Hypoalbuminemia Independently Predicts Surgical Site Infection, Pneumonia, Length of Stay, and Readmission After Total Joint Arthroplasty

The Journal of Arthroplasty 31 (2016) 15–21

Contents lists available at ScienceDirect

The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org

Hypoalbuminemia Independently Predicts Surgical Site Infection, Pneumonia, Length of Stay, and Readmission After Total Joint Arthroplasty Daniel D. Bohl, MD, MPH a, Mary R. Shen, MS b, Erdan Kayupov, MSE a, Craig J. Della Valle, MD a a b

Department of Orthopaedic Surgery, Rush University Medical Center, Chicago, Illinois University of North Carolina School of Medicine, Chapel Hill, North Carolina

a r t i c l e

i n f o

Article history: Received 12 May 2015 Accepted 12 August 2015 Keywords: malnutrition hypoalbuminemia surgical site infection periprosthetic joint infection pneumonia total joint arthroplasty

a b s t r a c t This study investigates the association between preoperative hypoalbuminemia, a marker for malnutrition, and complications during the 30 days after total joint arthroplasty. Patients who underwent elective primary total hip and knee arthroplasty as part of the American College of Surgeons National Surgical Quality Improvement Program were identified. Outcomes were compared between patients with and without hypoalbuminemia (serum albumin concentration b 3.5 g/dL) with adjustment for patient and procedural factors. A total of 49603 patients were included. In comparison to patients with normal albumin concentration, patients with hypoalbuminemia had a higher risk for surgical site infection, pneumonia, extended length of stay, and readmission. Future efforts should investigate methods of correcting nutritional deficiencies prior to total joint arthroplasty. If successful, such efforts could lead to improvements in short-term outcomes for patients. © 2016 Elsevier Inc. All rights reserved.

Total joint arthroplasty (TJA) is one of the most common surgical procedures performed in the United States, and the demand for TJA procedures is predicted to grow [1,2]. Although TJA is a safe, elective procedure, a small percentage of TJA procedures do result in major complications [3–6]. Although many previously identified risk factors for complications after TJA cannot be altered, other risk factors may potentially be modified prior to surgery [3–6]. One potential risk factor that is appealing for preoperative intervention is malnutrition [7]. Serologic laboratory values, anthropometric measurements, and standardized assessments can all be used as markers for malnutrition [7]. In particular, hypoalbuminemia (serum albumin concentration b3.5 g/dL) is one of the simplest and most widely used markers for malnutrition [7,8]. Preoperative malnutrition has been identified in a number of studies as a risk factor for surgical site infection (SSI) and delayed wound healing after TJA [9–18]. For example, in a prospective study of 213 patients, malnutrition (defined by an abnormal triceps skin fold) was associated with development of SSI [9]. Similarly, in a retrospective evaluation of 375 patients undergoing revision TJA for aseptic failure, malnutrition (defined by abnormal albumin concentration, lymphocyte

One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements, refer to http://dx.doi.org/10.1016/j.arth.2015.08.028 Reprint requests: Craig J. Della Valle, MD, Department of Orthopaedic Surgery, Rush University Medical Center, 1611 W. Harrison St, Suite 300, Chicago, IL 60612. http://dx.doi.org/10.1016/j.arth.2015.08.028 0883-5403/© 2016 Elsevier Inc. All rights reserved.

count, or transferrin concentration) was associated with development of acute postoperative periprosthetic joint infection [10]. There have been fewer investigations into the association of malnutrition with other postoperative complications. One example of such an investigation is a prospective study of 2161 patients undergoing TJA [11]. In this study, malnutrition (defined by abnormal albumin or transferrin concentrations) was associated with hematoma formation, renal complications, and cardiac complications. However, this study may have lacked the sample size to sufficiently evaluate the full range of potential complications, many of which are rare events. The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) is a prospective surgical registry that samples patients from community and academic centers nationwide [19]. The program prospectively identifies patients undergoing major surgical procedures, including TJA, and tracks them for 30 days for the development of postoperative complications. The program also prospectively collects preoperative laboratory data, including preoperative serum albumin concentration. In this context, regarding patients undergoing primary TJA, the present study uses the ACS-NSQIP to determine (1) preoperative associations with hypoalbuminemia and (2) the association between preoperative hypoalbuminemia and several specific postoperative complications. Methods A retrospective analysis of prospectively collected data was conducted. Patients who underwent primary total hip arthroplasty (THA) or total knee arthroplasty (TKA) were identified as part of the ACS-NSQIP

D.D. Bohl et al. / The Journal of Arthroplasty 31 (2016) 15–21

in 2011 to 2013. Patients were identified using primary Current Procedural Terminology codes 27130 for primary THA and 27447 for primary TKA. The additional associated Current Procedural Terminology code fields and International Classification of Disease, Ninth Revision diagnosis code field were then used to exclude patients who were not definitively undergoing elective primary TJA. Specifically, patients whose cases involved prosthesis revision, hardware removal, major ligament reconstruction, additional unrelated procedures, acute trauma, or preoperative infection were excluded. Patients described as undergoing surgery nonelectively by the ACS-NSQIP were excluded. Finally, patients without preoperative serum albumin laboratory values were excluded. The ACS-NSQIP collects baseline data on demographics, comorbidities, and preoperative laboratory values [19]. Demographic data includes age, sex, and body mass index (BMI). Comorbidity data include diabetes mellitus, congestive heart failure, dyspnea on exertion, hypertension, end-stage renal disease, and chronic obstructive pulmonary disease. Preoperative serum laboratory values include hematocrit and serum albumin concentration. Insulin-dependent diabetes mellitus was defined through individual review of patient medical records as diabetes for which a patient uses insulin as a medication. Non–insulindependent diabetes, on the other hand, was defined through individual review of patient medical records as diabetes for which a patient uses only noninsulin medication for control. Patients not taking any medication are considered to not have diabetes. For the present study, anemia was defined as preoperative hematocrit below 36 for women or 41 for men (2 SDs below the mean for each sex) [20]. Consistent with previous studies [7], hypoalbuminemia was defined as serum albumin concentration less than 3.5 g/dL. The ACS-NSQIP follows up patients prospectively during the first 30 postoperative days for the development of postoperative complications [19]. Patients were considered to have had a serious complication occur if any of the following occurred during the first 30 postoperative days: systemic sepsis (either with or without shock), myocardial infarction, stroke, cardiac arrest, pulmonary embolism, mortality, coma more than 24 hours, or unplanned intubation (either unplanned intubation after the procedure or ventilator requirement for N48 hours after surgery). Patients were considered to have had any complication occur if any of the following occurred during the first 30 postoperative days: any of the serious complications, deep vein thrombosis, graft/prosthesis/flap failure, wound dehiscence, SSI, peripheral nerve injury, renal insufficiency (either acute renal failure or progressive renal insufficiency), urinary tract infection, or pneumonia. The ACS-NSQIP also follows up patients for postoperative hospital length of stay and unplanned readmission to any hospital [19]. In order to limit the effects of outliers on the analysis, postoperative hospital length of stay was capped at 30 days. The level of significance was set at α = .05 (P b .05). All tests were 2 tailed. Of note, the ACS-NSQIP does not provide information on surgeon or institution, so adjustment for clustering by these factors was not possible. First, bivariate and multivariate Poisson regression with robust error variance [21] was used to identify risk factors for hypoalbuminemia among the following demographic, comorbidity, and laboratory characteristics: procedure type (primary THA or TKA), age (18-39, 40-49, 5059, 60-69, 70-79, 80-89, or ≥ 90 years), sex (male or female), BMI (≤18.5, 18.5-25, 25-30, 30-35, 35-40, 40-45, 45-50, or ≥50 kg/m 2), diabetes mellitus (no diabetes mellitus, non–insulin-dependent diabetes mellitus, insulin-dependent diabetes mellitus), congestive heart failure, dyspnea on exertion, hypertension, end-stage renal disease, chronic obstructive pulmonary disease, current smoking status, and anemia. Next, 30-day postoperative outcomes were compared between patients with hypoalbuminemia and patients with normal serum albumin concentration. These comparisons were conducted both before and after multivariate adjustment for the demographic, comorbidity, and laboratory characteristics listed in the prior paragraph. Bivariate and multivariate Poisson regression with robust error variance was used to

compare rates of any complications, serious complications, specific complications, and unplanned readmission. Only specific complications with an incidence of at least 0.1% were individually analyzed. Bivariate and multivariate linear regression was used to compare postoperative length of stay. Required ACS-NSQIP statement: “The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.”

Results A total of 101523 patients were initially identified, of whom 49 603 (48.9%) had preoperative serum albumin laboratory values available for analysis. These 49 603 patients constituted the study population. Of these patients, 19 975 (40.3%) underwent primary THA and 29 628 (59.7%) underwent primary TKA. The prevalence of hypoalbuminemia was 4.0% (95% confidence interval [CI], 3.8%-4.1%; Fig. 1). In the multivariate analysis (Table), hypoalbuminemia was independently associated with age strata (P b .001), female sex (P b .001), BMI strata (P b .001), diabetes mellitus (P b .001), dyspnea on exertion (P b .001), end-stage renal disease (P b .001), chronic obstructive pulmonary disease (P b .001), current smoker status (P b .001), and anemia (P b .001). For age and BMI, these associations were bimodal, with increased rates of hypoalbuminemia at both extremes (Fig. 2A and B; Table). For diabetes mellitus, the prevalence was highest for patients with insulin-dependent diabetes mellitus (Fig. 2C; Table). All subsequent analyses were adjusted for the demographic, comorbidity, and laboratory characteristics listed in Table. Both adjusted and unadjusted values are presented. In comparison to patients with normal serum albumin concentration, patients with hypoalbuminemia had a higher risk for occurrence of any complications (7.3% vs 4.0%; unadjusted relative risk [RR], 1.8 [95% CI, 1.6-2.2; P b .001]; adjusted RR, 1.5 [95% CI, 1.2-1.7; P b .001]; Fig. 3). Similarly, patients with hypoalbuminemia had a higher risk for occurrence of serious complications (2.1% vs 1.2%; unadjusted RR, 1.8 [95% CI, 1.3-2.5; P b .001]; adjusted RR, 1.4 [95% CI, 1.0-1.9; P = .042]).

Distribution of Preoperative Serum Albumin Concentration 30

Percent of Patients (%)

16

Normal Serum Albumin Concentration 47,639 Patients (96.0%)

20

10

Hypoalbuminemia 1,964 Patients (4.0%)

0 2

3

4

5

6

Serum Albumin Concentration (g/dL) Fig. 1. Distribution of preoperative serum albumin concentration. In total, 47639 patients (96.0%) had normal serum albumin concentration, whereas 1964 patients (4.0%) had hypoalbuminemia (serum albumin concentration b3.5 g/dL).

D.D. Bohl et al. / The Journal of Arthroplasty 31 (2016) 15–21

17

Table Prevalence and Risk Factors for Hypoalbuminemia. Statistical Comparisons Prevalence of Hypoalbuminemia Prevalence Overall Procedure Primary THA Primary TKA Age (y) 18-39 40-49 50-59 60-69 70-79 80-89 ≥90 Sex Male Female BMI (kg/m2) ≤18.5 18.5-25 25-30 30-35 35-40 40-45 45-50 ≥50 Diabetes mellitus No diabetes mellitus NIDDM IDDM Congestive heart failure No Yes Dyspnea on exertion No Yes Hypertension No Yes End-stage renal disease No Yes COPD No Yes Current smoker No Yes Anemiab No Yes

95% CI

4.0%

3.8-4.1

3.8% 4.0%

3.6%-4.1% 3.8%-4.3%

Multivariate Comparisonsa

Bivariate Comparisons RR

95% CI

Ref. 1.1

– 1.0-1.1

P

RR

95% CI

.262

.352 Ref. 1.0

1.0-1.1

1.3 1.1 Ref. 1.0 1.3 1.8 3.5

0.8-1.9 0.9-1.4 – 0.9-1.2 1.1-1.5 1.6-2.2 2.5-4.9

Ref. 1.6

– 1.4-1.7

2.6 1.2 1.0 Ref. 1.2 1.6 1.7 2.8

1.9-3.6 1.0-1.4 0.8-1.1 – 1.1-1.4 1.3-1.9 1.4-2.2 2.2-3.6

Ref. 0.8 1.4

– 0.7-0.9 1.2-1.6

Ref. 1.5

– 0.9-2.4

Ref. 1.6

– 1.4-1.8

b.001 4.8% 3.9% 3.3% 3.4% 4.0% 6.3% 13.3%

2.9%-6.7% 3.1%-4.8% 3.0%-3.7% 3.1%-3.6% 3.7%-4.4% 5.6%-6.9% 9.4%-17.2%

1.4 1.2 Ref. 1.0 1.2 1.9 3.9

1.0-2.2 0.9-1.5 – 0.9-1.2 1.1-1.4 1.6-2.2 2.9-5.4

3.3% 4.4%

3.1%-3.6% 4.1%-4.6%

Ref. 1.3

– 1.2-1.4

12.4% 4.9% 3.3% 3.2% 4.0% 5.1% 5.8% 9.9%

8.5%-16.3% 4.3%-5.4% 3.0%-3.5% 2.9%-3.5% 3.6%-4.4% 4.4%-5.8% 4.7%-7.0% 7.7%-12.1%

3.8 1.5 1.0 Ref. 1.2 1.6 1.8 3.1

2.8-5.3 1.3-1.7 0.9-1.1 – 1.1-1.4 1.3-1.8 1.4-2.2 2.4-3.9

3.8% 3.9% 8.0%

3.6%-4.0% 3.4%-4.4% 6.8%-9.2%

Ref. 1.0 2.1

– 0.9-1.2 1.8-2.5

3.9% 11.2%

3.8%-4.1% 6.0%-16.4%

Ref. 2.8

– 1.8-4.5

3.7% 7.9%

3.5%-3.8% 7.0%-8.8%

Ref. 2.1

– 1.9-2.4

3.2% 4.4%

2.9%-3.4% 4.2%-4.6%

Ref. 1.4

– 1.3-1.5

b.001

b.001

b.001

b.001

b.001

b.001

b.001

b.001

.099

b.001

b.001

b.001

.136 Ref. 1.1

1.0-1.2

Ref. 4.2

– 3.2-5.7

Ref. 1.4

– 1.1-1.6

Ref. 1.6

– 1.4-1.8

Ref. 3.0

– 2.8-3.3

b.001 3.9% 30.0%

3.7%-4.1% 21.1%-38.7%

Ref. 7.7

– 5.7-10.3

3.8% 7.9%

3.6%-4.0% 6.7%-9.0%

Ref. 2.1

– 1.8-2.4

3.8% 5.2%

3.6%-4.0% 4.6%-5.8%

Ref. 1.4

– 1.2-1.5

2.8% 8.2%

2.6%-2.9% 7.7%-8.8%

Ref. 3.0

– 2.7-3.3

P

b.001

b.001

b.001

b.001

b.001

b.001

b.001

Abbreviations: COPD, chronic obstructive pulmonary disease, IDDM, insulin-dependent diabetes mellitus; NIDDM, non–insulin-dependent diabetes mellitus; Ref., reference. a Adjusted for all other demographic, comorbidity, and laboratory characteristics listed in this table. b Anemia was defined as preoperative hematocrit below 36 for women or 41 for men.

A total of 9 specific complications had rates higher than 0.1% and were individually analyzed (Fig. 4). After adjustment, patients with hypoalbuminemia had a higher risk for SSI (2.29% vs 0.96%; adjusted RR, 2.0; 95% CI, 1.5-2.8; P b .001) and pneumonia (1.27% vs 0.30%; adjusted RR, 2.5; 95% CI, 1.6-4.0; P b .001). Mean postoperative length of stay was longer for patients with hypoalbuminemia (3.52 vs 3.10 days; unadjusted difference, 0.42 days [95% CI, 0.34-0.49 days; P b .001]; adjusted difference, 0.20 days [95% CI, 0.12-0.27; P b .001]; Fig. 5). Of note, of the 2232 patients with length of stay longer than 5 days, 463 (20.7%) had one of the specific complications analyzed above. Similarly, the rate of unplanned readmission to the hospital was higher for patients with hypoalbuminemia (6.3% vs 3.5%; unadjusted RR, 1.8 [95% CI, 1.5-2.2; P b .001]; adjusted RR, 1.4 [95% CI, 1.2-1.7; P b .001]; Fig. 6).

Discussion Prior studies have linked malnutrition to delayed wound healing and SSI after TJA [9–18]. However, few studies have investigated the association between malnutrition and other postoperative complications [11]. Consistent with the literature, the present study finds that hypoalbuminemia, a marker for malnutrition, is independently associated with a 2-fold increase in the rate of SSI. The present study also demonstrates independent associations between hypoalbuminemia and pneumonia, hospital length of stay, and hospital readmission. Malnutrition may predispose patients to SSI through several mechanisms [7,22]. The first mechanism involves impairment in wound healing via diminished fibroblast proliferation and collagen synthesis. The second mechanism involves impairment in the ability of the

18

D.D. Bohl et al. / The Journal of Arthroplasty 31 (2016) 15–21

Risk Factors for Hypoalbuminemia B

A Hypoalbuminemia, by Age

Hypoalbuminemia, by BMI 20

Percent of Patients with Hypoalbuminemia (%)

Percent of Patients with Hypoalbuminemia (%)

20

15

10

5

0

15

10

5

0 18-39

40-49

50-59

60-69

70-79

80-89

90

Age (Years) BMI (kg/m2)

C Hypoalbuminemia, by Diabetic Status No Diabetes Mellitus Non-Insulin-Dependent Diabetes Mellitus Insulin-Dependent Diabetes Mellitus 0

5

10

15

Percent of Patients with Hypoalbuminemia (%) Fig. 2. Risk factors for hypoalbuminemia. A and B, Age and BMI had bimodal associations with hypoalbuminemia, with patients at both extremes having elevated risk. C, Patients with insulin-dependent diabetes mellitus had elevated risk for hypoalbuminemia. A-C, Corresponding statistical associations are reported in Table.

immune system to fight infection, at least in part through lymphocytopenia. Pneumonia was the only complication in addition to SSI that was found to be independently associated with hypoalbuminemia. Specifically, pneumonia was more than twice as likely for patients with hypoalbuminemia, even after risk adjustment. Malnutrition could potentially increase the risk for pneumonia through the same immune impairment mechanism as for SSI. To our knowledge, this association has not been previously described.

The increased rates of SSI and pneumonia—as well as the nonstatistically significant changes in rates of other complications—all contributed to increased rates of the composite outcomes: “any complications” and “serious complications.” Both of these composite outcomes, which are commonly studied in analyses of risk-adjusted registry data [4,5,23–25], were increased about 1.5-fold in patients with hypoalbuminemia, even after risk adjustment. Given the increases in these composite outcomes, it is expected that there would be

Rates of Any Complications and Serious Complications Unadjusted RR = 1.8, p < 0.001 Adjusted RR = 1.5, p < 0.001

Any Complications

Unadjusted RR = 1.8, p < 0.001 Adjusted RR = 1.4, p = 0.042

Serious Complications 0

5

10

15

20

Percent of Patients with Any or Serious Complications Normal Serum Albumin Concentration Hypoalbuminemia

Fig. 3. Rates of any complications and serious complications. In comparison to patients with normal serum albumin concentration, patients with hypoalbuminemia had a higher risk for occurrence of any complications (7.3% vs 4.0%; unadjusted RR, 1.8 [95% CI, 1.6-2.2; P b .001]; adjusted RR, 1.5 [95% CI, 1.2-1.7; P b .001]). Similarly, patients with hypoalbuminemia had a higher risk for occurrence of serious complications (2.1% vs 1.2%; unadjusted RR, 1.8 [95% CI, 1.3-2.5; P b .001]; adjusted RR, 1.4 [95% CI, 1.0-1.9; P = .042]). Error bars represent standard errors.

D.D. Bohl et al. / The Journal of Arthroplasty 31 (2016) 15–21

19

Rates of Specific Complications Rate (%) 0

Wound Dehiscence Deep Vein Thrombosis

1

0.75% 0.71% 1.06% 1.63%

Pulmonary Embolism

0.45% 0.51%

Myocardial Infarction

0.24% 0.51%

Unplanned Intubation

0.2

0.4

1

2.5

Adjusted Relative Risk

5

10

2.29% 0.30%

1

2.5

5

0.9

1.5 *

1.2

10

1.2

1.3

1.8

3.1 *

1.8

2.4 *

2.0 *

4.2 *

1.27%

0.4

1.0

4.7 *

0.96%

0.2

0.9

2.1 *

0.18% 0.56%

0.1

0.9

1.1

0.17% 0.46%

Surgical Site Infection Pneumonia

0.1

0.16% 0.15%

Urinary Tract Infection

Renal Insufficiency

Unadjusted Relative Risk 3

2

2.5 *

Normal Serum Albumin Hypoalbuminemia Fig. 4. Rates of specific complications. A total of 9 specific complications had rates higher than 0.1% and were individually analyzed. After adjustment, patients with hypoalbuminemia had a higher risk for SSI (2.29% vs 0.96%; adjusted RR, 2.0; 95% CI, 1.5-2.8; P b .001) and pneumonia (1.27% vs 0.30%; adjusted RR, 2.5; 95% CI, 1.6-4.0; P b .001). Error bars represent 95% CIs. Asterisks (*) denote statistical difference (P b .05).

corresponding increases in other important hospital metrics, such as postoperative length of stay and unplanned hospital readmission. After adjustment, patients with hypoalbuminemia had longer postoperative stays of about one-fifth of a day. This finding has considerable implications when one considers that the average day in a US hospital costs more than $4000 [26]. The hospital readmission rate in patients with hypoalbuminemia was increased 1.4-fold. This finding has considerable implications in the setting of bundled payments, pay-forperformance, and public reporting of physician statistics [27–32]. The logical extension of the observed associations between preoperative hypoalbuminemia and postoperative complications is the attempt to correct preoperative malnutrition prior to surgery [33,34]. Optimization of nutritional status prior to major surgery has been demonstrated to improve outcomes in several studies [34–37]. For example, among a consecutive series of patients admitted for abdominal surgery as part of

Postoperative Length of Stay Percent of Patients Remaining in Hospital (%)

100

75 Normal Serum Albumin: Mean LOS = 3.10 Days Hypoalbuminemia: Mean LOS = 3.52 Days

50 Unadjusted Difference = 0.42 Days, p < 0.001 Adjusted Difference = 0.20 Days, p < 0.001

25

0 0

5

10

15

20

25

30

Days After Procedure Fig. 5. Postoperative length of stay. Mean postoperative length of stay was longer for patients with hypoalbuminemia (3.52 vs 3.10 days; unadjusted difference, 0.42 days [95% CI, 0.34-0.49]; adjusted difference, 0.20 days [95% CI, 0.12-0.27]; P b .001).

a multicenter, prospective cohort study, the complication rate was lower in a group that received comprehensive nutritional supplementation than in a control group (25.6% vs 50.6%) [35]. Similarly, in randomized trials of geriatric hip fracture patients, nutritional supplementation has been shown to reduce the total rate of complications [36,37]. However, the extent to which nutritional supplementation might help in the setting of more modest nutritional deficiencies and elective orthopedic procedures like TJA is still unknown [7]. Moreover, although hypoalbuminemia is a marker for malnutrition, specifically protein-energy deficiency, it may not be the direct cause of complications. In fact, previous research demonstrates that simply restoring albumin levels may not improve outcomes [38]. Correspondingly, it is currently not clear whether correction of hypoalbuminemia could lead to improvements in surgical outcomes. It is likely that hypoalbuminemia is simply associated with other nutritional deficiencies that lead to the poor outcomes. What is clear is that various forms of malnutrition that may include or simply be associated with hypoalbuminemia are strongly associated with outcomes. Hence, additional work is needed to clarify the extent to which hypoalbuminemia is a direct cause vs a correlate of poor surgical outcomes, and hence, whether correction of hypoalbuminemia or correction of the additional associated deficiencies will reduce the rates of postoperative adverse events. Finally, the present study characterized risk factors for hypoalbuminemia among a TJA population. Risk factors included both young and old age, both low and high BMI, female sex, and a range of comorbidities. The associations of hypoalbuminemia with both low and high BMI support previous findings [10]. These associations demonstrate the difficulty with using BMI as a proxy for malnutrition/hypoalbuminemia and suggest that clinicians should consider screening all patients for hypoalbuminemia regardless of BMI. This study is strengthened by use of the ACS-NSQIP, which has gained a high degree of acceptance in both the orthopedic [23,25,39,40] and general surgery [41–43] literature as a valid and reliable data source for risk-adjusted clinical investigations. Data from ACS-NSQIP are prospectively collected for the purpose of quality improvement and clinical research (unlike data contained in administrative databases, which are

20

D.D. Bohl et al. / The Journal of Arthroplasty 31 (2016) 15–21

Unplanned Hospital Readmission

Unadjusted RR = 1.8, p < 0.001 Adjusted RR = 1.4, p < 0.001

Readmission Within 30 Days of Surgery

0

2

4

6

8

10

12

14

16

Percent of Patients Readmitted Within 30 Days of Surgery (%) Normal Serum Albumin Concentration Hypoalbuminemia

Fig. 6. Unplanned hospital readmission. The rate of readmission within 30 days of surgery was higher for patients with hypoalbuminemia (6.3% vs 3.5%; unadjusted RR, 1.8 [95% CI, 1.5-2.2; P b .001]; adjusted RR, 1.4 [95% CI, 1.2-1.7; P b .001]). Error bars represent standard errors.

retrospectively collected from hospital discharge claims) [24,44]. Importantly, the demographic, comorbidity, and laboratory characteristics collected by the ACS-NSQIP were chosen by the American College of Surgeons to be the most relevant for risk adjustment of complication rates. Hence, the use of those characteristics to adjust analyses in the present study should minimize confounding of study results. The weaknesses of this study are also related to the ACS-NSQIP. Perhaps most importantly, the ACS-NSQIP does not collect information on orthopedic-specific outcomes such as return to function, change in pain, periprosthetic fracture, or patient satisfaction. Relatedly, outcomes extend only to the 30th postoperative day. Also, even in the setting of risk adjustment using ACS-NSQIP demographic, comorbidity, and laboratory characteristics, there remains the potential that unmeasured characteristics confound study results. Finally, one method of measuring nutrition is the delayed hypersensitivity skin test [45]. Unfortunately, patients in the ACS-NSQIP are not routinely given this test, so such data are unavailable to us. One additional point of discussion involves the extent to which albumin serves as a marker for malnutrition [46,47]. Hypoalbuminemia was initially identified as a marker for protein-energy malnutrition in association with kwashiorkor [48]. Serum albumin concentration has since been used widely as a clinical marker for nutritional status, despite the fact that serum albumin concentration may be correlated as strongly with inflammation as it is with malnutrition [46,47,49]. Such inflammation can be a direct result of obesity, among other processes, potentially explaining at least a portion of the hypoalbuminemia observed in obese patients. There remains considerable discussion within the field of nutrition regarding the significance of abnormal serum albumin concentrations in clinical studies and clinical practice [46,47,49]. Nevertheless, hypoalbuminemia remains the most widely recognized marker for malnutrition [7,8]. As demonstrated here, hypoalbuminemia prior to TJA is independently associated with specific infectious complications, increased length of stay, and readmission. Future work should focus on translating these findings into clinical interventions that can be tested for effectiveness. If such interventions can be shown to improve nutritional metrics and/or reduce complication rates, it is possible that they could become a routine part of preoperative care.

References 1. Fingar KR, Stocks C, Weiss AJ, et al. Most frequent operating room procedures performed in U.S. hospitals, 2003–2012. Agency for Healthcare Research and Quality; 2014. 2. Kurtz S, Ong K, Lau E, et al. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am 2007; 89(4):780. 3. Pulido L, Parvizi J, Macgibeny M, et al. In hospital complications after total joint arthroplasty. J Arthroplast 2008;23(6):139. 4. Belmont Jr PJ, Goodman GP, Waterman BR, et al. Thirty-day postoperative complications and mortality following total knee arthroplasty: incidence and risk factors among a national sample of 15,321 patients. J Bone Joint Surg Am 2014;96(1):20. 5. Belmont Jr PJ, Goodman GP, Hamilton W, et al. Morbidity and mortality in the thirtyday period following total hip arthroplasty: risk factors and incidence. J Arthroplast 2014;29(10):2025.

6. Illingworth KD, El Bitar YF, Banerjee D, et al. Inpatient mortality after primary total hip arthroplasty: analysis from the National Inpatient Sample Database. J Arthroplast 2014; 30(3):369–73. 7. Cross MB, Yi PH, Thomas CF, et al. Evaluation of malnutrition in orthopaedic surgery. J Am Acad Orthop Surg 2014;22(3):193. 8. Russell M, McAdams M, Matarese L, et al. Laboratory monitoring of nutritional status. Contemporary nutrition support practice: a clinical guide; 1998 47. 9. Font-Vizcarra L, Lozano L, Rios J, et al. Preoperative nutritional status and postoperative infection in total knee replacements: a prospective study of 213 patients. Int J Artif Organs 2011;34(9):876. 10. Yi PH, Frank RM, Vann E, et al. Is potential malnutrition associated with septic failure and acute infection after revision total joint arthroplasty? Clin Orthop Relat Res 2015;473(1):175. 11. Huang R, Greenky M, Kerr GJ, et al. The effect of malnutrition on patients undergoing elective joint arthroplasty. J Arthroplast 2013;28(8):21. 12. Maier G, Horas K, Seeger J, et al. Is there an association between periprosthetic joint infection and low vitamin D levels? Int Orthop 2014;38(7):1499. 13. Peersman G, Laskin R, Davis J, et al. Infection in total knee replacement: a retrospective review of 6489 total knee replacements. Clin Orthop Relat Res 2001;392:15. 14. Del Savio GC, Zelicof SB, Wexler LM, et al. Preoperative nutritional status and outcome of elective total hip replacement. Clin Orthop Relat Res 1996;326:153. 15. Gherini S, Vaughn BK, Lombardi Jr AV, et al. Delayed wound healing and nutritional deficiencies after total hip arthroplasty. Clin Orthop Relat Res 1993;293:188. 16. Greene KA, Wilde AH, Stulberg BN. Preoperative nutritional status of total joint patients. Relationship to postoperative wound complications. J Arthroplasty 1991;6(4):321. 17. Jaberi FM, Parvizi J, Haytmanek CT, et al. Procrastination of wound drainage and malnutrition affect the outcome of joint arthroplasty. Clin Orthop Relat Res 2008;466(6):1368. 18. Pratt WB, Veitch JM, McRoberts RL. Nutritional status of orthopedic patients with surgical complications. Clin Orthop Relat Res 1981;155:81. 19. American College of Surgeons National Surgical Quality Improvement Program. User Guide for the 2013 Participant Use Data File; 2014[Available at: http://www.acsnsqip. org. Accessed 2014 Dec]. 20. Chitale R, Campbell PG, Yadla S, et al. International Classification of Disease Clinical Modification 9 modeling of a patient comorbidity score predicts incidence of perioperative complications in a nationwide inpatient sample assessment of complications in spine surgery. J Spinal Disord Tech 2012;28(4):126–33. 21. Zou G. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol 2004;159(7):702. 22. Seibert DJ. Pathophysiology of surgical site infection in total hip arthroplasty. Am J Infect Control 1999;27(6):536. 23. Bohl DD, Basques BA, Golinvaux NS, et al. Extramedullary compared with intramedullary implants for intertrochanteric hip fractures: thirty-day outcomes of 4432 procedures from the ACS NSQIP database. J Bone Joint Surg Am 1871;96(22):2014. 24. Bohl DD, Fu MC, Golinvaux NS, et al. The “July effect” in primary total hip and knee arthroplasty: analysis of 21,434 cases from the ACS-NSQIP database. J Arthroplast 2014;29(7):1332. 25. Duchman KR, Gao Y, Pugely AJ, et al. Differences in short-term complications between unicompartmental and total knee arthroplasty: a propensity score matched analysis. J Bone Joint Surg Am 2014;96(16):1387. 26. Shamji MF, Parker S, Cook C, et al. Impact of body habitus on perioperative morbidity associated with fusion of the thoracolumbar and lumbar spine. Neurosurgery 2009; 65(3):490. 27. Rana AJ, Bozic KJ. Bundled payments in orthopaedics. Clin Orthop Relat Res 2015; 473(2):422. 28. Lucas DJ, Haider A, Haut E, et al. Assessing readmission after general, vascular, and thoracic surgery using ACS-NSQIP. Ann Surg 2013;258(3):430. 29. Lucas DJ, Pawlik TM. Readmission after surgery. Adv Surg 2014;48:185. 30. Bozic KJ, Ward L, Vail TP, et al. Bundled payments in total joint arthroplasty: targeting opportunities for quality improvement and cost reduction. Clin Orthop Relat Res 2014;472(1):188. 31. Clement RC, Kheir MM, Derman PB, et al. What are the economic consequences of unplanned readmissions after TKA? Clin Orthop Relat Res 2014;472(10):3134. 32. Mednick RE, Alvi HM, Krishnan V, et al. Factors affecting readmission rates following primary total hip arthroplasty. J Bone Joint Surg Am 2014;96(14):1201. 33. Dempsey DT, Mullen JL, Buzby GP. The link between nutritional status and clinical outcome: can nutritional intervention modify it? Am J Clin Nutr 1988;47(2 Suppl.):352.

D.D. Bohl et al. / The Journal of Arthroplasty 31 (2016) 15–21 34. Evans DC, Martindale RG, Kiraly LN, et al. Nutrition optimization prior to surgery. Nutr Clin Pract 2014;29(1):10. 35. Jie B, Jiang ZM, Nolan MT, et al. Impact of preoperative nutritional support on clinical outcome in abdominal surgical patients at nutritional risk. Nutrition 2012;28(10):1022. 36. Delmi M, Rapin CH, Bengoa JM, et al. Dietary supplementation in elderly patients with fractured neck of the femur. Lancet 1990;335(8696):1013. 37. Eneroth M, Olsson UB, Thorngren KG. Nutritional supplementation decreases hip fracture-related complications. Clin Orthop Relat Res 2006;451:212. 38. Vanek VW. The use of serum albumin as a prognostic or nutritional marker and the pros and cons of IV albumin therapy. Nutr Clin Pract 1998;13(3):110. 39. Schilling PL, Hallstrom BR, Birkmeyer JD, et al. Prioritizing perioperative quality improvement in orthopaedic surgery. J Bone Joint Surg Am 1884;92(9):2010. 40. Molina CS, Thakore RV, Blumer A, et al. Use of the National Surgical Quality Improvement Program in orthopaedic surgery. Clin Orthop Relat Res 2014;473(5):1574–81. 41. Ingraham AM, Richards KE, Hall BL, et al. Quality improvement in surgery: the American College of Surgeons National Surgical Quality Improvement Program approach. Adv Surg 2010;44:251. 42. Ghaferi AA, Birkmeyer JD, Dimick JB. Variation in hospital mortality associated with inpatient surgery. N Engl J Med 2009;361(14):1368.

21

43. Shiloach M, Frencher Jr SK, Steeger JE, et al. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg 2010;210(1):6. 44. Bohl DD, Russo GS, Basques BA, et al. Variations in data collection methods between national databases affect study results: a comparison of the nationwide inpatient sample and National Surgical Quality Improvement Program databases for lumbar spine fusion procedures. J Bone Joint Surg Am 2014;96(23):e193. 45. Linn BS. Delayed hypersensitivity skin testing in nutritional assessment. Am Surg 1987;53(11):628. 46. Don BR, Kaysen G. Serum albumin: relationship to inflammation and nutrition. Semin Dial 2004;17(6):432. 47. Fuhrman MP. The albumin-nutrition connection: separating myth from fact. Nutrition 2002;18((2):199. 48. Reeds PJ, Laditan AA. Serum albumin and transferrin protein-energy malnutrition. Their use in the assessment of marginal undernutrition and the prognosis of severe undernutrition. Br J Nutr 1976;36(2):255. 49. Moshage HJ, Janssen JA, Franssen JH, et al. Study of the molecular mechanism of decreased liver synthesis of albumin in inflammation. J Clin Invest 1987;79(6): 1635.