Social Vulnerability Index is Associated With Increased Emergency Department Utilization and Persistent Opioid Use After Elective Lumbar Decompression Surgery

Abstract

Introduction

Social determinants of health have been shown to influence various outcome measures after elective orthopaedic procedures, including spine surgery. However, there has been limited research investigating the influence between social determinants of health and healthcare resource utilization after elective spine surgery. The purpose of this study was to investigate the influence of social vulnerability on opioid use, healthcare utilization, and patient-report outcome measures (PROMs) after elective lumbar decompression procedures.

Methods

Patients undergoing elective 1- or 2- level lumbar decompression procedures between 2013 and 2018 at a single tertiary medical center were retrospectively identified. Various demographic variables, including the ZIP code of residence for each patient were reported. The Centers for Disease Control and Prevention (CDC) Social Vulnerability Index (SVI) scores were calculated for each patient. Social vulnerability was defined as combined socioeconomic and housing vulnerability scores in the upper quartile. Several outcome measures were compared between patients with or without social vulnerability. The outcomes of interest included healthcare utilization and patient-reported outcome measures – pain interference (PI) and physical function (PF) within 1 year postoperatively. Healthcare utilization metrics included opioid use, emergency department visits, urgent care visits, MRI studies, pain management referrals, and revision surgeries. The minimal clinically important difference (MCID) was defined as 8 for both PI and PF scores.

Results

A total of 216 patients were included in the cohort −58 patients met criteria for social vulnerability. Patients with social vulnerability were more likely to be female and ASA class 2. Otherwise, patient demographics and medical comorbidities were similar between the 2 groups. Patients with social vulnerability were more likely to utilize the emergency department within 1-year postoperatively and to have persistent opioid use at both 6 months and 1 year postoperatively. On multivariate regression, SVI was independently associated with pre-operative opioid use, post-operative emergency department utilization, and persistent opioid use after surgery. Relative improvement in PI and PF scores and rates of MCID attainment were similar between groups.

Conclusions

Patients with social disparities were more likely to utilize the emergency department and have persistent opioid use at 6 months and 1 year postoperatively from elective lumbar decompression procedures.

Keywords

lumbar decompression orthopaedic

Introduction

Social disparities have implications for timely access to healthcare and outcomes. Various factors, including socioeconomic status, housing stability, access to transportation, racial disparities, and language barriers, can significantly affect the quality of medical care and patient outcomes.^4-7 There has been increasing interest in the relationship between these social factors and outcomes after surgery, including elective spine surgery. Identifying social barriers that may negatively influence surgical outcomes is an important first step in providing quality healthcare to this vulnerable patient population.

Prior studies have investigated the influence of various social disparities on perioperative outcomes, patient-reported outcomes, and pain management after elective spine surgery.^7,12,15,19 Socioeconomic disparities have been associated with worse patient reported outcome measures after spine surgery.¹⁵ Social disparities have been associated with hospital length-of-stay, discharge disposition, and 30-day readmission.⁷ Furthermore, lower household income has been associated with higher rates of 3-month opioid script renewal after spine surgery.² However, there is a lack of research investigating the influence of social disparities on healthcare resource utilization after elective spine surgery.

Patients experiencing particular social disparities may engage with and perceive the healthcare system in varied and unpredictable ways, especially following major surgical procedures. The purpose of this study is to investigate the relationship between social vulnerability and both healthcare resource utilization and patient-reported outcome measures (PROMs) after elective lumbar decompression procedures at an urban academic center.

Methods & Materials

Data Source

This study was approved by the Institutional Review Board (IRB) at our respective institution. Patients who underwent elective lumbar spine surgery between November 1, 2013, and September 30, 2018, at a single academic institution were retrospectively identified using a deidentified patient-reported outcome measures (PROMs) database. All patients underwent surgery by 1 of 3 fellowship-trained orthopaedic spine surgeons.

Inclusion/Exclusion Criteria

One- and two-level, posterior-based lumbar decompression procedures (eg, laminectomy, laminotomy with discectomy) were included. Only primary elective procedures were included in the analysis. Any procedures performed for tumor, infection, or traumatic indications were excluded. Patients with less than 1 year of follow-up were also excluded. Patients with incomplete follow-up data, including PROMs, were excluded from the analysis. Any revision procedures were excluded from the analysis.

Social Vulnerability Index

The Center for Disease Controls’ Social Vulnerability Index (SVI) was reported for each patient.¹ The SVI is a location-based metric that is used to quantify demographic and socioeconomic factors associated with social vulnerability.¹ The CDC reports this data, which is publicly accessible, on their website.¹ The overall SVI is subcategorized into 4 distinct scores: socioeconomic status, household characteristics, racial and ethnic minority status, and housing type and transportation.¹ Within each subcategory, specific measures are recorded, as summarized in Table 1. The overall SVI score and 4 subcategory scores are reported on a scale from 0-1, with higher numbers indicating more social vulnerability.

Table 1.

Summary of the Social Vulnerability Index Categories

SVI metric	SVI subcategories
I. Socioeconomic status	a. Below 150% poverty
	b. Unemployment
	c. Housing cost burden
	d. No high school diploma
	e. No health insurance
II. Household characteristics	a. Aged 65 & older
	b. Aged 17 & younger
	c. Civilian with a disability
	d. Single-parent households
	e. English language proficiency
III. Racial & ethnic minority status	a. Hispanic or latino (of any race)
	b. Black or African American
	c. Asian
	d. American Indian or Alaska native
	e. Native Hawaiian or pacific Islander
	f. Two or more races
	g. Other races
IV. Housing type & transportation	a. Multi-Unit structures
	b. Mobile homes
	c. Crowding
	d. No vehicle
	e. Group quarters

Groups and Variables of Interest

Social vulnerability was defined as a home address in a ZIP code with “socioeconomic status” or “household characteristics” SVI scores in the upper quartile nationally.¹ The other patients were included in the control group. Race/ethnicity, insurance status, and insurance payor type were also reported and included as variables of interest.

Several patient demographics, medical and surgical variables were reported for each patient. Sex, age, body mass index (BMI), Charlson comorbidity index (CCI), and several medical comorbidities were reported for each patient. Age was categorized into the following groups: 18-39 years, 40-49 years, 50-59 years, 60-69 years, 70-79 years, and 80+ years. BMI was categorized into the following groups: underweight (<18.5 kg/m²), normal (18.5-24.9 kg/m²), overweight (25.0-29.9 kg/m²), obese class I (30.0-34.9 kg/m²), obese class II (35.0-39.9 kg/m²), and obese class III (40.0+ kg/m²). Charlson comorbidity index (CCI) was defined as previously described.²⁴ Medical comorbidities of interest included HIV/AIDS, cerebrovascular disease, chronic pulmonary disease, congestive heart failure, diabetes with or without chronic complications, hemiplegia/paraplegia, malignancy, metastatic solid tumor, mild and moderate or severe liver disease, myocardial infarction, peptic ulcer disease, peripheral vascular disease, renal disease, and rheumatic disease.

Various healthcare utilization metrics within the first year postoperatively were reported and compared between groups. These metrics included emergency department visits, urgent care visits, persistent opioid use, number of opioid prescriptions, post-operative lumbar spine magnetic resonance imaging (MRI) studies, pain management referrals, physical therapy referrals, nerve blocking medication use (eg, gabapentin), epidural injection use, and any revision surgery. Persistent opioid use was defined as any new opioid prescriptions prescribed at the follow-up time point of interest. Pre-operative opioid use within 60 days of surgery was also recorded.

Patient-reported outcome measures (PROMs) – pain interference (PI) and physical function (PF) were also recorded at preoperative baseline and 6 weeks, 3 months, and 12 months postoperatively. The relative change in these metrics from baseline were recorded at each follow-up time point. The absolute scores and the mean change in scores were compared between the groups. The minimal clinically important difference (MCID) was defined as 8 for both PI and PF scores, as previously described in the literature.²⁹ The percentage of patients achieving MCID at each time point was also reported and compared between groups.

Statistical Analysis

Continuous variables were compared using unpaired t-tests. Categorical variables were compared using Chi-squared analysis or Fisher exact tests when appropriate. Multivariate logistic regression was used to identify any potential variables associated with various outcomes of interest, including pre-operative opioid use, non-home discharge, emergency department visits, urgent care visits, persistent opioid use, post-operative MRI lumbar spine studies, pain management referrals, physical therapy referrals, nerve blocking medication use, and epidural injection use. Additionally, multivariate logistic regression was used to identify any potential variables associated with achieving MCID in the PF or PI scores at 1 year postoperatively. Statistical significance was defined as P < 0.05. All statistical analyses were undertaken using IBM SPSS Version 29.0 (Armonk, NY: IBM Corp).

Results

A total of 216 patients were included in the final cohort, with 58 patients meeting criteria for social vulnerability. Patients in the social vulnerability group were more likely to be female (55.2% vs 39.9, P = 0.045) and ASA Class II (70.7% vs 54.5%, P = 0.031). Patients in the social vulnerability group were also more likely to have race/ethnicity reported as Black/African-American (12.1% vs 4.4%, P = 0.043) or Hispanic/Latino (13.8% vs 3.8%, P = 0.008), and less likely to have race/ethnicity reported as White/Caucasian (65.6% vs 87.3%, P < 0.001). Patients in the social vulnerability group were more likely to have Medicaid (8.6% vs 1.9%, P = 0.020) and less likely to have private insurance (63.8% vs 72.2%, P < 0.001) as the insurance payor status. Other factors, such as age, BMI, Charlson comorbidity index, and the rate of various medical comorbidities, were similar between groups (Table 2).

Table 2.

Patient Demographics and Medical Comorbidities

	Control [n = 158]	SVI vulnerable [n = 58]	p
Sex
Male	95 (60.1%)	26 (44.8%)	0.045
Female	63 (39.9%)	32 (55.2%)
Age
18-39 years	35 (22.2%)	9 (15.5%)	0.283
40-49 years	19 (12.0%)	11 (19.0%)	0.191
50-59 years	28 (17.7%)	16 (27.6%)	0.111
60-69 years	46 (29.1%)	11 (19.0%)	0.134
70-79 years	27 (17.1%)	7 (12.1%)	0.369
80+ years	3 (1.9%)	4 (6.9%)	0.085
Race/Ethnicity
Asian	4 (2.5%)	4 (6.9%)	0.132
Black/African-American	7 (4.4%)	7 (12.1%)	0.043
Caucasian/White	138 (87.3%)	38 (65.6%)	<0.001
Hispanic/Latino	6 (3.8%)	8 (13.8%)	0.008
Not reported	3 (1.9%)	1 (1.7%)	0.932
Insurance payor
Medicaid	3 (1.9%)	5 (8.6%)	0.020
Medicare	41 (25.9%)	16 (27.6%)	0.808
Private	114 (72.2%)	37 (63.8%)	<0.001
BMI
Underweight	2 (1.3%)	0 (0.0%)	>0.999
Normal	43 (27.2%)	16 (27.6%)	0.957
Overweight	58 (36.7%)	27 (46.6%)	0.189
Obese class I	31 (19.6%)	10 (17.2%)	0.693
Obese class II	17 (10.8%)	3 (5.2%)	0.292
Obese class III	7 (4.4%)	2 (3.4%)	>0.999
ASA class
I	35 (22.2%)	9 (15.5%)	0.283
II	86 (54.4%)	41 (70.7%)	0.031
III	37 (23.4%)	8 (13.8%)	0.123
IV	0 (0.0%)	0 (0.0%)	-
Charlson comorbidity index (CCI)
0	35 (22.2%)	12 (20.7%)	0.817
1	26 (16.5%)	9 (15.5%)	0.868
2	37 (23.4%)	11 (19.0%)	0.485
3+	60 (38.0%)	26 (44.8%)	0.361
Medical comorbidities
HIV/AIDS	6 (3.8%)	1 (1.7%)	0.677
Cerebrovascular disease	9 (5.7%)	3 (5.2%)	>0.999
Chronic pulmonary disease	23 (14.6%)	10 (17.2%)	0.592
Congestive heart failure	6 (3.8%)	2 (3.4%)	>0.999
Diabetes with chronic complications	4 (2.5%)	3 (5.2%)	0.390
Diabetes without chronic complications	14 (8.9%)	8 (13.8%)	0.288
Hemiplegia/paraplegia	5 (3.2%)	2 (3.4%)	>0.999
Malignancy	14 (8.9%)	6 (10.3%)	0.749
Metastatic solid tumor	2 (1.3%)	1 (1.7%)	>0.999
Mild liver disease	13 (8.2%)	7 (12.1%)	0.429
Moderate or severe liver disease	0 (0.0%)	1 (1.7%)	0.269
Myocardial infarction	1 (0.6%)	1 (1.7%)	0.468
Peptic ulcer disease	2 (1.3%)	1 (1.7%)	>0.999
Peripheral vascular disease	7 (4.4%)	2 (3.4%)	>0.999
Renal disease	5 (3.2%)	1 (1.7%)	>0.999
Rheumatic disease	9 (5.7%)	4 (6.9%)	0.751

bold values indiciate statistical significance (p < 0.05).

Across the entire cohort, the mean socioeconomic status SVI score was 0.299 ± 0.257, which is better (ie, less social vulnerability) than the national average. However, the range of the socioeconomic SVI score was wide (0.002 to 0.9984). The mean household characteristics SVI score was 0.274 ± 0.248 (range: 0.003 – 0.998, better than (ie, less social vulnerability) than the national average. The mean racial & ethnic minority status SVI score was 0.520 ± 0.238 (range: 0.020 – 0.098), near the national average. The mean housing type & transportation SVI score was 0.535 ± 0.274 (range: 0.001 – 09.999), near the national average (Table 3).

Table 3.

Summary of SVI Scores for the Entire Cohort

	Mean +/− STD^a	Range
Socioeconomic status SVI score	0.299 ± 0.257	[0.002 – 0.9984]
Household characteristics SVI score	0.274 ± 0.248	[0.003 – 0.9984]
Racial & ethnic minority status SVI score	0.520 ± 0.238	[0.020 – 0.9984]
Housing type & transportation SVI score	0.535 ± 0.274	[0.001 – 0.9994]

^aSTD – standard deviation.

Socially vulnerable patients were more likely to utilize the emergency department in the first year postoperatively (25.9% vs 13.9%, P = 0.039). They were also more likely to have persistent opioid use at 6 months (17.2% vs 7.0%, P = 0.024) and 1-year postoperatively (25.9% vs 10.1%, P = 0.003). There were no significant difference in non-home discharges, urgent care visits, post-operative MRI studies, pain management referrals, physical therapy referrals, nerve blocking medication use, epidural injection use, or revision surgery within the first year post-operatively between the 2 groups (Table 4).

Table 4.

Healthcare Utilization Metrics Within 1-Year Post-operatively

	Control [n = 158]	SVI vulnerable [n = 58]	P
Pre-operative opioid use
Within 60 days	26 (16.5%)	15 (25.9%)	0.118
Non-home discharge	3 (1.9%)	2 (3.4%)	0.502
Emergency department visit
At 90 days	10 (6.3%)	3 (5.2%)	>0.999
At 180 days	16 (10.1%)	9 (15.5%)	0.272
At 365 days	22 (13.9%)	15 (25.9%)	0.039
Urgent care visits
At 90 days	9 (5.7%)	0 (0.0%)	0.117
At 180 days	14 (8.9%)	2 (3.4%)	0.246
At 365 days	20 (12.7%)	4 (6.9%)	0.329
Persistent opioid use
At 180 days	11 (7.0%)	10 (17.2%)	0.024
At 365 days	16 (10.1%)	15 (25.9%)	0.003
Number of opioid Rx
At 90 days	1.2 ± 1.3	1.4 ± 2.3	0.463
At 180 days	1.3 ± 1.5	2.1 ± 5.6	0.265
At 365 days	1.6 ± 2.3	3.3 ± 11.2	0.272
Post-op MRI lumbar spine
At 90 days	12 (7.6%)	6 (10.3%)	0.526
At 180 days	21 (13.3%)	7 (12.1%)	0.813
At 365 days	38 (24.1%)	11 (19.0%)	0.429
Pain management referral
At 90 days	3 (1.9%)	0 (0.0%)	0.566
At 180 days	10 (6.3%)	2 (3.4%)	0.521
At 365 days	17 (10.8%)	6 (10.3%)	0.930
Physical therapy referral
At 90 days	119 (75.3%)	39 (67.2%)	0.235
At 180 days	122 (77.2%)	41 (70.7%)	0.323
At 365 days	127 (80.4%)	43 (74.1%)	0.321
Nerve blocking medication use
At 90 days	75 (47.5%)	30 (51.7%)	0.579
At 180 days	79 (50.0%)	31 (53.4%)	0.653
At 365 days	82 (51.9%)	33 (56.9%)	0.514
Epidural injection use
At 90 days	3 (1.9%)	0 (0.0%)	0.566
At 180 days	4 (2.5%)	4 (6.9%)	0.215
At 365 days	13 (8.2%)	9 (15.5%)	0.116
Revision surgery
At 365 days	8 (5.1%)	0 (0.0%)	0.112

bold values indiciate statistical significance (p < 0.05).

On multivariate logistic regression, SVI Overall Vulnerability and ASA class were independently associated with pre-operative opioid use and emergency department visits at 1 year postoperatively. SVI Overall Vulnerability, SVI Housing & Transportation Vulnerability, preoperative opioid use, ASA class, and CCI were all independently associated with persistent opioid use at 1 year-postoperatively. (Table 5).

Table 5.

Multivariate Logistic Regression Identifying Independent Risk Factors for Various 1-Year Post-operative Healthcare Utilization Metrics

	Risk ratio [95% C.I.]	P
Pre-operative opioid use
SVI vulnerability	2.509 [1.093 – 5.759]	0.030
ASA class	2.868 [1.459 – 5.640]	0.002
Non-home discharge	-	-
Emergency department visits
SVI overall vulnerability	2.394 [1.114 – 5.148]	0.025
ASA class	2.419 [1.314 – 4.453]	0.005
Urgent care visits	-	-
Persistent opioid use
SVI overall vulnerability	3.315 [1.283 – 8.562]	0.013
SVI housing & transportation vulnerability	2.671 [1.104 – 6.462]	0.029
Preoperative opioid use	4.469 [1.689 – 11.825]	0.003
ASA class	2.719 [1.259 – 5.869]	0.011
Charlson comorbidity index	1.286 [1.069 – 1.547]	0.008
Post-op MRI lumbar spine	-	-
Pain management referral	-	-
Nerve blocking medication use (e.g. Gabapentin)
Preoperative opioid use	3.941 [1.777 – 8.743]	<0.001
Epidural or other spinal injection	-	-

bold values indiciate statistical significance (p < 0.05).

Patients with preoperative opioid use (within 60 days of surgery) were more likely to have persistent opioid use at 6 months (24.4% vs 6.3%, P < 0.001) and 365 days postoperatively (31.7% vs 10.3%, P < 0.001). Patients with preoperative opioid use also had higher numbers of opioid prescriptions at all time points (Table 6). These patients were also more likely to have nerve blocking medication use at all time points (Table 6).

Table 6.

Healthcare Utilization Metrics Within 1-Year Post-operatively for Patients With and Without Preoperative Opioid Use

	Control [n = 175]	Pre-operative opioid use [n = 41]	P
Non-home discharge	3 (1.7%)	2 (4.9%)	>0.999
Emergency department visit
At 90 days	9 (5.1%)	4 (9.8%)	0.276
At 180 days	19 (10.9%)	6 (14.6%)	0.496
At 365 days	27 (15.4%)	10 (24.4%)	0.170
Urgent care visits
At 90 days	9 (5.1%)	0 (0.0%)	0.213
At 180 days	14 (8.0%)	2 (4.9%)	0.742
At 365 days	20 (11.4%)	4 (9.8%)	>0.999
Persistent opioid use
At 180 days	11 (6.3%)	10 (24.4%)	<0.001
At 365 days	18 (10.3%)	13 (31.7%)	<0.001
Number of opioid Rx
At 90 days	0.9 ± 0.9	3.0 ± 2.9	<0.001
At 180 days	0.9 ± 1.0	4.0 ± 6.5	0.002
At 365 days	1.1 ± 1.5	6.0 ± 13.2	0.012
Post-op MRI lumbar spine
At 90 days	14 (8.0%)	4 (9.8%)	0.751
At 180 days	21 (12.0%)	7 (17.1%)	0.384
At 365 days	35 (20.0%)	14 (34.1%)	0.052
Pain management referral
At 90 days	3 (1.7%)	0 (0.0%)	>0.999
At 180 days	10 (5.7%)	2 (4.9%)	>0.999
At 365 days	18 (10.3%)	5 (12.2%)	0.779
Physical therapy referral
At 90 days	126 (72.0%)	32 (78.0%)	0.432
At 180 days	129 (73.7%)	34 (82.9%)	0.217
At 365 days	135 (77.1%)	35 (85.4%)	0.247
Nerve blocking medication use
At 90 days	77 (44.0%)	28 (68.3%)	0.005
At 180 days	80 (45.7%)	30 (73.2%)	0.002
At 365 days	83 (47.4%)	32 (78.0%)	<0.001
Epidural injection use
At 90 days	2 (1.1%)	1 (2.4%)	0.470
At 180 days	6 (3.4%)	2 (4.9%)	0.649
At 365 days	17 (9.7%)	5 (12.2%)	0.636
Revision surgery
At 365 days	6 (3.4%)	2 (4.9%)	0.649

bold values indiciate statistical significance (p < 0.05).

The baseline PI and PF PROMs scores were similar between the 2 groups. The absolute PI and PF scores at 6 weeks, 3 months, and 1 year post-operatively were similar between the 2 groups. The mean change and 6 weeks, 3 months, and 1 year post-operatively were similar between the 2 groups. The mean change and the percentage of patients achieving MCID in PI and PF scores were similar between the 2 groups at all time points (Table 7). Of note, patients with persistent opioid use at 1 year postoperatively were less likely to achieve MCID at 1 year postoperatively for PI (19.2% vs 27.7%, P = 0.371) and PF (15.4% vs 31.5%, P = 0.097) scores, although this did not achieve statistical significance.

Table 7.

Patient Reported Outcome Measures (PROMs) in SVI Vulnerable Patients

	Control [n = 158]	SVI vulnerable [n = 58]	P
Pain interference (PI)
Baseline	14.6 ± 6.1	14.9 ± 4.1	0.733
6 weeks	8.6 ± 4.5	10.0 ± 4.6	0.082
3 months	9.3 ± 3.9	10.7 ± 8.5	0.177
12 months	10.3 ± 3.3	10.9 ± 3.9	0.343
Physical function (PF)
Baseline	10.3 ± 5.8	9.8 ± 2.8	0.582
6 weeks	11.9 ± 7.0	11.9 ± 6.7	0.995
3 months	14.5 ± 7.3	14.1 ± 5.0	0.741
12 months	18.6 ± 13.2	17.5 ± 10.6	0.596
ΔPI
6 weeks	−6.0 ± 6.9	−5.0 ± 6.1	0.371
3 months	−5.3 ± 7.2	−4.3 ± 9.5	0.476
12 months	−4.3 ± 9.5	−4.0 ± 5.6	0.834
ΔPF
6 weeks	1.6 ± 5.9	2.1 ± 7.1	0.656
3 months	4.2 ± 6.0	4.3 ± 5.0	0.915
12 months	8.3 ± 12.9	7.6 ± 10.8	0.758
Achieving MCID
Pain interference (PI)
6 weeks	42 (37.8%)	16 (35.6%)	0.789
3 months	38 (34.2%)	17 (37.8%)	0.675
12 months	26 (23.4%)	15 (33.3%)	0.203
Physical function (PF)
6 weeks	9 (8.1%)	6 (13.3%)	0.316
3 months	21 (18.9%)	12 (26.7%)	0.283
12 months	31 (18.9%)	14 (31.1%)	0.691

On multivariate regression, baseline PF and PI scores were independently associated with MCID achievement in PF and PI scores, respectively. Higher baseline pain interference scores – signifying more significant pain interference – were associated with PI MCID achievement postoperatively. Lower baseline physical function scores – signifying reduced physical function – were postoperatively associated with PF MCID achievement. No other patient or surgical variables were associated with MCID achievement (Table 8).

Table 8.

Logistic Regression Identifying Factors Associated With Achieving MCID at 1 year

	Risk ratio [95% C.I.]	P
PF MCID
Baseline PF score	0.883 [0.781 – 0.997]	0.044
PI MCID
Baseline PI score	1.712 [1.393 – 2.104]	<0.001

bold values indiciate statistical significance (p < 0.05).

Discussion

In this cohort, patients with social vulnerability were more likely to utilize the emergency department within 1-year postoperatively. Additionally, patients with social vulnerability were nearly 3 times more likely to have persistent opioid use at 1-year postoperatively compared to the control group. Although preoperative opioid use was associated with persistent postoperative opioid use, social vulnerability was independently associated with persistent opioid use on multivariate regression. Additionally, on multivariate logistic regression analysis, SVI was independently associated with preoperative opioid use and postoperative emergency department utilization. The relative improvement in PROMs (ie, PI and PF scores) and rates of MCID achievement were similar between groups. On multivariate analysis, only baseline PROMs were associated with MCID achievement in this cohort – patients with better baseline PROMs were more likely to achieve MCID.

The association between social determinants of health and different outcome measures after spine surgery have been previously described.^{7,11,15,16,23,25,28} Table 9 provides a summary of recent literature that have investigated the association between different social variables and outcomes after various spine surgery procedures. Social disadvantage has been associated with perioperative outcomes,^11,15,25 perioperative opioid use,²⁸ hospital readmission,^7,23 and clinical outcomes^3,11 after spine surgery. However, there is a lack of research investigating the relationship between social determinants of health and how patients utilize the health system after spine surgery. Additionally, few studies have investigated the CDC SVI metric as a specific predictor of outcome measures and healthcare utilization after spine surgery.

Table 9.

Summary of Existing Literature on Social Determinants of Health and Outcomes After Spine Surgery

Study	Level of evidence	Social Determinant(s) and geographic metrics evaluated	Summary of findings
Shin et al. (2024)²³	III	Reported SDH: Race/ethnicity, language, religion, marital status, insurance coverage plan	Black patients demonstrated a greater LOS. Insured patients had a shorter LOS and lower readmission rates
Shin et al. (2024)²³	III		The most important social variables for predicting LOS and 90-day hospital readmission were insurance status and religion
Tang et al. (2024)²⁵	III	Reported race and ethnicity, insurance, household income, and safety-net hospital (SNH) treatment	For spine tumor admissions, SDH predicted surgical intervention, presenting severity, and perioperative outcomes
Wague et al. (2023)²⁸	III	American community survey (ACS) data	Unemployment, low physical activity level, and lower community median income were associated with preoperative opioid use and longer-term opioid use postoperatively
		Community median income
		Reported age, gender, race, preferred language, ethnicity, work status, religious belief, insurance status, marital status
Holbert et al. (2022)⁷	III	Self-reported social determinant of health (SDH) survey	Self-reported SDoH were associated with length-of-stay, discharge disposition, 30-day ED visits, and 30-day readmission after various decompression and/or fusion procedures
Buck et al. (2022)³	III	Reported race/ethnicity, educational attainment, employment status, insurance payer	Individuals with ≥3 SDH had a 67%, ≥3 SDH65%, 71%, 65%, and 46% decrease in the odds of a clinically meaningful outcome in back and leg pain, disability, quality of life, and patient satisfaction, respectively
Mohanty et al (2022, 1)¹⁵	III	Geographic Information systems (ArcGIS)	Racial and socioeconomic disparities were associated with adverse postoperative outcomes after various spine surgery procedures
Mohanty et al (2022, 1)¹⁵	III	Area deprivation index (ADI)
Mohanty et al (2022, 2)¹⁶	III	Geographic Information systems (ArcGIS)	Socioeconomically disadvantaged patients and patients from areas of high social deprivation have a higher risk of readmission following various spine surgery procedures
		Area deprivation index (ADI)
		U.S. Census Bureau American community survey data
Khalid et al. (2021)¹¹	III	ICD-9/10 codes to identify SDH disparities: Economic, education, social, environmental	SDH affect outcomes and are associated with an increased risk of symptomatic pseudarthrosis and postoperative complications after single-level lumbar fusion

The Area Deprivation Index (ADI) and SVI are 2 frequently referenced socioeconomic deprivation indices.²¹ The SVI was originally developed by the CDC to identify vulnerable communities that may need additional support during natural disasters, hazardous events, and disease outbreaks.²¹ However, its use has been extended more broadly in health policy and healthcare outcome studies.²⁶ The SVI is based on 15 indicators within 4 categories: socioeconomic status, house composition and disability, minority status and language, and housing type and transportation. The ADI was developed to quantify social deprivation in different neighborhoods using 17 U.S. census indicators of income, education, employment, and housing quality.²¹ In contrast to the SVI, the ADI does not take race/ethnicity into consideration.²⁶ Both the SVI and ADI have been correlated with various healthcare outcome measures,^6,17 although research has suggested that the 2 indices do not always correlate with each other and that they can not necessarily be used interchangeably. For instance, 1 study found that the ADI’s inclusion of unstandardized housing price items can lead to misclassification of certain urban environments as “not deprived” in comparison to SVI, which excluded these items.²¹ Additionally, 2 recent studies have suggested that SVI may provide better discriminative performance and granularity relative to ADI in identifying disadvantaged patients at risk for poor outcomes.^20,31

The current study demonstrated direct association of SVI and postoperative healthcare utilization. In our cohort, patients with social vulnerability were nearly twice as likely to utilize the emergency department within 1 year postoperatively. Prior research has suggested that patients in the immediate postoperative period often visit the emergency department, but there is significant variability in these utilization patterns.¹³ Durand et al. reported a 9% rate of 90-day emergency department visits after outpatient spine surgery, with roughly 10% occurring on the same day of surgery.⁵ The most common reasons for ED presentation reported were pain, chest pain, headache, abdominal pain, and limb pain.⁵ Jain et al. previously reported that risk factors for ED utilization after primary lumbar spine fusion included having Medicare insurance, anterior lumbar fusion, drug abuse/dependence, age less than 50 years old, African American ethnicity, and various medical comorbidities.⁹ Additionally, Holbert et al. found that certain social factors were associated with decreased risk of 30-day emergency department visits and hospital readmission.⁷ Emergency department utilization is associated with a significant cost burden. Wiley et al. reported that the mean direct hospital cost of an ED visit after elective lumbar spine surgery was $1,971, and the total cost to the institution over 5 years was $5.1 million.³⁰ Furthermore, emergency department utilization after lumbar spine surgery has been associated with lower patient Satisfaction and Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) scores.²⁰ However, the study did not investigate the influence of socioeconomic factors on emergency department utilization. Our findings demonstrated that social disadvantage, as defined by SVI, is associated with increased emergency department utilization after elective lumbar spine surgery. Developing a specific plan to address postoperative concerns, particularly for patients with social barriers, may be an opportunity to improve the quality of care and reduce costs for patients.

Persistent opioid use before and after spine surgery presents a significant challenge for patients, providers, and the healthcare system.³³ Persistent opioid use after spine surgery has been associated with poor outcomes, including hospital readmission, emergency department utilization, and wound complications.^10,14 The influence of persistent opioid use on PROMs after spine surgery remains unclear, with some research suggesting worse outcomes and some suggesting similar outcomes in certain metrics.^3,11,27 In our cohort, patients with social vulnerability were nearly 3 times more likely to have persistent opioid use at 1 year postoperatively. Additionally, social vulnerability was associated with preoperative opioid use within 60 days of surgery, although this was not statistically significant. Although preoperative opioid use was associated with persistent postoperative opioid use, social vulnerability was independently associated with persistent opioid use on multivariate regression at 1 year postoperatively. This suggests that social determinants of health may be independent risk factors for postoperative opioid use. In a previous study, Wague et al. previously demonstrated that unemployment and lower community median income were strongly associated with opioid use after lumbar spine surgery.²⁸ Futhermore, Wague et al. found that Black patients more likely to use opioids before surgery, and that preoperative opioid use was associated with persistent opioid use postoperatively.²⁸ Additionally, Schoenfeld et al. reported that lower socioeconomic status was significantly associated with a decreased likelihood of discontinuing opioid use after spine surgery in a military health system study.²² Our study found that the SVI metric can be independently predictive of persistent opioid use postoperatively. Identifying this risk preoperatively may help guide preoperative discussion around patient expectations and goals for opioid use in the postoperative setting. The authors view SVI and preoperative opioid use as reasonable screening tools to identify patients that may be at risk for opioid dependence postoperatively. Furthermore, patients with both social vulnerability and preoperative opioid use may be at even higher risk for persistent opioid use after surgery. It may be pertinent to involve a specific pain management team to help guide pre- and post-operative pain control, minimize the risk of opioid dependence, and set expectations prior to surgery.

Other social determinants of health, such as race/ethnicity, religious status, and insurance payor status have also been investigated as potential predictors of outcomes after spine surgery.^3,23,25,28 In our study, patients in the social vulnerability group were more likely to have race/ethnicity reported as Black/African-American or Hispanic/Latino, and less likely to have race/ethnicity reported as White/Caucasian. Patients in the social vulnerability group were also more likely to have Medicaid and less likely to have private insurance as the insurance payor status. However, on multivariate analysis, neither race/ethnicity nor insurance payor status were associated with perioperative outcomes, healthcare utilization metrics, or PROMs in this cohort. Other studies have found association between race/ethnicity and insurance payor status and outcomes after spine surgery, however. For example, Shin et al. reported that Black patients experienced a greater LOS after surgery and that insured patients had shorter LOS and lower 90-day readmission rates after surgery.²³ Tang et al. reported that the presenting severity of symptoms of patients with spine tumors was elevated among non-White and non-private insurance patients.²⁵ The authors also found that Black race, Medicaid, and lower income were all associated with decreased odds of favorable discharge disposition.²⁵ Others have found that opioid prescribing patterns have been different among patients with different race/ethnicity. For example, Wague et al. found that Black patients were 6.3 times more likely to use opioids before surgery, with preoperative opioid use being associated with persistent opioid use after surgery.²⁸ Although our study did not find any association between race/ethnicity and insurance payor status and outcome measures, it should be noted that the majority of patients in our cohort were Caucasian/White (69.9%) and with private insurance (81.5%). Therefore, the study could certainly be underpowered to investigate the influence of race/ethnicity and insurance status on outcomes and healthcare utilization metrics after spine surgery.

This study also compared PROMs and MCID achievement in patients with social vulnerability. In our cohort, SVI was not associated with baseline PF and PI PROMs, overall improvement in PF and PI PROMs, nor the rate of MCID achievement. On subgroup analysis, all patients with persistent opioid use trended towards lower PROMs, but this did not achieve statistical significance. This may be due to the fact that there was an overall wide range of PROM scores and a relatively small number of patients with persistent opioid use, decreasing statistical power. Prior studies have had some conflicting results regarding the association between social variables and PROMs after spine surgery. Of note, prior literature has suggested that socioeconomic disadvantage may be a barrier itself to completion of PROM surveys after lumbar spine fusion.⁸ Additionally, others have found that racial and socioeconomic disparities were associated with worse global physical health and VAS pain scores both pre- and post-operatively after spine surgery.¹⁵ There is a lack of research investigating social determinants of health and patient-reported outcomes in spine surgery. Given the multifactorial nature of social barriers, further research investigating the relationship between specific social metrics and PROMs would be useful.

Although not specifically investigated in this study, we believe that there are multiple target areas for improvement in patient care. The authors view that the SVI metric could be used as a screening tool to identify patients in clinic that may be at higher risk for certain poor outcome measures after surgery. Once identified, these patients could be placed on a high-risk protocol that would involve multidisciplinary discussions with other team members to ensure efficient and effective care in the perioperative period. Involving social work providers and home health agencies prior to surgery would likely be beneficial for these patients. Specifically, it would be useful for home health nursing and/or home physical therapists to be setup prior to surgery, or at least prior to discharge from the hospital. Home health care can be a tool to provide consistent care and also detect issues or challenges as they arise in a more efficient and timely manner. For patients with pre-operative opioid use and/or social vulnerability, it would be useful to involve either the pain management or physiatry services to help guide a post-operative pain control regimen, with the goal of limiting long-term opioid use after surgery. Having a non-surgical provider contact for pain control management may be useful for these patients, as follow-up and continued contact with these providers may be easier. For patients who are traveling long distances for surgery, establishing care with 1 of these providers closer to their residency may be beneficial.

Patient experience is a challenging metric to measure, but it remains a crucial component of healthcare, particularly after a major surgery.²⁷ For example, understanding and meeting patient expectations is an important goal for successful outcomes in spine surgery.^18,32 Although it was not specifically investigated in this study, the authors believe that social determinants of health likely play a significant role in patient experience in the perioperative period. For example, cultural and language barriers may play a significant role in understanding discharge instructions, expectations, and even PROM surveys. Although discharge instructions are translated to native language, there are often not in-person interpreters available to answer specific questions and communicate more effectively. As another example, transportation and housing barriers likely play a significant role in postoperative care – particularly at tertiary care centers, where patients may travel several miles for a complex spine surgery. These patients could certainly have challenges with post-operative follow-up visits and receiving more urgent perioperative care when issues arise. It is possible that these patients may ultimately turn to other healthcare providers in their area, whom may not have access to the same electronic health record and may not understand appropriate subspecialty-specific postoperative care. This could potentially lead to over-utilization of healthcare resources, unnecessary tests, and substandard care for patients. The exact influence and nuances of these barriers to care would be an important area of investigation for future studies on this topic.

This study does have limitations that should be considered in the context of its findings. Firstly, the study is a retrospective review of prospectively collected data, which is inherently subject to selection bias. All patients were treated at a tertiary care center, which may select for a certain patient population. Additionally, we did not collect data on how long patients had symptoms prior to surgery, which could certainly influence perioperative opioid use and potentially long-term outcomes. This would be an important variable to address in future studies. Patients without complete PROMs data at 1 year postoperatively were not included in the analysis, which can certainly introduce selection bias. Other authors have indeed reported that social determinants of health impact completion of PROMs – with the majority of patients completing PROMs residing in wealthier communities.⁸ Further studies to identify specific barriers to PROMs and rectify this challenge within this patient population would be of value. Additionally, patients in the lower SVI cohort had some minor differences in other demographic variables – these patients were more likely to be female and to have ASA class 2. Efforts were made to control for these potential confounding variables using multivariate regression. Nonetheless, the results should be considered in the context that some baseline demographic variables were different between groups. Lastly, efforts were made to search for utilization of healthcare resources at outside institutions using the “Care Everywhere” tool in the electronic medical record. However, it is possible that episodes of postoperative care at non-participating institutions would not be detected. Nonetheless, the data discussed herein provide insight into healthcare utilization trends and highlight potential targets to optimize healthcare equity in socially vulnerable patients undergoing elective spine surgery.

Conclusions

Patients with social vulnerability – as defined by an upper quartile score on the SVI socioeconomic and housing subcategories – were more likely to utilize the emergency department and have persistent opioid use at 1 year postoperatively from elective lumbar decompression procedures. Social vulnerability was independently associated with postoperative emergency department utilization and persistent opioid use on multivariate analysis. Further prospective studies are warranted to investigate the relationship between specific social disparities and healthcare utilization after elective spine surgery.

Footnotes

ORCID iD

Daniel Herrera

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

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

References

CDC/ATSDR SVI overview.

Barrie

Montgomery

Ogwumike

, et al. Household income as a predictor for surgical outcomes and opioid use after spine surgery in the United States. Glob Spine J. 2023;13(8):2124-2134.

Buck

Rethorn

Garcia

Cook

Gottfried

. The collective influence of social determinants of health on individuals who underwent lumbar spine revision surgeries: a retrospective cohort study. World Neurosurg. 2022;165:e619-e627.

Divi

Koss

Schmaltz

Loeb

. Language proficiency and adverse events in US hospitals: a pilot study. Int J Qual Health Care. 2007;19(2):60-67.

Durand

Badin

Ortiz-Babilonia

Musharbash

Raad

Jain

. Emergency department visits after outpatient spine surgery. Spine. 2022;47(14):1011-1017.

Higginbotham

Segovia

Rohm

Anderson

Breitenstein

. Social vulnerability index and health outcomes in the United States: a systematic review. Fam Community Health. 2025;48(2):81-96.

Holbert

Andersen

Stone

Pipkin

Turcotte

Patton

. Social determinants of health influence early outcomes following lumbar spine surgery. Ochsner J. 2022;22(4):299-306.

Issa

Lee

Toci

, et al. The role of socioeconomic factors as barriers to patient reported outcome measure completion following lumbar spine fusion. Spine J. 2023;23(10):1531-1539.

Jain

Brock

Phillips

Weaver

Khan

. 30-day emergency department visits after primary lumbar fusion: incidence, causes, risk factors, and costs. Clin Spine Surg. 2019;32(3):113-119.

10.

Jain

Phillips

Weaver

Khan

. Preoperative chronic opioid therapy: a risk factor for complications, readmission, continued opioid use and increased costs after one-and two-level posterior lumbar fusion. Spine. 2018;43(19):1331-1338.

11.

Khalid

Maasarani

Nunna

, et al. Association between social determinants of health and postoperative outcomes in patients undergoing single-level lumbar fusions: a matched analysis. Spine. 2021;46(9):E559-E565.

12.

Khan

Huang

Maeder-York

, et al. Racial disparities in outcomes after spine surgery: a systematic review and meta-analysis. World Neurosurg. 2022;157:e232-e244.

13.

Kocher

Nallamothu

Birkmeyer

Dimick

. Emergency department visits after surgery are common for medicare patients, suggesting opportunities to improve care. Health Aff. 2013;32(9):1600-1607.

14.

Lee

Armaghani

Archer

, et al. Preoperative opioid use as a predictor of adverse postoperative self-reported outcomes in patients undergoing spine surgery. JBJS. 2014;96(11):e89.

15.

Mohanty

Harowitz

Lad

Rouhi

Casper

Saifi

. Racial and social determinants of health disparities in spine surgery affect preoperative morbidity and postoperative patient reported outcomes: retrospective observational study. Spine. 2022;47(11):781-791.

16.

Mohanty

Lad

Casper

Sheth

Saifi

. The impact of social determinants of health on 30 and 90-day readmission rates after spine surgery. JBJS. 2022;104(5):412-420.

17.

Morenz

Liao

Hayes

. Area-level socioeconomic disadvantage and health care spending: a systematic review. JAMA Netw Open. 2024;7(2):e2356121.

18.

Rampersaud

Canizares

Perruccio

, et al. Fulfillment of patient expectations after spine surgery is critical to patient satisfaction: a cohort study of spine surgery patients. Neurosurgery. 2022;91(1):173-181.

19.

Reyes

Bajaj

Alvandi

Kurapaty

Patel

Divi

. Impact of social determinants of health in spine surgery. Curr Rev Musculoskelet Med. 2023;16(1):24-32.

20.

Rollings

Noppert

Griggs

Ibrahim

Clarke

. Comparing deprivation vs vulnerability index performance using medicare beneficiary surgical outcomes. JAMA Surg. 2024;159(12):1404-1413.

21.

Rollings

Noppert

Griggs

Melendez

Clarke

. Comparison of two area-level socioeconomic deprivation indices: implications for public health research, practice, and policy. PLoS One. 2023;18(10):e0292281.

22.

Schoenfeld

Nwosu

Jiang

, et al. Risk factors for prolonged opioid use following spine surgery, and the association with surgical intensity, among opioid-naive patients. JBJS. 2017;99(15):1247-1252.

23.

Shin

Razzouk

Thomas

, et al. Social determinants of health and disparities in spine surgery: a 10-year analysis of 8,565 cases using ensemble machine learning and multilayer perceptron. Spine J. 2024;25(10):2236-2252.

24.

Sundararajan

Henderson

Perry

Muggivan

Quan

Ghali

. New ICD-10 version of the Charlson comorbidity index predicted in-hospital mortality. J Clin Epidemiol. 2004;57(12):1288-1294.

25.

Tang

Ayala

Feler

, et al. Social determinants of health and outcome disparities in spine tumor surgery. Part 1: an analysis of 6.6 million nationwide admissions. J Neurosurg Spine. 2024;41(6):677-688.

26.

Tipirneni

Lantz

Karmakar

. Association of Area Deprivation Index Vs. Social Vulnerability Index and COVID-19 Outcomes in Michigan. University of Michigan, Institute for Healthcare Policy & Innovation; 2021.

27.

Tsai

Orav

Jha

. Patient satisfaction and quality of surgical care in US hospitals. Ann Surg. 2015;261(1):2-8.

28.

Wague

O'Donnell

Rangwalla

El Naga

Gendelberg

Berven

. Impact of social determinants of health on perioperative opioid utilization in patients with lumbar degeneration. N Am Spine Soc J. 2023;14:100221.

29.

Weiner

Snavely

Johnson

Hsu

Patel

. Impact of preoperative opioid use on postoperative patient-reported outcomes in lumbar spine surgery patients. Clin Spine Surg. 2021;34(3):E154-E159.

30.

Wiley

Carreon

Djurasovic

, et al. Economic analysis of 90-day return to the emergency room and readmission after elective lumbar spine surgery: a single-center analysis of 5444 patients. J Neurosurg Spine. 2020;34(1):89-95.

31.

Yang

Shaikh

Akoto

, et al. Social vulnerability index provides greater granularity compared with the area deprivation index when assessing outcomes following elective lumbar fusion. Spine. 2024;49(23):1676-1684.

32.

Yee

Adjei

Ford

Finkelstein

. Do patient expectations of spinal surgery relate to functional outcome? Clin Orthop Relat Res. 2008;466:1154-1161.

33.

Yerneni

Nichols

Abecassis

Karras

Tan

. Preoperative opioid use and clinical outcomes in spine surgery: a systematic review. Neurosurgery. 2020;86(6):E490-E507.