Increased risk of complications in patients with hypoalbuminemia undergoing revision total hip arthroplasty

Abstract

Objective

The relationship between hypoalbuminemia and complications after revision total hip arthroplasty (THA) has not been established. We hypothesize that hypoalbuminemia is associated with complications in patients undergoing revision THA.

Methods

The ACS-NSQIP database was queried for patients undergoing revision THA. Chi square and regression analyses were used to assess the relationship between hypoalbuminemia, demographics, other comorbidities, and complications.

Results

Hypoalbuminemia is associated with an increased risk of reoperation, bleeding complications, surgical site infections, non-routine discharge, medical complications, and surgical complications.

Conclusions

Albumin levels should be considered in the preoperative planning of patients undergoing revision THA for possible nutritional optimization.

1. Introduction

Primary total hip arthroplasty (THA) is a common procedure in the United States with favorable success rates and high patient satisfaction.¹ Component failure due to infection, dislocation, osteolysis, or loosening, however, often necessitates revision THA.² Due to increases in survivorship, functional demand, and the aging population in the United States, the number of primary THAs is expected to increase by 174% by 2030, and the number of revision THAs is projected to double by 2026.3, 4, 5 Patients undergoing revision procedures are at a greater risk of adverse events compared to those undergoing primary procedures. These complications, often the reason for revision after the index surgery, include mechanical loosening, infection, dislocation, osteolysis, and polyethylene wear.^6,7 Additionally, these patients are subject to longer operative times, prolonged hospital admissions, and higher readmission rates.8, 9, 10 Indeed, these complications place patients at more than double the risk of surgical site infections (2.9% vs. 1.18%) and triple the risk of perioperative mortality (0.97% vs. 0.30%) compared to those undergoing primary THA^11,12 while incurring an increasing burden on the healthcare system.⁶

Mitigation of risk factors is the goal of preoperative planning. As the number of people predisposed to complications increases, the heightened importance of perioperative patient optimization has become the focus of many investigations in the past decade.13, 14, 15, 16 While many prognostic factors are well documented in the primary joint replacement population,^17,18 the effect of malnutrition is not as well defined in the context of revision THA.

Hypoalbuminemia, defined as serum albumin <3.5 g/dl, is a widely accepted marker for patient malnourishment.^19,20 Its short half-life allows for detection of acute changes in nutrition²¹ and prevalent use in surgical planning. It is associated with a multitude of increased surgical complications across a variety of surgical specialties, including infection, postoperative pneumonia, increased length of stay, and greater readmission rates.^19,20,22 In patients undergoing joint arthroplasty, it has been shown to increase mortality nearly six-fold.^18,23 The risk and perioperative complications associated with hypoalbuminemia for patients undergoing revision THAs, however, have not yet been determined. This information may have implications in choosing appropriate surgical candidates, the timing of elective surgery, and patient discussions. We therefore aimed to assess the association of hypoalbuminemia with likelihood of reoperation and perioperative complications following revision THA.

2. Methods

2.1. Dataset

The American College of Surgeons National Surgical Quality Improvement Program (NSQIP) was used to obtain our study population. NSQIP is a national database that is validated, risk-adjusted, and outcomes based and used to improve the quality of surgical care.24, 25, 26 Data is collected by a trained surgical clinical reviewer from the preoperative through 30-day postoperative period using a HIPAA-complaint methodology. It reports data on over 130 clinical variables, including preoperative comorbidities and postoperative outcomes up to 30 days after surgery.

2.2. Patient selection

The NSQIP 2005–2016 database was queried to identify patients undergoing revision hip revision arthroplasty as one of the top 3 procedures performed using CPT codes 27137, 27134, and 27138. As our study does not qualify as human subjects research, Institutional Review Board (IRB) exemption was obtained. There were no sources of funding for this study.

2.3. Definitions and outcomes

The primary outcome of this study was unplanned reoperation. Additionally, we investigated bleeding complications, surgical site infections, discharge destination, medical, and surgical complications. Medical complications were defined as presence of either pneumonia, reintubation, pulmonary embolism, ventilator dependence >48 h postoperatively, renal insufficiency, acute renal failure, urinary tract infection, cerebrovascular accident, cardiac arrest, myocardial infarction, DVT, sepsis, or septic shock. Non-routine discharge was defined as discharge to a location other than the patient's home. Surgical complications were defined as either surgical site infections, wound disruption, or bleeding. Hypoalbuminemia was defined as an albumin level <3.5 g/dl, and obesity was defined as BMI≥30.

2.4. Statistical analysis

Bivariate analysis was performed using Pearson's Chi Square test to determine risk factors associated with each outcome. If a variable was significantly associated with an outcome on bivariate analysis, it was included in the regression analysis for that outcome. Demographic and significant variables were then analyzed with binary logistic regression. The results were recorded as odds ratios (OR) and 95% confidence intervals (95% CI) and significance was defined as p < 0.05. Statistical Package for Social Science (SPSS) (International Business Machines, Corp., Armonk, NY), version 23 was used for data management and statistical analysis.

3. Results

A total of 14,431patients were included in our final analysis. The majority of patients were between the ages of 61 and 80 (51.9%), Caucasian (77.7%), and female (54.7%) (Table 1).

Table 1.

Demographic data.

Variable	Frequency
Age
18-40	400 (2.8%)
41-60	4281 (29.7%)
61-80	7485 (51.9%)
>80	2212 (15.3%)
Race
Caucasian	11213 (77.7%)
African American	997 (6.9%)
Hispanic	239 (1.7%)
Other	228 (1.6%)
Unknown	1716 (11.9%)
Female	7888 (54.7%)

Hypoalbuminemia was present in 10.2% of patients. Non-routine discharge was the most common complication, followed by surgical complications and bleeding complications (Table 2).

Table 2.

Perioperative complications frequency.

Bleeding Complications	4014 (27.8%)
Surgical Site Infection	641 (4.4%)
Non-Routine Discharge	4728 (32.8%)
Unplanned Re-Operation	917 (6.4%)
Medical Complication	831 (5.8%)
Surgical Complication	4430 (30.7%)

After accounting for demographic and significant clinicopathologic variables, Hypoalbuminemia was associated with a 50% increased likelihood of reoperation. The magnitude of increased risk was similar to that associated with impaired functional status, hyponatremia, and obesity (Table 3). Hypoalbuminemia also was associated with an increased likelihood of bleeding complications, surgical site infections, non-routine discharge, unplanned reoperation, medical complications, and surgical complications (Table 4).

Table 3.

Predictors of unplanned reoperation.

Variable	P Value	OR (95% CI)
Age
18-40	Reference
41-60	.300	0.72 (0.38–1.35)
61-80	.660	0.87 (0.46–1.63)
>80	.360	0.73 (0.37–1.44)
Female	.203	1.14 (0.93–1.39)
Race
Caucasian	Reference
African American	.905	0.98 (0.67–1.42)
Hispanic	.545	0.80 (0.38–1.66)
Other	.565	0.76 (0.31–1.91)
Unknown	.254	0.77 (0.49–1.21)
Hypoalbuminemia	.000	1.55 (1.22–1.98)
Preoperative Transfusion	.671	0.88 (0.48–1.61)
Systemic Sepsis	.028	1.64 (1.06–2.55)
Corticosteroid Use	.531	1.13 (0.77–1.64)
Smoker	.006	1.46 (1.12–1.91)
Open Wound	.286	0.76 (0.46–1.26)
Obese	.000	1.56 (1.26–1.92)
Hyponatremia	.002	1.59 (1.19–2.13)
Hypertension	.069	1.23 (0.98–1.54)
Impaired Functional Status	.004	1.56 (1.15–2.13)
Use of Emergency Services	.485	1.19 (0.73–1.96)
COPD	.238	1.23 (0.87–1.73)
Bleeding Disorder	.080	1.37 (0.96–1.95)

Table 4.

Predictors of secondary complications.

	Bleeding Complications	Surgical Site Infection	Non-Home Discharge Destination	Medical Complications	Surgical Complications
Variable	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Age
18-40	Reference	Reference	Reference	Reference	Reference
41-60	0.53 (0.2–1.37)	1.17 (0.13–10.37)	1.88 (1.08–3.28)*	0.29 (0.06–1.36)	0.83 (0.57–1.22)
61-80	0.65 (0.26–1.65)	1.5 (0.18–12.47)	5.13 (2.95–8.89)*	1.05 (0.26–4.14)	0.98 (0.67–1.44)
>80	0.76 (0.28–2.04)	0.88 (0.09–8.58)	18.65 (10.52–33.06)*	1.69 (0.39–7.34)	1.34 (0.89–2.02)
Female	1.1 (0.79–1.53)	1.15 (0.59–2.24)	1.54 (1.37–1.74)*	1.29 (0.75–2.21)	1.22 (1.09–1.37)*
Race
Caucasian	Reference	Reference	Reference	Reference	Reference
African American	1.33 (0.68–2.6)	0.65 (0.14–2.95)	1.39 (1.11–1.75) *	1.21 (0.39–3.78)	1.37 (1.11–1.7)*
Hispanic	0.66 (0.26–1.67)	0.43 (0.05–3.47)	1.49 (0.98–2.24)	0.33 (0.04–2.8)	1.37 (0.94–1.99)
Other	2.38 (0.5–11.36)	N/A	0.87 (0.52–1.45)	0 (0–0)	1.44 (0.91–2.28)
Unknown	1.95 (1.05–3.61)*	0.67 (0.15–3.04)	0.64 (0.5–0.83) *	0.84 (0.29–2.49)	1.21 (0.97–1.51)
Hypoalbuminemia	2.14 (1.42–3.21)*	2.87 (1.36–6.05)*	2.07 (1.77–2.43) *	2.4 (1.32–4.37)*	2.37 (2.05–2.74)*
Preoperative Transfusion	2.25 (0.86–5.93)	N/A	1.96 (1.26–3.05) *	0.48 (0.09–2.47)	2.96 (2–4.39)*
Systemic Sepsis	0.89 (0.43–1.82)	0.15 (0.02–1.21)	1.67 (1.14–2.44) *	1.77 (0.69–4.52)	1.26 (0.91–1.74)
Corticosteroid Use	1.05 (0.55–2)	0.94 (0.26–3.4)	N/A	2.57 (1.13–5.85)*	1.16 (0.93–1.44)
Smoker	0.64 (0.38–1.05)	0.92 (0.37–2.29)	1.08 (0.9–1.3)	N/A	0.82 (0.69–0.98)*
Open Wound	0.98 (0.48–2.02)	1.03 (0.3–3.52)	1.47 (1.03–2.08) *	1.06 (0.42–2.7)	1.37 (1.01–1.87)*
On Dialysis	N/A	N/A	2 (1.06–3.78) *	N/A	0.72 (0.41–1.27)
Obese	1.15 (0.81–1.63)	2.68 (1.31–5.46)*	1.02 (0.89–1.15)	N/A	1.01 (0.9–1.15)
Hyponatremia	1.05 (0.62–1.77)	1.5 (0.61–3.67)	1.25 (1.01–1.56) *	1.65 (0.82–3.3)	1.34 (1.1–1.64)*
Myocardial Infarction	N/A	N/A	N/A	2.15 (0.33–14.03)	N/A
Congestive Heart Failure	N/A	N/A	N/A	0.7 (0.12–4.14)	1.18 (0.73–1.91)
Hypertension	N/A	N/A	1.22 (1.07–1.39) *	0.63 (0.35–1.11)	1.09 (0.96–1.24)
Impaired Functional Status	0.62 (0.38–1.01)	2.35 (1.07–5.16)*	2.84 (2.22–3.65) *	0.93 (0.48–1.8)	1.4 (1.14–1.72)*
Use of Emergency Services	N/A	N/A	1.75 (1.2–2.56) *	1.12 (0.39–3.22)	0.95 (0.76–1.21)
Preoperative Dyspnea	1.12 (0.65–1.92)	1.82 (0.7–4.74)	1.03 (0.8–1.33)	0.42 (0.15–1.18)	1.06 (0.9–1.25)
Diabetes	1.07 (0.67–1.7)	1.12 (0.47–2.66)	1.36 (1.15–1.62) *	1.9 (0.96–3.74)	0.47 (0.24–0.94)*
Disseminated Cancer	N/A	N/A	N/A	0.53 (0.05–5.63)	1.43 (0.45–4.62)
Pneumonia	N/A	N/A	N/A	1.5 (0.09–25.49)	1.4 (1.12–1.77)*
COPD	N/A	N/A	1.49 (1.17–1.91) *	2.25 (0.95–5.32)	N/A
Chemotherapy	1.65 (0.37–7.37)	4.01 (0.62–25.94)	N/A	N/A	N/A
Bleeding Disorder	0.83 (0.4–1.72)	0.48 (0.1–2.19)	1.4 (1.09–1.81) *	2.39 (1–5.72)	1.66 (1.33–2.09)*
Recent Weight Loss >10%	N/A	N/A	1.81 (0.92–3.59)	0.2 (0.01–2.72)	1.32 (0.72–2.43)
Ascites	N/A	N/A	2.87 (0.52–15.8)	N/A	1.58 (0.37–6.76)

* Significantly associated with complication, defined as p < 0.05.

N/A Not applicable due to non-significance on bivariate analysis with complication, defined as p > 0.05.

4. Discussion

Preoperative nutritional status can help clinicians anticipate surgical complications. Hypoalbuminemia has been previously shown to be associated with an increased risk of surgical site infections, pneumonia, longer hospital stays, and greater readmission rates in both primary total joint arthroplasties (TJAs) as well as other non-orthopaedic surgeries.^19,20,22,27 We hypothesized that similar trends between hypoalbuminemia and perioperative complications would exist in the revision THA patient population. Our results support this hypothesis, demonstrating an association between hypoalbuminemia and increased reoperation rate as well as bleeding complications, surgical site infections, non-home discharges, medical, and surgical complications in patients undergoing revision THA (Fig. 1).

Fig. 1.

Odds ratios and 95% confidence intervals of associations between hypoalbuminemia and postoperative complications.

Malnutrition has been previously shown to impair immune function, decrease collagen synthesis, and delay tissue healing.²⁸ Therefore, it is not surprising that malnutrition has also been shown to be associated with an increased risk of surgical site infections, increased length of hospital stay, and higher hospital costs.^29,30 Albumin is directly involved in the wound healing process, but its levels may also have implications as to other nutritional deficits, such as vitamin and other protein deficiencies needed for adequate wound healing. In primary arthroplasties, hypoalbuminemia has been associated with a nearly six-fold increase in mortality and a 50% increase in hospital readmissions.18, 19, 20^,23 In revision total knee arthroplasties, Kamath et al. reported an increased risk of deep surgical site infections, organ space surgical site infections, pneumonia, urinary traction infections, and sepsis.³¹ The risk of unplanned intubation, need for blood transfusion, mechanical ventilation for >48 h, and the development of acute renal failure was also substantially greater.³¹

Like hypoalbuminemia, a great deal of other patient risk factors described for revision THAs are potentially modifiable. Roth et al. reported an association between elevated BMI and increased rates of superficial infection, pneumonia, deep vein thrombosis, wound dehiscence, and urinary tract infections in patients undergoing revision THA.¹¹ Smoking has also been associated with an increased risk of deep infection and reoperation following revision THA.³² Mahadevan et al. reported an association between anemia and longer hospital stays in these patients,²⁹ and there has been an established association between revision joint arthroplasties due to sepsis and low albumin levels.³³ The findings of our investigation further add to the growing body of literature highlighting the impact of modifiable preoperative comorbidities on complication rates and short-term outcomes of revision THA.

Nutritional supplementation has been shown to reduce complications for geriatric hip fracture patients undergoing orthopaedic surgery.^19,34,35 Given that revision THAs are frequently needed secondary to periprosthetic hip fractures in the elderly population, postoperative supplementation in patients with hypoalbuminemia may be warranted.^35,36 For such high acuity injuries, surgery cannot be delayed for preoperative optimization for a long enough period of time to see clinically significant rises in albumin. The European Society for Clinical Nutrition and Metabolism (ESPEN) already recommends oral nutritional supplementation in geriatric patients after hip fracture and orthopaedic surgery.³⁷ Their recommendation is based off of the reported observation that voluntary oral intake in the elderly is oftentimes insufficient to meet the increased demands following surgery. In a group that already has additional comorbidities and an attenuated biologic healing capacity, they are at risk of deteriorating or having impaired recoveries. The ESPEN recommendation is largely based on a Cochrane review of 41 studies and 3881 that found that oral multi-nutrient supplements started before or soon after surgery may prevent complications within the first 12 months after hip fracture, despite a poor quality of evidence.³⁸ Additionally, Dwyer et al. found that calorie loading 48 h before total hip arthroplasty in the elderly reduced length of stay by three days.³⁹ Although there remain many questions regarding the duration, route, composition, and time to surgery surrounding the use of nutritional supplementation in the geriatric hip fracture population undergoing orthopaedic surgery, there is a strong case that it may improve outcomes. Therefore, future research should aim to investigate its utility in the malnourished revision total arthroplasty patients in conjunction with other parameters, such as prealbumin, lymphocyte count, and transferrin.

There are several important limitations to our study. The NSQIP database is limited by the accuracy and completeness of the data provided. Not all procedures performed at an institution are selected and added to the database, and NSQIP only tracks 30-day morbidity and mortality. Additionally, it does not provide severity of complications or specific indications for surgery. However, there is unlikely a systemic difference in the distribution of these limitations between types of procedures, and thus conclusions are unlikely to be affected by any minor discrepancies that may exist. Additionally, while albumin is a widely used marker of malnutrition, it is not specific. There are numerous possible etiologies of hypoalbuminemia including synthesis deficits or increased wasting or degradation.⁴⁰ It can be the result of hepatic dysfunction, extravascular protein loss, or increased degradation in catabolic states of injury or stress.^30,41 However, even if hypoalbuminemia is the result of a disease process other than malnutrition, it is important that this is identified prior to surgery. Furthermore, other means of measuring nutrition, such as prealbumin and transferrin, are being investigated at this time as well, which together with albumin can provide valuable information to optimize patients before surgery.

As the United States population continues to be more active in the later stages of life, revision THAs are expected to become more prevalent. Understanding the influence of hypoalbuminemia on outcomes is critical to maximizing patient functionality and quality of life. As hypoalbuminemia is associated with risk of increased complications in revision THA, we believe it warrants consideration in the preoperative planning of these patients. The benefits of preoperative nutritional supplementation should be further investigated.

Funding

None.

Author contribution

Nicole D. Rynecki: Conceptualization, Project administration, Original draft, Review & editing. Dominick V. Congiusta: Data curation, Formal analysis, Methodology, Project administration, Software, Supervision, Validation, Original draft, Review & editing. Michael Fields: Conceptualization, Original draft, Review & editing. Rushi Patel: Original draft, Review & editing. Michael M. Vosbikian: Conceptualization, Project administration, Original draft, Review & editing. Irfan H. Ahmed: Conceptualization, Project administration, Original draft, 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.