1 Introduction

Sickle cell disease (SCD) is a hemoglobinopathy characterized by homozygosity for the sickle hemoglobin gene (HbS) that affects approximately 5 % of the global population.^1,2 The highest prevalence of sickle cell disease is found in sub-Saharan Africa.^1–3 However, estimates from 2022 indicate that the United States has an SCD population of 89,079, with 2–3 million individuals carrying the HbS gene.^1–3

Osteonecrosis is among the most prevalent orthopaedic complications in SCD, with documented incidence rates between 3 and 50 % in specific SCD populations.^4,5 Typically manifesting in the third or fourth decade of life, this debilitating condition is closely associated with recurrent vascular occlusion and resulting oxygen deprivation and tissue ischemia experienced by patients with SCD.^4,5 While several joints may be affected, the femoral head is the most common site of osteonecrosis (ONFH). Conservative management through non-weight bearing and activity modification, as well as joint-preserving procedures, are commonly used to manage early-stage ONFH.⁶ The currently available literature on joint-preserving procedures, such as core decompression, however, have failed to demonstrate consistently beneficial outcomes, particularly in skeletally mature patients with later-stage ONFH.^5,7,8 Total joint arthroplasty remains the only definitive option for many patients.^7,9

Patients with sickle cell disease (SCD) often experience poorer postoperative outcomes in both orthopaedic and general surgical procedures.^3,9–12 Total hip arthroplasty (THA) is typically indicated for patients with SCD who suffer from intractable pain due to osteonecrosis of the femoral head (ONFH),⁷ have failed conservative management, or are dealing with end-stage osteoarthritis (OA) of the hip. Although THA offers significant benefits, including pain relief and improved functional status, it also presents considerable perioperative risks for this vulnerable population. Overall complication rates following THA in patients have been reported to be as high as 67 %, with intraoperative blood loss and sickle cell-related complications such as acute chest syndrome or vaso-occlusive crises commonly occurring.^13–15

The primary aim of this study is to characterize the risks of postoperative complications following THA in a sickle cell disease population compared to a matched non-SCD cohort. As a secondary aim, this study explores differences in medical costs and length of stay between these populations. Lastly, a subgroup analysis is performed to determine the impact of cemented versus press-fit THA designs on postoperative outcomes to contribute to ongoing efforts to determine the optimal fixation technique for this patient population. We hypothesize that patients with sickle cell disease will have higher rates of postoperative complications, as well as longer lengths of stay and total hospital charges. Additionally, we hypothesize there will be no difference in postoperative inpatient complication rates between patients with SCD receiving cemented versus press-fit implant designs.

2 Methods

3 Results

3.1

3.1 Patient demographics & comorbidities

A weighted sample of 2,830,040 total patients who underwent hip arthroplasty between the years 2015-Q4 and 2020-Q4 was initially included in this study, 2535 of whom had sickle cell disease (SCD). Propensity score matching based on patient age, gender, and race resulted in 5010 total patients, with 2475 (49 %) and 2535 (51 %) patients in the non-SCD and SCD cohorts, respectively (Fig. 1). Patient demographic information for the matched patient cohorts is detailed in Table 1.

Close

Fig. 1 Patient cohort selection flowchart.

Export to PPT

Table 1 Matched patient demographics.

		Sickle Cell Cohort (SD∗, %)	Control Cohort (SD∗, %)	p-value
Age		40.5 (16.3∗)	40.7 (16.5∗)	0.343
Gender	Male	1100 (43.4)	1065 (43.0)	0.796
	Female	1435 (56.6)	1410 (57.0)
Race	White	105 (4.1)	120 (4.8)	0.958
	Black	2090 (82.4)	2075 (83.8)
	Hispanic	150 (5.9)	150 (6.1)
	Asian or Pacific Islander	25 (1.0)	25 (1.0)
	Native American	10 (0.4)	10 (0.4)
	Other	95 (3.7)	95 (3.8)
Elective Admission		2040 (80.5)	2195 (88.7)	<0.001
Non-Elective Admission		485 (19.1)	270 (10.9)
Hospital Census Division	New England	140 (5.5)	120 (4.8)	<0.001
	Middle Atlantic	430 (17.0)	285 (11.5)
	East North Central	300 (11.8)	360 (14.5)
	West North Central	100 (3.9)	75 (3.0)
	South Atlantic	765 (30.2)	765 (30.9)
	East South Central	240 (9.5)	275 (11.1)
	West South Central	335 (13.2)	270 (10.9)
	Mountain	60 (2.4)	115 (4.6)
	Pacific	165 (6.5)	210 (8.5)

Univariate analysis of postoperative outcomes between the matched cohorts revealed significantly elevated occurrences of periprosthetic fracture (3.4 % vs. 2.0 %, p = 0.004), periprosthetic dislocation (2.6 % vs 0.8 %, p < 0.001), periprosthetic mechanical dysfunction (1.8 % vs. 0.8 %, p = 0.002), death (0.2 % vs. 0.0 %, p = 0.027), PE (0.4 % vs. 0.0 %, p = 0.002), AKI (6.1 % vs. 2.4 %, <0.001), anemia (28.2 % vs. 22.2 %, p < 0.001), and blood transfusion (27.0 % vs. 6.7 %, p < 0.001) (Table 2).

Table 2 Matched preoperative comorbidities and univariate postoperative outcomes.

		Sickle Cell Cohort (%)	Control Cohort (%)	p-value
Charlston Comorbidity Index	1	1530 (60.4)	1350 (54.5)	<0.001
	2	575 (22.7)	640 (25.9)
	3	210 (8.3)	250 (10.1)
	4	80 (3.2)	90 (3.6)
	5	60 (2.4)	85 (3.4)
	6	25 (1.0)	30 (1.2)
	7	20 (0.8)	20 (0.8)
	8 or More	35 (1.4)	10 (0.4)
Obesity		310 (12.2)	715 (28.9)	<0.001
Nicotine Use Disorder		595 (23.5)	800 (32.3)	<0.001
Periprosthetic Fracture		85 (3.4)	50 (2.0)	0.004
Periprosthetic Dislocation		65 (2.6)	20 (0.8)	<0.001
Periprosthetic Mechanical Dysfunction		45 (1.8)	20 (0.8)	0.002
Periprosthetic Infection		120 (4.7)	40 (1.6)	<0.001
Superficial SSI		0 (0.0)	0 (0.0)	0
Deep SSI		0 (0.0)	5 (0.2)	0.024
Wound Dehiscence		5 (0.2)	10 (0.4)	0.18
Died During Hospitalization		5 (0.2)	0 (0.0)	0.027
DVT		0 (0.0)	0 (0.0)	0
PE		10 (0.4)	0 (0.0)	0.002
PNA		35 (1.4)	20 (0.8)	0.52
AKI		155 (6.1)	60 (2.4)	<0.001
Anemia		715 (28.2)	550 (22.2)	<0.001
Blood Transfusion		685 (27.0)	165 (6.7)	<0.001

4 Discussion

This study aimed to delineate the distinct risk profiles for postoperative complications between patients with and without SCD while also quantifying the differences in medical costs and lengths of stay between the two cohorts. Additionally, we investigated the variations in short-term inpatient outcomes among patients with SCD undergoing THA with either cemented or press-fit, uncemented implants. We initially analyzed 2,830,040 patients who underwent hip arthroplasty between 2015 and 2020, including 2535 with SCD. Propensity score matching revealed that patients with SCD had significantly higher rates of complications such as periprosthetic fractures, periprosthetic mechanical dysfunction, periprosthetic infections, and deep vein thrombosis. Multivariate analysis further identified SCD as an independent risk factor for higher healthcare costs and longer hospital length of stay. A subanalysis comparing cemented to press-fit implant designs further revealed that the use of cemented implants in patients with SCD was associated with higher risks of periprosthetic fractures and infections.

Individuals with SCD face unique health challenges that heighten their susceptibility to orthopaedic conditions that require THA, as well as an elevated risk for subsequent complications. A systematic review by Kenanidis et al. highlighted the increased risk of complications in patients with SCD undergoing THA, with sickle cell crises and transfusion reactions being the most common, along with acute coronary syndrome, though rare, posing a serious threat.¹⁹ Perfetti et al. further demonstrated that patients with SCD undergoing THA have a 152 % higher overall risk of complications compared to controls, with increased incidences of pneumonia, surgical site issues, urinary tract infections, and wound hemorrhage.¹⁵ Moreover, patients with SCD required more frequent blood transfusions, underscoring their greater medical complexity.¹⁵ Our study reports similar findings, with significantly higher rates of periprosthetic fractures, mechanical dysfunction, periprosthetic infections, deep vein thrombosis, pneumonia, acute kidney injury, anemia, and blood transfusions in the unmatched SCD cohort. Even after matching, patients with SCD exhibited increased rates of periprosthetic dislocation, deep surgical site infections, mortality, and pulmonary embolism.

Given that sickle cell disease (SCD) patients experience more frequent complications, healthcare providers must understand the associated costs to allocate resources effectively. One key factor driving these costs is the length of hospital stay. Kenanidis et al. found that patients with SCD undergoing total knee arthroplasty (TKA) had longer hospital stays and higher charges than matched controls.¹⁹ Similarly, Perfetti et al. reported that, between 1988 and 2010, patients with SCD had an average hospital stay of 6.92 days compared to 3.83 days for patients without SCD, with significantly higher costs, averaging $51,240 versus $37,802.¹⁵ Our study showed that patients with SCD undergoing total hip arthroplasty (THA) had a mean stay of 4.94 days, compared to 3.00 days for patients without SCD.¹⁵ We also found that hospital charges for patients with SCD averaged $94,449, compared to $77,783 for the control group, indicating a widening cost gap over the past decade. While there has been some progress in reducing the difference in length of stay between patients with SCD and controls, this has not corresponded to a reduction in overall costs.¹³

The multivariable subanalysis comparing patients with SCD receiving cemented versus non-cemented, press-fit implants revealed cemented implants to be an independent risk factor for periprosthetic infection and periprosthetic fracture. In patients with SCD, periprosthetic fractures may be more common with cemented THA due to compromised bone quality from chronic ischemia and the heat generated during cement polymerization, which can further weaken bone.²⁰ The increased risk of infection is possibly related to the immune dysfunction in SCD and the cement acting as a foreign material that can harbor bacteria, especially in patients with impaired blood supply.^21,22 Non-cemented prostheses, which rely on bone ingrowth, may reduce these risks by minimizing stress shielding and bacterial colonization.

Our study has several important limitations. Inherent to studies utilizing administrative databases, we relied on the assumption that there were no coding or data entry errors in the database, which may not be true. Furthermore, the data provided by the NIS are limited to the duration of a patient's hospital stay, thus excluding complications that arise after discharge from the analysis. Additionally, total hip arthroplasty surgery may be indicated for the management of osteoarthritis or osteonecrosis. We considered both indications for THA in our analysis, potentially confounding the results. While our study provides a comprehensive view of THA outcomes across all indications, an osteonecrosis-specific analysis, with age-specific outcomes, could yield more nuanced insights and may be a potential avenue for future research. Finally, as a national database, analyses utilizing the NIS provide results that may not directly apply to specific regions; however, the large volume of data enhances the generalizability and reliability of the findings.

5 Conclusions

This study demonstrated that patients with sickle cell disease had significantly higher rates of complications after total hip arthroplasty, including periprosthetic fractures, mechanical dysfunction, periprosthetic infections, and deep vein thrombosis. Multivariate analysis confirmed SCD as an independent risk factor for increased healthcare costs, longer hospital stays, and higher charges. Additionally, patients with SCD receiving cemented implants faced higher risks of periprosthetic fractures and infections, as well as other complications, compared to those with press-fit implants, who experienced better outcomes and lower costs. These findings underscore the need for careful selection of surgical techniques or alternative treatment modalities to mitigate risks and optimize care for patients with SCD.

CRediT authorship contribution statement

Michael Miskiewicz: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Project administration. Rafael Madera: Writing – review & editing. Ilan Pesselev: Writing – review & editing. James Gallagher: Writing – review & editing. David Komatsu: Supervision, Writing – review & editing. James Nicholson: Supervision, Writing – review & editing.

Patients consent

This study retrieved data from a publicly-available national database that only offers de-identified patient information. Informed consent and institutional review board approval was therefore not necessary.

Ethical statement

This study did not involve the use of human or animal subjects. Data derived from the database used in this study contained only de-identified patient information. This study was exempt from informed consent or our institutional review board processing. Confidentiality of participant data was maintained throughout the research process. This study was conducted in accordance with the ethical guidelines of our institution.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.