Skip to main content
Download PDF

Cost-effectiveness of Transforaminal epidural steroid injections for patients with ACUTE sciatica: a randomized controlled trial

BMC Musculoskeletal Disorders volume 25, Article number: 247 (2024) Cite this article

Abstract

Background

Transforaminal epidural injections with steroids (TESI) are increasingly being used in patients sciatica. The STAR (steroids against radiculopathy)-trial aimed to evaluate the (cost-) effectiveness of TESI in patients with acute sciatica (< 8 weeks). This article contains the economic evaluation of the STAR-trial.

Methods

Participants were randomized to one of three study arms: Usual Care (UC), that is oral pain medication with or without physiotherapy, n = 45); intervention group 1: UC and transforaminal epidural steroid injection (TESI) 1 ml of 0.5% Levobupivacaine and 1 ml of 40 mg/ml Methylprednisolone and intervention group 2: UC and transforaminal epidural injection (TEI) with 1 ml of 0,5% Levobupivacaine and 1 ml of 0.9% NaCl (n = 50). The primary effect measure was health-related quality of life. Secondary outcomes were pain, functioning, and recovery. Costs were measured from a societal perspective, meaning that all costs were included, irrespective of who paid or benefited. Missing data were imputed using multiple imputation, and bootstrapping was used to estimate statistical uncertainty.

Results

None of the between-group differences in effects were statistically significant for any of the outcomes (QALY, back pain, leg pain, functioning, and global perceived effect) at the 26-weeks follow-up. The adjusted mean difference in total societal costs was €1718 (95% confidence interval [CI]: − 3020 to 6052) for comparison 1 (intervention group 1 versus usual care), €1640 (95%CI: − 3354 to 6106) for comparison 2 (intervention group 1 versus intervention group 2), and €770 (95%CI: − 3758 to 5702) for comparison 3 (intervention group 2 versus usual care). Except for the intervention costs, none of the aggregate and disaggregate cost differences were statistically significant. The maximum probability of all interventions being cost-effective compared to the control was low (< 0.7) for all effect measures.

Conclusion

These results suggest that adding TESI (or TEI) to usual care is not cost-effective compared to usual care in patients with acute sciatica (< 8 weeks) from a societal perspective in a Dutch healthcare setting.

Trial registration

Dutch National trial register: NTR4457 (March, 6th, 2014).

Peer Review reports

Background

Sciatica is characterized by pain radiating to the leg following one of the lumbosacral nerve roots [1]. Other than pain, patients may also experience sensory symptoms and/or weakness of the involved myotome. Approximately 85% of sciatica cases are caused by mechanical compression of the nerve root by a herniated intervertebral disc [2]. The annual incidence of lumbosacral radicular syndrome in the Netherlands has been estimated at 9 per 1000 patient-years [3] and the annual prevalence has been estimated at 17.2 per 1000 patient-years [3]. The prognosis of sciatica is generally described as favourable: within 3 months, 75% of patients are expected to reach bearable pain levels and can resume their work without surgery [4, 5]. Nonetheless, a UK-based study of patients seeking primary care for back-related leg pain, including lumbosacral radicular syndrome, showed that only 55% of patients with lumbosacral radicular syndrome had more than 30% reduction in disability 1 year after their first visit to primary care [6].

In addition to these patient related problems, sciatica poses a major economic burden. Although there are no recent specific cost data for sciatica, in 2017, the total healthcare cost of lower back pain in general (including sciatica) in the Netherlands was estimated to be 937 million [7]. This equals 1.07% of the total expenditure on health care in the Netherlands. Indirect costs due to absenteeism and lost productivity while being at work (i.e., presenteeism) were not included in this cost estimate; and hence the total societal cost will be even higher. Understanding the cost-effectiveness of different management strategies for sciatica, such as medication, physiotherapy, and transforaminal epidural steroid injections (TESI), is important to prevent high healthcare and socioeconomic costs. This requires formal assessments of the best available evidence on the cost-effectiveness of interventions and, where necessary, undertaking economic studies if there is a lack of good quality evidence [8].

TESI is increasingly used in patients with sciatica [9, 10]. In 2021 we described a survey among 80 neurologists (including residents) and 44 anesthesiologists. The results of this survey showed that 40–60% of neurologists think that TESIs are effective in 40% of injected patients and that 23/44(52%) of anesthesiologists think that TESIs are effective in 60–80% of the injected patients [11]. This seems to contradict the current evidence. Four recent systematic reviews and meta-analyses found that TESI only slightly reduced leg pain and disability compared to placebo at short-term follow-up (4–6 weeks) in patients with sciatica, but not at long-term follow-up (> 3 months) [12,13,14,15]. According to the GRADE [16], the quality of evidence is low to moderate. Recently, our research group finalized the STeroids against Radiculopathy (STAR) trial, a pragmatic two-center, randomized clinical trial (RCT) assessing the clinical effectiveness of TESI against sciatica [17]. Our results do not support TESI as a standard treatment for patients with acute sciatica (< 8 weeks). Although it is debated as to whether trial-based economic evaluations should still be performed if positive clinical effects are lacking [18], one should be aware that a lack of statistical differences between therapies does not necessarily mean that they are identical. The International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Cost Effectiveness Analysis Randomized Clinical Trial (CEA-RCT) task force therefore recommends that researchers perform CEA if positive clinical results are lacking [19]. The aim of this study was to evaluate the cost-effectiveness of TESI in patients with acute sciatica who participated in the STAR-trial. Within this trial TESI was compared to a (1) treatment regimen of medication only (usual care) and (2) to transforaminal epidural injection with local anesthetic and saline solution (TEI), from a societal perspective.

Methods

Study design

An economic evaluation was conducted alongside the STAR trial [17, 20, 21], an RCT evaluating the effectiveness of TESI in patients with acute sciatica. The RCT was conducted in two Dutch hospitals, the Zaans Medisch Centrum (Zaandam) and OLVG Teaching Hospital (Amsterdam), between January 13, 2016, and October 24, 2019.

The following three groups were compared:

  • Usual Care (UC): Oral pain medication with or without physiotherapy

  • Intervention group 1: Usual care and TESI of 1 ml of 0.5% Levobupivacaine and 1 ml of 40 mg/ml Methylprednisolone

  • Intervention group 2: Usual care and transforaminal epidural injection with 1 ml of 0.5% Levobupivacaine and 1 ml NaCI 0.9%

We analyzed three pairwise comparisons:

  • Comparison 1 (main comparison): Intervention group 1 versus Usual Care

  • Comparison 2: Intervention group 1 versus Intervention group 2

  • Comparison 3 (for completeness): Intervention group 2 versus Usual Care

The RCT was approved by the Medical research Ethics Committees United, Nieuwegein, The Netherlands (registration number NL45805.100.15). The protocol was registered in the Dutch Trial Register number NTR4457 (6/03/2014). The CONSORT statement was followed for reporting [22]. See appendix 1.

Participants

Eligible patients had sciatic symptoms < 8 weeks and were seen by a neurologist at one of the two study centres upon referral by their general practitioners (GPs). Additional inclusion criteria were as follows: a) age between 18 and 75 years; b) magnetic resonance imaging (MRI)-confirmed disc herniation with nerve root impingement causing clinical symptoms; c) pain experienced on average over the last week rated on a numerical rating scale (NRS)(> 4/10); d) good understanding of the Dutch language; and e) Internet access to complete online questionnaires [21]. The exclusion criteria were a) severe weakness of the legs (Medical Research Council ((MRC) score < 3); b) spinal surgery < 1 year at the symptomatic lumbar level; c) lumbar spinal stenosis or spondylolisthesis as the cause of radicular pain diagnosed by MRI; d) pregnancy; and e) severe comorbidity (e.g., cancer) [21].

Randomization and blinding

Written informed consent was obtained of all patients before entering the study (see Appendix 2 for consent form). Randomization was performed by the study coordinator (BTM) or by one of the two trial nurses using ALEA® software (NKI-AVL, Netherlands). Alea® generated a random schedule of blocks with a maximum size of six. The participants who received a transforaminal injection were blinded to the type of injection.

Intervention

The procedure was similar for both intervention arms). Participants were brought to a fluoroscopy room and placed in a prone position on the procedure table. Fluoroscopy was used to localize MRI-confirmed disc herniation. Target identification and needle entry were performed according to international procedures [23]. First, the skin was prepped using chlorhexidine. Second, injections were administered using a 22-gauge 100 mm facet tipped needle (Pajunk RGN™). The correct needle position was confirmed by the injection of 0.5–1.5 cc of Joversol 300 mg/ml contrast material (Optiray™ 300, Mallinckrodt). Once an image was obtained demonstrating contrast material spreading into the epidural space medial to a line connecting the ipsilateral lumbar vertebral pedicles, the injection was performed at the level of the herniated disc. The injections were not repeated.

The study participants in intervention group 1 received 1 ml of 0,5% levobupivacaine, followed by 1 ml of 40 mg/ml methylprednisolone in an opaque syringe. The study participants in intervention group 2 received 1 ml 0,5% levobupivacaine followed by 1 ml of 0.9% NaCl. The total volume of the two injections was the same (2 ml).

After epidural injection, washout of the contrast fluid was observed on an X-ray image. The image was saved. Finally, the needle was removed, and the patient was brought to the recovery area.

Usual care

Patients in all groups used analgesics registered during the trial using online questionnaires [24]. In the case of kinesiophobia, patients were permitted to visit a physiotherapist. There were no restrictions on the use of analgesics or physiotherapy in any group, and their use was monitored. Pain medication, with or without physiotherapy, was planned at the discretion of the attending physician and according to the patient’s personal needs.

Effect measures

The primary effect measure was health-related quality of life. At baseline, 3 and 6 weeks, and 3 and 6 months after the start of the intervention, the patients’ health states were assessed using the Euroqol-5 dimensions- 3 levels (EQ-5D-3L) [25]. The EQ-5D-3L comprises the following five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension had three levels: no problems, problems, and serious problems. The patient indicated her health state by ticking the box next to the most appropriate statement for each of the five dimensions. Each health state was converted into a utility score using the Dutch tariff [26]. Hence, the utility values estimated in this study were indicative of a Dutch person’s value or desirability of patients’ health states. Values ranged between − 0.33 (0 is equivalent to death; negative values are worse than death), and 1 (full health). The EQ-5D-3L is commonly used in cost-utility analyses and, for that reason, applied in this economic evaluation as well [27].

Secondary outcomes were back and leg pain (average previous week) (measured using a 10-cm numerical rating scale [NRS] [28]), physical functioning (measured using the Dutch version of the Roland-Morris Disability Questionnaire [29, 30]), and global perceived recovery (GPR). The latter was measured on a 7-point Likert scale [31] ranging from ‘completely recovered’ to ‘worse than ever,’ which was dichotomized into success (categories ‘completely’ and ‘much recovered’) and non-success (categories ‘slightly recovered,’ ‘no change, ‘slightly worse’, ‘much worse’ and ‘worse than ever’).

Costs measures

Resource use was assessed using cost questionnaires administered at three and 6 months. In line with Dutch guidelines, costs were assessed from a societal perspective, meaning that all costs were included, irrespective of who paid or benefited [32, 33]. Intervention costs were estimated using the data acquired from the accounting records of the two participating clinics. Data on other healthcare utilization, informal care, unpaid productivity, and absenteeism due to back pain were collected using 3-monthly self-reported web-based cost questionnaires [20]. Healthcare utilization included primary care (e.g., general practitioner care, physiotherapy, manual therapy, chiropractic care, and exercise therapy), secondary care (e.g., hospitalization, and diagnostic and therapeutic interventions), and the use of prescribed and over-the-counter medications. Healthcare utilization was valued using Dutch standard costs and the prices of professional organizations if standard costs were not available. Medication use was valued using prices derived from http://www.medicijnkosten.nl. Informal care included care by family, friends, and other volunteers and was valued according to the proxy good method using an estimate of the hourly cost of a housekeeper. Absenteeism was measured using the Productivity and Disease Questionnaire (PRODISQ) [34]. Absenteeism was valued using sex-specific price weights in accordance with the friction cost approach (friction period = 12 weeks). Unpaid productivity costs included all hours of volunteer work and domestic and educational activities that participants were not able to perform owing to their sciatica; these were also valued using the aforementioned proxy method. All costs were converted to 2020 Euros using consumer price indices. Discounting of costs and effects was not necessary because the follow-up period was 6 months.

Sample size

Sample sizes were calculated based on a power of 0.9 and a two-sided alpha of 0.05 [20, 21]. These calculations indicated that 48 patients were needed per arm to detect a clinically relevant between-group MD of two points on the 0–10 NRS for leg and back pain (SD = 30). For physical functioning, 22 patients were needed per arm to detect a clinically relevant between-group MD of four points (SD = 4). For dichotomized GPR, 79 patients were needed per arm to detect a clinically relevant between-group difference of 20% [35, 36]. Anticipating a 10% loss to follow-up, 264 patients were included (n = 88 per arm).

In accordance with the guidelines of the ‘European Medicines Agency,’ we will only consider one intervention effective over another if statistically significant and clinically relevant differences are found between them for all co-primary outcomes [37].

Cost-effectiveness and utility analyses

A cost-effectiveness analysis (CEA) and a cost-utility analysis (CUA) were conducted. In the CEA, total costs were related to improvements in back pain, leg pain, functioning, and global perceived effect. In the cost-utility analyses (CUA), total costs were related to the QALYs gained during follow-up. All analyses were performed using intention-to-treat. Baseline characteristics were compared between the intervention and control groups. Missing data for the economic evaluation were handled using multivariate imputation with chained equations. The imputation model included all available cost and effect measure values as well as variables differing between groups at baseline, variables related to the ‘missingness’ of data, and variables related to the outcomes. Ten complete datasets were created so that the loss of efficiency would be less than 5% [18].

The mean between-group cost differences were calculated for total and disaggregated costs (intervention costs, healthcare costs, informal care costs, absenteeism costs, presenteeism costs, and unpaid productivity costs). To determine the mean incremental difference in the cost and effect between the intervention and control groups, we used seemingly unrelated regression (SUR). SUR runs two regressions to determine incremental cost and incremental effect differences simultaneously, adjusting for any potential correlation between costs and effects. The regression for determining the incremental cost difference was adjusted for baseline values (if available) and confounding variables (age, sex, body mass index [BMI], severity of back and leg pain at baseline, and work status).

Incremental cost-effectiveness ratios (ICERs) were calculated by dividing the difference in total costs by the difference in QALYs adjusted for confounders. Uncertainty surrounding the cost differences and ICERs was estimated using bias-corrected and accelerated (BCA) bootstrapping (5000 replications) and graphically presented in cost-effectiveness planes and cost-effectiveness acceptability curves (CEACs). The latter indicates the probability of an intervention condition being cost-effective compared with the control condition at different values of willingness to pay (further referred to as the ceiling ratio). In these analyses, SUR analyses and BCA bootstrapping were nested in multiple imputations, meaning that multiple imputations were used to generate 10 complete datasets, after which the SUR and BCA bootstrapping methods were applied to each of the complete datasets [20, 38]. The intermediate results per completed dataset were pooled using Rubin’s rules [39].

Economic evaluations were performed using STATA (V16) (Stata Corp., College Station, TX, USA).

Sensitivity analysis

A predetermined sensitivity analysis (SA) was performed to assess the robustness of the results by comparing the friction cost approach with the human capital approach (SA1). Furthermore, complete-case analysis (SA2) and sensitivity analysis using the healthcare perspective (SA3) were performed.

Results

Study participants

During the study period, 1564 (922 in Amsterdam and 642 in Zaandam) adults with sciatica (regardless of duration) and nerve root compression on MRI were observed. Of these, 141 patients had acute sciatica and were willing to participate in the study. After providing informed consent, 45 patients were assigned to the control group, 46 to intervention group-1, and 50 to intervention group-2. The baseline characteristics are presented in Table 1 and were comparable between the groups. Complete data on all measurements was obtained from 35 (of 45) in the control group, 32 (of 46) in intervention-group 1 (‘steroids’) and 35 (of 50) in intervention-group 2 (‘anesthetic only’). Figure 1 provides an overview of the randomisation between the three study groups and the follow up of participants during 6 months follow-up. All participants were included in the final analysis.

Table 1 Baseline variables of included patients
Full size table
Fig. 1
figure 1

STAR-trial: enrolment, randomization and follow-up

Full size image

Effects

None of the between-group differences in effects were statistically significant, for any of the outcomes (QALY, back pain, leg pain, functioning, and global perceived recovery) at the 26-weeks follow-up [17].

Costs

The mean cost per patient in the various study groups and the unadjusted mean cost differences between the groups are shown in Table 2. For both intervention groups, the cost per participant was estimated to be €486 per participant. The total societal costs were €21,724 (SEM 2461) per participant for intervention group 1, €21,337 (SEM2087) per participant for intervention group 2, and €21,400 (SEM 2165) for the usual care group.

Table 2 Mean cost per patient in the various study groups, and unadjusted mean cost differences between groups
Full size table

The adjusted mean difference in total societal costs was €1718 (95% confidence interval [CI], − 3020 to 6052) for comparison 1 (intervention group 1 versus usual care), €1640 (95%CI: − 3354 to 6106) for comparison 2 (intervention group 1 versus intervention group 2), and €770 (95%CI: − 3758 to 5702) for comparison 3 (intervention group 2 versus usual care). Except for intervention costs, none of the aggregate and disaggregate cost difference was statistically significant.

Cost-effectiveness

The results of the cost-effectiveness analyses, including differences in costs, differences in effects, ICERs, and distributions of bootstrapped cost-effect pairs across the four quadrants of the CE plane, can be found in Table 3 for all comparisons.

Table 3 Cost-effectiveness analysis results (main analysis)
Full size table

At 6 months, the ICER for QALYs was €234,478 for comparison 1 (intervention group 1 versus usual care), indicating that the additional societal cost in intervention group 1 compared to usual care was €234,478 per QALY gained. This ICER shows that the intervention was -on average- “more costly” and “more effective” than usual care, which was also the case for leg pain and perceived recovery. ICERs for back pain and physical functioning, on the other hand, showed that the intervention was on average “more costly” and “less effective,” indicating that it was dominated by usual care.

For comparison 2 (intervention group 1 versus intervention group 2), ICERs for QALYs and physical functioning indicated that intervention 1 was dominated intervention 2 for these outcomes (i.e. on average “more costly” and “less effective”), while ICERs for back and leg pain showed that intervention 1 was on average “more costly” and “more effective.” It should be noted that outcomes for self-perceived recovery are lacking because all participants in both groups indicated recovery after 6 months.

For comparison 3 (intervention group 2 versus usual care), ICERs indicated that the intervention was dominated by usual care for back pain and physical functioning (i.e. on average “more costly” and “less effective”), while it was “more costly” and “more effective” for QALYs, leg pain, and self-perceived recovery.

The CEACs in Figs. 2, 3, 4, 5, and 6 show that the maximum probability of both interventions being cost-effective compared to usual care was low (< 0.7) for all effect measures. A probability of cost-effectiveness of 0.7 means that if the intervention is implemented, it will indeed be cost-effective in 70% of cases, whereas in 30% of cases it will not.

Fig. 2
figure 2

Cost-effectiveness acceptability curve indicating the probability of the intervention conditions’ being cost-effectiveness compared with control for different ceiling ratios (€) for quality-adjusted life-years

Full size image
Fig. 3
figure 3

Cost-effectiveness acceptability curve indicating the probability of the intervention conditions’ being cost-effectiveness compared with control for different ceiling ratios (€) for back-pain

Full size image
Fig. 4
figure 4

Cost-effectiveness acceptability curve indicating the probability of the intervention conditions’ being cost-effectiveness compared with control for different ceiling ratios (€) for leg pain

Full size image
Fig. 5
figure 5

Cost-effectiveness acceptability curve indicating the probability of the intervention conditions’ being cost-effectiveness compared with control for different ceiling ratios (€) for physical functioning

Full size image
Fig. 6
figure 6

Cost-effectiveness acceptability curve indicating the probability of the intervention conditions’ being cost-effectiveness compared with control for different ceiling ratios (€) for self-perceived recovery

Full size image

Sensitivity analyses

In line with the main analysis, between-group differences in total costs and effects were not significant in any of the sensitivity analyses (Appendix 3). The complete-case analysis showed a small statistically significant effect on GPR for comparison 1 (0.04 (95%CI 0.02 to 0.07)). However, the overall conclusion of this study does not change when any of the assumptions of the sensitivity analyses are used.

Adverse effects

No adverse events were observed. None of the participants withdrew from the trial out of safety measures.

Discussion

Principal findings

In this study, the cost and cost-effectiveness of adding an invasive treatment, that is, TESI to usual care in the treatment of acute sciatica (< 8 weeks) in the secondary care setting was assessed. The results suggest that adding TESI to usual care is not cost-effective compared with usual care alone in patients with acute sciatica from a societal perspective in a Dutch healthcare setting. That is, the maximum probability for adding invasive treatment TESI of being cost-effective in comparison to the control was low for all possible outcomes (< 0.7) and all comparisons. Comparisons 2 and 3 show comparable results.

Sensitivity analyses confirmed these results, although a small positive effect on GPR was found in the complete-case analysis for comparison 1 (∆E 0.04 (95%CI 0.02–0.07)). This discrepancy with the main analysis is likely caused by the selective dropout of participants, and as multiple imputation was used in the main analysis to handle missing data, we consider these results to be more valid.

All in all, these results suggest that TESI or TEI in additional to usual care (oral pain medication) is not cost-effective compared with usual care alone from a societal perspective in acute sciatica patients in the Dutch healthcare setting.

Comparison to the literature

Price et al [40] conducted a prospective, double-blind randomized trial in the UK assessing the (cost-)effectiveness of TESI versus a placebo injection of normal saline into the interspinous ligament in 228 participants with sciatica during a 12-month follow-up. The most important outcomes were the number needed to treat to realize a 75% improvement in pain relief and functional status at 3 weeks, which was 11.4 (p = 0.017), and the cost per QALY for one epidural steroid injection which was £25,746 from a provider’s perspective and £31,904 from a purchaser’s perspective. Given the fact that in this trial there was no clinical benefit of TESI over placebo between 6 and 52 weeks, the authors concluded that TESIs were not cost-effective. However, Price et al did not assess the costs of informal care as and productivity losses. In addition Price et al assumed that both treatment groups received similar pain medication that did not differ between groups. As a result costs of pain medication were not explicitly measured.

Spijker et al [41] performed a pragmatic, randomized, controlled, single-blinded trial in Dutch general practices (n = 63) and compared one segmental epidural steroid injection containing 80 mg triamcinolone (intervention) to usual care (taking analgesics as needed, and maintaining normal daily activities as much as possible) during 52 weeks. Mean total costs were €4414 in the intervention group and €5121 in the control group, a difference that was mostly due to differences in lost productivity. The point estimate for the ICER was - €730, meaning that a one point decrease on the NRS back pain scale in one patient during the course of 1 year was associated with a saving of €730 compared with usual care. The cost-effectiveness acceptability curve (CEAC) showed that without additional investments, the probability that epidural corticosteroid injections are cost-effective was more than 80%. Spijker et al concluded that the effect on pain and disability of epidural corticosteroid injections in sciatica is small, but significant (contrary to this RCT that found no clinically relevant differences between groups [16, 17]), and at lower costs and recommended that ‘policymakers could consider segmental epidural steroid injections as an additional treatment option’. The difference in results between our trial and that of Spijker et al [41] could be explained by differences in the way both studies handled missing data (i.e. multiple imputation in our study versus a complete-case analysis in the study of Spijker et al) and baseline imbalances (i.e. regression-adjustment in our study versus no adjustment in the economic evaluation of Spijker et al) as well as the absence of informal care, unpaid productivity, and presenteeism costs in the study of Spijker et al.

Strengths and limitations

The major strength of the STAR trial is its pragmatic design, meaning that its set-up resembled real-life routine practice conditions as much as possible [42]. Thus, the STAR trial enabled us to evaluate TESI against acute sciatica under circumstances directly in line with clinical practice, making the current results generalizable to Dutch clinical practice. The current analyses were also conducted using state of the art methods. That is, multiple imputation was used for handling missing data, regression-based adjustment for handling baseline imbalances, non-parametric bootstrapping for handling the skewed nature of cost data, and seemingly unrelated regression for handling the correlation between costs and effects [43]. This is important, as previous research indicates that using less optimal methods may notably impact results and might even impact on the conclusions of trial-based economic evaluations [44]. Another strength is that not only QALYs, functional status, and pain intensity were used as an outcome measure in the economic evaluation, but also perceived recovery as measured by the GPR. The similarity in results from four different outcome measures gave confidence in the robustness of results. Moreover, varies sensitivity analyses were performed that showed the robustness of results as well. All these attributes support the validity of the findings observed in this study.

This trial also has limitations. One limitation is that we used one particular injection technique, i.e. TESI. Other epidural injection techniques, such as caudal epidural approach [45] or echography- guided transforaminal approach [46], might further reduce costs and improve the cost-effectiveness of epidural injections. The latter techniques were not chosen because in The Netherlands there is a strong preference for TESI (usual care within the Pain Department). A second limitation is the use of self-reported retrospective cost questionnaires that may have introduced recall bias and/or “social desirability bias. However, as it seems unlikely that recall bias or the degree to which participants gave socially desirable answers systematically differed between groups, it is not expected that self-report biased the results. A third limitation concerns the missing data. We used multiple imputation (MI) to deal with this limitation. MI is considered a powerful statistical tool, as it accounts for uncertainty by replicating the incomplete dataset multiple times and substituting the missing values with plausible ones [44]. A fourth limitation was the use of the three-level version of the EQ-5D rather than the 5-level version to measure QALYs [47]. This is a possible limitation, as the EQ-5D-3L was found to have significantly higher floor effects for the health dimension pain/discomfort and ceiling effects for the health dimensions mobility, self-care, pain/discomfort, anxiety/depression, compared with the EQ-5D-5L. However, as the floor and ceiling effects are likely to be equal in all study groups, we do not expect our use of the EQ-5D-3L to have severely biased our results. As a fifth limitation, opioid use was analysed on an aggregate level, and no detailed on the type, number, and dosage of opioids that the patients used was collected. When measuring pain, it is important to measure the medication that affects pain. We recommend future studies to measure this on a detailed level.

Conclusion

Although a common treatment among patients with sciatica due to an MRI-confirmed herniated lumbar disc, evidence from our trial-based economic evaluation suggests that TESI in the acute phase (< 8 weeks) cannot be considered cost-effective compared to usual care from a societal perspective in a Dutch healthcare setting during a 6 months follow up period. Therefore, the current status of this treatment in the Dutch healthcare setting should be reconsidered.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Abbreviations

AE:

Adversed effect

AR:

Adversed reaction

CI:

Confidence interval

CEA:

Cost Effectiveness Analysis

CEAC:

Cost-effectiveness acceptability curve

CUA:

Cost-utility analysis

EQ-5D-3L:

Euroqol-5 dimensions- 3 levels

EQ-5D-5 L:

Euroqol-5 dimensions- 5 levels

FDA:

Food and drug administration

GP:

General practitioner

GPR:

Global perceived recovery

GRADE:

Grades of recommendation, assessment, development and evaluation

ICER:

Incremental cost-effectiveness ratios

MD:

Mean difference

MI:

Multiple Imputation

MRC:

Medical research council

MRI:

Magnetic resonance imaging

NRS:

Numerical rating scale

NTR:

Nederlands trial register

Prodisc:

Productivity and disease questionnaire

Qaly:

Quality adjusted life year

RCT:

Randomized controlled trial

RDQ:

Roland-Morris disability questionnaire

SUR:

Seemingly unrelated regression (SUR)

TEI:

Transforaminal epidural injection

TESI:

Transforaminal epidural injections with steroids

VAS:

Visual analogous scale

References

  1. Ropper AH, Zafonte RD. Sciatica. N Engl J Med 2015 26;372(13):1240–8.

  2. El Barzouhi A, Verwoerd AJ, Peul WC, Verhagen AP, Lycklama À Nijeholt GJ, Van der Kallen BF, Koes BW, Vleggeert-Lankamp CL; Leiden–The Hague Spine Intervention Prognostic Study Group. Prognostic value of magnetic resonance imaging findings in patients with sciatica. J Neurosurg Spine. 2016;24(6):978–85.

  3. Spijker-Huiges A, Groenhof F, Winters JC, van Wijhe M, Groenier KH, van der Meer K. Radiating low back pain in general practice: incidence, prevalence, diagnosis, and long-term clinical course of illness. Scand J Prim Health Care. 2015;33(1):27–32.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Konstantinou K, Dunn KM. Sciatica: review of epidemiological studies and prevalence estimates. Spine (Phila Pa 1976). 2008;33(22):2464–72.

  5. Vroomen PC, de Krom MC, Knottnerus JA. Predicting the outcome of sciatica at short-term follow-up. Br J Gen Pract. 2002;52(475):119–23.

  6. Konstantinou K, Dunn KM, Ogollah R, Lewis M, van der Windt D, Hay EM. Prognosis of sciatica and back-related leg pain in primary care: the ATLAS cohort. Spine J. 2018;18:1030–40.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Available from: https://www.vzinfo.nl/nek-en-rugklachten/zorguitgaven.

  8. Fitzsimmons D, Phillips CJ, Bennett H, Jones M, Williams N, Lewis R, et al. Cost-effectiveness of different strategies to manage patients with sciatica. Pain. 2014;155(7):1318–27.

    Article  PubMed  Google Scholar 

  9. Available from: https://digital.nhs.uk/data-and-information/data-tools-and-services/data-services/hospital-episode-statistics.

  10. Manchikanti L, Pampati V, Hirsch JA. Retrospective cohort study of usage patterns of epidural injections for spinal pain in the US fee-for-service Medicare population from 2000 to 2014. BMJ Open. 2016;6(12):e013042.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Ter Meulen B, Overweg C, Feenstra T, Brouwer B, Terheggen M, van Dongen H, et al. Diagnosis and treatment of sciatica in the Netherlands: a survey among neurologists and anesthesiologists. Eur Neurol. 2021;84(4):219–29.

    Article  PubMed  Google Scholar 

  12. De Bruijn TM, de Groot IB, Miedema HS, Haumann J, Ostelo RWJG. Clinical relevance of epidural steroid injections on lumbosacral radicular syndrome-related symptoms: systematic review and Meta-analysis. Clin J Pain. 2021;37(7):524–37.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Verheijen EJA, Bonke CA, Amorij EMJ, Vleggeert-Lankamp CLA. Epidural steroid compared to placebo injection in sciatica: a systematic review and meta-analysis. Eur Spine J. 2021;30(11):3255–64.

    Article  CAS  PubMed  Google Scholar 

  14. Oliveira CB, Maher CG, Ferreira ML, Hancock MJ, Oliveira VC, McLachlan AJ, et al. Epidural corticosteroid injections for lumbosacral radicular pain. Cochrane Database Syst Rev. 2020;4(4):CD013577.

    PubMed  Google Scholar 

  15. Manchikanti L, Knezevic E, Knezevic NN, Sanapati MR, Thota S, Abd-Elsayed A, et al. Epidural injections for lumbar radiculopathy or sciatica: a comparative systematic review and Meta-analysis of Cochrane review. Pain Phys. 2021;24(5):E539–54.

    Google Scholar 

  16. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ (Clin res ed). 2008;336(7650):924–6.

    Article  Google Scholar 

  17. Ter Meulen BC, van Dongen JM, Maas E, van de Vegt MH, Haumann J, Weinstein HC, et al. Effect of Transforaminal Epidural Corticosteroid Injections in Acute Sciatica: A Randomized Controlled Trial. Clin J Pain. 2023;39(12):654–62.

  18. Maas ET, Juch JNS, Ostelo RWJG, Groeneweg JG, Kallewaard JW, Koes BW, et al. Cost-effectiveness of radiofrequency denervation for patients with chronic low Back pain: the MINT randomized clinical trials. Value Health. 2020;23(5):585–94.

    Article  PubMed  Google Scholar 

  19. Ramsey SD, Willke RJ, Glick H, Reed SD, Augustovski F, Jonsson B, et al. Cost-effectiveness analysis alongside clinical trials II-An ISPOR Good Research Practices Task Force report. Value Health. 2015;18(2):161–72.

  20. Ter Meulen BC, Maas ET, Vyas A, van der Vegt M, de Priester K, de Boer MR, et al. Treatment of acute sciatica with transforaminal epidural corticosteroids and local anesthetic: design of a randomized controlled trial. BMC Musculoskelet Disord. 2017;18(1):215.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Ter Meulen BC, van Dongen JM, van der Vegt M, Weinstein HC, Ostelo RWJG. STeroids against radiculopathy (STAR) trial: a statistical analysis plan. Trials. 2021;22(1):80.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Schulz KF, Altman DG, Moher D. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMC Med. 2010;8:18.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Van Boxem K, Cheng J, Patijn J, van Kleef M, Lataster A, Mekhail N, et al. Pain Pract. 2010;10(4):339–58.

    Article  PubMed  Google Scholar 

  24. Survalyzer - to survey and analyze. Utrecht. 2018. p. https://survalyzer.com/.

  25. Aaronson NK, Acquadro C, Alonso J, Apolone G, Bucquet D, Bullinger M, et al. International quality of life assessment (IQO-LA) project. Qual Life Res. 1992;1:349–51.

    Article  CAS  PubMed  Google Scholar 

  26. Lamers LM, Stalmeier PF, McDonnell J, Krabbe PF, van Busschbach JJ. Kwaliteit van leven meten in economische evaluaties: het Nederlands EQ-5D-tarief [Measuring the quality of life in economic evaluations: the Dutch EQ-5D tariff]. Ned Tijdschr Geneeskd. 2005;149(28):1574–8. Dutch.

  27. Kind P. The Euroqol instrument: an index of health-related quality of life. In: Spilker B, editor. Quality of life and pharmacoeconomics in clinical trials. Philadelphia, PA: Lippincott-Raven Publishers; 1996. p. 191–201.

    Google Scholar 

  28. Childs JD, Piva SR, Fritz JM. Responsiveness of the numeric pain rating scale in patients with low back pain. Spine (Phila Pa 1976). 2005;30(11):1331–4.

  29. Gommans IHB, Koes BW, van Tulder MW. Validiteit en responsiviteit Nederlandstalige Roland Disability Questionnaire. Vragenlijst naar functionele status bij patiënten met lage rugpijn. Ned Tijdschr Fys. 1997;107:28–33.

  30. Roland M, Fairbank J. The Roland-Morris questionnaire and the Oswestry disability questionnaire. Spine. 2000;25:3115–24.

    Article  CAS  PubMed  Google Scholar 

  31. Kamper SJ, Maher CG, Herbert RD, Hancock MJ, Hush JM, Smeets RJ. How little pain and disability do patients with low back pain have to experience to feel that they have recovered? Eur Spine J. 2010;19(9):1495–501.

  32. Swan Tan S, Bouwmans-Frijters C, Hakkaart-van Roijen L. Handleiding voor kostenonderzoek: methoden en referentieprijzen voor economische evaluaties in de gezondheidszorg. Tijds. gezondheids.wetenschappen. 2012;90:367–72.

  33. Goossens ME, Rutten-van Mölken MP, Vlaeyen JW, van der Linden SM. The cost diary: a method to measure direct and indirect costs in cost-effectiveness research. J Clin Epidemiol. 2000;53(7):688–95.33.

    Article  CAS  PubMed  Google Scholar 

  34. Koopmanschap MA. PRODISQ: a modular questionnaire on productivity and disease for economic evaluation studies. Expert Rev Pharmacoecon Outcomes Res. 2005;5(1):23.

    Article  PubMed  Google Scholar 

  35. Ferreira ML, Herbert RD, Ferreira PH, Latimer J, Ostelo RW, Grotle M, et al. The smallest worthwhile effect of nonsteroidal anti-inflammatory drugs and physiotherapy for chronic low back pain: a benefit-harm trade-off study. J Clin Epidemiol. 2013;66(12):1397–404.

    Article  PubMed  Google Scholar 

  36. Christiansen DH, de Vos Andersen NB, Poulsen PH, Ostelo RW. The smallest worthwhile effect of primary care physiotherapy did not differ across musculoskeletal pain sites. J Clin Epidemiol. 2018;101:44–52.

  37. https://www.ema.europa.eu/en/documents/scientific-guideline/draft-guideline-multiplicity-issues-clinical-trials_en.pdf.

  38. Van Dongen JM, El Alili M, Varga AN, Guevara Morel AE, Jornada Ben A, Khorrami M, et al. What do national pharmacoeconomic guidelines recommend regarding the statistical analysis of trial-based economic evaluations? Expert Rev Pharmacoecon Outcomes Res. 2020;20(1):27–37.

    Article  PubMed  Google Scholar 

  39. Rubin DB. Multiple imputation for nonresponse in surveys, vol. 81. John Wiley & Sons; 2004.

    Google Scholar 

  40. Price C, Arden N, Coglan L, et al. Cost-effectiveness and safety of epidural steroids in the management of sciatica. Health Technol Assess. 2005;1:58.

    Google Scholar 

  41. Spijker-Huiges A, Vermeulen K, Winters JC, van Wijhe M, van der Meer K. Costs and Cost-effectiveness of Epidural Steroids for Acute Lumbosacral Radicular Syndrome in General Practice. Spine. 2014;39(24):2007–12.

  42. Hohenschurz-Schmidt D, Kleykamp BA, Draper-Rodi J, Vollert J, Chan J, Ferguson M, et al. Pragmatic trials of pain therapies: a systematic review of methods. Pain. 2022;163(1):21–46.

    Article  PubMed  Google Scholar 

  43. El Alili M, Vrijens B, Demonceau J, Evers SM, Hiligsmann M. A scoping review of studies comparing the medication event monitoring system (MEMS) with alternative methods for measuring medication adherence. Br J Clin Pharmacol. 2016;82(1):268–79.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Mutubuki EN, El Alili M, Bosmans JE, Oosterhuis T, J Snoek F, Ostelo RWJG, et al. The statistical approach in trial-based economic evaluations matters: get your statistics together! BMC Health Serv Res. 2021;21(1):475.

  45. Klunklin K, Sangsin A, Leerapun T. Efficacy and safety of ultrasound-guided caudal epidural steroid injection in patients with low back pain and sciatica. J Back Musculoskelet Rehabil. 2022;35(2):317–22.

    Article  PubMed  Google Scholar 

  46. Manchikanti L, Knezevic E, Knezevic NN, Vangala BP, Sanapati MR, Thota S, et al. A comparative systematic review and Meta-analysis of 3 routes of Administration of Epidural Injections in lumbar disc herniation. Pain Phys. 2021;24(6):425–40.

    Article  Google Scholar 

  47. EQ-5D-5L is: Herdman M, Gudex C, Lloyd A, et al. Development and preliminary testing of the new five-level version of EQ- 5D (EQ- 5D- 5L). Qual Life Res. 2011;20:1727–36.

Download references

Acknowledgements

The authors acknowledge the help of doctors and trial nurses of the Zaans MC, Zaandam and OLVG Hospital Amsterdam for performing the study. The authors acknowledge W. Kraan from the Datamanagement Group VU for his contribution to the study software. We thank Alan Jenks for his contribution to the writing of this manuscript.

Funding

The OLVG-hospital paid for the randomization software (ALEA), CRF software and web based questionnaires.

Author information

Authors and Affiliations

  1. Department of Neurology at OLVG Teaching Hospital, Amsterdam, The Netherlands

    Bastiaan C. ter Meulen & Henry C. Weinstein

  2. Department of Epidemiology and Data Sciences, Amsterdam UMC, Vrije Universiteit, Amsterdam Movement Sciences Research Institute Musculoskeletal Health, De Boelelaan 1089a, 1081, HV, Amsterdam, The Netherlands

    Bastiaan C. ter Meulen, Esther T. Maas, Raymond W. J. G. Ostelo & Johanna M. van Dongen

  3. Department of Health Sciences, Faculty of Science, Vrije Universiteit Amsterdam and the Amsterdam Movement Sciences Research Institute, De Boelelaan 1085, 1081, HV, Amsterdam, The Netherlands

    Raymond W. J. G. Ostelo

  4. Department of Pain Medicine and Anesthesiology Zaans MC, Zaandam, The Netherlands

    Rien van der Vegt

  5. Department of Pain Medicine and Anesthesiology, OLVG, Amsterdam, The Netherlands

    Johan Haumann

Authors
  1. Bastiaan C. ter Meulen
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Esther T. Maas
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Rien van der Vegt
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Johan Haumann
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Henry C. Weinstein
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Raymond W. J. G. Ostelo
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Johanna M. van Dongen
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

RO and HW are the principal investigators. BTM is the coordinator of the trial and is responsible for data management of the trial. He designed the study protocol together with RO, HW and EM. MvdV and JH are responsible for the section on transforaminal injections. RO calculated the sample size. EM wrote the section on economic evaluation. All authors read and corrected draft versions of the manuscript and approved the final manuscript.

Corresponding author

Correspondence to Bastiaan C. ter Meulen.

Ethics declarations

Ethics approval and consent to participate

The RCT was approved by the Medical research Ethics Committees United, Nieuwegein, The Netherlands (registration number NL45805.100.15). The protocol was registered at the Dutch Trial Register, number NTR4457 (6/03/2014). Written informed consent was obtained of all patients before entering the study. There is no identifiable information (image, face, name etc.) of participants revealed in this submission.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1.

Appendix I CONSORT-statement.

Additional file 2.

Appendix II Informed consent form.

Additional file 3.

Appendix III Cost-effectiveness analysis results (Sensitivity analyses).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

ter Meulen, B.C., Maas, E.T., van der Vegt, R. et al. Cost-effectiveness of Transforaminal epidural steroid injections for patients with ACUTE sciatica: a randomized controlled trial. BMC Musculoskelet Disord 25, 247 (2024). https://0-doi-org.brum.beds.ac.uk/10.1186/s12891-024-07366-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://0-doi-org.brum.beds.ac.uk/10.1186/s12891-024-07366-5

Keywords

BMC Musculoskeletal Disorders

ISSN: 1471-2474

Contact us