Sage Journals: Discover world-class research

Abstract

Background:

Chronic opioid use has been linked to adverse surgical outcomes, yet its impact on ankle fracture surgery remains underexplored. This study evaluates the association between preoperative opioid use and postoperative health care utilization in patients undergoing surgical fixation for an isolated ankle fracture.

Methods:

A retrospective cohort study was conducted at a level 1 trauma center over a 2-year period. Patients who underwent surgical fixation for an isolated ankle fracture were identified using Current Procedural Terminology (CPT) codes. Those with concurrent orthopaedic injuries or insufficient follow-up data were excluded. Patients were categorized as long-term opioid users (>3 months of opioid use) or opioid-naïve. Electronic medical records were reviewed to obtain data on patient demographics, comorbidities, injury characteristics, length of stay, opioid prescriptions, and 90-day postoperative outcomes, including emergency department (ED) visits, readmissions, and reoperations. Bivariate and multivariate analyses were used to assess the impact of opioid use on postoperative complications and health care utilization.

Results:

Of the 303 patients analyzed, 40 (13.2%) were long-term opioid users. Compared with opioid-naïve patients, opioid users had significantly longer hospital stays (5.3 ± 6.1 vs 2.6 ± 2.9 days, P < .001) and higher rates of 90-day ED visits (0.55 ± 1.11 vs 0.24 ± 0.61, P = .010), unplanned readmissions (0.33 ± 0.69 vs 0.076 ± 0.33, P < .001), and unplanned reoperations (0.18 ± 0.59 vs 0.05 ± 0.26, P = .026). Pain-related complaints were the most common reason for postoperative ED visits among opioid users. Logistic regression identified psychiatric history and hospital length of stay as independent predictors of increased readmissions, and hospital length of stay an independent predictor of postoperative ED visits. Opioid use alone was not predictive of readmissions or ED visits.

Conclusion:

Preoperative opioid use is linked to longer hospital stays and increased postoperative health care utilization on unadjusted analyses. In multivariable models, psychiatric history and hospital length of stay, rather than opioid use alone, were independent predictors of readmissions and ED visits, suggesting that addressing psychiatric comorbidities and optimizing pain management strategies may improve outcomes in this patient population.

Level of Evidence:

Level III, retrospective cohort study.

Keywords

opioids ankle fracture health care utilization

Introduction

Since the onset of the opioid epidemic in the United States, more than 645 000 deaths have been attributed to opioid overdose. In 2022, synthetic opioids were implicated in 90% of opioid-related overdose deaths.⁴ Despite ongoing efforts, opioid use remains prevalent, with approximately one-third of Americans having received an opioid prescription within the previous 2 years, and 4.1% of US adults reporting nonmedical opioid use.^16,17 Annual overdose deaths have risen from 12 940 in 2003 to 79 355 in 2023.⁹ Beyond overdose mortality, long-term opioid use negatively affects multiple physiological systems, including immune function, cardiovascular health, sleep, and gastrointestinal function. In surgical populations, opioid use has been linked to higher postoperative complication rates, particularly in abdominal surgery.^11,24

In orthopaedic surgery, increasing attention has been given to the effects of substance use on postoperative outcomes. A recent study by Maxson et al¹² examined the effect of marijuana use on postoperative outcomes for tibial shaft fracture patients, finding no differences in postoperative complication rates. Opioid use, conversely, has been linked to poor surgical outcomes. Studies on arthroscopic rotator cuff repair, total shoulder arthroplasty, and knee surgery have implicated preoperative long-term opioid use in increased postoperative complications and health care utilization,^7,22,27 leading to calls to wean patients from opioids to optimize them prior to surgery.¹³ Given the established risks of preoperative opioid use, it is important to understand these risks stratified by type of surgery.

Ankle fractures are one of the most common orthopaedic injuries, with an estimated incidence of 4.22 per 10 000 person-years in the United States, with up to 25% requiring surgery.^5,23 The few existing studies examining preoperative opioid use in this population largely focus on patients with a formal diagnosis of opioid use disorder (OUD). These studies suggest that preoperative opioid use negatively affects postoperative recovery and increases the risk of prolonged opioid dependence following ankle fracture surgery.^1,2,19,21 Research on open lower extremity fractures indicates that patients with OUD experience more complex hospital stays and higher 90-day complication rates than opioid-naive patients.²⁰ However, limited evidence exists regarding the specific effects of long-term opioid use on the postoperative course of isolated ankle fractures.

This exposure-based, retrospective cohort study aimed to determine whether long-term opioid use at the time of injury is associated with increased postoperative health care utilization following surgical fixation of isolated ankle fractures. Improved understanding of these outcomes may allow surgeons to make more informed decisions regarding surgical decision-making and perioperative care.

Methods

Study Design and Population

This was a retrospective, exposure-based cohort study. Ethical approval was obtained from our Institutional Review Board prior to initiation of the study. All patients who underwent open repair of an ankle fracture at a single level 1 trauma center over a 24-month period (October 1, 2021, to October 1, 2023) were retrospectively identified using Current Procedural Terminology (CPT) codes (27814, 27766, 27769, 27792, 27829, 27822, 27823) and screened for inclusion. Exclusion criteria were (1) any concurrent orthopaedic injury (polytrauma, associated foot, pilon, or tibial shaft injuries, contralateral injuries, or upper limb injuries), (2) lacking at least 2 weeks of documented follow-up data, and (3) patients without records clearly documenting their drug use or lack thereof. Patients with systemic comorbidities were not excluded because of homogenous distribution between groups. We hypothesized that patients with long-term preoperative opioid use would experience increased health care utilization within 90 days of surgery compared with opioid-naïve patients. Primary outcome measures included 90-day emergency department (ED) visits, unplanned readmissions, and unplanned reoperations. Routine postoperative follow-up typically occurred at approximately 2 weeks, 6 weeks, and 3 months after surgery, allowing for assessment of wound healing, progression of weightbearing, and postoperative complications.

Variables

Retrospective chart review via electronic medical records was performed to extract patient data. Data included patient demographics (sex, age, body mass index, race, insurance status), preexisting comorbidities (diabetes, vascular disease, cardiac disease, pulmonary disease, renal disease, rheumatoid disease, smoking status, previous psychiatric diagnoses, positive history of chronic pain), and opioid use at time of injury. Patients were divided into 2 cohorts based on opioid use at presentation:

Opioid users: Those with a documented electronic medical record history of long-term (≥3 months) opioid use at time of presentation, including prescription use (eg, oxycodone, hydrocodone, tramadol), maintenance therapy (methadone, buprenorphine), or illicit substances (eg, heroin, fentanyl)

Opioid-naïve: Those without any documented history of opioid use or those who had been taking opioids for ≤3 months

In-hospital data included fracture pattern, open vs closed injury, Gustilo-Anderson grade, presence of syndesmotic injury, initial external fixation, administration of preoperative nerve block, use of postoperative pain catheter, length of initial hospital stay, discharge status, opioid prescriptions from the ED, and total morphine milligram equivalents (MMEs) prescribed at discharge. Fracture pattern was assigned to one of 4 categories based on intraoperative findings: lateral malleolar, medial malleolar, bimalleolar, and trimalleolar. All open fractures were assigned a Gustilo-Anderson grade (1, 2, 3a, 3b, or 3c) depending on intraoperative findings. MMEs at discharge were calculated as number of tablets times the dosage of the medication in milligrams converted to morphine equivalents using the Ohio Board of Pharmacy Oral MME Conversion Table.¹⁸ Ninety-day follow-up data included number of ED visits, number of unplanned readmissions, number of unplanned reoperations, development of postoperative infection, number of clinic visits, number of telephone messages, and number of virtual chart messages. Reasons for postoperative ED visits and readmissions were grouped into one of 5 categories based on chief complaint and hospital course: (1) pain-related, (2) infection and wound concerns, (3) medical or systemic issue, (4) trauma or reinjury, and (5) surgical hardware concerns.

Our primary outcomes were number of postoperative ED visits, unplanned readmissions, and unplanned reoperations within 90 days of the index procedure. Exploratory outcomes included rates of postoperative infection, duration of follow-up, number of clinic visits, number of telephone encounters, and number of chart messages within 90 days of index procedure.

Statistical Analysis

Descriptive statistics summarized patient demographics and clinical characteristics. Categorical variables were reported as frequencies (percentages) and compared using χ² analysis or Fisher exact tests, as appropriate. Bonferroni-adjusted post hoc χ² was used to assess pairwise comparisons. Continuous variables were presented as means ± SDs and assessed for normality using Wilcoxon rank-sum tests. Comparisons between 2 groups were performed using Student t tests for normally distributed variables.

To determine whether preoperative opioid use independently predicted 90-day readmissions or ED visits, a binary logistic regression was used. This type of analysis allows us to examine how multiple factors together influence the likelihood of a binary outcome (eg, whether or not a patient was readmitted), while holding the other variables constant.

Eight clinically relevant variables were entered into each regression model: preoperative opioid use, tobacco use, fracture pattern, open vs closed fracture, baseline ambulatory status, employment status, psychiatric history, and hospital length of stay. Model fit was assessed using the Hosmer-Lemeshow goodness-of-fit test, which evaluates how well the predicted results from the model match the actual observed outcomes. P values greater than .05 indicate good model fit, with agreement between observed and predicted outcomes.

The threshold for significance was set at α = 0.05 for all statistical testing. Missing data points were excluded from analysis. All statistical analyses were conducted using IBM SPSS Statistics for Windows, version 29.0.2.0 (IBM Corp).

Results

Patient Demographics

A total of 523 patients were initially identified through electronic medical records, and a total of 303 patients were included in the analysis (Figure 1). Among them, 40 patients were long-term opioid users at the time of presentation and 263 were not. Of those with documented opioid use, 70.0% reported prescription use, 12.5% were actively undergoing treatment for OUD with maintenance medications, and 17.5% reported regular, illicit, intravenous drug use. There were no significant differences in age or sex distribution between opioid-naïve (50.0 ± 19.1 years, 40.6% male) and opioid-using (52.4 ± 16.7 years, 37.5% male) groups. There were no significant differences in rates of diabetes (16.3% vs 22.5%), vascular disease (0.7% vs 2.5%), cardiac disease (11.8% vs 12.5%), pulmonary disease (4.1% vs 10.0%), renal disease (4.9% vs 7.5%), or rheumatoid disease (1.5% vs 5.0%) between opioid-naïve and opioid-using cohorts. Both groups had similar body mass indexes (30.9 ± 7.5 for opioid-naïve patients vs 33.2 ± 8.0 for opioid-using patients). There were no significant differences in racial or ethnic makeup between the 2 cohorts.

Figure 1.

Flow diagram illustrating the original cohort of patients initially identified by Current Procedural Terminology (CPT) code, the number of patients filtered by each exclusion criteria, and the final pool of patients reviewed.

Opioid users were more likely to report current smoking at the time of injury (47.5% vs 23.6%, P = .001). They were also more likely to have positive histories of chronic pain (27.5% vs 4.6%, P < .001) and existing psychiatric diagnoses (32.5% vs 15.6%, P = .009), specifically mood disorders (27.5% vs 11.4%, P = .006). Opioid users were less likely to report employment at the time of presentation (25.0% vs 56.7%, P < .001) or independent ambulation at baseline (70.0% vs 88.6%, P = .003), and more likely to use assistive devices for ambulation (25.0% vs 9.9%, P = .010) (Table 1).

Table 1.

Baseline Demographic Characteristics for Both Opioid-Naïve Patients and Long-term Opioid Users.^a

	Opioid-naïve (n = 263)	Opioid users (n = 40)	P Value
Sex			.702
Male	107 (40.6)	15 (37.5)
Female	156 (59.3)	25 (62.5)
Age, mean ± SD	50.0 ± 19.1	52.4 ± 16.7	.466
Body mass index, mean ± SD	30.9 ± 7.5	33.2 ± 8.0	.078
Race			.088
White	178 (67.7)	26 (65.0)
Black	62 (23.6)	11 (27.5)
Hispanic	13 (4.9)	1 (2.5)
Asian	7 (2.7)	0 (0)
Multiracial	2 (0.7)	0 (0)
Other/not listed	1 (0.4)	2 (5.0)
Diabetes	43 (16.3)	9 (22.5)	.368
Vascular disease	2 (0.7)	1 (2.5)	.347
Cardiac disease	31 (11.8)	5 (12.5)	.798
Pulmonary disease	11 (4.1)	4 (10.0)	.121
Renal disease	13 (4.9)	3 (7.5)	.453
Rheumatoid disease	4 (1.5)	2 (5.0)	.181
Tobacco use			.005*
Current	62 (23.6)	19 (47.5)	.001*
Former	49 (18.6)	7 (17.5)	.846
Never	152 (57.8)	14 (35.0)	.007*
Employed at time of injury	149 (56.7)	10 (25.0)	<.001*
Baseline ambulatory status			<.006*
Independent	233 (88.6)	28 (70.0)	<.003*
Assistive device	26 (9.9)	10 (25.0)	<.010*
Nonambulatory	4 (1.5)	2 (5.0)	.181
Chronic pain	12 (4.6)	11 (27.5)	<.001*
Any psychiatric diagnosis	41 (15.6)	13 (32.5)	.009*
Mood disorder	30 (11.4)	11 (27.5)	.006*
Anxiety disorder	17 (6.4)	6 (15.0)	.058
Psychotic disorder	3 (1.1)	1 (2.5)	>.999
Stress/trauma disorder	4 (1.5)	1 (2.5)	>.999
Neurodevelopmental disorder	3 (1.1)	0 (0)	>.999

Unless otherwise noted, values are n (%).

Statistically significant (P < .05) values are in bold.

Injury Characteristics

Trimalleolar fractures were the most common fracture pattern among both groups. While there was a significant difference in fracture pattern distribution (P = .040), pairwise comparisons of individual fracture types were insignificant after Bonferroni correction. Opioid users had slightly higher rates of open fractures (22.5% vs 14.4%, P = .190), but this difference was not statistically significant. Opioid users had significantly longer hospital stays (5.3 ± 6.1 vs 2.6 ± 2.9 days, P < .001). There were no differences in rates of syndesmotic injury, external fixation, or time from presentation to definitive surgery. Short-course opioid prescriptions from the ED and total MMEs at discharge were similar between groups (Table 2).

Table 2.

Injury Characteristics and Discharge Information of Opioid-Naïve Patients and Opioid Users.^a

	Opioid-Naïve	Opioid Users	P Value
Fracture type			.040*
Lateral malleolar	34 (12.9)	4 (10.0)	.631
Medial malleolar	6 (2.2)	3 (7.5)	.102
Bimalleolar	107 (40.7)	9 (22.5)	.028
Trimalleolar	116 (44.1)	24 (60.0)	.060
Syndesmotic injury	71 (26.9)	9 (22.5)	.548
Syndesmotic stabilization	84 (31.9)	16 (40.0)	.312
Open fracture	38 (14.4)	9 (22.5)	.190
Gustilo type			.377
1	7 (18.4)	3 (7.5)
2	19 (50.0)	4 (10.0)
3a	12 (31.6)	2 (5.0)
3b	0 (0)	0 (0)
3c	0 (0)	0 (0)
External fixation	11 (4.1)	3 (7.5)	.352
Days in external fixator, mean ± SD	16 ± 13	25 ± 10	.290
Days from presentation to definitive surgery, mean ± SD	3.6 ± 5.8	4.4 ± 8.9	.052
Given short course of opioids from ED	50 (19.0)	4 (10.0)	.190
Length of stay, mean ± SD	2.6 ± 2.9	5.3 ± 6.1	<.001*
Nerve block	240 (91.3)	34 (85.0)	.210
Pain catheter	134 (50.9)	22 (55.0)	.633
Disposition status			.109
Home	206 (78.3)	25 (62.5)
SNF	46 (17.5)	11 (27.5)
IPR	8 (3.0)	3 (7.5)
Left AMA	1 (0.4)	1 (2.5)
MMEs at discharge, mean ± SD	212 ± 124	200 ± 151	.615

Abbreviations: AMA, against medical advice; ED, emergency department; MMEs, morphine milligram equivalents; IPR, inpatient rehabilitation; SNF, skilled nursing facility.

Unless otherwise noted, values are n (%).

Statistically significant (P < .05) values are in bold.

Complications and Health care Utilization

Opioid users had significantly higher rates of 90-day health care utilization. They had more ED visits (0.55 ± 1.11 vs 0.24 ± 0.61 visits, P = .010), unplanned readmissions (0.33 ± 0.69 vs 0.076 ± 0.33, P < .001), and unplanned reoperations (0.18 ± 0.59 vs 0.05 ± 0.26, P = .026). All unplanned re-operations were for wound- or infection-related causes. The overall percentage of patients experiencing unplanned readmission was significantly higher among opioid users (20.0% vs 5.7%, P < .001). No significant differences were observed in postoperative infection rates or reasons for unplanned reoperations.

Pain-related complaints were the most common reason for postoperative ED visits among opioid users (40.9% vs 28.1%), while infections and wound concerns were the most frequent reasons for unplanned readmissions (53.8% vs 45.0%). Opioid users had significantly higher average number of telephone encounters (0.60 ± 1.05 vs 0.24 ± 0.94, P = .031) and chart messages (0.40 ± 1.10 vs 0.05 ± 0.24, P < .001) within the 90-day follow up period (Table 3).

Table 3.

Rates of 90-Day Postoperative Complication and 90-Day Health Care Utilization Between Opioid-Dependent and Opioid-Naïve Patients.^a

	Opioid-Naïve (n = 263)	Opioid Users (n=40)	P Value	95% CI of the Difference
	Opioid-Naïve (n = 263)	Opioid Users (n=40)	P Value	Lower	Upper
Total number of ED visits within 90 d	64	22
Mean number of ED visits, mean ± SD	0.24 ± 0.61	0.55 ± 1.11	.010*	−0.539	−0.074
Reason for ED visit			.816
Pain or medication related	18 (28.1)	9 (40.9)
Infection and wound related	16 (25.0)	5 (22.7)
Medical or systemic issue	26 (40.6)	7 (31.8)
Trauma or reinjury	3 (4.7)	1 (4.5)
Surgical hardware concerns	1 (1.6)	0 (0.0)
Total number of unplanned readmissions within 90 d	20	13
Number of unplanned readmissions, mean ± SD	0.076 ± 0.33	0.33 ± 0.69	<.001*	−3.81	−0.117
Reason for unplanned readmission			.671
Pain or medication related	0 (0.0)	0 (0)
Infection and wound related	9 (45.0)	7 (53.8)
Medical or systemic issue	10 (50.0)	6 (46.2)
Trauma or reinjury	1 (5.0)	0 (0.0)
Surgical hardware concerns	0 (0.0)	0 (0.0)
Total number of unplanned reoperations within 90 d	14	7		−.229	−0.014
Number of unplanned reoperations, mean ± SD	0.05 ± 0.26	0.18 ± 0.59	.026*
Reason for unplanned reoperation			>.999
Infection and wound related	14 (100)	7 (100)
Postoperative infection	23 (8.7)	4 (10.0)	>.999
Patients with any ED visit within 90 d	46 (17.5)	11 (27.5)	.131
Patients with any unplanned readmission within 90 d	15 (5.7)	8 (20.0)	.001*
Patients with any unplanned reoperation within 90 d	12 (4.6)	4 (10.0)	.152
Duration of follow-up, d, mean ± SD	202 ± 187	177 ± 165	.422
Number of clinic visits, mean ± SD	2.8 ± 1.4	2.7 ± 1.7	.500
Number of telephone encounters, mean ± SD	0.24 ± 0.94	0.60 ± 1.05	.031*
Number of chart messages, mean ± SD	0.05 ± 0.24	0.40 ± 1.10	<.001*

Abbreviation: ED, Emergency department.

Unless otherwise noted, values are n (%).

Statistically significant (P < .05) values are in bold.

Two multivariable logistic regression models were constructed to identify independent predictors of (1) unplanned hospital readmission within 90 days and (2) ED visits within 90 days of discharge.

After controlling for fracture characteristics, comorbidities, and social factors, a prior psychiatric diagnosis and longer initial hospital stay were the only variables significantly associated with increased odds of unplanned 90-day readmission. Specifically, patients with a documented psychiatric diagnosis were more than 6 times more likely to experience readmission compared to those without (OR = 6.24, 95% CI 2.20-17.66, P < .001), and each additional hospital day increased the odds of readmission by approximately 14% (OR = 1.14, 95% CI 1.02-1.27, P = .025). For 90-day ED visits, length of initial hospital stay was the only significant predictor. Each additional hospital day increased the likelihood of an ED visit by roughly 17% (OR = 1.17, 95% CI 1.06-1.28, P = .001). Although psychiatric history and tobacco use approached significance (P = .068 and P = .090, respectively), these associations did not reach statistical significance. Preoperative opioid use itself was not a significant independent predictor of either readmission (OR = 1.77, P = .33) or ED visits (OR = 0.91, P = .84) once other clinical and psychosocial factors were considered (Tables 4 and 5). The Hosmer-Lemeshow goodness-of-fit test indicated good model calibration (P = .83 for unplanned readmissions; P = .15 for ED visits), suggesting no significant deviation between observed and predicted outcomes.

Table 4.

Results From Binomial Logistic Regression Data Predicting the Likelihood of Unplanned Readmission Within 90 Days of Surgery.^a

	B	SE	Wald χ²	Df	P	Odds Ratio	95% CI for Odds Ratio
	B	SE	Wald χ²	Df	P	Odds Ratio	Lower	Upper
Fracture pattern
Unimalleolar			1.015	2	.602
Bimalleolar	0.813	1.147	0.503	1	.478	2.255	0.238	21.359
Trimalleolar	1.083	1.123	.929	1	.335	2.952	0.327	26.676
Open fracture	0.616	0.595	1.074	1	.300	1.852	0.577	5.944
Length of initial stay	0.128	0.057	5.006	1	.025*	1.137	1.016	1.272
Baseline ambulatory status
Independent			0.847	2	.655
Assisted (cane or walker)	0.556	0.606	0.842	1	.359	1.743	0.532	5.712
Nonambulatory	0.197	1.117	0.031	1	.860	1.218	0.136	10.877
Preoperative opioid use	0.571	0.584	0.955	1	.328	1.769	0.563	5.558
Current tobacco use	−0.459	0.607	0.573	1	.449	0.632	0.192	2.074
Unemployed at presentation	−0.859	0.667	1.659	1	.198	0.424	0.115	1.565
Previous psychiatric diagnosis	1.831	0.532	11.890	1	<.001*	6.238	2.204	17.657
Constant	−4.498	1.193	14.215	1	<.001*	0.011

Abbreviations: B, constant coefficient; Df, degrees of freedom; ED, emergency department; SE, standard error.

Included variables were selected on the basis of clinical relevance.

Statistically significant (P < .05) values are in bold.

Table 5.

Results From Binomial Logistic Regression Data Predicting the Likelihood of ED Visit Within 90 Days of Discharge.^a

	B	SE	Wald χ²	Df	P	Odds Ratio	95% CI for Odds Ratio
	B	SE	Wald χ²	Df	P	Odds Ratio	Lower	Upper
Fracture pattern
Unimalleolar			3.577	2	.167
Bimalleolar	−0.450	0.464	0.940	1	.332	0.638	0.257	1.584
Trimalleolar	−0.878	0.475	3.416	1	.065	0.416	0.164	1.054
Open fracture	0.447	0.406	1.212	1	.271	1.563	0.706	3.462
Length of initial stay	0.154	0.048	10.425	1	.001*	1.167	1.063	1.282
Baseline ambulatory status
Independent			0.205	2	.903
Assisted (cane or walker)	0.154	0.484	0.101	1	.751	1.166	0.451	3.012
Nonambulatory	−0.237	0.993	0.057	1	.811	0.789	0.113	5.520
Preoperative opioid use	−0.097	0.467	0.043	1	.835	0.907	0.363	2.266
Current tobacco use	0.599	0.353	2.871	1	.090	1.820	0.910	3.637
Unemployed at presentation	−0.275	0.375	0.539	1	.463	0.759	0.364	1.584
Previous psychiatric diagnosis	0.697	0.382	3.330	1	.068	2.008	0.950	4.245
Constant	−1.757	0.505	12.104	1	<.001*	0.173

Abbreviations: B, constant coefficient; Df, degrees of freedom; ED, emergency department; SE, standard error.

Included variables were selected on the basis of clinical relevance.

Statistically significant (P < .05) values are in bold.

Discussion

Our findings indicate that long-term preoperative opioid users experience significantly higher postoperative health care utilization than nonusers. Specifically, these patients had longer hospital stays, a higher number of 90-day ED visits, unplanned readmissions, unplanned reoperations, telephone messages, and electronic chart messages. No significant differences were observed in rates of postoperative infection between the 2 groups. Opioid users also demonstrated a higher baseline burden of psychosocial comorbidity, with significantly greater rates of chronic pain and psychiatric diagnoses, as well as poorer preinjury functional status. They were more likely to require an assistive device for ambulation and to be discharged to skilled nursing or inpatient rehabilitation facilities, despite having comparable rates of systemic medical diseases such as diabetes, vascular, cardiac, pulmonary, renal, and rheumatoid disease. These findings suggest that increased health care utilization among opioid users is more closely related to baseline functional and psychosocial factors rather than underlying systemic illness and that long-term opioid use may function primarily as a marker of this higher-risk profile. Notably, the rate at which opioids were prescribed from the ED and total MMEs prescribed at discharge were similar between groups. Multivariate analysis further demonstrated that a history of psychiatric diagnosis and longer initial hospital stay were associated with increased odds of 90-day readmission, whereas preoperative opioid use alone was not an independent predictor of readmission or ED utilization.

These results align with previous studies examining the impact of opioid use in ankle fracture patients. Oladeji et al¹⁹ analyzed a database of 61 424 patients and found that increasing preoperative opioid consumption correlated with higher ED visits, unplanned readmissions, and postoperative health care costs, suggesting a dose-dependent relationship. However, although Oladeji et al¹⁹ reported increased postoperative surgical site infections among opioid users, our study did not find significant differences in infection rates. Similarly, in a retrospective analysis of patients undergoing elective primary joint replacements or spinal fusion, Menendez et al¹⁴ demonstrated that preoperative opioid abuse is associated with greater likelihood of postoperative morbidity, mortality, and health care resource utilization. They concluded that surgeons should work with patients to stop high-risk opioid use before surgery and decline to perform elective surgery altogether on patients who misuse them. Among 17 811 patients with open lower extremity fractures, Peluso et al²⁰ found 45% higher odds of 90-day all-cause readmission and longer lengths of stay in patients with a formal diagnosis of opioid-use disorder, calling for the increased use of addiction medicine services prior to surgery.

Our results confirm the well-established link between long-term opioid use, chronic pain, and comorbid mental health disorders.²⁵ Prior research has demonstrated that orthopaedic surgery patients with mood disorders experience higher health care costs, longer hospital stays, and increased all-cause postoperative mortality.³ Given the high prevalence of psychiatric comorbidities in opioid users, these factors may introduce a source of confounding in our data. In our analysis, opioid users were more than twice as likely to have documented histories of any psychiatric diagnosis or mood disorder, and more than 6 times more likely to have a documented history of chronic pain. Importantly, multivariate analysis revealed that a history of psychiatric diagnosis, rather than opioid use itself, was associated with increased odds of 90-day readmissions and ED visits, suggesting that this subset of patients may potentially benefit from increased use of multidisciplinary management involving psychiatry, social work, and pain medicine specialists.

Additionally, heightened pain perception may contribute to increased health care utilization among opioid users. More than 40% of all ED visits among opioid users were for postoperative pain, compared with 28% among opioid-naïve patients. Prolonged opioid administration has been found to induce upregulation of nociceptive neurons in the dorsal horn and downregulation of the descending pain-modulating pathway, resulting in an enhanced pain response.^8,10,26 Clinically, this can lead to increased postoperative health care utilization. Oladeji et al¹⁹ found a dose-dependent relationship between preoperative opioid dose and pain-related admissions in ankle fracture patients. Similarly, Dasinger et al⁶ reported a strong association between preoperative opioid use and pain-related hospital admissions across various surgical populations. Although discharge MMEs were similar between groups, this likely reflects a standardized institutional prescribing protocol rather than individualized dosing. Equivalent postoperative dosing despite preexisting opioid tolerance may lead to inadequate pain control and drive additional pain-related ED visits or readmissions.²¹ Psychiatric comorbidities further modulate pain perception, with studies showing that lower extremity fracture patients with anxiety or depression report higher levels of postoperative pain and increased ED visits.¹⁵

Several limitations should be considered when interpreting our findings. This was a single-institution, retrospective cohort study with a relatively narrow scope, focused specifically on isolated ankle fractures. Although this design limits generalizability to other institutions and fracture types, it also allowed for a more homogenous patient population and minimized confounding from polytrauma and other complex injuries. Although patients with systemic comorbidities were included, rates of diabetes, vascular disease, cardiac disease, pulmonary disease, renal disease, and rheumatoid disease were comparable between opioid-using and opioid-naïve groups, suggesting that these factors were unlikely to bias the observed outcomes. As with all retrospective studies, potential inaccuracies in documentation and reporting bias are possible. Our minimum follow-up period of 14 days may not fully capture long-term complications or health care utilization received elsewhere. Additionally, the nature of patient-reported social histories may introduce reporting bias; however, our observed opioid use rate of 12.9% aligns with rates reported elsewhere.^19,21 Our long-term opioid use cohort was heterogeneous, including patients using prescribed opioids, those undergoing maintenance therapy for OUD, and those using illicit opioids. Because of limited sample size, we were unable to stratify outcomes by opioid type or use pattern. This is an important limitation, as these subgroups may differ in health care access, physiologic tolerance, pain perception, and engagement with postoperative care. Furthermore, the number of long-term opioid users (n = 40) relative to the number of variables included in the multivariable models raises the possibility of overfitting and wide CIs, so these regression results should be interpreted as preliminary findings. Future studies with larger opioid-exposed cohorts should explore these distinctions to better tailor perioperative management strategies. This study only assessed opioids prescribed at discharge, not longitudinally throughout the 90-day follow-up period, preventing us from drawing meaningful conclusions about postoperative opioid use. Given the retrospective nature of this study, we also did not evaluate patient-reported outcomes, which remains an important area for future investigation.

The findings in this study highlight the significant burden that long-term opioid use places on the health care system. Despite similar rates of postoperative complications, the higher incidence of unplanned readmissions, ED visits, reoperation, and patient-provider communication among opioid users illustrate the susceptibility of this population of patients to suboptimal surgical outcomes, strengthening the case for perioperative opioid weaning programs and integration of interdisciplinary management such as addiction medicine, pain control, and psychiatry into orthopaedic care. Future work should focus on clarifying the mechanisms underlying increased rates of health care utilization to establish more effective management strategies for this difficult patient population.

Conclusion

In our single level 1 trauma center study of patients who underwent open repair of an ankle fracture, we found that long-term preoperative opioid use was associated with increased postoperative health care utilization, including longer hospital stays and higher rates of 90-day unplanned readmissions, ED visits, and reoperations, but not rates of postoperative infection. In this cohort, psychiatric history and longer initial hospital stay, rather than opioid use alone, were independent predictors of readmission and ED utilization, underscoring the importance of multidisciplinary approaches to mental health and pain management in this population. Further, high-quality prospective studies are needed to elucidate the mechanisms underlying increased health care utilization among opioid users.

Supplemental Material

sj-pdf-1-fao-10.1177_24730114251405246 – Supplemental material for Chronic Preoperative Opioid Use Is Associated With Increased Health Care Utilization Following Ankle Fracture Surgery

Supplemental material, sj-pdf-1-fao-10.1177_24730114251405246 for Chronic Preoperative Opioid Use Is Associated With Increased Health Care Utilization Following Ankle Fracture Surgery by Isaac C. Hale, Samuel Gerak, Paul McMillan, Zeping Wang, Logan Lake and Richard Laughlin in Foot & Ankle Orthopaedics

Footnotes

ORCID iD

Isaac C. Hale, BS

Ethical Considerations

This study was approved by the University of Cincinnati Institutional Review Board (OS20007 IRB 2020-0116).

Consent to Participate

The requirement for informed consent to participate was waived by the institutional review board.

Consent for Publication

Not applicable.

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. Disclosure forms for all authors are available online.

Data Availability Statement

The data supporting the findings of this research are available from the corresponding author upon request.

References

Badin

Ortiz-Babilonia

Gupta

, et al. Opioid use for operative foot and ankle fractures: a systematic review. Foot Ankle Orthop. 2022;7(4):2473011421S00570. doi:10.1177/2473011421S00570

Broggi

Oladeji

Spenser

Kadakia

Bariteau

JT.

Risk factors for prolonged opioid use after ankle fracture surgery. Foot Ankle Spec. 2023;16(5):476-484. doi:10.1177/19386400211029123

Brown

Alas

Bortz

, et al. Patients with psychiatric diagnoses have increased odds of morbidity and mortality in elective orthopedic surgery. J Clin Neurosci. 2021;84:42-45. doi:10.1016/j.jocn.2020.11.049

Centers for Disease Control and Prevention. Understanding the opioid overdose epidemic. Published November 1, 2024. Accessed December 27, 2024. https://www.cdc.gov/overdose-prevention/about/understanding-the-opioid-overdose-epidemic.html

Court-Brown CM, Caesar B. Epidemiology of adult fractures: a review. Injury. 2006;37(8):691-697. doi:10.1016/j.injury.2006.04.130

Dasinger

Graham

Wahl

, et al. Preoperative opioid use and postoperative pain associated with surgical readmissions. Am J Surg. 2019;218(5):828-835. doi:10.1016/j.amjsurg.2019.02.033

Farley

Wilson

Spencer

, et al. Preoperative opioid use is a risk factor for revision surgery, complications, and increased resource utilization after arthroscopic rotator cuff repair. Am J Sports Med. 2020;48(13):3339-3346. doi:10.1177/0363546520960122

Hay

White

Bochner

Somogyi

Semple

Rounsefell

Hyperalgesia in opioid-managed chronic pain and opioid-dependent patients. J Pain. 2009;10(3):316-322. doi:10.1016/j.jpain.2008.10.003

Kaiser Family Foundation. Opioid deaths fell in mid-2023, but progress is uneven and future trends are uncertain. Published September 23, 2024. Accessed March 27, 2025. https://www.kff.org/mental-health/issue-brief/opioid-deaths-fell-in-mid-2023-but-progress-is-uneven-and-future-trends-are-uncertain/

10.

Koppert

Schmelz

The impact of opioid-induced hyperalgesia for postoperative pain. Best Pract Res Clin Anaesthesiol. 2007;21(1):65-83. doi:10.1016/j.bpa.2006.12.004

11.

Kotlińska-Lemieszek

Żylicz

Less well-known consequences of the long-term use of opioid analgesics: a comprehensive literature review. Drug Des Devel Ther. 2022;16:251-264. doi:10.2147/DDDT.S342409

12.

Maxson

Rapaport

Covarrubias

, et al. Marijuana use and complication risk after tibia shaft fracture fixation. J Orthop Trauma. 2025;39(3):137-143. doi:10.1097/BOT.0000000000002945

13.

McAnally

Rationale for and approach to preoperative opioid weaning: a preoperative optimization protocol. Perioper Med (Lond). 2017;6:19. doi:10.1186/s13741-017-0079-y

14.

Menendez

Ring

Bateman

BT.

Preoperative opioid misuse is associated with increased morbidity and mortality after elective orthopaedic surgery. Clin Orthop Relat Res. 2015;473(7):2402-2412. doi:10.1007/s11999-015-4173-5

15.

Momtaz

Jahn

Eskenazi

, et al. The impact of anxiety and depression on postoperative pain management and emergency room visits in pediatric patients with surgically treated lower extremity fractures: a propensity-matched cohort analysis. J Pediatr Orthop. 2025;45(5):251-257. doi:10.1097/BPO.0000000000002910

16.

National Institute on Alcohol Abuse and Alcoholism (NIAAA). Rates of nonmedical prescription opioid use and opioid use disorder double in 10 years. 2016. Accessed December 27, 2024. https://www.niaaa.nih.gov/news-events/news-releases/rates-nonmedical-prescription-opioid-use-and-opioid-use-disorder-double-10-years?utm_source=chatgpt.com

17.

NORC at the University of Chicago. One-third of Americans have received an opioid prescription in the past two years. Published September 27, 2018. Accessed December 27, 2024. https://www.norc.org/research/library/one-third-of-americans-have-received-an-opioid-prescription-in-t.html

18.

Ohio Automated Rx Reporting System. Morphine equivalent daily dose conversion table. Updated August 16, 2017. Accessed March 10, 2025. https://www.ohiopmp.gov/Documents/MorphineEquivalentDailyDoseConversionTable.pdf

19.

Oladeji

Broggi

Spencer

Hurt

Hernandez-Irizarry

The impact of preoperative opioid use on complications, readmission, and cost following ankle fracture surgery. Injury. 2021;52(8):2469-2474. doi:10.1016/j.injury.2021.06.013

20.

Peluso

Araya

Patel

, et al. How is preoperative opioid use associated with readmissions and outcomes in lower extremity trauma? Clin Orthop Relat Res. 2025;483(5):918-927. doi:10.1097/CORR.0000000000003346

21.

Reiners

Vang

Rivard

Bostrom

Nguyen

MP.

Association of preoperative opioid use with postoperative opioid use and patient-reported outcomes in ankle fracture fixation patients. Foot Ankle Orthop. 2022;7(4):24730114221139787. doi:10.1177/24730114221139787

22.

Ridenour

Kowalski

Yadavalli

, et al. Preoperative opioid use is associated with persistent use, readmission and postoperative complications after arthroscopic knee surgery. Arthrosc J Arthrosc Relat Surg. 2021;37(5):1567-1572. doi:10.1016/j.arthro.2020.12.187

23.

Scheer

Newman

Zhou

, et al. Ankle fracture epidemiology in the United States: patient-related trends and mechanisms of injury. J Foot Ankle Surg. 2020;59(3):479-483. doi:10.1053/j.jfas.2019.09.016

24.

Song

Tang

Roses

, et al. Opioid use disorder is associated with complications and increased length of stay after major abdominal surgery. Ann Surg. 2021;274(6):992. doi:10.1097/SLA.0000000000003697

25.

Sullivan

Edlund

Steffick

Unützer

Regular use of prescribed opioids: association with common psychiatric disorders. Pain. 2005;119(1-3):95-103. doi:10.1016/j.pain.2005.09.020

26.

White

JM.

Pleasure into pain: the consequences of long-term opioid use. Addict Behav. 2004;29(7):1311-1324. doi:10.1016/j.addbeh.2004.06.007

27.

Wilson

Farley

Gottschalk

Daly

Wagner

ER.

Preoperative opioid use is an independent risk factor for complication, revision, and increased health care utilization following primary total shoulder arthroplasty. J Shoulder Elbow Surg. 2021;30(5):1025-1033. doi:10.1016/j.jse.2020.08.007

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.01 MB