Hypoalbuminemia as a Predictor of Mortality, Disability, and Readmission in Patients Undergoing Spine Surgery

A Retrospective Analysis

Zane Randell, BA; Brook Martin, MPH, PhD; Nathan Hendrickson, MD; Darrel Brodke, MD; Ryan Spiker, MD; Brandon Lawrence, MD; Nicholas Spina, MD

Disclosures

Spine. 2023;48(18):1300-1307.

Abstract and Introduction

Abstract

Study Design: This was a retrospective cohort study.

Objective: The objective of this study was to clarify the association between preoperative albumin status and mortality and morbidity in lumbar spine surgery.

Summary of Background Data: Hypoalbuminemia is a known marker of inflammation and is associated with frailty. Hypoalbuminemia is an identified risk factor for mortality following spine surgery for metastases, yet has not been well studied among spine surgical cohorts outside of metastatic cancer.

Materials and Methods: We identified patients with preoperative serum albumin laboratory values who underwent lumbar spine surgery at a US public university health system between 2014 and 2021. Demographic, comorbidity, and mortality data were collected along with preoperative and postoperative Oswestry Disability Index (ODI) scores. Any cause readmission within 1 year of surgery was recorded. Hypoalbuminemia was defined as <3.5 g/dL in serum. We examined the Kaplan-Meier survival plots based on serum albumin. Multivariable regression models were used to identify the association between preoperative hypoalbuminemia with mortality, readmission, and ODI, while controlling for age, sex, race, ethnicity, procedure, and Charlson Comorbidity Index.

Results: Of 2573 patients, 79 were identified as hypoalbuminemic. Hypoalbuminemic patients had a significantly greater adjusted risk of mortality through 1 year (odds ratio=10.2; 95% CI: 3.1–33.5; P<0.001), and 7 years (hazard ratio=4.18; 95% CI: 2.29–7.65; P<0.001). Hypoalbuminemic patients had ODI scores 13.5 points higher (95% CI: 5.7–21.4; P<0.001) at baseline. Adjusted readmission rates were not different between groups through 1 year (odds ratio=1.15; 95% CI: 0.5–2.62; P=0.75) or through full surveillance (hazard ratio=0.82; 95% CI: 0.44–1.54; P=0.54).

Conclusions: Preoperative hypoalbuminemia was strongly associated with postoperative mortality. Hypoalbuminemic patients did not have demonstrably worse outcomes in their functional disability beyond 6 months. Within the first 6 months following surgery, the hypoalbuminemic group improved at a similar rate to the normoalbuminemic group despite having a greater preoperative disability. However, causal inference is limited in this retrospective study.

Introduction

Increased annual volume of lumbar spine surgeries^[1,2] intensifies the need to identify patients at risk for postoperative mortality and morbidity.^[3,4] Hypoalbuminemia is commonly associated with mortality and morbidity in surgery^[5] but has not been studied in spine surgery outside of spinal metastatic cancer.^[6]

As a negative acute-phase protein, plasma albumin levels drop when the body is in an inflammatory state associated with chronic or critical illness.^[7] While the etiology behind reductions in plasma albumin are multifactorial and not entirely understood, hypoalbuminemia is a risk factor for surgical complications.^[8–12] The Spinal Oncology Research Group (SORG), through its machine learning algorithm, has identified hypoalbuminemia to be a critical risk factor for estimating mortality in patients undergoing surgery for spinal metastasis.^[13] However, less is known about hypoalbuminemia as a predictor of readmission and disability in lumbar spine surgery patients.

Common risk factors for mortality and morbidity in spine surgery, including race, age, and sex, are unmodifiable.^[3,14–16] Serum albumin, however, may be modifiable^[7] and may allow surgeons to better council patients on when, and whether, to proceed with surgery. Accordingly, this study assesses the relationship between preoperative hypoalbuminemia and postoperative mortality, readmission, and Oswestry Disability Index (ODI)^[17] in patients undergoing lumbar spine surgery.

1 2 3 4

Next Section

Table 1. Characteristics of Lumbar Spine Surgery Cases by Preoperative Albumin Level
	n (%)		P
	>3.5 g/dL	≤3.5 g/dL	P
N	2458	79	—
Age [mean (SD)]	61.2 (14.4)	59.3 (16.2)	0.319
Preoperative albumin level [mean (SD)]	4.3 (0.3)	3.0 (0.4)	<0.001
Sex			0.128
Male	1261 (51.3)	46 (58.2)
Female	1197 (48.7)	33 (41.8)
Hispanic			0.608
No	2313 (95.0)	72 (92.3)
Yes	122 (5.0)	6 (7.7)
Race			0.030
White	2222 (90.4)	63 (79.8)
Black	25 (1.0)	3 (3.8)
Other	211 (8.6)	13 (16.5)
Procedures			0.602
Decompression without fusion	837 (34.1)	29 (36.7)
Fusion	1583 (64.4)	49 (62.0)
Osteotomy or corpectomy	37 (1.5)	1 (1.3)
Surgical Invasiveness Index	1.2 (0.9)	1.3 (0.8)	0.372
Charlson Comorbidity Index score	2.0 (2.5)	3.3 (3.0)	<0.001
Individual comorbidities
Myocardial infarction	164 (6.7)	6 (7.6)	0.651
Congestive heart failure	204 (8.3)	13 (16.5)	0.022
Peripheral vascular disease	311 (12.7)	18 (22.8)	0.016
Cardiovascular disease	250 (10.2)	11 (13.9)	0.262
Dementia	40 (1.6)	2 (2.5)	0.380
COPD	656 (26.8)	24 (30.4)	0.519
Rheumatic disease	213 (8.7)	10 (12.7)	0.224
Peptic ulcer disease	109 (4.5)	6 (7.6)	0.103
Liver disease, mild	278 (11.4)	16 (20.3)	0.020
Liver disease, severe	2 (<0.1)	0 (0)	1.00
Diabetes without complication	302 (12.3)	16 (20.3)	0.055
Diabetes with complications	320 (13.1)	9 (11.4)	0.865
Paraplegia	125 (5.1)	9 (11.4)	0.035
Renal disease	266 (10.9)	14 (17.7)	0.067
Cancer without metastases	228 (9.3)	7 (8.9)	0.845
Cancer with metastases	54 (2.2)	7 (8.9)	0.002
HIV or AIDS	22 (0.9)	3 (3.8)	0.041

AIDS indicates acquired immunodeficiency syndrome; COPD, chronic obstructive pulmonary disease; HIV, human immunodeficiency virus.

Table 2. Indications for Surgery by Albumin Status
	n (%)		P
	>3.5 g/dL	≤3.5 g/dL	P
N	2458	79	—
Spinal diagnosis (not mutually exclusive)
Axial pain, disk degeneration	1340 (54.5)	45 (57.0)	0.667
Disk herniation	1002 (48.5)	13 (20.0)	<0.001
Disk herniation with neurological involvement	856 (41.4)	8 (12.3)	<0.001
Stenosis	1610 (77.9)	44 (67.7)	0.051
Spondylolisthesis	937 (45.4)	25 (38.5)	0.272
Scoliosis	396 (19.2)	12 (18.5)	0.887
Fracture or dislocation	72 (3.5)	15 (23.1)	<0.001
Spinal cord injury	14 (0.7)	3 (4.6)	0.014
Congenital or other anomaly	64 (3.1)	15 (23.1)	<0.001
Osteoporosis	13 (0.6)	1 (1.5)	0.353
Surgical aftercare	909 (44.0)	41 (63.1)	0.002

Table 3. Multivariable Regression Model for Readmission and Mortality, by Preoperative Albumin Status
	Odds ratio (95% CI)
	1-yr mortality	1-yr readmission
Hypoalbumin (reference=no)	1.0	1.0
Yes	10.2 (3.1–33.5)*	1.15 (0.50–2.62)
Age	1.08 (1.01–1.15)*	1.00 (0.99–1.01)
Sex (reference=male)	1.0	1.0
Female	0.54 (0.23–1.28)	0.90 (0.67–1.21)
Race (reference=White)	1.0	1.0
Black	1.0 (NA)	1.73 (0.56–5.37)
Other	0.79 (0.35–1.77)	0.78 (0.43–1.43)
Hispanic (reference=no)	1.0	1.0
Yes	3.09 (1.27–7.49)*	1.68 (0.85–3.31)
Charlson Comorbidity Index score	1.09 (0.93–1.28)*	1.07 (1.02–1.14)*
Procedure (reference=decompression)	1.0	1.0
Fusion	2.78 (0.83–9.32)	0.71 (0.47–1.07)
Osteotomy or corpectomy	1.00 (NA)	2.05 (0.85–4.92)
Surgical Invasiveness Index	0.50 (0.24–1.05)	1.03 (0.88–1.21)

Excludes patients with metastatic cancer.
NA indicates not available.
*P<0.05.

Table 4. ODI Response Rate by Months, as Percentage of Total Cohort
Months	ODI response rate (%)
2	68
4	58
6	31
8	29
10	12
12	15
Any	85

ODI indicates Oswestry Disability Index.

Table 5. Multivariable Regression Model for Oswestry Disability Index by Preoperative Albumin Status
	Oswestry disability index (score)
Preoperative hypoalbuminemia	14.0 (5.4; 22.6)*
Postoperative time
2 mo	−7.5 (−8.4; −6.6)*
4 mo	−15.6 (−16.5; −14.6)*
6 mo	−17.8 (−19.0; −16.6)*
8 mo	−18.3 (−19.5; −17.1)*
10 mo	−13.4 (−15.1; −11.6)*
12 mo	−16.9 (−18.5; −15.3)*
Interaction
Hypoalbuminemia×2 mo	−5.0 (−14.9; 4.8)
Hypoalbuminemia×4 mo	−1.2 (−13.2; 10.8)
Hypoalbuminemia×6 mo	−13.8 (−31.0; 3.4)
Hypoalbuminemia×8 mo	−15.8 (−26.8; −4.5)*
Hypoalbuminemia×10 mo	2.9 (−12.6; 18.5)
Hypoalbuminemia×12 mo	−4.6 (−20.2; 11.1)
Sex (reference=male)	1.0
Female	1.4 (0.0; 2.7)*
Race (reference=White)	1.0
Black	9.2 (2.3; 16.2)*
Other	3.8 (1.06; 6.59)*
Hispanic (reference=no)	1.0
Yes	3.6 (0.2; 7.1)*
Charlson Comorbidity Index score (reference=0)	1.0
1	2.4 (0.6; 4.2)* 2.7 (0.7; 4.6)*
2	5.5 (3.8; 7.1)*
3+
Procedure (reference=decompression)	1.0
Fusion	2.9 (1.01; 4.8)*
Osteotomy or corpectomy	9.6 (4.5; 14.6)*
Surgical Invasiveness Index	−0.13 (−0.82; 0.57)
Constant	39.4 (37.6; 41.2)*
Random effects parameters
SD (constant)	12.8 (12.3; 13.4)
SD (residual)	11.6 (11.4; 11.9)