Article Text
PDF
Abstract
Introduction Socioeconomic status affects the treatment of patients with low back pain and/or neck pain. We examined the relationship between socioeconomic status (occupation and household income level) and treatments such as chronic opioid use and interventional procedures among these patients.
Methods Data from the National Health Insurance Service database in South Korea were used in this population-based cross-sectional study. Approximately 2.5% of adult patients diagnosed with low back pain and/or neck pain between 2010 and 2019 were selected using a stratified random sampling technique and included in the analysis.
Results We analyzed the data of 5,861,007 patients with low back pain and/or neck pain in total. Among them, 4.9% were chronic opioid users and 17.7% underwent interventional procedures. Healthcare workers and unemployed individuals had 18% lower and 6% higher likelihood of chronic opioid use compared with office workers, respectively. Those with a very low household income had 18% higher likelihood of chronic opioid use than those with a poor household income. Other workers and unemployed individuals had 4% and 8% higher likelihood of undergoing interventional procedures than office workers, respectively. Healthcare workers had 5% lower likelihood of undergoing interventional procedures than office workers. Patients with middle, high, and very poor household incomes had a higher likelihood of undergoing interventional procedures, while those in the very high household income group had a lower likelihood of undergoing interventional procedures than those with poor household incomes.
Conclusions Socioeconomic status factors are associated with treatment in patients with low back pain and/or neck pain.
- analgesics, opioid
- back pain
- neck pain
- pain management
- chronic pain
Data availability statement
Data are available on reasonable request. Data are available on reasonable request to corresponding author.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
The treatment of low back pain (LBP) and/or neck pain (NP) is costly, which could result in financial burden.
WHAT THIS STUDY ADDS
The number of opioid users and patients who underwent interventional procedures for the treatment of LBP and/or NP increased from 2010 to 2019 (over 10 years). Socioeconomic status (SES) factors, such as occupation, household income level, and residence, were associated with treatment among these patients.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This is the first study to show that SES affects the treatment of patients with LBP and/or NP.
Introduction
Low back pain (LBP) and neck pain (NP) are critical public health problems that have been the leading causes of disability for several years.1 2 Approximately one-third of the adult population in the USA has experienced LBP and/or NP.3 Approximately half of the patients who have LBP and/or NP report pain and disability that require healthcare.4 Thus, LBP and NP are significant and important public health issues.
Disparity in healthcare quality according to socioeconomic status (SES) has been emphasized as an important public health issue.5 Previous studies have shown that SES factors are associated with healthcare utilization in various healthcare systems.6–8 A cohort study in China reported that body pain is influenced by SES factors, such as occupation and education level.9 A cross-sectional study in Japan reported significant socioeconomic inequalities in LBP among older individuals.10 SES might affect treatment in patients with LBP and/or NP because the treatment of LBP and/or NP is costly, which could result in financial burden.11 12 However, information regarding the relationship between SES and the treatment of LBP and/or NP is still lacking.
We hypothesized that SES factors might affect treatment, such as chronic opioid use and receipt of intervention procedures, among patients with LBP and/or NP. Thus, we aimed to investigate the trends in treatment and examine the relationship between SES factors and treatment.
Methods
Study design
Data source
The National Health Insurance Service (NHIS) database was used as a data source. As it is the sole public health insurance system in South Korea, all individuals in South Korea should be registered in the NHIS, including the military, children, or unemployed persons. In addition, foreigners are obliged to subscribe to health insurance services if they stay in South Korea for more than 6 months. The NHIS database includes information regarding all disease diagnoses and prescriptions for any drug and/or procedure. The International Classification of Diseases and Related Health Issues 10th Revision (ICD-10) is used to register disease diagnoses in the NHIS database. The NHIS transitioned ICD-9 codes to ICD-10 codes from 2010 to 2014 based on a previous study13 and provided disease diagnosis information from 2010 to 2014 using ICD-10 codes for researchers.
Study population
Adult patients (aged ≥20 years) who were diagnosed with neck and/or back pain (NBP) and who visited an outpatient clinic or hospital in South Korea from January 1, 2010 to December 31, 2019 (10 years) were included in this study. The ICD-10 codes of G54.1, G54.3, G54.4, G57.0–G57.12, M43.2–M43.5, M43.8, M43.9, M45–M49, M49.2–M49.89, M51–M51.9, M53, M53.2–M54, M54.1–M54.18, M54.3–M54.9, M99, and M99.1–M99.9 were used to extract patients with LBP. The ICD-10 codes G54.2, M50–M50.93, M53.0, M54.0–M54.09, and M54.2 were used to extract patients with NP. Over 20,000,000 adult patients visited the outpatient clinic or were admitted to the hospital with a diagnosis of LBP or NP during the 2010–2019 period. A stratified random sampling technique,14 considering age and sex as exclusive strata for sampling, was used to extract the 2.5% of adult patients diagnosed with LBP and/or NP between 2010 and 2019, which would be used as the data sample included in the analysis. By using this sampling strategy, we ensured that the distribution of age and proportion of sex in the sample of patients with LBP and/or NP was similar to the entire population with LBP and/or NP. Finally, 6,252,780 patients were sampled using a stratified random sampling method using SAS V.9.4 (SAS Institute). Next, we excluded 391,401 patients diagnosed with cancer and 372 patients who underwent surgery, because cancer or surgery might affect the treatment pattern for LBP or NP. Ultimately, 5,861,007 patients with LBP and/or NP were included in this study, as shown in figure 1.
Flow chart depicting the patient selection process. LBP, low back pain; NP, neck pain.
SES (independent variable)
For SES-related information, data on occupation, household income level, and residence were collected. All patients were classified into four groups according to their occupation: office workers, healthcare workers, other workers, and unemployed. Information on household income level was included in the NHIS database to determine the insurance premiums of the South Korean population. However, individuals who are too poor to pay their insurance premiums or have difficulty supporting themselves financially are assigned to the Medical Aid Program, a government program developed to cover almost all medical expenses to help reduce the burden of medical costs on individuals. Except for the patients in the Medical Aid Program, all patients were divided into four groups using the quartile ratio. Then, we classified the total patients into five groups (four groups using quartile ratio and one group (Medical Aid Program group)). We also defined the five groups as the very low (medical aid program group), low (Q1 in lowest), middle (Q2), high (Q3), and very high (Q4 in highest) household income groups. The residences of all patients were classified into two groups: urban areas (Seoul and other metropolitan cities) and rural areas (all other areas).
Opioid use and interventional procedures (dependent variables)
This study had two objectives. First, we examined the trends in treatment, such as chronic opioid use and receipt of interventional procedures. Second, we investigated the factors associated with chronic opioid use and receipt of interventional procedures to identify whether SES affected the treatment of patients with LBP and/or NP. Similar to most previous studies,15 16 we defined patients who were prescribed opioids regularly and continuously for ≥90 days as chronic opioid users and considered other patients as non-users. Specifically, the duration of prescription of chronic opioid users had to be ≥90 days regardless of the daily dose or type of opioid. The opioid prescription information was completely collected from the NHIS database regardless of the setting because physicians from all outpatient clinics, long-term facility care centers, and acute setting hospitals need to register all opioid prescription information in the NHIS database. In addition, opioid prescription includes a combination of tramadol and acetaminophen. In South Korea, when a certain drug is prescribed by a physician, the relevant primary diagnosis should be registered in the NHIS database for the patient to receive financial coverage for treatment costs. In this study, we included the opioid prescription data with registered ICD-10 codes of LBP and/or NP as the primary diagnosis. Therefore, opioids prescribed in this study might be associated with LBP and/or NP. Spinal nerve plexus, root, and ganglion and epidural nerve block were included as interventional procedures, and the list of interventional procedures is presented in online supplemental table S1. However, radiofrequency ablation or trigger point injections were not included in the interventional procedures. The codes of the procedures should be registered in the NHIS database by the physicians who performed the procedures to have the treatment costs covered by the NHIS.
Supplemental material
Covariates
Age and sex were collected as demographic information as these are known to affect characteristics of chronic pain.17 Data on underlying disability were collected because pain might lead to almost unnoticeable changes in disability.18 All individuals with disabilities were registered in the NHIS database to receive benefits from South Korea’s social welfare system. Disabilities were divided into 15 types: physical and brain lesion disabilities; visual disturbances; hearing and speech disabilities; autism; intellectual, mental, renal, heart, and respiratory disorders; hepatopathies; facial disfigurements; intestinal and urinary fistulae; and epilepsy. In addition, disability was divided into six severity grades. The diagnosis and determination of disability was performed strictly according to the law by a specialty physician in each field. Finally, two severity groups were considered (grades 1–3: severe disability; grades 4–6: mild to moderate disability). In addition, long-term (≥90 days) gabapentin, pregabalin, paracetamol, and non-steroidal anti-inflammatory drug (NSAID) prescription information was collected as covariates. Non-opioid medication prescriptions, such as gabapentin, pregabalin, paracetamol, and NSAIDs, were collected because treatment with opioids was not superior to treatment with non-opioid medications among patients with chronic back pain.19
Statistical analysis
The clinicopathological characteristics of the study population are presented as mean values with SD for continuous variables and frequencies with percentages for categorical variables. For comparison of clinicopathological characteristics between chronic opioid users and non-users, t-test and χ2 tests were used for continuous variables and categorical variables, respectively. Next, we performed univariable and multivariable logistic regression analyses for chronic opioid use or receipt of interventional procedures in patients with LBP or NP. For multivariable logistic regression modeling, all covariates were included for adjustment. Moreover, we performed subgroup analyses according to the presence of underlying disability, which may or may not be from LBP and/or NBP, because it could contribute to the utilization rate of medical facilities. The results were presented as ORs with 95% CIs, and the Hosmer-Lemeshow test was used to confirm that the goodness of fit of the models was appropriate. There was no issue of multicollinearity because intervariable variance inflation factors were below 2.0 in the multivariable models. All statistical analyses were performed using R software (V.4.0.3; R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was set at p<0.05.
Results
Study population
The comparison of clinicopathological characteristics between opioid users and non-users among 5,861,007 patients with LBP and/or NP is presented in table 1. Of these, 284,860 (4.9%) were chronic opioid users, while 5,576,147 (95.1%) were non-users. In addition, 17.7% (1,039,302/5,861,007) of patients underwent interventional procedures for the treatment of LBP and/or NP. Online supplemental table S2 shows the comparison of clinicopathological characteristics between chronic opioid users and non-users among 5,861,007 patients with LBP and/or NP. Online supplemental table S3 shows the comparison of clinicopathological characteristics between patients who underwent and those who did not undergo interventional procedures among 5,861,007 patients with LBP and/or NP.
Clinicopathological characteristics of 5,861,007 patients with LBP and/or NP
Online supplemental figures S1 and S2 show the proportion of opioid users and patients who underwent interventional procedures, respectively. The proportion of opioid users was 1.5% (6681/460,164) in 2010, and it gradually increased to 6.2% (43,835/705,558) in 2019. The proportion of patients who underwent interventional procedures was 12.6% (57,907/460,104) in 2010, and it gradually increased to 22.1% (155,875/705,558) in 2019.
Opioid use for LBP and/or NP
Online supplemental table S4 shows the results of univariable logistic regression analysis for chronic opioid use. Table 2 (SES-related variables) and online supplemental table S5 (other covariates) show the results of the multivariable logistic regression model for chronic opioid use. The healthcare worker group had an 18% lower likelihood of chronic opioid use (OR 0.82, 95% CI 0.76 to 0.88; p<0.001), while the unemployment group had a 6% higher likelihood of chronic opioid use (OR 1.06, 95% CI 1.03 to 1.10; p=0.001) than the office worker group. The very poor household income group had an 18% higher likelihood of chronic opioid use (OR 1.18, 95% CI 1.13 to 1.24; p<0.001) than the poor household income level group. Those living in rural areas had a 7% lower likelihood of chronic opioid use (OR 0.93, 95% CI 0.91 to 0.95; p<0.001) than those living in urban areas. Online supplemental tables S6 and S7 show the results of the multivariable regression model for chronic opioid use in patients without and with disability among those with LBP and/or NP, respectively.
Adjusted associations between socioeconomic status and chronic opioid use
Interventional procedures for LBP and/or NP
Online supplemental table S8 shows the results of the univariable logistic regression model for the receipt of intervention procedures. Table 3 (SES-related variables) and online supplemental table S9 (other covariates) show the results of the multivariable logistic regression model for the receipt of interventional procedures. Other workers and unemployed individuals had a 4% (OR 1.04, 95% CI 1.03 to 1.04; p<0.001) and 8% (OR 1.08, 95% CI 1.08 to 1.09; p<0.001) higher likelihood of undergoing interventional procedures, while healthcare workers had a 5% (OR 0.95, 95% CI 0.94 to 0.97; p<0.001) lower likelihood of undergoing interventional procedures than office workers, respectively. Patients in the middle (OR 1.01, 95% CI 1.00 to 1.02; p=0.001), high (OR 1.02, 95% CI 1.01 to 1.02; p<0.001), and very poor household income groups (OR 1.11, 95% CI 1.10 to 1.12; p<0.001) had a higher likelihood of undergoing interventional procedures, while those in the very high household income group (OR 0.94, 95% CI 0.93 to 0.95; p<0.001) had a lower likelihood of undergoing interventional procedures than those in the poor household income group, respectively. Moreover, among such patients, those living in rural areas had an 11% lower likelihood of undergoing interventional procedures (OR 0.89, 95% CI 0.89 to 0.89; p<0.001) than those living in urban areas. Online supplemental tables S10 and S11 show the results of a multivariable regression model for the receipt of interventional procedures in patients without or with disability, respectively.
Adjusted associations between socioeconomic status and receipt of interventional procedure
Discussion
This population-based cross-sectional study showed that the number of opioid users and patients who underwent interventional procedures for the treatment of LBP and/or NP increased from 2010 to 2019 (over 10 years). SES factors, such as occupation, household income level, and residence, were associated with treatment. To our knowledge, this is the first study to show that SES affects the treatment of patients with LBP and/or NP.
Most importantly, the participants in the medical aid program group, who are too poor to pay their insurance premiums or have difficulty supporting themselves financially, tend to take more opioids and undergo interventional procedures for the treatment of LBP and/or NP. A previous Swedish cohort study reported that individuals receiving social welfare had a higher risk of incident opioid overdose.20 Moreover, a cross-sectional study in the UK reported that lower SES was associated with increased opioid prescription.21 Furthermore, a recent cohort study in the USA reported that low SES was associated with fatality due to opioid overdose.22 Similarly, in line with these findings, we report that the medical aid program group in South Korea who have a low economic status have a higher risk of chronic opioid use than other individuals.
Interestingly, healthcare workers tended to use opioids less often whereas those in the unemployed group tended to use more opioids than office workers. Healthcare workers, such as physicians or nurses, may be more aware of the risk and side effects of opioid prescription than those in other groups, and this perception might have influenced the low opioid usage among healthcare workers with LBP and/or NP. Moreover, the relationship between job loss and mortality due to opioid overdose is a crucial public issue.23 A recent cohort study reported that the unemployment rate and economic hardship were both associated with increased opioid abuse.24 Another cohort study conducted in the UK reported that unemployment appears to have an impact on increased opioid prescriptions.25 In addition to previous studies,24 25 we showed that unemployment could be an associated factor for opioid use in patients with LBP and/or NP.
The association of SES with the receipt of interventional procedures is an important finding. The interventional procedure for the treatment of LBP and/or NP are costly in South Korea,26 and the economic status could affect the receipt of interventional procedures. A recent study by Manchikanti et al reported that lumbar facet joint nerve blocks in the treatment of chronic LBP shows clinical effectiveness and cost utility atUS$$2654.08 for the direct costs of the procedures.27 Therefore, the SES of patients with LBP and/or NP should be considered in health policy.
Among SES factors, living in rural areas was associated with decreased chronic opioid use and interventional procedures for the treatment of LBP and/or NP. The relatively poor access to healthcare facilities might explain this phenomenon, as an access barrier was previously reported in rural older adults’ use of pain management.28 The access and cost could be barriers to pharmacological therapy, particularly for persons who live in rural areas where treatments are not available.29 The Centers for Disease Control and Prevention clinical practice guideline for prescribing opioids for pain in 2022 recommended that rurality should be considered in opioid prescribing due to health inequity.30 Therefore, further studies are needed to examine whether pain in patients living in rural areas is underestimated or neglected.
This study had some limitations. First, we used a stratified random sampling technique using age and sex for data extraction. Thus, there may have been differences between the sampled patients with LBP and/or NP and those in the population in South Korea. Second, we did not assess the severity of LBP or NP in this study. The duration, stage, and severity of pain in patients with LBP and/or NP may affect their treatment pattern. Third, some important factors, such as body mass index, alcohol consumption, and smoking history, were not used in this study because they were not included in the NHIS database. Fourth, the generalizability of the results in this study might be limited because the health insurance systems in many countries are different. Finally, unmeasured and residual confounders might have existed in this study, which might have affected our findings.
Conclusions
The number of chronic opioid users and patients who underwent interventional procedures for the treatment of LBP and/or NP increased from 2010 to 2019 in South Korea. SES factors, such as occupation, household income level, and residence, were associated with treatment among patients with LBP and/or NP.
Data availability statement
Data are available on reasonable request. Data are available on reasonable request to corresponding author.
Ethics statements
Patient consent for publication
Ethics approval
For this population-based cross-sectional study, all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Declaration of Helsinki of 1975, as revised in 2008. All procedures involving human patients were approved by the Institutional Review Board (IRB) of the Seoul National University Bundang Hospital (IRB approval number: X-2105-685-901). The requirement for informed consent was waived by the IRB because the data were analyzed anonymously after masking the individual and sensitive information of the study population. The medical record technician from the Big Data Center in the NHIS extracted and provided the data for this study following approval of the study protocol by the NHIS Ethics Committee (NHIS approval number: NHIS-2021-1-615).
References
Footnotes
Presented at The abstract of this article will be presented in the 48th Annual Regional Anesthesiology and Acute Pain Medicine Meeting on April 20, 2023–April 22, 2023.
Contributors TKO designed the study, analysed the data, interpreted the data, and drafted the manuscript. I-AS contributed to the study conceptualization, acquisition of data, and review of the manuscript. I-AS accepts full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish as guarantor. All authors read and approved the final manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.