictive analyses depend on the model. Second, multiple factors represented in Table 3 were predictive in only one study or a few more. Third, none of the studies included completed an external validation process, confirming the model’s predictivity in another population []. However, we recall that the goal of this scoping review was not to provide evidence-based guidelines for developing ready-to-use prediction models in any context, nor to assess study quality and bias in support of clinical interventions. Rather, the goal was to review the existing practices in order to help healthcare professionals, researchers, and policy makers develop (or improve) specific prediction models in specific situations, having access to similar real-world studies (musculoskeletal population, facility setting, rehabilitation pathway, data management, local regulation, factors most relevant, and factors available). Any time a certain predictor in a certain context seems relevant to the needs of the reader, it is possible to read the methodological details of each study and take them into careful consideration before replying.

5. Conclusions

The idea behind real-world patient stratification for outcome prediction is that we should not seek the optimal outcome (in this case, inpatient rehabilitation LOS) for any patient, in any situation, but the realistic rehabilitation outcome for a certain type of patient, in a certain clinical, social, and organizational context, driving the most appropriate clinical and policy arrangements.
Age, caregiver presence, presence of comorbidities, sex, BMI, RAPT composite score, pre-operative CROMS, pre-operative PROMs, and post-operative Barthel Index were predictive to some degree of inpatient rehabilitation LOS in more than one study. The studies were fairly distributed between retrospective and prospective, with multicentric databases more spread among the latter. Data collection occurred in acute hospitals more than in specialized rehabilitation facilities.
  • Using comprehensive models supported by electronic health records and powerful information technologies;
  • Analyzing specific inpatient rehabilitation LOS as distinguished from surgical ward rehabilitation;
  • Using institutional registries, and including specific rehabilitation factors in these registries;
  • Promoting vocabulary standardization (including inpatient rehabilitation LOS across different countries) and federated data sharing;
    The above can strongly enhance the predictivity of models investigating rehabilitation outcomes and support earlier discharge from acute wards when appropriate.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/app152211957/s1, File S1: Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist.

Author Contributions

Conceptualization, F.P. and G.B.; methodology, F.P. and G.B.; investigation, F.P.; writing—original draft preparation, F.P.; writing—review and editing, F.P.; supervision, G.B. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported and funded by the Italian Ministry of Health, “Ricerca Corrente”.

Institutional Review Board Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
ADLActivities of Daily Living
ASAAmerican Society of Anesthesiologists
BIBarthel Index
BMIBody Mass Index
CCICharlson Comorbidity Index
CROMsClinician-Reported Outcome Measures
EQ-5DEuroQol-5 Dimensions
FIMFunctional Independence Measure
GDS Geriatric Depression Scale
HHSHarris Hip Score
HOOS-JRHip dysfunction and Osteoarthritis Outcome Score–Joint Replacement
IKSSInternational Knee Society Score
IPAQInternational Physical Activity Questionnaire
KOOS-JRKnee dysfunction and Osteoarthritis Outcome Score–Joint Replacement
KSSKnee Society Score
LEASLower Extremity Activity Scale
LOSLength Of Stay
MMSEMini-Mental State Examination
MOSS-SSMedical Outcomes Study Social Support Scale
NICENational Institute for health and Care Excellence

References

  1. Nham, F.H.; Patel, I.; Zalikha, A.K.; El-Othmani, M.M. Epidemiology of primary and revision total knee arthroplasty: Analysis of demographics, comorbidities and outcomes from the national inpatient sample. Arthroplasty 2023, 5, 18. [Google Scholar] [CrossRef]
  2. Patel, I.; Nham, F.; Zalikha, A.K.; El-Othmani, M.M. Epidemiology of total hip arthroplasty: Demographics, comorbidities and outcomes. Arthroplasty 2023, 5, 2. [Google Scholar] [CrossRef]
  3. Simões, J.L.; Soares, S.; Sa-Couto, P.; Lopes, C.; Magina, D.; Melo, E.; Voegeli, D.; Bolhão, I. The Influence of Presurgical Factors on the Rehabilitation Outcome of Patients Following Hip Arthroplasty. Rehabil. Nurs. 2019, 44, 189–202. [Google Scholar] [CrossRef] [PubMed]
  4. Singh, J.A.; Yu, S.; Chen, L.; Cleveland, J.D. Rates of Total Joint Replacement in the United States: Future Projections to 2020–2040 Using the National Inpatient Sample. J. Rheumatol. 2019, 46, 1134–1140. [Google Scholar] [CrossRef] [PubMed]
  5. Lin, J.J.; Kaplan, R.J. Multivariate analysis of the factors affecting duration of acute inpatient rehabilitation after hip and knee arthroplasty. Am. J. Phys. Med. Rehabil. 2004, 83, 344–352. [Google Scholar] [CrossRef]
  6. Centers for Disease Control and prevention. Public Health and Aging: Projected Prevalence of Self-Reported Arthritis or Chronic Joint Symptoms Among Persons Aged >65 Years --- United States, 2005—2030; Atlanta: Morbidity and Mortality Week Report; CDC: Washington, DC, USA, 2003. Available online: https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5221a1.htm (accessed on 7 October 2025).
  7. Weingarten, S.; Riedinger, M.; Conner, L.; Siebens, H.; Varis, G.; Alter, A.; Ellrodt, A.G. Hip replacement and hip hemiarthroplasty surgery: Potential opportunities to shorten lengths of hospital stay. Am. J. Med. 1994, 97, 208–213. [Google Scholar] [CrossRef]
  8. Kehlet, H. Fast-track hip and knee arthroplasty. Lancet 2013, 381, 1600–1602. [Google Scholar] [CrossRef]
  9. Riga, M.; Altsitzioglou, P.; Saranteas, T.; Mavrogenis, A.F. Enhanced recovery after surgery (ERAS) protocols for total joint replacement surgery. SICOT J. 2023, 9, E1. [Google Scholar] [CrossRef] [PubMed]
  10. Dawson-Bowling, S.J.; Jha, S.; Chettiar, K.K.; East, D.J.; Gould, G.C.; Apthorp, H.D. A multidisciplinary enhanced recovery programme allows discharge within two days of total hip replacement; three- to five-year results of 100 patients. Hip Int. 2014, 24, 167–174. [Google Scholar] [CrossRef]
  11. Jensen, C.B.; Troelsen, A.; Foss, N.B.; Nielsen, C.S.; Lindberg-Larsen, M.; Gromov, K. 10-year evolution of day-case hip and knee arthroplasty: A Danish nationwide register study of 166,833 procedures from 2010 to 2020. Acta Orthop. 2023, 94, 178–184. [Google Scholar] [CrossRef]
  12. Lan, R.H.; Samuel, L.T.; Grits, D.; Kamath, A.F. Contemporary Outpatient Arthroplasty Is Safe Compared with Inpatient Surgery: A Propensity Score-Matched Analysis of 574,375 Procedures. J. Bone Jt. Surg. Am. 2021, 103, 593–600. [Google Scholar] [CrossRef] [PubMed]
  13. Cheung, A.; Fu, H.; Cheung, M.H.; Chan, W.K.V.; Chan, P.K.; Yan, C.H.; Chiu, K.Y. How well do elderly patients do after total knee arthroplasty in the era of fast-track surgery? Arthroplasty 2020, 22, 16. [Google Scholar] [CrossRef]
  14. Berg, U.; W-Dahl, A.; Rolfson, O.; Nauclér, E.; Sundberg, M.; Nilsdotter, A. Influence of fast-track programs on patient-reported outcomes in total hip and knee replacement (THR/TKR) at Swedish hospitals 2011–2015: An observational study including 51,169 THR and 8,393 TKR operations. Acta Orthop. 2020, 91, 306–312. [Google Scholar] [CrossRef]
  15. Vanni, F.; Foglia, E.; Pennestrì, F.; Ferrario, L.; Banfi, G. Introducing enhanced recovery after surgery in a high-volume orthopaedic hospital: A health technology assessment. BMC Health Serv. Res. 2020, 20, 773. [Google Scholar] [CrossRef]
  16. Sarpong, N.O.; Boddapati, V.; Herndon, C.L.; Shah, R.P.; Cooper, H.J.; Geller, J.A. Trends in Length of Stay and 30-Day Complications After Total Knee Arthroplasty: An Analysis From 2006 to 2016. J. Arthroplast. 2019, 34, 1575–1580. [Google Scholar] [CrossRef]
  17. Siciliani, L.; Wen, J.; Gaughan, J. Living alone and provider behaviour in public and private hospitals. J. Health Econ. 2025, 102, 103016. [Google Scholar] [CrossRef]
  18. Majeed, M.U.; Williams, D.T.; Pollock, R.; Amir, F.; Liam, M.; Foong, K.S.; Whitaker, C.J. Delay in discharge and its impact on unnecessary hospital bed occupancy. BMC Health Serv. Res. 2012, 12, 410. [Google Scholar] [CrossRef]
  19. LeBrun, D.G.; Nguyen, J.T.; Fisher, C.; Tuohy, S.; Lyman, S.; Gonzalez Della Valle, A.; Ast, M.P.; Carli, A.V. The Risk Assessment and Prediction Tool (RAPT) Score Predicts Discharge Destination, Length of Stay, and Postoperative Mobility After Total Joint Arthroplasty. J. Arthroplast. 2023, 38 (Suppl. S2), S121–S129. [Google Scholar] [CrossRef] [PubMed]
  20. Zalikha, A.K.; Court, T.; Nham, F.; El-Othmani, M.M.; Shah, R.P. Improved performance of machine learning models in predicting length of stay, discharge disposition, and inpatient mortality after total knee arthroplasty using patient-specific variables. Arthroplasty 2023, 5, 31. [Google Scholar] [CrossRef] [PubMed]
  21. Kurmis, A.P.; Ianunzio, J.R. Artificial intelligence in orthopedic surgery: Evolution, current state and future directions. Arthroplasty 2022, 4, 9. [Google Scholar] [CrossRef]
  22. Skov, C.D.; Lindberg-Larsen, M.; Wiil, U.K.; Varnum, C.; Schmal, H.; Jensen, C.M.; Holsgaard-Larsen, A. Impact of motivational feedback on levels of physical activity and quality of life by activity monitoring following knee arthroplasty surgery-protocol for a randomized controlled trial nested in a prospective cohort (Knee-Activity). BMC Musculoskelet. Disord. 2024, 25, 778. [Google Scholar] [CrossRef]
  23. Negrini, F.; de Sire, A.; Lazzarini, S.G.; Pennestrì, F.; Sorce, S.; Arienti, C.; Vitale, J.A. Reliability of activity monitors for physical activity assessment in patients with musculoskeletal disorders: A systematic review. J. Back Musculoskelet. Rehabil. 2021, 34, 915–923. [Google Scholar] [CrossRef]
  24. Pennestrì, F.; Banfi, G. Primary Care of the (Near) Future: Exploring the Contribution of Digitalization and Remote Care Technologies through a Case Study. Healthcare 2023, 11, 2147. [Google Scholar] [CrossRef]
  25. Cordani, C.; Perillo, S.; Corbetta, D.; Sarasso, E.; Agosta, F.; Filippi, M.; Mazzali, A.G.; Pennestrì, F. Developing Physiotherapy in Primary Health Care: A First Snapshot from the Italian Metropolitan City of Milan. Healthcare 2024, 12, 1628. [Google Scholar] [CrossRef] [PubMed]
  26. Shaikh, H.J.F.; Botros, M.; Ramirez, G.; Thirukumaran, C.P.; Ricciardi, B.; Myers, T.G. Comparable performance of machine learning algorithms in predicting readmission and complications following total joint arthroplasty with external validation. Arthroplasty 2023, 5, 58. [Google Scholar] [CrossRef]
  27. Pennestrì, F.; Cabitza, F.; Picerno, N.; Banfi, G. Sharing reliable information worldwide: Healthcare strategies based on artificial intelligence need external validation. Position paper. BMC Med. Inform. Decis. Mak. 2025, 25, 56. [Google Scholar] [CrossRef] [PubMed]
  28. Lu, Y.; Khazi, Z.M.; Agarwalla, A.; Forsythe, B.; Taunton, M.J. Development of a Machine Learning Algorithm to Predict Nonroutine Discharge Following Unicompartmental Knee Arthroplasty. J. Arthroplast. 2021, 36, 1568–1576. [Google Scholar] [CrossRef]
  29. Seeber, G.H.; Wijnen, A.; Lazovic, D.; Bulstra, S.K.; Dietz, G.; van Lingen, C.P.; Stevens, M. Effectiveness of rehabilitation after a total hip arthroplasty: A protocol for an observational study for the comparison of usual care in the Netherlands versus Germany. BMJ Open 2017, 7, e016020. [Google Scholar] [CrossRef] [PubMed]
  30. Coudeyre, E.; Eschalier, B.; Descamps, S.; Claeys, A.; Boisgard, S.; Noirfalize, C.; Gerbaud, L. Transcultural validation of the Risk Assessment and Predictor Tool (RAPT) to predict discharge outcomes after total hip replacement. Ann. Phys. Rehabil. Med. 2014, 57, 169–184. [Google Scholar] [CrossRef]
  31. Colquhoun, H.L.; Levac, D.; O’Brien, K.K.; Straus, S.; Tricco, A.C.; Perrier, L.; Kastner, M.; Moher, D. Scoping reviews: Time for clarity in definition, methods, and reporting. J. Clin. Epidemiol. 2014, 67, 1291–1294. [Google Scholar] [CrossRef]
  32. Centre for Reviews and Dissemination. Systematic Reviews. In CRD’s Guidance for Undertaking Reviews in Health Care; University of York: York, UK, 2009; Available online: https://www.york.ac.uk/media/crd/Systematic_Reviews.pdf (accessed on 7 October 2025).
  33. European Commission. Directive 2011/24/EU. Off. J. Eur. Union 2025. Available online: https://eur-lex.europa.eu/eli/dir/2011/24/oj/eng (accessed on 7 October 2025).
  34. The Swedish Knee Arthroplasty Register. Annual Report 2020; Svenska Knäprotesregistret: Lund, Sweden, 2020; Available online: https://www.myknee.se/en/ (accessed on 7 October 2025).
  35. Istituto Superiore di Sanità. Registro Italiano Artroprotesi. 2025. Available online: https://riap.iss.it/riap/it/ (accessed on 7 October 2025).
  36. American College of Surgeons. ACS National Surgical Quality Improvement Program. 2025. Available online: https://www.facs.org/quality-programs/data-and-registries/acs-nsqip/ (accessed on 7 October 2025).
  37. Tornese, D.; Robustelli, A.; Ricci, G.; Rancoita, P.M.V.; Maffulli, N.; Peretti, G.M. Predictors of postoperative hospital length of stay after total knee arthroplasty. Singapore Med. J. 2024, 65, 68–73. [Google Scholar] [CrossRef]
  38. Pennestrì, F.; Tosto, V.; Pelosi, C.; Grippa, D.; Negrini, S.; Kiekens, C.; Sarasso, E.; Banfi, G.; Cordani, C.; The PREPARE Project Group. Predictive Factors of Inpatient Rehabilitation Stay and Post-Discharge Burden of Care After Joint Replacement for Hip and Knee Osteoarthritis: A Retrospective Study on 1678 Patients. Appl. Sci. 2024, 14, 11993. [Google Scholar] [CrossRef]
  39. Dauty, M.; Smitt, X.; Menu, P.; Dubois, C. Which factors affect the duration of inpatient rehabilitation after total knee arthroplasty in the absence of complications? Ann. Phys. Rehabil. Med. 2009, 52, 234–245. [Google Scholar] [CrossRef]
  40. Skov, C.D. Motivational Feedback Following Total or Unicompartmental Knee Arthroplasty (KneeActivity). 2023. Available online: https://clinicaltrials.gov/study/NCT06005623?term=NCT06005623.&rank=1 (accessed on 7 October 2025).
  41. Masiero, S.; Avesani, R.; Pegoraro, S.; Frigo, A.C.; Bonaldo, L. Predictive factors for functional recovery in patients admitted to subacute care nursing home in Italy. Aging Clin. Exp. Res. 2011, 23, 74–77. [Google Scholar] [CrossRef]
  42. Raad, M.; Amin, R.M.; El Abiad, J.M.; Puvanesarajah, V.; Best, M.J.; Oni, J.K. Preoperative Patient Functional Status Is an Independent Predictor of Outcomes After Primary Total Hip Arthroplasty. Orthopedics 2019, 42, e326–e330. [Google Scholar] [CrossRef]
  43. Mont, M. Neuromuscular Electrical Stimulation (NMES) for Improving Outcomes Following Knee Arthroplasty (TKA). 2016. Available online: https://clinicaltrials.gov/study/NCT03044028?term=NCT03044028&rank=1 (accessed on 7 October 2025).
  44. Kirk, A.G.; Burge, A.T.; Ekegren, C.L.; Liew, S.M.; Kimmel, L.A. Physical Activity in the Acute Hospital Following Elective Lower Limb Joint Arthroplasty. N. Z. J. Physiother. 2022, 50, 42–49. [Google Scholar] [CrossRef]
  45. Poitras, S.; Au, K.; Wood, K.; Dervin, G.; Beaulé, P.E. Predicting hospital length of stay and short-term function after hip or knee arthroplasty: Are both performance and comorbidity measures useful? Int. Orthop. 2018, 42, 2295–2300. [Google Scholar] [CrossRef] [PubMed]
  46. Ciapini, G.; Cervi, V.; Dolfi, L.; Parchi, P.; Lisanti, M. In-Depth Oral Presentations and Oral Communications. J. Orthopaed Traumatol. 2013, 14 (Suppl. S1), S106. [Google Scholar]
  47. National Guideline Centre (UK). Evidence Review for Inpatient Hip Knee Postoperative Rehabilitation: Joint Replacement (Primary): Hip Knee Shoulder: Evidence Review; NICE Guideline, No. 157; National Institute for Health and Care Excellence (NICE): London, UK, 2020. Available online: https://www.ncbi.nlm.nih.gov/books/NBK561415/ (accessed on 7 October 2025).
  48. Maiorano, E.; Bodini, B.D.; Cavaiani, F.; Pelosi, C.; Sansone, V. Length of stay and short-term functional outcomes after total knee arthroplasty: Can we predict them? Knee 2017, 24, 116–120. [Google Scholar] [CrossRef] [PubMed]
  49. Dall, G.F.; Ohly, N.E.; Ballantyne, J.A.; Brenkel, I.J. The influence of pre-operative factors on the length of in-patient stay following primary total hip replacement for osteoarthritis: A multivariate analysis of 2302 patients. J. Bone Jt. Surg. Br. 2009, 91, 434–440. [Google Scholar] [CrossRef]
  50. Menendez, M.E.; Schumacher, C.S.; Ring, D.; Freiberg, A.A.; Rubash, H.E.; Kwon, Y.M. Does "6-Clicks" Day 1 Postoperative Mobility Score Predict Discharge Disposition After Total Hip and Knee Arthroplasties? J. Arthroplast. 2016, 31, 1916–1920. [Google Scholar] [CrossRef] [PubMed]
  51. Robbins, S.M.; Rastogi, R.; McLaughlin, T.L. Predicting acute recovery of physical function following total knee joint arthroplasty. J. Arthroplast. 2014, 29, 299–303. [Google Scholar] [CrossRef]
  52. Bergman, L.; Guenna Holmgren, A.; Nilsson, U.; Amirpour, A.; Conte, H.; Eckerblad, J. The bumpy road to recovery: Older adults’ experiences during the first year after hip replacement surgery—A longitudinal qualitative study. BMC Geriatr. 2025, 25, 483. [Google Scholar] [CrossRef]
  53. Krakers, S.M.; Woudsma, S.; van Dartel, D.; Vermeer, M.; Vollenbroek-Hutten, M.M.R.; Hegeman, J.H.; On Behalf Of The Up Amp Go After A Hip Fracture Group. Rehabilitation of Frail Older Adults after Hip Fracture Surgery: Predictors for the Length of Geriatric Rehabilitation Stay at a Skilled Nursing Home. J. Clin. Med. 2024, 13, 4547. [Google Scholar] [CrossRef] [PubMed]
  54. Sikora-Klak, J.; Gupta, A.; Bergum, C.; Zarling, B.; Markel, D.C. The Evaluation of Comorbidities Relative to Length of Stay for Total Joint Arthroplasty Patients. J. Arthroplast. 2017, 32, 1085–1088. [Google Scholar] [CrossRef] [PubMed]
  55. den Hartog, Y.M.; Mathijssen, N.M.; Hannink, G.; Vehmeijer, S.B. Which patient characteristics influence length of hospital stay after primary total hip arthroplasty in a ‘fast-track’ setting? Bone Jt. J. 2015, 97, 19–23. [Google Scholar] [CrossRef]
  56. Kun, Y.; Pogoda, T. Orthopaedic patient outcomes following interdisciplinary inpatient rehabilitation. Int. J. Ther. Rehabil. 2013, 20, 361–366. [Google Scholar]
  57. Lewis, Z.H.; Hay, C.C.; Graham, J.E.; Lin, Y.L.; Karmarkar, A.M.; Ottenbacher, K.J. Social Support and Actual Versus Expected Length of Stay in Inpatient Rehabilitation Facilities. Arch. Phys. Med. Rehabil. 2016, 97, 2068–2075. [Google Scholar] [CrossRef]
  58. Tal-Akabi, A.; Schmid, S.; Taeymans, J. Determinants of inpatient rehabilitation length of stay and discharge modality after hip and knee replacement surgery in Switzerland—A retrospective observational study. Swiss Med. Wkly. 2013, 143, w13832. [Google Scholar] [CrossRef]
  59. Takagawa, S.; Kobayashi, N.; Yukizawa, Y.; Oishi, T.; Tsuji, M.; Misumi, T.; Inaba, Y. Identifying factors predicting prolonged rehabilitation after simultaneous bilateral total knee arthroplasty: A retrospective observational study. BMC Musculoskelet. Disord. 2021, 22, 368. [Google Scholar] [CrossRef]
  60. Frontera, W.R.; Cordani, C.; Décary, S.; DEGroote, W.; Del Furia, M.J.; Feys, P.; Jette, A.M.; Kiekens, C.; Negrini, S.; Oral, A.; et al. Relevance and use of health policy, health systems and health services research for strengthening rehabilitation in real-life settings: Methodological considerations. Eur. J. Phys. Rehabil. Med. 2024, 60, 154–163. [Google Scholar] [CrossRef]
  61. Bozic, K.J.; Wagie, A.; Naessens, J.M.; Berry, D.J.; Rubash, H.E. Predictors of discharge to an inpatient extended care facility after total hip or knee arthroplasty. J. Arthroplast. 2006, 21 (Suppl. S2), 151–156. [Google Scholar] [CrossRef]
  62. Husted, H.; Holm, G.; Rud, K.; Bach-Dal, C.; Hansen, H.C.; Andersen, K.L.; Kehlet, H. Indlaeggelsesvarighed ved primaer total hofte- og knaealloplastik i Danmark 2001–2003 [Length of stay after primary total hip and knee arthroplasty in Denmark, 2001–2003]. Ugeskr. Laeger 2006, 168, 276–279. [Google Scholar]
  63. Vincent, H.K.; Alfano, A.P.; Lee, L.; Vincent, K.R. Sex and age effects on outcomes of total hip arthroplasty after inpatient rehabilitation. Arch. Phys. Med. Rehabil. 2006, 87, 461–467. [Google Scholar] [CrossRef]
  64. Winemaker, M.; Petruccelli, D.; Kabali, C.; de Beer, J. Not all total joint replacement patients are created equal: Preoperative factors and length of stay in hospital. Can. J. Surg. 2015, 58, 160–166. [Google Scholar] [CrossRef]
  65. Oldmeadow, L.B.; McBurney, H.; Robertson, V.J. Predicting risk of extended inpatient rehabilitation after hip or knee arthroplasty. J Arthroplast. 2003, 18, 775–779. [Google Scholar] [CrossRef] [PubMed]
  66. Cohen, E.; Reid, D.B.C.; Quinn, M.; Walsh, D.; Raducha, J.; Hubbard, L.; Froehlich, J. Modifying the RAPT Score to Reflect Discharge Destination in Current Practice. Arthroplast. Today 2020, 7, 17–21. [Google Scholar] [CrossRef]
  67. Sconza, C.; Respizzi, S.; Grappiolo, G.; Monticone, M. The Risk Assessment and Prediction Tool (RAPT) after Hip and Knee Replacement: A Systematic Review. Joints 2019, 7, 41–45. [Google Scholar] [CrossRef]
  68. Bieganowski, T.; Bloom, D.A.; Kugelman, D.N.; Davidovitch, R.I.; Schwarzkopf, R.; Rozell, J.C. Comparison of Two Previously Validated Risk Assessment and Prediction Tool Models in Extended Length of Stay Patients Following total Hip Arthroplasty. J. Arthroplast. 2024, 39, 2520–2524.e1. [Google Scholar] [CrossRef] [PubMed]
  69. Dibra, F.F.; Parvataneni, H.K.; Gray, C.F.; Vasilopoulos, T.; Prieto, H.A. The Risk Assessment and Prediction Tool Accurately Predicts Discharge Destination After Revision Hip and Knee Arthroplasty. J. Arthroplast. 2020, 35, 2972–2976. [Google Scholar] [CrossRef] [PubMed]
  70. Dibra, F.F.; Silverberg, A.J.; Vasilopoulos, T.; Gray, C.F.; Parvataneni, H.K.; Prieto, H.A. Arthroplasty Care Redesign Impacts the Predictive Accuracy of the Risk Assessment and Prediction Tool. J. Arthroplast. 2019, 34, 2549–2554. [Google Scholar] [CrossRef] [PubMed]
  71. Gabor, J.A.; Feng, J.E.; Schwarzkopf, R.; Slover, J.D.; Meftah, M. Machine Learning with Electronic Health Record Data Outperforms a Risk Assessment Prediction Tool in Predicting Discharge Disposition After Total Joint Arthroplasty. Orthopedics 2022, 45, e211–e215. [Google Scholar] [CrossRef]
  72. Edusei, E.; Grossman, K.; Payne, A.; Perez, B.; Inneh, I.; Nelson, C.L.; Slover, J. Impact of social support and pain coping ability on length of stay and discharge disposition following hip and knee arthroplasty: A prospective study. Bull. Hosp. Jt. Dis. 2017, 75, 134–139. [Google Scholar]
  73. University of Valencia. Multi-Variate Prediction Model of Total Knee Replacement Outcome. 2021. Available online: https://clinicaltrials.gov/study/NCT03894514?term=NCT03894514.&rank=1 (accessed on 7 October 2025).
  74. Keeney, T.; Kumar, A.; Erler, K.S.; Karmarkar, A.M. Making the Case for Patient-Reported Outcome Measures in Big-Data Rehabilitation Research: Implications for Optimizing Patient-Centered Care. Arch. Phys. Med. Rehabil. 2022, 103, S140–S145. [Google Scholar] [CrossRef]
  75. Langenberger, B. Who will stay a little longer? Predicting length of stay in hip and knee arthroplasty patients using machine learning. Intell.-Based Med. 2023, 8, 100111. [Google Scholar] [CrossRef]
  76. Campagner, A.; Milella, F.; Banfi, G.; Cabitza, F. Second opinion machine learning for fast-track pathway assignment in hip and knee replacement surgery: The use of patient-reported outcome measures. BMC Med. Inform. Decis. Mak. 2024, 24 (Suppl. S4), 203. [Google Scholar] [CrossRef]
  77. Katzman, J.L.; Cardillo, C.; Schaffler, B.C.; Schwarzkopf, R.; Rozell, J.C. Can Preoperative Patient-Reported Outcome Measures Predict Clinical Outcomes following Total Knee Arthroplasty? J. Arthroplast. 2025, online ahead of print. [Google Scholar] [CrossRef]
  78. Guida, S.; Vitale, J.A.; Swinnen, E.; Beckwée, D.; Bargeri, S.; Pennestrì, F.; Castellini, G.; Gianola, S. Effects of Prehabilitation with Advanced Technologies in Patients with Musculoskeletal Diseases Waiting for Surgery: Systematic Review and Meta-Analysis. J. Med. Internet Res. 2024, 26, e52943. [Google Scholar] [CrossRef]
  79. De Klerk, T.C.; Dounavi, D.M.; Hamilton, D.F.; Clement, N.D.; Kaliarntas, K.T. Effects of home-based prehabilitation on pre- and postoperative outcomes following total hip and knee arthroplasty. Bone Jt. Open 2023, 4, 315–328. [Google Scholar] [CrossRef]
  80. van Aalst, M.J.; Oosterhof, J.; Nijhuis-van der Sanden, M.W.; Schreurs, B.W. Can the length of hospital stay after total hip arthroplasty be predicted by preoperative physical function characteristics? Am. J. Phys. Med. Rehabil. 2014, 93, 486–492. [Google Scholar] [CrossRef]
  81. Willis, R.; Edris, R.; Akehurst, S.; Ferraro, F.V. Do pre- and post- rehabilitation influence hospital length of stay and patient recovery following total hip replacement in the UK? A systematic scoping review and Delphi study. Int. J. Orthop. Trauma Nurs. 2025, 58, 101209. [Google Scholar] [CrossRef]
  82. Su, W.; Zhou, Y.; Qiu, H.; Wu, H. The effects of preoperative rehabilitation on pain and functional outcome after total knee arthroplasty: A meta-analysis of randomized controlled trials. J. Orthop. Surg. Res. 2022, 17, 175. [Google Scholar] [CrossRef]
  83. Almeida, G.J.; Khoja, S.S.; Zelle, B.A. Effect of prehabilitation in older adults undergoing total joint replacement: An Overview of Systematic Reviews. Curr. Geriatr. Rep. 2020, 9, 280–287. [Google Scholar] [CrossRef]
  84. Garrison, S.R.; Schneider, K.E.; Singh, M.; Pogodzinski, J. Preoperative Physical Therapy Results in Shorter Length of Stay and Discharge Disposition Following Total Knee Arthroplasty: A Retrospective Study. J. Rehabil. Med. Clin. Commun. 2019, 23, 1000017. [Google Scholar] [CrossRef] [PubMed]
  85. Wallace, I.J.; Worthington, S.; Felson, D.T.; Jurmain, R.D.; Wren, K.T.; Maijanen, H.; Woods, R.J.; Lieberman, D.E. Knee osteoarthritis has doubled in prevalence since the mid-20th century. Proc. Natl. Acad. Sci. USA 2017, 114, 9332–9336. [Google Scholar] [CrossRef]
  86. Bradley, B.M.; Griffiths, S.N.; Stewart, K.J.; Higgins, G.A.; Hockings, M.; Isaac, D.L. The effect of obesity and increasing age on operative time and length of stay in primary hip and knee arthroplasty. J. Arthroplast. 2014, 29, 1906–1910. [Google Scholar] [CrossRef]
  87. Fu, H.; Wang, C.; Afzal, I.; Kader, D.; Asopa, V.; Sochart, D. The effect of BMI and other factors on post-operative length of stay: A multivariable regression analysis of 860 patients undergoing total knee arthroplasty. Knee 2023, 42, 82–89. [Google Scholar] [CrossRef]
  88. Lozano, L.M.; Tió, M.; Rios, J.; Sanchez-Etayo, G.; Popescu, D.; Sastre, S.; Basora, M. Severe and morbid obesity (BMI ≥ 35 kg/m(2)) does not increase surgical time and length of hospital stay in total knee arthroplasty surgery. Knee Surg. Sports Traumatol. Arthrosc. 2015, 23, 1713–1719. [Google Scholar] [CrossRef] [PubMed]
  89. Issa, K.; Pivec, R.; Kapadia, B.H.; Shah, T.; Harwin, S.F.; Delanois, R.E.; Mont, M.A. Does obesity affect the outcomes of primary total knee arthroplasty? J. Knee Surg. 2013, 26, 89–94. [Google Scholar] [CrossRef] [PubMed]
  90. Mahoney, F.I.; Barthel, D.W. Functional Evaluation: The Barthel Index. Md. State Med. J. 1965, 14, 61–65. [Google Scholar]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Article Metrics

Citations

Article Access Statistics

Journal Statistics
Multiple requests from the same IP address are counted as one view.
Download PDF
Qualitative Analysis of a Blockchain-Based System Adoption for Academic Credentials Verification That Complies with the GDPR: GAVIN Project
Previous
Fleet Monitoring of Bridge Using Direct Calculation of Moving Reference Influence Line
Next