INTRODUCTION
Preoperative chronic opioid use is strongly linked to negative surgical [1] and functional outcomes [2], including prolonged hospitalization [3], higher healthcare utilization and costs [4], higher risk of surgical revision [5,6], delayed return to work, elevated pain interference [7], increased pain intensity [8], and greater disability in patients undergoing spine surgery [2,8]; despite these risks, preoperative screening for opioid use and misuse remains uncommon [9]. Notably, 38% of patients self-report opioid use that is undocumented in the electronic health record [9], suggesting the need for targeted opioid screening to identify high-risk patients.
Back and neck pain are highly prevalent conditions contributing significantly to disability and healthcare utilization worldwide, making them common reasons for seeking pain management [10,11]. Consequently, opioid analgesics have been widely prescribed for the management of chronic spine-related pain in the United States [11,12]. Commonly prescribed opioids for pain include medications such as hydrocodone, oxycodone, tramadol, morphine, and hydromorphone [13,14]. Despite the known risks associated with long-term opioid use, the demand persists, partly due to the severe, often debilitating nature of spine pain and the variable effectiveness of alternative nonopioid treatments [15]. While interventions, such as local injections and nerve ablations, are frequently utilized, evidence suggests they may have limited long-term efficacy for many patients with chronic low back pain [15], underscoring the continued reliance on pharmacological approaches, including opioids. This landscape underscores the critical need to understand opioid use patterns, including misuse, in patients undergoing spine surgery.
It is important to clarify the distinction between opioid misuse, opioid use disorder (OUD), and persistent postoperative opioid use (PPOU). Opioid misuse refers to the use of opioids other than as directed or prescribed, regardless of whether adverse effects are present [16,17]. Adverse effects are required for a diagnosis of OUD, which is a medical condition characterized by a problematic pattern of opioid use leading to clinically significant impairment or distress, diagnosed based on criteria reflecting compulsive use, impaired control, negative consequences, and craving; importantly, it is not based on the duration of opioid use [18,19]. PPOU is characterized by the duration of opioid use after surgery, commonly defined as at least 3 months postoperative [20]. Many studies operationalize this as a least one filled opioid prescription within 90 days postoperative and then at least one additional filled prescription between 91 and 180 days postoperative [21]. although some extend the latter period to 365 days postoperative [21,22]. Some definitions further clarify based on preoperative opioid use. For example, the 2018 POQI-4 consensus defined PPOU as opioid use for at least 60 days between postoperative days 90–365 in opioid naive patients, and as increased opioid use relative to the preoperative 90-day period in nonnaive patients during the same postoperative timeframe [22].
Preoperative opioid misuse specifically correlates with worse postoperative outcomes in elective orthopaedic surgery patients [23]. Opioid misuse is highly prevalent in this patient population, and historically, orthopaedic patients have been prescribed more opioids than other surgical patients, compounding the associated risks [24]. Beyond increased postoperative mortality, morbidity, mood disorders, and other opioid misuse-associated adverse events noted in orthopaedic patients, patients undergoing spine surgery exhibit the highest prevalence of PPOU (≥3 months) when compared to those undergoing other orthopaedic procedures [23]. Additional research is needed to differentiate the impact of preoperative opioid misuse from general preoperative chronic opioid use on patient-reported postoperative outcomes after spine surgery.
PPOU (≥3 months) is a recognized outcome in both preoperatively opioid-naive and opioid-tolerant patients undergoing lumbar fusion [25]. Understanding the biopsychosocial risk factors for PPOU could enable clinicians to implement targeted screening and risk stratification when prescribing opioids for postoperative pain control [26,27]. A 2022 meta-analysis of over 80,000 patients from 12 studies found that preoperative opioid use, substance abuse, psychiatric disorders, smoking history, arthritis, and a higher American Society of Anesthesiologists score increased the risk for prolonged opioid use following lumbar fusion [25]; however, the specific role of opioid misuse in the development of PPOU after spine surgery remains unclear.
Large database studies have identified risk factors for opioid abuse and dependence during the perioperative period, including younger age, Black or African American race, substance abuse, psychiatric disorders, smoking history, and public insurance (Medicaid and Medicare) [23]; however, many of these studies are retrospective, limiting their ability to assess disease severity before evaluating the impact of opioid misuse on surgical outcomes [28]. Additionally, most research has focused on how preoperative opioid use broadly affects spine surgery outcomes, rather than distinguishing opioid misuse as a distinct risk factor [1-8]. These studies often group patients with varying opioid consumption patterns—including prescribed use, misuse, and OUD—potentially resulting in an over- or underestimate of risk. By specifically examining opioid misuse, we aim to provide targeted insights into this high-risk group of patients and their unique postoperative challenges.
Research on patient-reported outcomes has demonstrated that preoperative opioid use is associated with worse spine surgery recovery, including poorer patient-reported physical and mental health [29], as well as smaller improvements in pain, function, and quality of life at one year compared to opioid-naive patients [7,30]. While these findings are significant, further research is needed to understand how preoperative opioid misuse specifically influences longitudinal pain trends and postoperative opioid cessation. Most studies assess patient-reported outcomes at baseline, 6 months, and 12 months postsurgery [31], which may not fully capture short-term postoperative pain fluctuations or provide a comprehensive understanding of pain trajectories over time. Increasing the frequency of longitudinal data collection could offer deeper insights into acute and subacute postoperative pain patterns over time.
This prospective cohort study assessed the relationship between preoperative opioid misuse and time to pain cessation and opioid discontinuation among patients undergoing elective spine surgery. Additionally, we examined variations in patient-reported pain, depression, anxiety, physical function, and sleep over one year after surgery, comparing patients with and without preoperative opioid misuse.
MATERIALS AND METHODS
1. Study Design and Eligible Participants
This prospective, observational study was conducted at Stanford Hospital. Patients scheduled for elective spine surgery were considered for inclusion. Eligibility criteria included: age ≥18 years, English proficiency, and ability and willingness to complete assessments. Exclusion criteria included: inability to complete longitudinal assessments due to education, cognitive ability, mental status, or medical limitations; pregnancy; suicidality (defined as a score >0 on question 9 of the Patient Health Questionnaire [PHQ-9], which assesses suicidal thoughts); or enrollment in another perioperative trial.
Eligible participants were enrolled via phone consent and completed preoperative baseline questionnaires prior to their elective spine surgery date. Enrollment occurred between July 12th, 2019, and November 19th, 2021. After undergoing surgery, participants reported changes in pain, opioid use, and psychological status through various postoperative follow-up assessments, administered weekly until week 24 and monthly until study completion at 1 year (52 weeks) postoperative.
The study was performed in accordance with the Declaration of Helsinki and approved by the Stanford University Institutional Review Board (IRB #43163) on April 18, 2018. Written informed consent was obtained from all study participants prior to participation
2. Assessments
Before surgery, participants completed an online questionnaire assessing demographics, pain, and opioid use, in addition to questions regarding disability compensation, litigation involvement, and worker’s compensation claims related to their spine pain.
Participants completed the modified Brief Pain Inventory (BPI) [32], referencing pain at the upcoming surgical site. Pain intensity was measured using a 10-point numerical rating scale (0=no pain, 10=worst pain imaginable) for worst, least, average, and current pain. The BPI also assessed pain interference with general activity, mood, walking ability, normal work, personal relationships, sleep, and enjoyment of life using a 0–10 scale (0=no interference to 10=complete interference). The modified BPI was administered at baseline and weekly from postoperative week 1 to week 24, and monthly thereafter until week 52.
Opioid use was assessed at baseline and weekly from postoperative week 1 to week 24, and monthly thereafter until week 52 via a 21-item two-section survey. In the first section—medication use— participants indicated whether they had taken opioid or nonopioid pain medications in the past week; for each opioid, they reported dosage, administration route (e.g., patch, tablet, film), and amount taken in the past 24 hours. In the second section—medication misuse and side effects—participants reported severity of medication side effects and reasons for opioid use beyond pain management, such as sleep aid, increased dosage beyond prescription, or use for anxiety or mood improvement.
At baseline, participants completed the Current Opioid Misuse Measure (COMM-17), a 17-item questionnaire evaluating behaviors associated with opioid misuse. Responses are rated on a 5-point scale (0=never, 1=seldom, 2=sometimes, 3=often, 4=very often). A total COMM-17 score of ≥9 is classified as positive for opioid misuse [33].
The PHQ-9 was also administered at baseline to assess depressive symptoms over the past 2 weeks. Nine symptoms are rated on a 4-point scale (0=not at all, 1=several days, 2=more than half the days, 3=nearly every day), with total scores categorized as minimal (1–4), mild (5–9), moderate (10–14), moderate severe (15–19), and severe (20–27) depression [34]. If any symptoms are present, participants then rate how difficult these symptoms have made it to perform work tasks, manage household responsibilities, or interact with others on a 4-point scale (1=not difficult at all, 2=somewhat difficult, 3=very difficult, 4=extremely difficult); this question is excluded from the PHQ-9 total score.
Additional mood assessments were conducted at baseline and every 4 weeks postoperatively from week 4 to week 52 using National Institutes of Health (NIH) Patient-Reported Outcomes Measurement Information System (PROMIS) measures: PROMIS depression evaluates cognitive effects of depression (e.g., hopelessness, worthlessness, and feelings of failure) [35]. PROMIS anxiety measures cognitive and somatic effects of anxiety (e.g., fear, dread, restlessness, and dizziness) [35]. PROMIS sleep disturbance assesses sleep quality and bedtime behaviors correlated with disturbance [36]. PROMIS physical function evaluates difficulty in performing specific manual tasks to understand limitations of function [37].
Each PROMIS measure used computerized adaptive testing, presenting 4–6 questions from validated test banks per measure. Responses are scored on a 5-point scale (1=never to 5=always) and summed to generate a T score for each measure (reference population mean=50, standard deviation=10) [35]. Higher T scores indicate worse symptoms for depression, anxiety, and sleep disturbance, whereas higher T scores for physical function indicated better functional ability.
3. Primary and Secondary Outcomes
The primary outcome was time to return to baseline opioid dose, defined as the first of 2 consecutive postoperative reports indicating a return to preoperative opioid use levels calculated as oral milligram morphine equivalents. Reponses to the opioid use questionnaire were used to calculate a 24-hour oral morphine equivalent dose at baseline and each postoperative time point from weeks 1 to 52. This information was derived from patient-reported opioid dosage, administration route, and amount taken in the past 24 hours.
The secondary outcomes included time to opioid cessation, defined as the first of 2 reports of no opioid use (defined as postoperative patient-reported denial of opioid use on the postoperative opioid sue questionnaire assessed from weeks 1 to 52); time to pain cessation, defined as the first of 2 reports of 0 out of 10 average pain intensity at the surgical site in the past 24 hours assessed via the modified BPI from postoperative weeks 1 to 52; and time to patient-reported recovery, defined as the first report of “yes” in response to the question, “Would you say that you are fully recovered from your surgery?” assessed via the modified BPI from postoperative weeks 1 to 52.
4. Statistical Analyses
A sample size estimation was performed a priori assuming 25% of the participants would report preoperative opioid misuse, with a 78% event rate for return to baseline opioid dose. A final sample size of 50 participants was calculated to achieve 81% statistical power to detect a hazard ratio (HR) of 0.35 (assuming proportional hazards), at a significance level of 0.05.
SAS ver. 9.4 (SAS Institute Inc.) was used for all analyses. We determined baseline characteristics for the entire cohort partitioned by preoperative opioid misuse. We presented preoperative categorical variables as numbers (percentages). Continuous variables were presented as mean (standard deviation) or median (interquartile range). To assess differences between groups, we used the t-test for continuous variables and the chi-square or Fisher exact test for categorical variables.
To assess the association of preoperative opioid misuse with time to opioid and pain cessation events, we constructed Cox regression models to estimate HRs and 95% confidence intervals (95% CI), adjusting for age, sex, operative region (cervical, lumbar, thoracic), number of spinal levels of the operation (1, 2, 3–5, 6–10, ≥11), and any preoperative opioid use. The models were examined via the PROC PHREG data procedure in SAS. The proportional hazards assumption was tested using Schoenfeld residuals, with non-violation confirmed by assessing their association with time via linear regression. A 2-sided p-value <0.05 was considered statistically significant.
The longitudinal changes in pain intensity, pain interference, depression, anxiety, sleep disturbance, and physical function for 52 weeks postsurgery were analyzed in a mixed model approach to linear regression for repeated measurements. These linear mixed-effects models, implemented via the PROC MIXED procedure in SAS, included time after surgery (weeks), preoperative opioid misuse status, and the interaction of these 2 factors as fixed explanatory variables. A random intercept was included to account for within-subject dependence, as individuals are assessed repeatedly over time. Analyses were carried out with all data from week 1 to week 52 for postoperative pain, reported as both current and average pain intensity at the surgical site over the past 24 hours on the BPI, and postoperative pain interference on the BPI. Scores were consolidated to months 1, 3, 6, 9, and 12 using a last observation carried forward approach for NIH PROMIS depression, anxiety, sleep disturbance, and physical function T scores. Optimal model fit and variance-covariance matrix structure was determined by comparing Akaike Information Criteria scores.
Missing data for linear effects models were handled in 2 ways. First, results were reported solely with the maximum likelihood estimation algorithm inherent to PROC MIXED implicitly accounting for missing data by including all available data in the analysis. Second, multiple imputation using the fully conditional specification method in SAS was conducted to create 20 complete data sets to impute missing postoperative current and average pain intensity, pain interference, and NIH PROMIS depression, anxiety, sleep disturbance, and physical function scores at each time point. Then combined imputed results were derived from 20 complete data sets with PROC MIANALYZE.
RESULTS
1. Demographics
A total of 52 participants undergoing spinal surgery completed presurgical and longitudinal assessments and were included in the analysis. Baseline characteristics and clinical features are summarized in Table 1, stratified by preoperative opioid misuse status.
Participants with and without preoperative opioid misuse were similar in age (56.2±21.7 years vs. 60.4±21.9 years, p=0.6) and sex distribution (46.2% vs. 31.8% male, p=0.3). Racial and ethnic diversity was limited, with 85% (n=33) of the COMM-Negative group identifying as Caucasian, while all patients in the COMM-Positive group (n=13, 100%) were Caucasian and non-Hispanic. Among sociodemographic factors, only annual income differed significantly, with a higher proportion of the COMM-Positive group reporting an income below $80,000.
Surgical characteristics were comparable between groups, with most procedures targeting the lumbar spine and involving single-level fusions. There was no difference between groups for the presence of metastatic disease or malignant tumor; and operative factors including instrumentation, fusion, operative time, and estimated blood loss. Additional details regarding primary surgical indication are listed in the supplement (Supplementary Table 1). Baseline opioid use was more prevalent in the COMM-Positive group (54% vs. 23%) but did not differ significantly between groups. Patients with preoperative opioid misuse reported significantly higher baseline pain interference (6.8±1.7 vs. 4.3±2.5, p=0.002) and lower physical function (32.4±4.0 vs. 36.8±9.7, p=0.03). Additionally, they exhibited significantly greater depressive and anxiety symptoms across multiple presurgical psychological measures, including the PHQ-9 and NIH PROMIS scores.
2. Opioid Use Outcomes
A total of 1,097 postoperative observations were analyzed. Forty-one participants returned to their presurgical opioid use levels, and 38 achieved opioid cessation (Table 2). Median time to baseline opioid dose was longer in the preoperative opioid misuse group (43 days; interquartile range [IQR], 27–74 days) compared to those without opioid misuse (14 days; IQR, 14–28 days) (Fig. 1A). Median time to opioid cessation was also prolonged in the positive COMM-Positive group (88 [IQR, 49–228] days vs. 30 [IQR, 28–42] days) (Fig. 1B). In multivariable analysis adjusting for age, sex, operative region, number of spinal levels, and any preoperative opioid use, a positive preoperative COMM-17 score was associated with a 65% reduction in the rate of return to baseline opioid dose (p=0.02) and a 75% reduction in the rate of opioid cessation (p=0.008). No significant associations were found between preoperative opioid misuse and time to pain cessation or patient-reported recovery.
3. Pain Trajectories
Weekly trajectories for postoperative current and average pain intensity and pain interference are shown in Fig. 2A–C, respectively. Both groups demonstrated a decrease in current pain intensity (slope=-0.03) and average pain intensity (slope=-0.4) over time, with no significant differences in initial intensity. At week 1, patients with preoperative opioid misuse reported similar current pain intensity (5.07; 95% CI, 4.00–6.15) to those without opioid misuse (4.37; 95% CI, 3.96–4.77). By week 52, pain levels were minimal in both the positive and negative preoperative COMM-17 groups, with current pain intensities of 0.78 (95% CI, -0.15 to 1.72) and 1.60 (95% CI, 1.18–2.03), respectively.
Average pain intensity was initially higher in the COMM-Positive group (7.15; 95% CI, 6.26–8.04) compared to the COMM-Negative group (4.90; 95% CI, 4.59–5.22). Pain scores converged by week 7 but began to diverge again around week 48, with the COMM-Positive group reporting slightly higher average pain intensity at week 52 (3.27; 95% CI, 2.55–4.00) compared to the COMM-Negative group (1.93; 95% CI, 1.64–2.23).
Pain interference was significantly higher in the preoperative opioid misuse group 1 week after surgery (7.0 [95% CI, 6.0–8.1] vs. 5.2 [95% CI, 4.8–5.7], p=0.009). Pain interference was severe at week 1 but declined to mild levels by week 11 in those without preoperative opioid misuse; patients with preoperative opioid misuse continued to report moderate pain interference until week 48, after which interference dropped to minimal levels [38]. On average, both groups demonstrated a decrease in pain interference over time (slope=-0.03, p=0.0002) with clinically significant reductions in pain interference one year after surgery.
Sensitivity analysis with multiple imputation yielded comparable results for trajectories of postoperative current and average pain intensity (Supplementary Table 2). However, there was no significant difference in pain interference between groups one week after surgery with imputation. The significant decrease in pain interference over time in both groups was also found when examining the multiple imputation results.
4. Psychological and Functional Outcomes
Trajectories for NIH PROMIS scores are shown in Fig. 3A–D. At postoperative month 1, patients without preoperative opioid misuse exhibited similar postoperative NIH PROMIS depression scores compared to baseline (46.7; 95% CI, 45.3–48.1) while patients with preoperative opioid misuse exhibited a clinically significant reduction in scores (50.1; 95% CI, 47.7–52.5). Patients with preoperative opioid misuse exhibited a rebound in depressive symptoms at 3 months which continued over the following year. At month 12, patients with preoperative opioid misuse re-ported depressive symptoms (54.1; 95% CI, 51.5–56.9) comparable to baseline levels, and those without preoperative opioid misuse (47.6; 95% CI, 46.2–49.1) also reported symptoms similar to baseline. On average, Depression scores remained constant over time in both groups.
Postoperative anxiety symptoms 1 month after surgery were reduced in both patients with and without preoperative opioid use respectively (52.8 [95% CI 50.1–55.6] vs. 46.7 [95% CI 45.1–48.3]). Similar to depressive symptoms, patients with preoperative opioid misuse exhibited a clinically significant reduction in anxiety 1 month after surgery with a rebound in symptoms at 3 months while patients without preoperative opioid misuse exhibited similar NIH PROMIS anxiety scores compared to baseline. At Month 12, COMM-Positive patients reported anxiety scores (56.0; 95% CI, 52.9–59.1) comparable to baseline levels, and COMM-Negative patients also reported anxiety scores (49.4; 95% CI, 47.8–51.1) that were similar to baseline.
Physical function scores at month 1 were similar between positive and negative preoperative COMM-17 participants (31.0 [95% CI, 29.1–32.8] vs. 31.6 [95% CI, 30.5–32.7]), comparable to the impaired physical function reported by patients with chronic lumbar spine pain [39]. After 12 months, both positive and negative preoperative COMM-17 groups reported improvements in physical function scores, but the scores were comparable to baseline (32.1 [95% CI, 29.5–34.7] vs 37.3 [95% CI, 35.9–38.7]). On average, both groups demonstrated increasing physical function over time (slope=0.47, p<0.0001).
Baseline sleep disturbance scores were comparable among participants with and without preoperative opioid use (50.9 [95% CI, 47.3–54.6] vs. 51.2 [95% CI, 49.1–53.2]). At 12 months, participants with and without preoperative opioid use reported scores that were comparable to preoperative levels of sleep disturbance (56.3 [95% CI, 53.3–59.1] vs. 52.1 [95% CI, 47.3–50.6]).
Sensitivity analysis with multiple imputation yielded similar results for trajectories of postoperative depression, physical function, and sleep disturbance (Supplementary Table 1). There was a moderately significant increase in postoperative anxiety scores across both groups over time with imputation. (slope=0.18, p=0.04).
DISCUSSION
Our study emphasizes preoperative opioid misuse, as measured by the COMM-17, as a significant independent risk factor for both delayed return to baseline opioid dose and delayed postoperative opioid cessation. Our adjusted hazard models accounted for any preoperative opioid use, which was not a significant independent risk factor when considered alongside preoperative opioid misuse. These findings suggest that prior asso-ciations between preoperative opioid use and PPOU may have been confounded by opioid misuse. This distinction is particularly relevant for targeted screening and risk reduction strategies. By differentiating opioid misuse from general opioid use, healthcare providers can more effectively allocate resources and interventions to prevent prolonged opioid use after surgery, mitigate opioid misuse, and reduce the risk of transitioning to OUD.
Previous studies have identified preoperative opioid use as a predictor of persistent opioid use postsurgery. A retrospective cohort study of 583 patients undergoing elective spine surgery for cervical or lumbar degenerative pathology found that any opioid prescription within 60 days before surgery was associated with opioid use at 6 months postoperatively [40]; however, no other dimensions of opioid use, including opioid misuse were captured. Similarly, a machine learning approach examining persistent opioid use 3 months after major spine surgery identified preoperative opioid use as one of several factors found to have the highest impact on the identification of PPOU [41]; although the study considered substance use, including active alcohol use, smoking history, buprenorphine history, and active marijuana use, preoperative opioid use was uni-dimensionally characterized as preoperative chronic opioid use [41]. A meta-analysis of 12 studies involving 80,935 lumbar fusion surgery patients reported a nearly sixfold increase in the odds of continued opioid refills 3 months postoperatively among those with preoperative opioid use, and this risk factor was determined to have the strongest association with prolonged opioid use after surgery; however, none of these studies examined preoperative opioid misuse as a distinct variable [25].
To our knowledge, our study is the first to explicitly characterize the impact of preoperative opioid misuse, as assessed by the COMM-17, on postoperative opioid use in spine surgery patients. Our findings suggest that future research is needed to further delineate the distinct risks associated with preoperative opioid misuse to optimize screening and targeted risk reduction interventions.
Postoperative surgical pain intensity trajectories were similar between patients with and without preoperative opioid misuse. Both groups reported comparable and significant reductions in pain over time, suggesting that patients reporting preoperative opioid misuse can still experience clinically significant reductions in pain intensity and pain interference after spine surgery. However, our survival analyses for pain cessation and patient-reported recovery were limited by the small percentage of patients reaching these endpoints. Future research is needed to further characterize the association between preoperative opioid misuse and persistent postsurgical pain.
Postoperative mental health trajectories of depression and anxiety symptoms demonstrated transient, clinically significant improvements in depression and anxiety symptoms among patients with preoperative opioid misuse one month after surgery, while patients with no preoperative opioid misuse report levels of anxiety and depression that were similar to their baseline levels. COMM-Negative patients reported levels of depression and anxiety below the average score for the United States (US) general population (T score=50), while COMM-Positive patients reported near normalization of depression and anxiety symptoms. Among COMM-Positive patients, the median time to baseline opioid dose was 43 days after surgery, and the median time to opioid cessation was 88 days after surgery. Thus, initial elevations in daily postoperative opioid dose may have provided transient improvements in mood among patients with preoperative opioid use, which serves to reinforce continued opioid use after surgery. At 3 months, after the majority of COMM-Positive patients discontinued opioids, patients experienced a rebound of worsened mood. As trajectories of pain intensity and pain interference demonstrated more rapid and sustained improvements after surgery, the mechanisms of this rebound in negative affect may be less linked to pain, but rather the withdrawal of opioids.
Opioid withdrawal is associated with a negative affective state. Preclinical models demonstrate that opioid withdrawal leads to heightened anxiety responses and despair-like behaviors among mice receiving chronic opioids [42]. In a cross-sectional study of 404 patients receiving long-term opioid therapy for chronic non-cancer pain, moderate to severe withdrawal symptoms were associated with anxiety, depression, and longer duration of opioid therapy [43]. Research has also demonstrated an association between discontinuation of long-term opioid therapy and increased risks of both overdose and suicide. A study of 1,394,102 US Veterans demonstrated death rates for overdose and suicide increased immediately after stopping prescription opioids with the incidence gradually decreasing over three to 12 months [44]. In a single-center study of 280 patients undergoing spine surgery, 33% of participants recalled experiencing one or more withdrawal symptoms related to opioid tapering at 3 months after surgery [45]. Our study adds to these findings capturing the marked rebound in negative affect 3 months after surgery among patients with preoperative opioid misuse undergoing spine surgery. Targeted interventions to reduce negative affect and treat opioid withdrawal may facilitate opioid dose reduction and tapering among patients undergoing spine surgery.
Longitudinal postoperative mood assessments among patients undergoing surgery reflect the trajectories of patients in our study without preoperative opioid misuse [46]. For example, a study of 1,771 patients undergoing elective surgical procedures reported relatively stable average PROMIS anxiety T scores at 1 month (45.6), 3 months (45.3), and 6 months (45.2) postoperatively with T scores less than the average for the US general population [47]. A study of 1,168 patients undergoing surgery with a duration of greater than one hour and overnight hospital admission, were assessed thrice-daily for postoperative anxiety and depression symptoms, but follow-up to was limited to 30 days after surgery [48]. The study did not find an association between preoperative opioid use and postoperative anxiety or depression, but results may have differed from our study due to the limited longitudinal follow-up which could not identify potential rebound points in the longer-term postoperative mental health trajectories. Our findings underscore the importance of higher resolution longitudinal monitoring in the early, subacute, and chronic postoperative periods to better understand the dynamic evolution of postoperative anxiety and depression and their relationship to preoperative opioid use and misuse.
Several limitations must be acknowledged. First, the small cohort size (n=52) coupled with limited racial and ethnic diversity (predominantly Caucasian participants), may restrict the generalizability of our findings to other populations. However, the observed prevalence of 25% screening positive for preoperative opioid misuse using the COMM-17 in this elective spine surgery cohort is concerning and warrants attention, as this is markedly higher than estimates of prescription opioid misuse in the general US adult population [49]. This finding, highlights the potential for patients presenting for spine surgery to constitute a particularly high-risk group for preoperative opioid misuse and provides novel, albeit preliminary, insights into their unique postoperative challenges, emphasizing the importance of targeted screening.
Second, the low percentage of participants achieving the pain cessation and patient-reported recovery endpoints during the study (21.1% and 30.8%, respectively) restricted our ability to comprehensively assess the impact of preoperative opioid misuse on persistent postsurgical pain and patient-reported recovery outcomes.
Third, while weekly data collection commenced at postoperative week 1 and continued for 1 year, the frequency of assessment may have constrained our ability to capture the more granular fluctuations in pain and psychological symptoms, particularly in the immediate postoperative period during the patient’s hospitalization. Furthermore, although valuable insights were gained over the 1-year observation period, a longer follow-up (e.g., ≥2 years) would strengthen the evidence by capturing later phase recovery patterns; however, beyond 1 year, the likelihood of confounding variables increases significantly, such as the development of new or unrelated pain conditions or patients undergoing subsequent surgical procedures, which could independently influence patient-reported outcomes and opioid use patterns and serve as competing risks in survival analyses.
CONCLUSION
This study highlights preoperative opioid misuse as a significant independent risk factor for prolonged elevations in postoperative opioid dose and delayed opioid cessation after elective spine surgery, while preoperative opioid use was not associated with either outcome. Thus, preoperative opioid misuse, rather than opioid use alone, may be the key risk factor for PPOU, emphasizing the need for targeted preoperative screening and intervention strategies. Moreover, patients with preoperative opioid misuse experienced transient normalization in depression and anxiety symptoms with a rebound by postoperative month 3, after the majority of patients had reached a return to their preoperative opioid dose; this suggests worsening mood may be a response to postoperative opioid withdrawal. Future development of targeted behavioral or pharmacological strategies to limit the rebound in negative affect may mitigate the impact of preoperative opioid misuse on PPOU.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Supplement
Print
