Predictive value of the hospital frailty risk score in perioperative complications of artificial hip and knee arthroplasty in elderly patients

Abstract

Background

Although artificial hip and knee arthroplasty has been widely used, the incidence of perioperative complications remains relatively high due to factors such as long operation time and large surgical incisions. However, the Hospital Frailty Risk Score (HFRS) clinical value for patients undergoing artificial hip and knee arthroplasty in China is not yet clear. This study aims to explore its clinical value in this population.

Objective

To explore predictive value of the HFRS in perioperative complications of artificial hip and knee arthroplasty in elderly patients.

Methods

Elderly patients who underwent artificial hip and knee arthroplasty in our hospital from March 2020 to March 2022 were selected as the study subjects. The patients were divided into the non-frail risk group (HFRS <5 points) and the frail risk group (HFRS ≥5 points) on the basis of the literature grading method. 5-factor modified frailty index (mFI-5) and Charlson Comorbidity Index (CCI) scores were assessed. General data including age, sex, ASA classification, preoperative course, and surgical time were collected through the electronic medical record system. The occurrence of postoperative deep venous thrombosis, periprosthetic infection, hematoma, anemia, and overall complications was recorded. Hospitalization days, surgical costs, and total treatment costs were also calculated for all patients.

Results

There were no significant differences in age, sex, ASA classification, preoperative course, or surgical time between the two groups (p > 0.05). The frail risk group had significantly higher mFI-5 and CCI scores (p < 0.05). This group also had a higher incidence of deep venous thrombosis and overall complications (p < 0.05), but similar rates of periprosthetic infection, hematoma, and anemia (p > 0.05). ROC analysis showed HFRS had the highest predictive value for postoperative complications (AUC = 0.851) compared to mFI-5 (0.786) and CCI (0.739). Surgical costs were similar (p > 0.05), but the frail group had longer hospital stays and higher total treatment costs (p < 0.05).

Conclusion

HFRS has better predictive value for perioperative complications in elderly patients undergoing artificial hip and knee arthroplasty compared to mFI-5 and CCI scores, and it can be used for predicting perioperative complications in elderly patients with this surgery.

Keywords

Hospital frailty risk score artificial hip and knee arthroplasty elderly patients complications predictive value

1 Introduction

Artificial hip and knee arthroplasty is an important surgical procedure in clinical practice for treating hip and knee joint pain, correcting joint deformities, and improving joint function.¹ It is also the most effective method for treating osteoarthritis. With the accelerated pace of social life and population aging, the number of elderly patients undergoing artificial hip and knee arthroplasty is gradually increasing.² Although artificial hip and knee arthroplasty has been widely used, the incidence of perioperative complications remains relatively high due to factors such as long operation time and large surgical incisions.³ A retrospective study conducted abroad⁴ showed that at least 80% of surgical patients experienced one or more complications during the perioperative period, including infection, deep vein thrombosis, postoperative hematoma, etc., which significantly affected the prognosis of elderly patients. Charlson Comorbidity Index (CCI) can reflect the overall immune and nutritional status of patients, and its application in predicting complications in surgical patients has been widely reported in clinical practices.^5,6 In recent years, foreign scholars have developed a new comorbidity index scoring system called the Hospital Frailty Risk Score (HFRS), which has been proven valuable in predicting postoperative complications in orthopedic surgery in foreign populations.⁷ However, its clinical value for patients undergoing artificial hip and knee arthroplasty in China is not yet clear. This study aims to explore its clinical value in this population.

2 Participants and methods

2.1 Research object

The elderly patients who underwent artificial hip and knee arthroplasty in our hospital from March 2020 to March 2022 were selected. Inclusion criteria were as follows: (1) age > 60 years; (2) meeting the surgical requirements for artificial hip and knee arthroplasty; (3) first-time artificial hip and knee arthroplasty. Exclusion criteria were as follows: (1) patients with malignant tumors; (2) abnormal mental status or cognitive impairment; (3) history of previous surgery on the same side of the hip or knee joint. All patients signed informed consent for the study and received approval from the hospital's medical ethics committee.

2.2 HFRS assessment

HFRS scoring is based on the International Classification of Diseases (ICD-10) in the electronic medical record system, which provides relatively standardized disease definitions and diagnoses, reducing the subjectivity in the clinical diagnostic process for physicians.⁸ Gilbert et al.⁹ developed the HFRS scoring system based on the ICD-10, using a modeling group of 22,139 elderly individuals and a validation group of 1,013,590 elderly individuals. By assigning scores to various clinical diagnoses in the ICD-10 and summing them up, HFRS score is obtained.⁹ A higher HFRS score indicates a greater frailty risk in patients. According to the literature grading method, patients are classified into the non-frail risk group (HFRS < 5 points) and the frail risk group (HFRS ≥ 5 points).

2.3. Data collection

2.3.1 CCI assessment

The patients were evaluated based on 5-factor modified frailty index (mFI-5)¹⁰ and the CCI,¹¹ as referenced in the literature. Both the mFI-5 and CCI scores indicate a higher frailty risk with higher scores.

2.3.2 Clinical data collection

Such general information as age, sex, American Society of Anesthesiologists (ASA) classification, preoperative course, and surgical duration were collected through the electronic medical record system. The incidence of complications in all patients from 1 day after operation to discharge was counted. The occurrence of postoperative deep vein thrombosis, periprosthetic infection, hematoma, anemia, and overall complications (excluding rare but severe complications such as pulmonary embolism, angina, myocardial infarction, and arrhythmias) were recorded.

The diagnosis of DVT was referred to the relevant literature,¹² lower extremity Doppler ultrasound examination combined with clinical diagnosis results as the standard..Periprosthetic infection was determined according to the results of pathogen culture, clinical symptoms and serological examination.

Hospital days, surgical costs, and total treatment costs were also documented for all patients.

2.4 Statistical analysis

The collected experimental data were analyzed with Statistic Package for Social Science (SPSS) 27.0 (IBM, Armonk, NY, USA). Normally distributed continuous variables were presented as X ± S, and independent samples t-test was used for comparisons. Categorical variables were presented as counts or rates, and comparisons were conducted using χ² test or Fisher's exact test. The receiver operating characteristic (ROC) curve was used to evaluate the predictive value of each scoring system for postoperative complications in elderly patients undergoing artificial hip and knee arthroplasty. All statistical tests were two-tailed, and a significance level of p < 0.05 was considered statistically significant.

3 Results

3.1 Comparison of general information between two groups

Finally, 98 patients were included in this study,there were 56 males and 42 females. There were no statistically significant differences in age, sex, ASA classification, preoperative course, and surgical duration between the two groups (p > 0.05), as shown in Table 1.

Table 1.
Comparison of general information between two groups.

Group Number of Cases Age (years) Sex (Male/Female) ASA classification (I/II/III) Preoperative Course (years) Surgical Duration (min)

Non-frail Risk Group 52 66.34 ± 6.25 30/22 32/16/4 2.31 ± 0.53 89.34 ± 26.58

Frail Risk Group 46 66.83 ± 6.18 26/20 25/16/5 2.38 ± 0.55 89.83 ± 27.10

t/χ² Value 0.389 0.014 0.606 0.641 0.090

p Value 0.698 0.907 0.739 0.523 0.928

Group	Number of Cases	Age (years)	Sex (Male/Female)	ASA classification (I/II/III)	Preoperative Course (years)	Surgical Duration (min)
Non-frail Risk Group	52	66.34 ± 6.25	30/22	32/16/4	2.31 ± 0.53	89.34 ± 26.58
Frail Risk Group	46	66.83 ± 6.18	26/20	25/16/5	2.38 ± 0.55	89.83 ± 27.10
t/χ² Value		0.389	0.014	0.606	0.641	0.090
p Value		0.698	0.907	0.739	0.523	0.928

3.2 Comparison of MFI-5 and CCI scores between two groups

The MFI-5 and CCI scores in the frail risk group were significantly higher than those in the non-frail group (p < 0.05), as shown in Table 2.

Table 2.
Comparison of MFI-5 and CCI scores between two groups (X ± S, score).

Group Number of Cases HERS MFI-5 CCI

Non-frail Risk Group 52 2.86 ± 1.80 1.80 ± 0.43 2.86 ± 0.59

Frail Risk Group 46 8.34 ± 3.17 2.38 ± 0.62 3.66 ± 0.70

t/χ² Value 10.675 5.431 6.138

P Value <0.001 <0.001 <0.001

Group	Number of Cases	HERS	MFI-5	CCI
Non-frail Risk Group	52	2.86 ± 1.80	1.80 ± 0.43	2.86 ± 0.59
Frail Risk Group	46	8.34 ± 3.17	2.38 ± 0.62	3.66 ± 0.70
t/χ² Value		10.675	5.431	6.138
P Value		<0.001	<0.001	<0.001

3.3 Comparison of postoperative complications between two groups

The frail risk group had a higher incidence of deep vein thrombosis and overall complications compared to the non-frail risk group, and the differences were statistically significant (p < 0.05). However, there were no statistically significant differences in the incidence of periprosthetic infection, hematoma, and anemia between the two groups (p > 0.05), as shown in Table 3.

Table 3.
Comparison of postoperative complications between two groups.

Group Number of Cases Deep vein thrombosis Periprosthetic Infection Hematoma Anemia Total

Non-frail Risk Group 52 15 1 1 4 21

Frail Risk Group 46 24 1 3 8 36

t/χ² Value 5.544 — — — 14.391

p Value 0.019 >0.999 0.339 0.217 <0.001

Group	Number of Cases	Deep vein thrombosis	Periprosthetic Infection	Hematoma	Anemia	Total
Non-frail Risk Group	52	15	1	1	4	21
Frail Risk Group	46	24	1	3	8	36
t/χ² Value		5.544	—	—	—	14.391
p Value		0.019	>0.999	0.339	0.217	<0.001

Note: - means Fisher's exact test

3.4 Comparison of predictive value of HFRS, MFI-5, and CCI scores for postoperative complications in artificial hip and knee arthroplasty

The results of ROC analysis showed that the area under the curve (AUC) for HFRS, MFI-5, and CCI scores in predicting postoperative complications in elderly patients undergoing artificial hip and knee arthroplasty were 0.851, 0.786, and 0.739, respectively. The HFRS had the largest AUC, as shown in Table 4 and Figure 1.

Table 4.
Comparison of predictive value of HFRS, MFI-5, and CCI scores for postoperative complications in artificial hip and knee arthroplasty.

Indicator AUC Sensitivity (%) Specificity (%) 95%CI P Value

HERS 0.851 83.55 82.34 0.815∼0.883 <0.001

MFI-5 0.786 75.64 73.86 0.759∼0.819 <0.001

CCI 0.739 71.09 69.86 0.713∼0.761 <0.001

Indicator	AUC	Sensitivity (%)	Specificity (%)	95%CI	P Value
HERS	0.851	83.55	82.34	0.815∼0.883	<0.001
MFI-5	0.786	75.64	73.86	0.759∼0.819	<0.001
CCI	0.739	71.09	69.86	0.713∼0.761	<0.001

Figure 1.

ROC curve.

3.5 Comparison of hospitalization conditions between two groups

There were no statistically significant differences in surgical costs between the two groups (p > 0.05). However, the frail risk group had longer hospitalization duration and higher total treatment costs compared to the non-frail risk group, and these differences were statistically significant (p < 0.05), as shown in Table 5.

Table 5.
Comparison of hospitalization conditions between two groups.

Group Number of Cases Hospitalization Duration (days) Surgical Costs (RMB) Total Treatment Costs (RMB)

Non-frail Risk Group 52 11.44 ± 3.12 2365.56 ± 435.89 62,345.34 ± 13,564.53

Frail Risk Group 46 13.61 ± 3.25 2371.86 ± 441.89 69,438.75 ± 14,653.85

t/χ² Value 3.370 0.071 2.488

P Value 0.001 0.944 0.015

Group	Number of Cases	Hospitalization Duration (days)	Surgical Costs (RMB)	Total Treatment Costs (RMB)
Non-frail Risk Group	52	11.44 ± 3.12	2365.56 ± 435.89	62,345.34 ± 13,564.53
Frail Risk Group	46	13.61 ± 3.25	2371.86 ± 441.89	69,438.75 ± 14,653.85
t/χ² Value		3.370	0.071	2.488
P Value		0.001	0.944	0.015

4 Discussion

Artificial hip and knee arthroplasty is a commonly performed orthopedic procedure with a high utilization rate. Apart from improving joint function, it can alleviate pain and has a high acceptance rate in clinical practice, particularly among the elderly population. However, clinical studies have found that the overall incidence of postoperative complications after artificial hip and knee arthroplasty can be as high as 40% to 80%, with a severe complication rate of approximately 13.4%.¹³ These complications not only impair organ function but can also be life-threatening in severe cases. Enhancing preoperative, intraoperative, and perioperative assessments of patients is of significant clinical importance for improving treatment outcomes.

HFRS has been partially applied in clinical settings. Kilkenny et al.¹⁴ analyzed hospital data and found that HFRS had good predictive value for outcomes in stroke patients after discharge. Ramos et al.¹⁵ demonstrated that HFRS had better predictive value for clinical outcomes in COVID-19 pneumonia patients compared to clinical frailty scores. Aitken et al.¹⁶ found, after adjusting for various clinical variables, that HFRS could relatively accurately predict the prognosis of elderly vascular surgery patients. In this study, HFRS was used as the basis for grouping. It was found that the MFI-5 and CCI scores were higher in the frail risk group compared to the non-frail risk group, indicating that these three scores have certain discriminatory value for frailty status in elderly patients and may be correlated to some extent.

Furthermore, this study found that the frail risk group had a higher incidence of deep vein thrombosis and overall complications compared to the non-frail risk group, indicating the value of HFRS in predicting the occurrence of deep vein thrombosis and complications in elderly patients undergoing artificial hip and knee arthroplasty, consistent with the findings of previous studies by foreign scholars.^7,17,18 In further analysis, this study showed that the AUC for predicting postoperative complications using HFRS, MFI-5, and CCI scores in elderly patients undergoing artificial hip and knee arthroplasty were 0.851, 0.786, and 0.739, respectively. The HFRS had the largest AUC, indicating that HFRS had better predictive value for elderly patients undergoing artificial hip and knee arthroplasty compared to MFI-5 and CCI scores. The reason for this could be that MFI-5 and CCI scores are commonly used comorbidity indices in previous studies. While they have the advantages of a smaller number of evaluation items and faster assessment, the development of these scales has been for many years, and some items may not be fully applicable to patients undergoing modern surgical treatments.¹⁹ On the other hand, HFRS is based on the internationally-recognized ICD-10 classification, making it more relevant to the diagnosis of diseases in the current patient population.

Researchers²⁰ have found that the hospitalization costs for patients undergoing artificial hip and knee arthroplasty mainly include consumables, medication, examination, surgical, treatment, and anesthesia fees. The occurrence of postoperative complications undoubtedly significantly increases the hospitalization days and the cost of hospitalization. The comparative results of this study found no difference in surgical costs between the two groups. However, the frail risk group had longer hospitalization duration and higher total treatment costs compared to the non-frail risk group, indicating that assessing frailty status using HFRS not only predicts the occurrence of complications but also reflects the length of hospital stay and overall treatment costs to some extent. Therefore, targeted intervention plans should be developed for patients with higher HFRS scores to reduce their hospitalization duration and overall treatment costs.

This study has several limitations. Firstly, it is a single-center, retrospective study with a relatively small sample size, which may limit the generalizability of the findings. Additionally, the study lacks long-term follow-up data to assess the enduring impact of perioperative complications, and further prospective, multi-center studies are needed to corroborate these results.

5 Conclusions

In conclusion, the results of this study demonstrate that HFRS has better predictive value for perioperative complications in elderly patients undergoing artificial hip and knee arthroplasty compared to MFI-5 and CCI scores. It can be used for predicting perioperative complications in this population. Further research is needed to expand the study population and conduct larger prospective studies to refine and supplement the existing conclusions.

Footnotes

Ethics approval and consent to participate

This study was approved by the Ethical Committee of The Affiliated Hospital, Southwest Medical University (approval no. KY-20-069). Signed written informed consents were obtained from the patients and/or guardians. This study was conducted in accordance with the Declaration of Helsinki and followed the ethical standards of China.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Shaikh

Hasan

Woo

, et al. Exposure to extended reality and artificial intelligence-based manifestations: a primer on the future of hip and knee arthroplasty. J Arthroplasty 2023; 38: 2096–2104.

Smolle

Horlesberger

Maurer-Ertl

, et al. Periprosthetic fractures of hip and knee-A morbidity and mortality analysis. Injury 2021; 52: 3483–3488.

Ohnuma

Raghunathan

Fuller

, et al. Trends in comorbidities and complications using ICD-9 and ICD-10 in total hip and knee arthroplasties. J Bone Joint Surg Am 2021; 103: 696–704.

Memtsoudis

Cozowicz

Bekeris

, et al. Peripheral nerve block anesthesia/analgesia for patients undergoing primary hip and knee arthroplasty: recommendations from the international consensus on anesthesia-related outcomes after surgery (ICAROS) group based on a systematic review and meta-analysis of current literature. Region Anesth Pain M 2021; 46: 971–985.

Badge

Churches

Naylor

, et al. Non-compliance with clinical guidelines increases the risk of complications after primary total hip and knee joint replacement surgery. Plos One 2021; 16: e0260146.

Poeran

Chan

Zubizarreta

, et al. Safety of tranexamic acid in hip and knee arthroplasty in high-risk patients. Anesthesiology 2021; 135: 57–68.

Meyer

Schwarz

Renkawitz

, et al. Hospital frailty risk score predicts adverse events in revision total hip and knee arthroplasty. Int Orthop 2021; 45: 2765–2772.

Krawczyk

Swiecicki

. ICD-11 vs. ICD-10 - a review of updates and novelties introduced in the latest version of the WHO international classification of diseases. Psychiatr Pol 2020; 54: 7–20.

Gilbert

Neuburger

Kraindler

, et al. Development and validation of a hospital frailty risk score focusing on older people in acute care settings using electronic hospital records: an observational study. Lancet 2018; 391: 1775–1782.

10.

Subramaniam

Aalberg

Soriano

, et al. New 5-factor modified frailty Index using American college of surgeons NSQIP data. J Am Coll Surgeons 2018; 226: 173–181.

11.

Borja

Connolly

Kvint

, et al. Charlson comorbidity index score predicts adverse post-operative outcomes after far lateral lumbar discectomy. Clin Neurol Neurosur 2021; 206: 106697.

12.

Iverson

Gomez

. Deep venous thrombosis: prevention and management. Clin Plast Surg 2013; 40: 389–398.

13.

Manning

Metcalf

Clark

, et al. Clinical characteristics, etiology, and initial management strategy of newly diagnosed periprosthetic joint infection: a multicenter, prospective observational cohort study of 783 patients. Open Forum Infect Di 2020; 7: ofaa068.

14.

Kilkenny

Phan

Lindley

, et al. Utility of the hospital frailty risk score derived from administrative data and the association with stroke outcomes. Stroke 2021; 52: 2874–2881.

15.

Ramos-Rincon

Moreno-Perez

Pinargote-Celorio

, et al. Clinical frailty score vs hospital frailty risk score for predicting mortality and other adverse outcome in hospitalised patients with COVID-19: spanish case series. Int J Clin Pract 2021; 75: e14599.

16.

Aitken

Lujic

Randall

, et al. Predicting outcomes in older patients undergoing vascular surgery using the hospital frailty risk score. Brit J Surg 2021; 108: 659–666.

17.

Meyer

Parik

Greimel

, et al. Hospital frailty risk score outperforms current risk stratification models in primary total hip and knee arthroplasty. J Arthroplasty 2021; 36: 1533–1542.

18.

Guo

Chen

. Risk factors for periprosthetic joint infection after primary artificial hip and knee joint replacements. J Infect Dev Countr 2020; 14: 565–571.

19.

Magan

Kayani

Chang

, et al. Artificial intelligence and surgical innovation: lower limb arthroplasty. Brit J Hosp Med 2020; 81: 1–7.

20.

Kremers

Leijtens

Camps

, et al. Evaluation of early wound leakage as a risk factor for prosthetic joint infection. J Am Assoc Nurse Pra 2019; 31: 337–343.