Shorter length of stay and significant cost savings with ambulatory surgery primary unilateral total knee arthroplasty in Asians using enhanced recovery protocols

Abstract

Introduction

Enhanced recovery after surgery (ERAS) in total knee arthroplasty (TKA) has reduced the length of stay (LOS) and cost of TKA in the Western population. Asians had been identified to be at higher odds of non-home discharge following TKA due to cultural differences. The efficacy of ERAS in TKA for Asian patients is less known. We aimed to investigate the efficacy of ERAS in reducing the LOS, transition to ambulatory surgery, improving home discharges, and reducing cost in an Asian population following TKA.

Methods

Retrospective analysis was performed on 634 TKA patients in 2017 (pre- ERAS) and 584 TKA patients who had undergone ERAS in 2022 in a tertiary hospital.

Results

Patients in 2022 (ERAS) were older (69 ± 7 vs. 68 ± 7 years old, p < 0.001) and had a higher proportion of patients with poorer function (p < 0.001). The LOS reduced from 5.4 days (95% CI:5.2–5.6) to 2.9 days (95% CI:2.7–3.2) (p < 0.001) with about 49 % of patients transitioning to ambulatory surgery and having a LOS of 1.4 days (95 %CI:1.3–1.5). The proportion of patients being discharged home in 2022 (78.9 %) was higher compared to 2017 (62.2 %) (p < 0.001). This saved the hospital 1817.4 inpatient ward bed days, which translated to S$2,124,540.60 of cost saving in a year, and up to S$2397.28 for the individual patient.

Conclusion

ERAS after TKA was able to safely achieve LOS comparable to the western population and allowed transition to ambulatory knee replacement in the Asian population. Consequently, this led to higher proportion of home discharges and achieved significant cost saving and hospital bed days.

1. Introduction

Total knee arthroplasty as a solution to the end stage osteoarthritis of the knee is increasingly common, with more than 640,000 procedures performed in the United States in a year, totalling approximately $10.2 billion.¹ Enhanced recovery after surgery (ERAS) has been increasingly adopted into the practice of total knee arthroplasty (TKA).² Systematic review and meta-analysis have shown that ERAS is safe, improved clinical and patient related outcomes, with potential reduction of cost between €109 and $20573.² Most studies, however, were performed in Western or Caucasian populations.²

While the reduction of length of stay (LOS) is a main contributor to the cost reduction,² there is also cost incurred on the patients should they require further care in rehabilitation or skilled nursing facilities. Previous analyses showed that Asians were at a higher risk of requiring rehabilitation or skilled nursing facilities after TKA.^3,4

To the best of our knowledge, there is limited literature on the cost savings of ERAS and its ability to discharge patients to home in an Asian setting. Lei et al. showed a reduction of the mean LOS from 12 to 10 days with unreported discharge destination, saving nearly S$750.⁵ There is room for improvement in these figures, considering the average LOS following TKA in the US and the UK have been reported to be 56 h and 3.9 days respectively.^6,7 Ding et al. showed a reduction in the average LOS from 4.9 to 3.8 days during the COVID-19 pandemic with ERAS protocols, but up to 68 % were still discharged to further rehabilitation facility.⁸ Xu et al. showed ambulatory surgery for TKA was possible in selected patients, however with no description of cost saving.⁹

Our institution started adopting ERAS for our knee replacements in late 2017 with our protocol following closely to the ERAS Society Consensus Statement recommendations.¹⁰ Our protocol continued to evolve over the next few years, throughout the COVID-19 pandemic and finally matured to a steady state in late 2021. We aimed to investigate the efficacy of ERAS in an Asian population, with respect to the LOS, transition to ambulatory surgery, discharge destination, and safety. Our secondary aim was to determine the cost saving. We hypothesized that ERAS will reduce the LOS, improve the rate of home discharges, and provide cost saving without increasing complications.

2. Methods

This was a retrospective cohort study in a tertiary public hospital located in an urban state within Southeast Asia. Institutional ethics approval was obtained prior to the study (NHG DSRB: 2023/00550). The reporting of this study follows the STROBE guidelines.

2.1. Patients

Patients ≥55 years old who were planned for primary unilateral TKA were included. Exclusion criteria included partial knee arthroplasty, bilateral TKA, and revision TKA.

All patients meeting the above criteria in 2017 and 2022 were included. Patients between 2018 and 2021 were not included in the study to reduce bias. During this period, the ERAS was constantly evolving in phases. Furthermore, there were additional institutional protocol considerations due to the COVID-19 pandemic before the programme eventually matured in October 2021. Inclusion of patients based on calendar year allowed a more accurate cost calculation.

2.2. Enhanced recovery after surgery (ERAS)

Our ERAS protocol followed the consensus statement provided by the ERAS Society¹⁰ Pre-operatively, there would be routine education, and medical optimisation alongside smoking and alcohol cessation.¹⁰ Perioperatively, there should be optimised fasting period, standard multimodal anaesthesia regime, with the use of local infiltrative analgesia, tranexamic acid, and multimodal prophylactic treatment for postoperative nausea, vomiting, and analgesia.¹⁰ Normothermia should be maintained, with balanced fluid management.¹⁰ Postoperatively, patients were allowed early return to normal diet, mobilised early, started on venous thromboembolism prophylaxis, and discharged upon achieving pre-determined criteria.¹⁰

Further institutional enhancements to the statements were made in phases as below. The eventual typical journey map of the patient would be better appreciated in Fig. 1.

Fig. 1.

Typical journey map of a patient undergoing total knee arthroplasty.

2.3. Care coordinator

Prior to the surgery, a care coordinator would provide personalised care planning and arrangement, education on discharge day, home environment and home safety education and equipment preparation. These scopes would not have been otherwise covered in the initial preoperative education and counselling, which focused on perioperative exercises.

2.4. Ambulatory surgery

The ambulatory surgery TKA initiative started from COVID 19 pandemic and continued post pandemic as its safety and efficacy were tried and tested.⁸ The selection criteria for Ambulatory Surgery (AS) TKA those undergoing primary unilateral TKA, American Association of Anaesthesia (ASA) score of ≤2, absence of renal impairment, and no allergies to NSAIDS, COX-2 inhibitors, and opioids. Patients who did not meet these criteria would be admitted as a same-day admission (SDA) to the ring-fenced arthroplasty ward. Patients in AS would typically stay 1 night, however it may be extended to a maximum of 2 nights if necessary.

2.5. Post-discharge teleconsultation and 24-h hotline

Following discharge, an Advanced Nurse Practitioner would perform a routine telephonic review of the patients at post-discharge day 1 or 2 and address any concerns from the patients. In addition, patients were provided with a 24-h hotline for patients to call or message for any other concerns. The hotline was linked to a smartphone, with video conferencing capabilities. The hotline was manned by the nurse in-charge on duty in the arthroplasty ward daily.

2.6. Early supported discharge for ambulatory surgery TKA patients

Previously, wound care would be performed at polyclinics or general practice clinics, while physiotherapy sessions would be held within the institution. With the enhancement, AS primary unilateral TKA patients were offered early supported discharge (ESD). With this, a physiotherapist would perform a home-based physiotherapy session between post discharge day 4–7 and a specialist nurse would perform home-based wound care including suture removal, where applicable, typically between post-operative day 10–14. Subsequently, patients continue their rehabilitation journey at collaborative community physiotherapy centres which may be closer to their homes.

2.7. Data collection and costs calculation

Data collected included age, gender, pre-operative level of ambulatory independence, the use of walking aids, pre-operative walking distance, transfusion rate, LOS, discharge destination, readmission within 30 days, and complications. These data were routinely collected as part of the institution’s registry for TKA. Complications were defined as deep vein thrombosis (DVT), superficial and deep wound infection, pulmonary embolism.

The proportion of patients admitted into general ward compared and into Ambulatory Surgery was retrieved. Cost saving was calculated in the form of bed days, and absolute dollars. Bed days for a particular period was calculated using the product of the average LOS and the total number of patients. Cost incurred by the hospital per day was obtained from the hospital’s value office. Patient’s daily treatment costs for AS and SDA were obtained based on the median bill from both groups. The cost of surgery, implant, laboratorial, and radiological investigations were excluded as these were would be relatively constant. The large variability in the bill size would arise from the LOS and the corresponding increase in the amount of treatment and therapy sessions received. We believe this would better reflect the effect of ERAS on the cost saving. Rates from 2022 were applied to 2017’s cohort for cost saving comparison to adjust for inflation.

2.8. Statistical analysis

No sample size calculation was performed as all available patients in the defined time period were included. Normality test of continuous data was assessed with Shapiro-Wilk test. Kruskal-Wallis test was used to analyse nonparametric data of multiple groups, with Mann-Whitney U test in post-hoc as required with adjustment of the p-value. Chi-square test or Fisher’s exact test was used for categorical data. Odds ratio for home discharges were analysed with binary logistic regression. P-value of <0.05 was deemed to be statistically significant.

3. Results

There were 634 patients in 2017 and 581 patients in 2022. Their demographics and pre-operative functional status are shown in Table 1. Compared to patients in 2017, patients in 2022 were significantly older (69 ± 7 vs. 68 ± 7 years old, p < 0.001), had a higher proportion of patients whom required supervision or assistance during ambulation, and not being a community ambulator (p < 0.001). Patients in 2022 also had a significantly larger proportion with a carer available at home (69.2 % vs. 56.7 %, p < 0.001).

Table 1.

Patient demographics from 2017 to 2022.

	2017 n = 634	2022 n = 581	p
Age	68 ± 7	69 ± 7	<0.001
Gender (%)			0.756
Female	67.5	68.2
Male	32.5	31.8
Ambulation dependence (%)			0.006
Independent with or without aid	99.6	97.9
Supervised or assisted	0.4	12
Walking aid (%)			0.098
Walking frame	1.8	2.7
Broad based quadstick	0.8	0.9
Narrow based quadstick	3.8	3.2
Walking stick	26.6	25.7
Crutches	0.0	0.3
Not required	65.5	63.3
Others	1.5	3.9
Walking distance (%)			<0.001
Homebound	7.9	8.7
Limited community	2.8	9.3
Community	89.3	82.0
Carer available (%)	56.7	69.2	<0.001

From Table 2, there was a statistically significant reduction in the average LOS from 5.4 days (95 % confidence interval (CI): 5.2–5.6) in 2017 to 2.9 days (95 % CI: 2.7–3.2) (p < 0.001). There was a statistically significant increase in the proportion of patients being discharged home in 2022 compared to 2017 (78.9 % and 62.2 % respectively, p < 0.001). Controlling for the age, level of ambulation dependence, walking distance, and the availability of carer at home, the odds of patients being discharged home in 2022 was 2.7 (95 % CI: 2.0–3.5) (p < 0.001) times higher compared to patients in 2017.

Table 2.

Length of stay and discharge destination in 2017 and 2022.

	2017 n = 634	2022 n = 581		p
Length of stay (95 % CI)	5.4 (5.2–5.6)	2.9 (2.7–3.2)		<0.001
		AS n = 281	SDA n = 300
Length of stay (95 % CI)		1.4 (1.3–1.5)	4.4 (4.0–4.7)	<0.001
Discharge destination (%)				<0.001
Home	62.2	78.9
CH or RF	37.8	21.1

AS – Ambulatory Surgery.

SDA – Same Day Admission.

CH – Community hospital.

RH – Rehabilitation facility.

In 2022, 281 patients were Ambulatory Surgery (AS) patients (49 %). The average LOS among the AS and SDA patients were 1.4 days (95 % CI: 1.3–1.5) and 4.4 days (95 % CI: 4.0–4.7) respectively (p < 0.001).

Transfusion rate reduced significantly from 8.9 % in 2017 to 1.3 % in 2022 (p < 0.001).

The readmission within 30 days reduced from 1.5 % in 2017 to 0.9 % in 2022, although it did not reach statistical significance (p = 0.258).

There was no statistically significant change in complication rate from 1.5 % in 2017 to 1.37 % in 2022 (p = 0.667). Deep vein thrombosis rate remained at 0.5 % in 2017 and 2022 (p = 0.915). Other complications encountered are shown in Table 3.

Table 3.

Complications in 2017 and 2022.

	2017	2022	p
Complications			0.667
Deep vein thrombosis	3	3
Cellulitis	2	1
Wound dehiscence	1	1
Prosthetic joint infection	1	2
Pulmonary embolism	0	1
Total	7	8

3.1. Cost saving

With the average reduction of LOS by 4 days and 1 day if patients were admitted into Ambulatory Surgery and general ward respectively, the implementation of the programme saved 1424 hospital bed day as seen in Table 4. This was a 45.4 % reduction from the year 2017. In particular, there were 1817.4 inpatient ward bed days saved. This was a 57.9 % reduction from the year 2017.

Table 4.

Savings in bed days. * - The number of patients reduced to match the load in 2022.

	2017	2022
	SDA	AS	SDA
Number of patients	581*	281	300
Average LOS	5.4	1.4	4.4
Total bed days	581 × 5.4  = 3137.4	1.4 × 281  = 393.4	4.4 × 300  = 1320.0
	2022 subtotal	1713.4
Total Hospital bed days saved		3137.4–1713.4 = 1424 bed days
Total Inpatient bed days saved		3137.4–1320 = 1817.4 bed days

LOS – Length of stay.

AS – Ambulatory Surgery.

SDA – Same Day Admission.

The median daily AS and SDA costs to a patient were S$ 803.40 and S$ 652.23 respectively. The daily cost to the hospital was S$ 1169.00. Applying these rates, the 1817 inpatient hospital bed days saved accrued to S$ 2,124,540.60 for the hospital.

Applying this rate to the average LOS of 5.4 days in 2017, the total inpatient ward charges would have cost S$ 3522.04 to the patient. Comparatively, an ERAS primary unilateral TKA patient would incur S$ 2869.81 as SDA with the LOS of 4.4 days, or only S$1124.76 as AS with the LOS of 1.4 days. This would translate to a saving of S$ 2397.28 to the patient who would go through the AS route, and S$652.23 for the patient who would go through the SDA route in 2022. The cost of ESD would be an additional S$360. As such, even if the patient opted for ESD following discharge from AS, there would still be a saving of S$2037.28.

4. Discussion

According to the systemic analysis for the Global Burden of Disease Study 2021, the knee is the most common site of osteoarthritis with a global age-standardised prevalence of 43,087 cases per 10,000.¹¹ In South Asia, East Asia, and Southeast Asia, knee osteoarthritis formed 55 %–70 % of the age-standardised prevalence of osteoarthritis regions.¹¹ It was estimated that knee osteoarthritis would affect 642 million individuals by 2050, a 74.9 % increase from 2020.¹¹ It recognised that there would be greater burden on the health system, and access to definitive treatment such as joint arthroplasty should be one of the items to be focused on to reduce the disease burden in the long term.¹¹

To meet this rising demand, it would be desirable to reduce the LOS and cost in total knee arthroplasty. The LOS had been successfully reduced in Western population with ERAS. For example, in the United States, the average LOS decreased from 76.6 h to 56.1,⁶ while in the United Kingdom, it decreased from 4.9 days to 3.9 days.⁷ Further systematic review of 38 studies found that 20 of them had significant reduction in the LOS, with an overall median LOS reduction of 6.5 days.² Only one of the studies reviewed were done in Asia, namely Malaysia, which showed no significant change in the LOS from 9.05 days to 9.93 days following an elective TKA.¹² In that study, there were 73 patients with ERAS and 79 patients in the conventional care model.¹² In China, 2687 patients who began ambulating within 24 h of surgery had a reduced LOS of 10 days, compared to 12 days in 3761 patients who ambulated after 24 h post-surgery.⁵ In Singapore, 342 TKA patients achieved an average LOS of 1.13 days following ERAS ambulatory surgery, and 396 patients had an average LOS of 4.12 days following ERAS inpatient, though there was no comparison to LOS from a pre-ERAS period.⁹ During the COVID-19 pandemic, the shortage of hospital beds, resources, and manpower was even more pressing such that any measures of reducing LOS would help the hospital cope with surging bed demands posted by COVID-19 patients then.⁸

Apart from reducing the LOS, there has been an increasing shift to achieve home discharge as this would cost 5 to 7 times less than discharges to non-home facilities.¹³ These facilities, such as rehabilitation or skilled nursing facility add to the overall cost and may undermine the success of value-based health care.¹³ Asians have been identified as having higher odds of non-home discharges compared to Caucasians, with various odds risk of 1.22, 2.41, and 3.33.^3,4,13 Nevertheless, the actual percentages of home discharges for Asians in these studies were 92.5 %,³ and 84.3 %.⁴ However, ERAS studies in Asia had not investigated the discharge destinations of patients.

Given the paucity of literature of ERAS in Asian population, our study was conducted. Our results showed that the LOS could be reduced to 2.9 days with successful implementation of a mature ERAS program, even in traditionally-conservative Asian population. In addition, 78.9 % were discharged home instead of non-home facilities. Of note, this was despite having a statistically significantly older group of patients with poorer ambulatory status in 2022 (and after adjusting for the availability of carer at home). Almost half of our knee replacements (49 %) were performed in the ambulatory surgery setting with an average LOS of 1.4 days.

We believe the success in reducing the LOS and increasing the proportion of home discharges from 2017 to 2022 was multifactorial, namely from the aspects of clinical, administrative, and psychosocial. Clinically, perioperative steps were optimised as per ERAS protocol, including establishing criteria-based discharge.⁸ Administratively, primary unilateral TKA was allowed to be performed as an AS case. The accelerated uptake of ambulatory surgery primary unilateral TKA stemmed from the COVID-19 pandemic period when hospital beds, operating resources were curtailed while ambulatory surgery manpower and operating facilities were preserved.

Home discharge was a complex issue with influence from the availability of home support, cultural preferences,³ socioeconomic status, and employment.⁴ Our dedicated care coordinator would discuss these aspects soon after listing the patient for surgery, allowing ample time to address any issues prior to surgery. Furthermore, the option of having home visits by nurses and physiotherapists, routine post discharge telephonic call by our nurse clinician and providing them a 24-hotline post-discharge provided patients with an ease of mind on being discharged home earlier. Although the home discharge rate was lower compared to the literature, those reported in the literature were more functionally independent, and were younger at the mean age of 53, which was known predictor of home discharge.^3,14

From safety perspective, the ERAS reduced complication rate, with no increase in readmission rate. This was consistent with the variety of results in the systematic review of various enhanced recovery programmes.² Our readmission rate was also lower than other Asian ERAS.⁹ We postulate that the availability of our 24-h hotline allowed a first-line triaging of patients into those truly required re-presentation to the hospital for further evaluation, should there be any concerns. There was a significant reduction in our transfusion rate with ERAS, consistent with the literature, and this could be largely attributed to the use of tranexamic acid.¹⁵

The programme also achieved significant cost saving. At a system level, there were 1817.4 inpatient bed days saved in a year, which equated to saving at least S$ 2,124,540.60. To the individual patient, this would be a saving of S$ 2397.28, which was nearly 70 %, if the patient met the criteria for primary unilateral TKA as a AS. This was consistent with the existing literature that showed successful reduction of cost with ERAS.² The savings were attributed to the reduction in room cost, meal cost, medications, laboratorial investigations, and physical therapy sessions.² There would also be more cost saving with patients going home instead of further rehabilitation facilities. A study in Australia comparing the cost of inpatient rehabilitation to home discharge after TKA found that the inpatient rehabilitation could cost $9500 in that episode, and that it was not associated with improved patient-reported scores.¹⁶

What had been obvious were the hospital bed days saved and the cost savings to the hospital and patient. Beyond the obvious, there were other intangible benefits that ERAS had conferred to our patients in terms of reduced need for transfusions, earlier discharge to the comfort of their own homes with reduced risk of healthcare associated infections, increased turnover for the hospital to operate on more patients and increased capacity for the hospital to accommodate more inpatient beds in the emergency setting. The additional costs of patients incurred by patients going to community hospitals or other rehabilitation facilities were also not factored into the total cost savings.

The strengths of our study include what we believed to be the first to report the effectiveness and cost saving of ERAS for primary unilateral TKA in a large cohort of Asian population. We described further administrative and social extensions of the ERAS and the feasibility of them to be implemented in phases at an institutional level. This study was limited by it being a retrospective study. Knowing the various subtype of Asians in the study may further help us understand the pattern of discharge destination.

5. Conclusions

In conclusion, ERAS after primary unilateral TKA in an Asian population was able to safely achieve LOS comparable to the western population. ERAS also allowed half of our primary unilateral knee replacements to be performed in the ambulatory setting, achieving higher proportion of home discharges without compromising safety and patient care. This study should provide confidence to other Asian institutions to safely adopt ERAS protocols and ambulatory surgery primary unilateral joint replacement.

CRediT authorship contribution statement

Wu Chean Lee: Conceptualization, Formal analysis, Methodology, Writing – original draft, Writing – review & editing. Eng Chuan Neoh: Data curation, Methodology, Resources, Writing – original draft, Writing – review & editing. Luak Pak Wong: Data curation, Methodology, Resources, Writing – original draft, Writing – review & editing. Kelvin Guoping Tan: Conceptualization, Formal analysis, Methodology, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.