Length of hospital stay after total knee arthroplasty among obese patients: A single-center study

Abstract

Knee osteoarthritis (OA) accounts for almost 80% of the worldwide burden of OA, with a pooled global prevalence of 22.9% (95% CI, 19.8–26.1%) in individuals aged 40 years and older. Total knee arthroplasty (TKA) is a highly effective procedure for treating knee OA. The length of stay (LOS) after TKA is an important measure of care quality, and predicting it can assist in optimizing patient recovery and healthcare strategies. This study aimed to evaluate the LOS and re-admission rates following TKA at Prince Metaab Bin Abdulaziz Hospital in Saudi Arabia. This retrospective, observational, analytical study included 194 patients who underwent primary TKA at Prince Metaab Bin Abdulaziz Hospital from January 2019 to June 2024. Data were collected from electronic health records and included demographics, LOS, mobility, blood transfusions, re-admissions, ASA grade, and postoperative complications. A negative binomial regression model was used to identify the factors associated with LOS. Statistical analysis was performed using SPSS version 27, with a P-value of < .05 considered significant. The median age of the patients was 63 years. Most patients were females (77%), while 72% and 26% were obese and overweight, respectively, and 15% were smokers. The median LOS was 4 days, and 47.4% of the patients stayed for 4–6 days. Re-admission occurred in 6.7% of cases, and the most common reasons being infection, joint pain, and knee stiffness. Postoperative complications occurred in 10% of the patients. A stepwise negative binomial regression model revealed that postoperative mobility (Exp(B) = 0.578, P < .001) was associated with shorter LOS, and ASA grade 2 patients were more likely to have a shorter LOS than ASA grade 3 patients (Exp(B) = 0. 749, P = .002). Dyslipidemia (Exp(B) = 1.522, P < .001) and smoking history (Exp(B) = 1.407, P < .001) were associated with a longer LOS. This study identified postoperative mobility, dyslipidemia, ASA grade, and smoking history as factors influencing the LOS after TKA. Improving early mobilization and managing comorbidities, such as dyslipidemia and smoking, may reduce LOS, enhance patient outcomes, and reduce healthcare costs.

Key points.

• Factors influencing length of hospital stay following TKR: Length of stay (LOS) is affected by factors such as with mobility after the operation, smoking, prevalent dyslipidemia and the ASA score.

• Targeting Modifiable Risks: Less time of stay may be achieved by appropriate smoking cessation intervention in combination with encouraging early mobilization of patients.

• Perioperative Management: Comorbidities are most likely to be present in patients with high ASA grades; therefore better management of these conditions is key to faster recovery and enhanced postoperative care.

1. Introduction

Osteoarthritis (OA) is a widespread chronic musculoskeletal condition that significantly affects patients and healthcare systems worldwide, contributing to considerable social and economic challenges.^[1] Knee OA, which accounts for nearly 80% of the global OA burden, is strongly associated with obesity and aging.^[2] The pooled global prevalence of knee OA was 16.0% (95% confidence interval [CI]: 14.3–17.8%) among individuals aged 15 years and above, rising to 22.9% (95% CI, 19.8–26.1%) for those aged 40 and over.^[3] In Saudi Arabia, knee OA ranges from 13% to 41%, with an average prevalence of 67.8% reported among individuals 40 years of age or above (16.2–71.4%).^[4]

Knee replacement surgery or arthroplasty was first performed in 1968. Since then, advances in materials and surgical techniques have markedly improved its efficacy. Total knee arthroplasty (TKA) is currently one of the most successful interventions for knee OA,^[5] offering a safe and cost-effective solution for pain relief, improved function, and enhanced quality of life in patients with end-stage knee arthritis.^[6] Research has consistently shown that many elective TKA patients can safely experience reduced lengths of stay (LOS) without increased post-discharge complications.^[7] LOS is the most significant cost factor in elective TKA^[8] and a key indicator of care quality.^[9]

Accurately predicting LOS is crucial because it enables surgeons to optimize patient care, plan for rapid recovery, and help patients manage their expectations. In addition, it can provide a foundation for more tailored healthcare policies. Focusing on modifiable clinical risk factors is essential for improving the recovery and discharge outcomes in TKA patients.^[10]

Reducing the duration of hospital stay while shifting more postoperative care to the home environment has notable advantages, including reduced costs associated with immediate post-surgical care.^[11] However, there are concerns that shorter hospitalizations could lead to increased complications or re-admissions, particularly among high-risk patients.^[12] Moreover, reducing hospital stay may burden patients, caregivers, and surgeons.^[13,14] This study aimed to assess LOS following primary TKA procedures conducted between January 2019 and June 2024 at Prince Metaab Bin Abdulaziz Hospital in Al-Jouf, Saudi Arabia. It also sought to analyze re-admission rates after discharge and identify the factors contributing to these re-admissions.

2. Materials and methods

2.1. Study design

This was a retrospective observational analytical study of individuals in the orthopedic surgery department who underwent primary TKA at Prince Metaab Bin Abdulaziz Hospital between January 2019 and June 2024 for the first time.

2.2. Sampling and recruitment

All patients who were diagnosed with primary osteoarthritis and underwent primary TKA during the specified period were included in the study. Patients with indications other than primary knee osteoarthritis or those who underwent revision knee surgery were excluded. In total, 194 patients met the inclusion criteria.

2.3. Data collection

Electronic health records provided relevant information, which was systematically compiled through a Google Form with 3 segments:

• The first part collected demographic details, namely age, sex, BMI, existing diseases, and history of tobacco smoking.

• The second part included particulars pertinent to admission parameters, such as duration of stay (in days), ambulation, hemoglobin level, and blood transfusion.

• The last segment collected data on re-admission, including rate, duration of stay post re-admission, causes of re-admission, ASA grade, and any other documented postoperative complications.

2.4. Statistical analysis

The dataset was prepared using Excel and later brought to SPSS version 27 (SPSS Inc., Chicago) for analysis. Normal distribution of continuous variables was assessed by displaying both histograms and the Kolmogorov-Smirnov test. Continuous variables were described as median and interquartile range, while categorical variables were described in terms of frequency and percentages. The relationship between length of hospital stay and demographic data was evaluated using a negative binomial regression model. A P-value < .05 was deemed to be clinically significant.

2.5. Ethical considerations

This study was approved by the Institutional Review Board of the Hail Health Cluster (IRB number:2024-58). Anonymity and confidentiality were maintained. Data were stored securely and accessible only to the research team, and all identifiers were removed prior to analysis.

3. Results

This study included 194 patients with a median age of 63 years. Most patients were females (77%), and only 23% were males. Only 2.1% of the patients had normal weight, while the majority were obese (72%) and overweight (26%). Approximately 15% of patients were smokers. Most patients had hypertension (49%), diabetes (37%), immunological (32%), cardiovascular (18%), neurological (15%), respiratory (13%), and renal (12%) diseases (Table 1).

Table 1.

Demographic characteristics of study participants.

Characteristic	N = 194
Age	63 (56, 72)
Gender
Female	150 (77%)
Male	44 (23%)
Body mass index
Normal	4 (2.1%)
Overweight	50 (26%)
Obese	140 (72%)
Comorbidities
Hypertension	96 (49%)
Diabetes	72 (37%)
Immunological disease	62 (32%)
Cardiovascular disease	34 (18%)
Neurological disease	30 (15%)
Respiratory disease	26 (13%)
Renal disease	23 (12%)
Dyslipidemia	16 (8.2%)
Anemia	16 (8.2%)
Hypothyroidism	18 (9.3%)
Other chronic disease	109 (56%)
History of smoking
No	165 (85%)
Yes	29 (15%)

The median LOS was 4 days. Approximately 38% stayed for 1–3 days, 47.4% and 14.9% stayed for 4–6 days and 7–16 days, respectively. Most patients were able to mobilize after surgery (96%). Only 12% required blood transfusion and 12% received blood postoperatively (Table 2).

Table 2.

Postoperative length of stay and complications among patients with total knee arthroplasty.

Characteristic	N = 194*
Length of stay in days	4.00 (2.00, 6.00)
The ability to mobility after surgery
No	7 (3.6%)
Yes	187 (96%)
Postoperative hemoglobin
No	170 (88%)
Yes	24 (12%)
Blood receive
No	171 (88%)
Yes	23 (12%)

^*Median (IQR); n (%).

Approximately 93.3% of the patients did not experience re-admission, while 6.7% were readmitted. The LOS after re-admission varied, with 93.3% of participants not requiring further hospitalization, 4.6% staying 1–3 days, and 2.06% staying 4–7 days. The reasons for re-admission included the need for emergency department admission (0.5%), infection (1.5%), joint pain (2.1%), knee stiffness (2.1%), and vascular surgery (0.5%). Regarding ASA grade, 6.7% of the patients were classified as grade 1, 65% as grade 2, and 28% as grade 3. Postoperative complications occurred in 10% of the patients, while the majority (90%) did not experience complications (Table 3).

Table 3.

Readmission rates, causes, and duration.

Characteristic	N = 194*
Re-admission
No	181 (93.3%)
Yes	13 (6.7%)
Length of stay after readmission
None	181 (93.3%)
1–3 d	9 (4.6%)
4–7 d	4 (2.06%)
Reasons for Re-admission
Need for Emergency Department admission	1 (0.5%)
Infection	3 (1.5%)
Joint pain	4 (2.1%)
Knee stiffness	4 (2.1%)
Vascular surgery	1 (0.5%)
Not admitted	181 (93.3%)
ASA grade
1	13 (6.7%)
2	127 (65%)
3	54 (28%)
Postoperative complication
No	174 (90%)
Yes	20 (10%)

^*Median (IQR); n (%).

An explanatory stepwise negative binomial regression model was used to determine the factors that influenced LOS. This is because of the overdispersed nature of the LOS data, and as a result, the negative binomial model was used instead of the Poisson regression model. The intercept is estimated to be 1.517 (SE = 0.3796, P < .001), and the expected log count of LOS if the other independent variables were at their baseline or reference level. This also means that the expected LOS is approximately 4. 558 days; Exp(B) = 4.558, 95% CI: 2.166 to 9.593).

In backward stepwise negative binomial regression analysis, variables with P-values above .10 were removed and those with a P-value below .05 were retained. The approach began with a comprehensive model that included all potential predictors: sex, mobility ability after surgery, postoperative hemoglobin, blood receive, diabetes, hypertension, dyslipidemia, anemia, hypothyroidism, cardiovascular disease, respiratory disease, renal disease, neurological disease, immunological disease, ASA grade, history of smoking, age, and BMI. The initial model thus indicated the effect of these variables, although not all of them significantly affected the LOS. In the stepwise method, the variables were removed individually because of their nonsignificance to the model fit and contribution to the model, including sex, postoperative hemoglobin, blood receive, diabetes, hypertension, anemia, hypothyroidism, cardiovascular disease, renal disease, neurological disease, immunological disease, age, and BMI.

The final model retained key predictors, including mobility after surgery, dyslipidemia, ASA grade, and smoking history. Namely, a higher postoperative mobilization level was related to a significantly lower LOS (Exp(B) = 0.578, P < .001)), whereas dyslipidemia was associated with an increased LOS (Exp(B) = 1.522, P < .001). Patients with ASA grade 2 were more likely to have a shorter LOS than those with ASA grade 3 (Exp(B) = 0.749, P = .002), whereas a history of smoking was associated with an increased LOS (Exp(B) = 1.407, P < .001). The last model proved to be the best fit because the number of deviances/degrees of freedom ratio was 1.271 and Pearson Chi-Square/df ratio of 1.332, which satisfies the data, indicates that the model has a good fit with the data. The Akaike Information Criterion of the final model was 866.94 less than that of the initial model (Akaike Information Criterion = 879). 68, indicating that it was a more efficient model (Table 4).

Table 4.

Association between length of stay in hospital and patients’ characteristics using negative binomial regression model n = 194.

Parameter	B	Std. error	95% Wald confidence interval		Hypothesis test			Exp(B)	95% Wald confidence interval for Exp(B)
			Lower	Upper	Wald Chi-Square	df	Sig.		Lower	Upper
(Intercept)	2.076	0.1525	1.777	2.375	185.279	1	.000	7.972	5.912	10.749
[Ability of mobility after surgery = 1.00]	−0.549	0.1428	−0.829	−0.269	14.771	1	.000	0.578	0.437	0.764
[Ability of mobility after surgery = 2.00]	0							1
[History of smoking = 1.00]	0.342	0.0848	0.175	0.508	16.208	1	.000	1.407	1.192	1.662
[History of smoking = 2.00]	0							1
[respiratory disease = 1.00]	−0.366	0.1170	−0.596	−0.137	9.806	1	.002	0.693	0.551	0.872
[respiratory disease = 2.00]	0							1
[ASA grade = 1.00]	0.072	0.1332	−0.189	0.333	0.294	1	.588	1.075	0.828	1.396
[ASA grade = 2.00]	−0.231	0.0760	−0.380	−0.082	9.240	1	.002	0.794	0.684	0.921
[ASA grade = 3.00]	0							1
[Dyslipidemia = 1.00]	0.420	0.1160	0.192	0.647	13.089	1	.000	1.522	1.212	1.910
[Dyslipidemia = 2.00]	0							1

4. Discussion

This cross-sectional study was conducted to investigate the factors associated with the length of stay (LOS) after total knee replacement (TKR) surgery at Prince Mettab Hospital. The study also involved 194 patients, with a mean age of 63 years, and 77% of the total patients were females. The majority of patients were classified as either obese (72%) or overweight (26%), and all patients had some form of comorbid disease, including hypertension (49%), diabetes (37%), and dyslipidemia. The median LOS was 4 days, with re-admission rates being relatively low at 6.7%). As such, TKR patients were discharged a few days from the hospital, and such factors are worth further investigation.

It has been shown through negative binomial regression analysis that there are several significant determinants of LOS. Postoperative mobility is one of the most important factors. Patients who by 2 weeks after surgery achieved a certain degree of mobility had a much shorter LOS, whereas patients who were more sedentary were more likely to have longer lengths of stay. This addition is in good agreement with the increasing amount of literature that provides evidence on how early postoperative management leads to enhanced recovery. Previous studies have demonstrated the significance of early mobilization post-surgery. In fact, LOS is improved by a decrease of about 1.8 days post surgical mobilization.^[15–17] These, however, come to the realization that it increases the recovery of the patients; however, it also decreases the incidence of post-surgical complications, such as thromboembolic illnesses and infections acquired within the hospital. Of the patients assessed, 96% were able to mobilize, resulting in a low median LOS of 4 days.

Smoking history is an additional determinant of a prolonged LOS. Smokers had a significantly longer hospital stay than non-smokers, which is consistent with the adverse effects of smoking on the process of healing in general. It has been established that smoking decreases the rate of recovery following surgery while also increasing the chance of surgical complications and causing unhealed wounds that elongate hospital stays.^[17] Indeed, it has been confirmed that both the short- and long-term effects of smoking on the cardiovascular and respiratory systems are well known to complicate recovery, hence increasing the length of hospital stay.^[18]

Dyslipidemia is one of the most common metabolic disorders and is associated with longer hospitalizations. Patients with dyslipidemia may have more complicated recoveries due to increased chances of cardiovascular complications, which are bound to fetch delays in the healing process and prolong the period of recovery.^[15] Another conclusion is the need to address this specific metabolic condition to minimize the negative impact of other diseases that might affect surgical outcomes.

LOS was also associated with the ASA, physical status classification which was regarded as one of the most significant predictors of LOS. Patients with a high ASA grade had poor outcomes, resulting in prolonged hospital stays, which is consistent with the literature that links high ASA grades to poor recoveries. In particular, those who were classified as ASA grade 2 had a lower chance of being hospitalized for an extended period upon discharge (Exp(B) = 0.794), while most ASA grades were associated with prolonged hospital stays associated with the presence of other diseases.^[15,17,18] This is similar to other studies conducted by other researchers, including that of Fahd on knee replacement; indeed, he was able to identify intraoperative blood loss, ASA grade, and operative time as key factors associated with the length of hospital stay post surgery.^[19] Similarly, a UK-based study found that older age, higher BMI, higher ASA grades, and the need for postoperative blood transfusions were key determinants of longer stays.^[16,20]

Interestingly, respiratory diseases, which are often considered risk factors for prolonged hospitalization, were managed effectively in this cohort, leading to shorter stays. This may be attributed to optimized perioperative care strategies that address respiratory conditions, thereby preventing complications and promoting faster recovery. The effective management of respiratory comorbidities could be a contributing factor to the overall shorter LOS observed in this study than in other settings.

In terms of postoperative complications, only 10% of the patients experienced significant issues, and 12% required blood transfusions. These rates were lower than those reported in studies conducted in central Saudi Arabia, where transfusion rates ranged from 29% to 35%.^[21–23] Postoperative blood transfusions have been associated with extended hospital stays in other studies, highlighting the importance of perioperative blood management strategies. Correcting hemoglobin levels and maintaining proper intraoperative hemostasis could potentially reduce the need for transfusions and, in turn, shorten the LOS.

The low re-admission rate (6.7%) observed in this study is noteworthy given the complexity of TKR surgery and the potential for complications. The most common reasons for re-admission included infection (1.5%), joint pain (2.1%), knee stiffness (2.1%), and vascular intervention (0.5%). In comparison, studies conducted in other countries, such as Australia and Scotland, have reported higher re-admission rates (3% and 6%, respectively).^[24,25]

5. Implications and recommendations

The findings of this study underscore the multifactorial nature of hospital stay after TKR surgery. Both modifiable risk factors, such as smoking history and postoperative mobility, and intrinsic patient characteristics, such as ASA grade and the presence of comorbidities, such as dyslipidemia, play pivotal roles in determining LOS.

Interventions aimed at targeting these modifiable risk factors could potentially reduce the duration of the hospital stay. For instance, smoking cessation programs should be integrated into the presurgical care of patients undergoing TKR, as smoking has been shown to delay recovery. Similarly, enhancing post-surgical mobility through early mobilization protocols could help patients recover more quickly and reduce their LOS. Enhanced recovery after surgery protocols, which emphasize early mobilization and efficient management of complications, could further decrease LOS and minimize the risk of hospital-acquired infections and thromboembolic events.

The presence of comorbidities such as dyslipidemia necessitates more personalized perioperative management strategies. Patients with high-risk profiles, including those with metabolic or cardiovascular disorders, may benefit from tailored recovery plans that address their specific needs. Improved perioperative care, including better management of respiratory and cardiovascular conditions, can significantly reduce complications and the LOS.

Future studies should focus on refining perioperative management strategies for high-risk patients and expanding the study to include larger and more diverse populations. Additionally, prospective follow-up studies are needed to capture the post-discharge complications that may occur in other healthcare settings.

6. Limitations

The primary limitation of this study is its cross-sectional design, which restricts its ability to capture long-term outcomes and complications that may arise after discharge. Furthermore, the lack of prospective follow-up limits the understanding of re-admissions that occur in different emergency departments or clinics outside the study setting. Another limitation is the small sample size, but the results are similar to what has been published. Future research should consider a longitudinal design to better assess long-term recovery and post-discharge complications.

7. Conclusion

In conclusion, this study demonstrated a relatively short median postoperative LOS of 4 days for patients undergoing TKR at Prince Mettab Hospital, with low complication and re-admission rates. Key predictors of LOS included postoperative mobility, smoking history, dyslipidemia, and ASA score. Interventions that target modifiable risk factors, such as smoking cessation and early mobilization, could further reduce the LOS and improve overall patient outcomes. Enhanced perioperative management of comorbidities, particularly in patients with higher ASA grades, is also essential for reducing recovery times and improving postoperative care.

Acknowledgments

We wish to express our appreciation to the entire hospital staff for their invaluable support throughout our research. Our deepest gratitude goes to Dr Saqer Al-Ruwaili, Dr Bandar Al-Maeen, and Dr Ahmed Al-Anzi. We are profoundly grateful for their contributions, guidance, and the time they devoted to making this work possible. We extend our gratitude to the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia, for their support in conducting this research.

Author contributions

Conceptualization: Ahmed Hassan Kamal, Abdulmajeed Salem Alsharari, Saif Farhan Alruwaili.

Data curation: Ahmed Hassan Kamal, Abdulmajeed Salem Alsharari, Faisal Musafiq Alruwaili, Abdulmohsen Nayef Alenzi, Saif Farhan Alruwaili, Abdulelah Essam Saba, Abdulaziz Saeed Almutlaq, Sultan Ayed Alanazi, Khalid Omar Alsharari.

Formal analysis: Ahmed Hassan Kamal, Abdulmajeed Salem Alsharari.

Funding acquisition: Ahmed Hassan Kamal.

Investigation: Faisal Musafiq Alruwaili, Abdulmohsen Nayef Alenzi, Saif Farhan Alruwaili, Abdulelah Essam Saba, Abdulaziz Saeed Almutlaq, Sultan Ayed Alanazi, Khalid Omar Alsharari.

Methodology: Ahmed Hassan Kamal, Abdulmajeed Salem Alsharari.

Project administration: Ahmed Hassan Kamal.

Software: Ahmed Hassan Kamal.

Supervision: Ahmed Hassan Kamal.

Validation: Faisal Musafiq Alruwaili, Abdulmohsen Nayef Alenzi, Abdulaziz Saeed Almutlaq, Sultan Ayed Alanazi, Khalid Omar Alsharari.

Visualization: Saif Farhan Alruwaili, Abdulelah Essam Saba.

Writing – original draft: Abdulmajeed Salem Alsharari.

Writing – review & editing: Ahmed Hassan Kamal.