Background: Persistent postoperative opioid use is a rising concern that may lead to higher risks of
opioid dependence and related adverse outcomes. With smaller incisions, minimally invasive surgery is
considered related to less postoperative pain and improved recovery. However, there is limited research
addressing persistent opioid use after minimally invasive surgery. In this study, we aimed to compare
persistent postoperative opioid use between traditional surgery and minimally invasive surgery using data
retrieved from the National Health Research Database in Taiwan.
Methods: From 2004–2016, a total of 121,127 patients who underwent herniorrhaphy; hepatectomy;
cholecystectomy; appendectomy; colorectal, gastric, gynecological, thoracic or renal surgeries with either
traditional or minimally invasive approaches were enrolled. We stratified the participants by surgery
type and compared the incidence of persistent opioid use over 90 days after surgery between the two
approaches. The adjusted odds ratios (aORs) with 95% confi dence intervals (CIs) were estimated using
multivariable logistic regression analysis.
Results: After adjusted for age, sex, hospital level, and comorbidities, traditional approach was
associated with higher odds for persistent postoperative opioid use than minimally invasive approach in
herniorrhaphy (aOR, 3.00; 95% CI, 1.08–8.37); cholecystectomy (aOR, 1.86; 95% CI, 1.38–2.52); gastric
(aOR, 1.91; 95% CI, 1.56–2.34), thoracic (aOR, 1.47; 95% CI, 1.09–1.99), and renal surgeries (aOR, 1.83;
95% CI, 1.13–2.98).
Conclusion: The minimally invasive approach was independently associated with a lower risk of
persistent postoperative opioid use after herniorrhaphy; cholecystectomy; gastric, thoracic, and renal
surgery compared to the traditional approach.

Keywords: analgesics, laparoscopic surgery, minimally invasive surgery, persistent postoperative opioid use,
thoracoscopic surgery, traditional surgery

Opioids are crucial components of pain treatment regimens, particularly for the treatment of acute pain following surgery. However, opioid usage can lead to side effects such as sedation, dizziness, nausea, vomiting, constipation, respiratory depression, and tolerance [1]. Persistent postoperative opioid use is commonly defined as the prolonged use of opioids 90 days after surgery in opioid-naïve patients or any increase in opioid dose after surgery in previous users [2]. It has become a growing concern recently, with a significant increase in hospital utilization and a higher risk of developing chronic pain, mental health disorders, and adverse outcomes. Long-term opioid consumption raises concerns regarding the increasing prevalence of opioid misuse and addiction [3,4]. Therefore, both identifying patients at risk and developing strategies to prevent their development are crucial for improving patient outcomes and reducing the healthcare burden.

Risk factors for patients to develop prolonged opioid use after surgery include sex, substance use, and certain underlying diseases [5,6]. The incidence varies depending on the type of surgery performed [7]. The highest rate of persistent opioid use was observed after thoracotomies and lumbar laminectomies. Minimally invasive surgery, associated with smaller incisions and less tissue trauma, usually results in less acute postoperative pain, faster recovery, and reduced chronic pain [8,9]. Consequently, patients who undergo minimally invasive surgery may have a lower incidence of persistent postoperative opioid use than those who undergo traditional open surgery. However, studies comparing persistent opioid use after a traditional surgical approach with that after a minimally invasive surgical approach are limited.

In this study, we hypothesized that the traditional surgical approach may be associated with increased persistent postoperative opioid use compared to minimally invasive surgery. We aimed to investigate and compare the incidence of persistent opioid use in patients undergoing traditional surgery with that in patients undergoing minimally invasive surgery by analyzing a nationwide longitudinal database of 2 million beneficiaries in Taiwan.

The data analyzed in this study were retrieved from the National Health Research Database (NHIRD). Taiwan launched a single-payer National Health Insurance (NHI) program on March 1, 1995. More than 99% of the population in Taiwan was covered by this program. The NHIRD is a database released for public research that provides encrypted patient identification numbers, sex, date of birth, date of admission, details of prescription filled, as well as the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes of diagnoses and procedures. We analyzed data corresponding to 2004–2016 and collected from a longitudinal database containing the claims data of two million individuals randomly sampled from the 23 million beneficiaries of the NHIRD.

Since all types of personal identification were encrypted for patient privacy, this study was granted an exemption from full ethical review, and the requirement for obtaining patient consent was waived by the Institutional Review Board of the Chi Mei Medical Centre (IRB-10812-E01). The authors comply with the Declaration of Helsinki (2013), relevant national laws, and regulations.

This retrospective population-based cohort study included patients who underwent surgery from 2005 to 2015. The enrollees have been traced backward for one year for underlying conditions to identify the occurrence of primary and secondary outcomes. Our objective was to compare the incidence of persistent postoperative opioid use between patients who underwent traditional surgery and those who underwent minimally invasive surgery. We targeted thoracic and abdominal surgeries that could be approached through a traditional open incision or less-invasive small incisions, with the assistance of videoscopy. By examining the associated ICD-9 codes, thoracic surgeries such as video-assisted thoracoscopic surgery (VATS) and thoracotomy, and abdominal surgeries such as cholecystectomy; appendectomy; herniorrhaphy; hepatic, gastric, colorectal, renal, and gynecologic surgeries, that could be performed by employing laparotomy and/or laparoscopy were identified. The corresponding ICD-9-CM codes are listed in Table A1 in the Appendix.

The exclusion criteria were as follows: (1) Patients who had intraoperative conversion from laparoscopy to laparotomy or conversion from VATS to open thoracotomy, (2) Multiple surgeries during a single admission, (3) Opioid or non-opioid analgesic users for longer than one month prior to surgery (to minimize the influence of underlying opioid dependence or chronic pain status), (4) Advanced cancer patients under radiotherapy or chemotherapy (to eliminate any potential impact of surgical pain on recovery), and (5) Patients who died within 180 days. (Figure 1)

Patients were classified into four age categories: < 30 years, 30–44 years, 45–59 years, and ≥ 60 years. Comorbidities were determined from the medical claims records that were documented one year before surgery such as dysthymic disorder (ICD-9-CM code 300.4), peripheral vascular diseases (ICD-9-CM codes 443.8–444.9), osteoporosis (ICD-9-CM codes 733.0), gout (ICD-9-CM code 274), malignancy (ICD-9-CM codes 140–208), headache (ICD-9-CM codes 307.81, 784.0, 346), diabetic neuropathy (ICD-9-CM codes250.6, 357.2), rheumatoid arthritis (ICD-9-CM code 714), and pressure ulcer (ICD-9-CM code 707). The hospital accreditation levels were also considered. According to the Taiwan Joint Commission on Hospital Accreditation, a local community hospital is a hospital with fewer than 30 beds that only provides primary medical services. Regional hospitals are hospitals that have 301–999 beds and provide secondary medical services. Medical centers have 1,000–2,500 beds and are responsible for most of the staff-training burden and tertiary medical services.

Figure 1. Flow Chart for the Selection Process of the Study Sample

The primary outcome was persistent postoperative opioid use for over 90 days after surgery, and the secondary outcome was the persistent use of all types of analgesics. Patients who filled one or more prescriptions for opioids (morphine, fentanyl, oxycodone, buprenorphine, hydromorphone, tramadol, codeine, or meperidine) from outpatient clinics from 90 to 120 days post-surgery were defined as persistent postoperative opioid users. Similarly, patients who filled prescriptions for all analgesics, including opioids and non-opioid painkillers (acetaminophen, nonsteroidal anti-inflammatory drugs, phenytoin, carbamazepine, pregabalin, and gabapentin), from outpatient clinics from 90 to 120 days post-surgery were defined as persistent analgesic users. The corresponding Anatomical Therapeutic Chemical codes are presented in Table A2. Furthermore, we identified persistent opioids or all kinds of analgesic use for more than 180 days (from 180 to 210 days) after surgery in a sensitivity analysis to determine whether there were similar results under different definitions.

The demographic characteristics of the target population, including age groups, sex, hospital level, and preexisting comorbidities, are presented as frequencies with percentages (Table 1). Potential confounding factors were included as covariates in the predicted model to evaluate the differences in the incidence of persistent opioid or analgesic use between the minimally invasive and traditional approaches for a similar surgery. Multivariate logistic regression was used to compare the likelihood of persistent opioid and analgesic use between the two groups based on estimated adjusted odds ratios (aORs) and 95% confidence intervals (CIs). All statistical analyses were performed using the SAS version 9.4 software (SAS Institute Inc., Cary, NC, USA).

A total of 121,127 patients who underwent targeted surgery between 2005 and 2015 were enrolled. Among these patients, 53,333 underwent traditional surgery, and 67,794 underwent minimally invasive surgeries, including thoracoscopic and laparoscopic surgeries. The demographic characteristics of the patients enrolled in this study and the distribution of surgery types in the sampled population are summarized in Table 1. Most patients received surgery in medical centers (50.95%) or regional hospitals (42.69%). Only 6.36% of cases were from local community hospitals. We stratified the patients based on the type of surgery and tracked filled pharmacy claims for opioids and all types of analgesics for one year after surgery. The incidence of persistent opioid and analgesic use over 90 and 180 days postoperatively are presented in Tables 2 and 3, respectively.

The highest incidence of persistent postoperative opioid use > 90 days after surgery was found after thoracic surgery (3.26%), followed by hepatectomy (2.80%) and colorectal surgery (2.00%), whereas the incidence of prolonged opioid use > 180 days post-surgery was 2.80%, 2.70%, and 2.30% after thoracic surgery and colorectal surgery, respectively. Among all surgical procedure subtypes, traditional thoracotomy was associated with the highest rate of persistent postoperative opioid use (4.66% over 90 days post-surgery and 3.33% over 180 days post-surgery). Significantly higher odds of persistent opioid use over 90 days after surgery was observed in the traditional approach than minimally invasive approach, including thoracic surgery (aOR, 1.47; 95% CI, 1.09–1.99), herniorrhaphy (aOR, 3.00; 95% CI, 1.08–8.37), cholecystectomy (aOR, 1.86; 95% CI, 1.38–2.52), gastric surgery (aOR, 1.91; 95% CI, 1.56–2.34), and renal surgery (aOR, 1.83; 95% CI, 1.13–2.98) after adjustment for age, sex, hospital level, and comorbidities. Over 180 days after surgery, significantly higher odds of persistent opioid use were still found in certain types of traditional surgery, including cholecystectomy (aOR, 2.00; 95% CI, 1.49–2.67), gastric surgery (aOR, 1.83; 95% CI, 1.50–2.22), and renal surgery (aOR, 2.23; 95% CI, 1.34–3.73) after adjustment for other covariates.

As for persistent postoperative analgesic use over 90 days, the highest overall incidence was found after renal surgery (27.52%), followed by thoracic surgery (26.26%) and cholecystectomy (26.00%). Over 180 days after surgery, the incidence of persistent analgesic use was 26.32% after renal surgery, followed by cholecystectomy (25.52%) and gynecological surgery (24.55%). Among all the subtypes of surgical procedures, traditional thoracic surgery was associated with the highest rate of persistent analgesic use over 90 days after surgery (29.5%), followed by traditional renal surgery (28.93%). Patients who underwent traditional renal surgery (27.22%) and traditional thoracic surgery (25.22%) also had the highest rates of persistent analgesic use over 180 days postoperatively. After adjustment for age, sex, hospital accreditation level, and comorbidities, certain traditional surgeries were associated with significantly higher odds of persistent analgesic use over 90 days postoperative, including thoracic surgery (aOR, 1.18; 95% CI, 1.03–1.35) and renal surgery (aOR, 1.33; 95% CI, 1.14–1.57) than minimally invasive approach. Higher odds of analgesic use were still noted180 days after traditional renal surgery (aOR, 1.25; 95% CI, 1.06–1.47). Over 180 days post-surgery, significantly higher odds of analgesic use were observed for herniorrhaphy (aOR, 1.22; 95% CI, 1.05–1.41), hepatectomy (aOR, 1.39; 95% CI, 1.06–1.84), and colorectal surgery (aOR, 1.30; 95% CI, 1.02–1.65) after adjustment for potential confounders.

In this retrospective cohort study, we found that opioid-naïve patients undergoing a traditional surgical approach were associated with higher odds for persistent postoperative opioid use than that of patients undergoing a minimally invasive approach in herniorrhaphy (aOR, 3.00; 95% CI, 1.08–8.37); cholecystectomy (aOR, 1.86; 95% CI, 1.38–2.52); thoracic (aOR, 1.47; 95% CI, 1.09–1.99), gastric (aOR, 1.91; 95% CI, 1.56–2.34), and renal surgeries (aOR, 1.83; 95% CI, 1.13–2.98) after adjusted for age, sex, hospital level, and comorbidities. Similar results were obtained when the definition of persistent postoperative opioid use was extended from > 90 days to > 180 days after surgery, except in cases of herniorrhaphy.

Persistent postoperative opioid use is an increasing concern related to substance dependence, overdose, disability, and even death. It is recognized as one of the most common postoperative complications in the U.S. [10]. The incidence varies, depending on the type of surgery performed. Stark et al. [11] described that 13.7% of patients after orthopedic surgery were found persistent opioid use, which could be up to 23.6% in those after spinal surgery. A systematic review that focused on cardiac surgery reported that the pooled rate of persistent postoperative opioid use in opioid-naïve patients was 5.7% [5]. For major surgeries, including open and minimally invasive surgeries, a population-based cohort study in Canada reported the incidence of persistent opioid use to be about 3% [12]. This is compatible with the incidence reported in our study, from 3.26% after thoracic surgery to 0.44% after gynecological surgery, but still much lower than the incidence of 5.9% to 6.5% reported after major surgery in the U.S. [13]. The discrepancy in persistent opioid use may stem from the relatively conservative narcotic policy in Taiwan, where the use of opioids is strictly regulated and reviewed closely by the authorities. In general, opioid prescriptions in Taiwan are much lower than those in high-opioid-consumption countries, such as the United States, Canada, and United Kingdom [14]. Although persistent opioid use may not necessarily be due to surgical pain, the incidence reported in our study aligns with the incidence of severe chronic pain described in a multicenter study conducted in Europe [15].

Thoracic surgery is a high-risk procedure for the development of moderate-to-severe chronic pain [16,17], which makes it a possible predisposing factor for prolonged opioid use. In our study, patients who underwent thoracic surgery reported the highest rate of persistent postoperative opioid use more than 90 days after surgery (4.66% after thoracotomy and 2.93% after thoracoscopic surgery) but still lower than the rates reported in Canada and the United States [12,13,18]. In addition to narcotic policies in different countries, the definition of persistent postoperative opioid use may contribute to a wide range of incidences [2]. Other factors may also have played a role. A retrospective study among veterans in the United States reported prolonged opioid use one year after surgery in 4.7% of patients without preoperative chronic opioid use, psychosocial diagnoses, medication use, or chronic pain [19]. Compared to thoracotomy, VATS causes less tissue damage, intercostal nerve injury, and acute pain [20]. While the protective effect of VATS on the development of chronic pain remains controversial [21,22], a randomized controlled trial suggested that VATS was associated with less postoperative pain during a 52-week follow-up [8]. Clarke et al. [12] found that thoracotomy was associated with 2.58 times higher odds of prolonged postoperative opioid use than open prostatectomy, whereas VATS was associated with 1.95 times higher odds. Another observational study using SEER-Medicare database compared VATS with thoracotomy directly. It found that VATS was associated with 0.75 odds of persistent opioid use for 3–6 months postoperatively than thoracotomy in a multivariable model [23]. In our study, thoracotomy was associated with higher odds of persistent opioid use for more than 90 days (aOR, 1.47; 95% CI, 1.09–1.99) postoperatively after adjusted for other covariates. The incidence and odds of persistent analgesic use for more than 90 days postoperatively were also significantly higher in patients who underwent thoracotomy as compared with those who underwent a minimally invasive approach (29.51% vs. 25.49%, P-value = 0.0025; aOR, 1.18; 95% CI, 1.03–1.35).

The literature on chronic pain after laparoscopic or open abdominal surgery is inconsistent [24]. There is scant evidence comparing laparoscopic surgery with the traditional approach and the occurrence of persistent postoperative opioid use. We found significantly higher odds of persistent opioid use for > 90 days after herniorrhaphy, cholecystectomy, gastric surgery, or renal surgery. As a common postoperative complication, chronic pain occurs more frequently after an open approach than after the laparoscopic technique [3,25,26]. We found a significantly higher incidence of opioid use for > 90 days and analgesic use for > 180 months after traditional herniorrhaphy than after a minimally invasive approach. This might imply the existence of chronic pain with gradually decreasing intensity from 90 days to 180 days after surgery. Regarding gastric surgery, elevated chronic opioid use after bariatric surgery, including both traditional and laparoscopic approaches, compared with non-surgical controls, was mentioned in a retrospective study [27]. The rate of persistent opioid use after traditional gastric surgery was significantly higher than that after the minimally invasive approach after 90 and 180 days with higher odds after adjusting for other covariates. For renal surgery, although Alper and Yüksel [28] demonstrated an equal risk of developing chronic pain in patients after laparoscopic or open nephrectomy, we found that traditional renal surgery was associated with higher odds of prolonged opioid and analgesic use > 90 and 180 days after surgery. Theoretically, a smaller surgical wound minimizes the extent of tissue injury and intensity of acute pain, which may mitigate the development of chronic pain [25]. However, not all minimally invasive surgical approaches are associated with a significantly lower incidence of persistent opioid or analgesic use. This inconsistency suggests that chronic pain could be multifactorial and that incisional pain is far from sufficient to predict persistent opioid or analgesic use.

One strength of this study was its nationwide sampling design. By analyzing the data of randomly sampled 2 million beneficiaries from the NHIRD, it is less likely that the study is biased due to specific healthcare units or practitioners. Another strength is the high coverage rate ensured by Taiwan NHI system, which is a single-payer program offering comprehensive medical care to more than 99% of Taiwan’s legal residents [29]. Thus, the loss to follow-up rate within one year after surgery was very low in this longitudinal database.

However, this study had some limitations. First, the inherent limitations associated with a retrospective observational study design cannot be overlooked. We could only observe and adjust for recorded covariates. Second, the accuracy of the records may be a concern. The possibility of coding errors cannot be excluded. After being reviewed and validated by auditors, most of the codes were confirmed to be accurate. Patients with missing or incomplete data were excluded from further analyses. Third, prescription claims may not necessarily reflect the exact opioid use. However, in Taiwan, long-term opioid use is strictly regulated by the relevant authorities. Patients who used opioids for > 14 days were registered with the National Bureauof Controlled Drugs. Each patient was assessed by the hospital’s opioid committee, and the study was approved by the Taiwan Food and Drug Administration. The duration of each opioid prescription is limited, and long-term opioid therapy should be re-evaluated and recorded for surveillance at least every four months. For patients with aberrant behavior, physicians must report to the hospital’s opioid committee to discontinue opioid treatment [30,31]. This also explains the low incidence of persistent postoperative use in the present study. Consequently, filled opioid prescription claims may reflect persistent opioid use. Fourth, persistent use of opioids or analgesics does not necessarily equal chronic post-surgical pain. Although we cannot claim that all the filled prescriptions of analgesics were absolutely for post-surgical pain, we have tried our best to minimize the influence of pain from other sources. Finally, the data obtained from the nationwide database in Taiwan, may not be directly generalizable to other countries due to differences in the composition of ethnicity and health insurance systems. Nevertheless, this study provides an overall picture of the comparison of persistent postoperative opioid use between traditional and minimally invasive approaches stratified by different surgeries in Taiwan.

In conclusion, traditional approaches to certain surgeries were associated with a higher incidence of persistent postoperative opioid use over 90 days after surgery than minimally invasive approaches, including thoracic surgery, herniorrhaphy, cholecystectomy, gastric surgery, and renal surgery. While the mechanism for persistent postoperative opioid use could be multifactorial, our findings suggest that different approaches for certain surgeries could be one of the factors. More research would be required to clarify the causal relationship between surgical approach and persistent opioid use further and supplement the strategies in the prevention of prolonged opioid use.

This study was based on data from the Taiwan National Health Insurance Research Database provided by the Taiwan Bureau of National Health Insurance, Taiwan Department of Health, Taipei, and managed by the National Health Research Institutes.

This work was funded by Chi Mei Medical Center (grant number: CMFHR10930).