- Email: [email protected]
Cost-effectiveness analysis of the pharmacological management of chronic low back pain with four leading drugs
Journal of Orthopaedic Science 24 (2019) 805e811
Contents lists available at ScienceDirect
Journal of Orthopaedic Science journal homepage: http://www.elsevier.com/locate/jos
Original Article
Cost-effectiveness analysis of the pharmacological management of chronic low back pain with four leading drugs Takashi Kaito a, *, Yukihiro Matsuyama b, Toshihiko Yamashita c, Mamoru Kawakami d, Kazuhisa Takahashi e, Munehito Yoshida f, Shiro Imagama g, Seiji Ohtori e, Toshihiko Taguchi h, Hirotaka Haro i, Hiroshi Taneichi j, Masashi Yamazaki k, Gen Inoue l, Kotaro Nishida m, Hiroshi Yamada n, Daijiro Kabata o, Ayumi Shintani o, Motoki Iwasaki p, Manabu Ito q, Naohisa Miyakoshi r, Hideki Murakami s, Kazuo Yonenobu t, Tomoyuki Takura u, Joji Mochida v, The Project Committee of the Japanese Society for Spine Surgery and Related Research (JSSR)1 a
Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, Suita, Osaka, Japan Division of Orthopedic Surgery, Hamamatsu University School of Medicine, Hamamatsu, Japan Department of Orthopaedic Surgery, Sapporo Medical University School of Medicine, Sapporo, Japan d Spine Care Center, Wakayama Medical University Kihoku Hospital, Katsuragi-cho, Japan e Department of Orthopaedic Surgery, Graduate School of Medicine, Chiba University, Chiba, Japan f Sumiya Orthopaedic Hospital. Wakayama, Japan g Department of Orthopaedics/Rheumatology, Nagoya University Graduate School of Medicine, Japan h Department of Orthopaedic Surgery, Yamaguchi Rosai Hospital, Sanyoonoda, Japan i Department of Orthopaedic Surgery, University of Yamanashi, Cyuo, Japan j Department of Orthopaedic Surgery, Dokkyo Medical University, Mibumachi, Japan k Department of Orthopaedic Surgery, University of Tsukuba, Tsukuba, Japan l Department of Orthopaedic Surgery, Kitasato University School of Medicine, Sagamihara, Japan m Department of Orthopaedic Surgery, University of the Ryukyus, Faculty of Medicine, Nishihara, Japan n Department of Orthopaedic Surgery, Wakayama Medical University, Wakayama, Japan o Department of Medical Statistics, Osaka City University Graduate School of Medicine and Faculty of Medicine, Osaka, Japan p Department of Orthopaedic Surgery, Osaka Rosai Hospital, Sakai, Japan q Department of Orthopaedic Surgery, National Hospital Organization, Hokkaido Medical Center, Sapporo, Japan r Department of Orthopedic Surgery, Akita University Graduate School of Medicine, Akita, Japan s Department of Orthopaedic Surgery, Nagoya City University, Graduate School of Medical Sciences, Nagoya, Japan t Osaka Yukioka College of Health Science, Japan u Department of Healthcare Economics and Health Policy, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan v Department of Orthopaedic Surgery, Japan Medical Alliance, Ebina General Hospital, Ebina, Japan b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 May 2019 Received in revised form 6 June 2019 Accepted 7 June 2019 Available online 21 June 2019
Background: Chronic low back pain is a major health problem that has a substantial effect on people's quality of life and places a significant economic burden on healthcare systems. However, there has been little cost-effectiveness analysis of the treatments for it. Therefore, the purpose of this prospective observational study was to evaluate the cost-effectiveness of the pharmacological management of chronic low back pain. Methods: A total of 474 patients received pharmacological management for chronic low back pain using four leading drugs for 6 months at 28 institutions in Japan. Outcome measures, including EQ-5D, the Japanese Orthopaedic Association (JOA) score, the JOA back pain evaluation questionnaire (BPEQ), the RolandeMorris Disability Questionnaire, the Medical Outcomes Study SF-8, and the visual analog scale, were investigated at baseline and every one month thereafter. The incremental costeutility ratio (ICUR) was calculated as drug cost over the quality-adjusted life years. An economic estimation was performed
* Corresponding author. Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, 2-2, Yamadaoka, Suita, Osaka 565-0871, Japan. Fax: þ81 6 6879 3559. E-mail address: [email protected] (T. Kaito). 1 The Project Committee of the Japanese Society for Spine Surgery and Related Research (JSSR) consists of all the authors in this manuscript. https://doi.org/10.1016/j.jos.2019.06.004 0949-2658/© 2019 The Japanese Orthopaedic Association. Published by Elsevier B.V. All rights reserved.
806
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
from the perspective of a public healthcare payer in Japan. Stratified analysis based on patient characteristics was also performed to explore the characteristics that affect cost-effectiveness. Results: The ICUR of pharmacological management for chronic low back pain was JPY 453,756. Stratified analysis based on patient characteristics suggested that the pharmacological treatments for patients with a history of spine surgery or cancer, low frequency of exercise, long disease period, low scores in lumbar spine dysfunction and gait disturbance of the JOA BPEQ, and low JOA score at baseline were not costeffective. Conclusions: Pharmacological management for chronic low back pain is cost-effective from the reference willingness to pay. Further optimization based on patient characteristics is expected to contribute to the sustainable development of a universal insurance system in Japan. © 2019 The Japanese Orthopaedic Association. Published by Elsevier B.V. All rights reserved.
1. Introduction Chronic low back pain (CLBP) is the leading cause of disability worldwide, having a substantial effect on patients' quality of life and placing a significant economic burden on healthcare systems [1,2]. Most CLBP is nonspecific in nature and multifactorial in etiology [3]. Psychological factors have also been linked to the aggravation and chronicity of LBP [4]. Therefore, multimodal and interdisciplinary treatment approaches including pharmacological management, cognitive behavioral therapy, physical therapy, exercise therapy, acupuncture, and patient education have been demonstrated to be effective and are recommended in the guidelines [5,6]. However, pharmacological management acting on nociceptive receptors or the central nervous system remains the first line of treatment by alleviating both physical pain and psychological distress [5]. A variety of options exist for the pharmacological management of CLBP, including acetaminophen, non-steroidal anti-inflammatory drugs (NSAIDs), muscle relaxants, weak and strong opioids, and antidepressants. Although acetaminophen and NSAIDS are first-line drugs, the superiority of any one drug in terms of longterm effectiveness and side effects remains unclear. In light of a declining birth rate and an aging population in Japan, efficient allocation of limited healthcare resources based on cost-effective analyses is an urgent issue. The Japanese government introduced the concept of cost-effectiveness into the decisionmaking process for reimbursement prices of healthcare technology in 2018. However, no cost-effective analysis of pharmacological management of CLBP exists in Japan, and such analyses are scarce even in the world [7e10]. Therefore, the purpose of this study was to evaluate the costeffectiveness of the pharmacological management of CLBP using four leading drugs to provide evidence for contributing to the sustainable development of universal insurance system in Japan. 2. Materials and methods This multicenter prospective observational study was performed by the Project Committee of the Japanese Society for Spine Surgery and Related Research from January 2014 to June 2016 at 28 institutions. The study was approved by the local ethics committee, and informed consent was obtained from all participants. 2.1. Inclusion criteria and exclusion criteria Patients from 20 to 85 years of age with chronic (>3 months) low back pain were considered for inclusion. Patients were excluded if they had 1) hemorrhagic gastrointestinal diseases; 2) serious cardiac, hepatic, or renal function disorders; 3) dementia or another psychiatric disease preventing the reliable recording of
study information; 4) a history of drug or alcohol dependence; 5) a malignant tumor within the previous 5 years; or 6) Brief Scale for Psychiatric Problems in Orthopaedic Patients (BS-POP) scores indicating psychiatric problems, which is 10 points or more for the physician version and 15 points or more for the patient version [11]. 2.2. Study design Patients received pharmacological management for CLBP with one or a combination of four leading drugs for six months. The drugs were loxoprofen (Loxonin®), celecoxib (Celecox®), acetaminophen (Calonal®), and a combination drug of tramadol and acetaminophen (Tramacet®) [12]. Information about prescription contents, medical fees, clinical outcomes, and adverse events was collected each month. Use of medications for pain (e.g., neuropathic pain) other than these four drugs did not change during the observation period. Consultation fees and examination costs, including radiographs, computed tomography scans, and magnetic resonance imaging, were not recorded. Patients who followed up for more than three months were included in the analysis. 2.3. Data collection Background data collected at baseline included age, gender, duration of CLBP, body mass index (BMI), history of malignant tumors, smoking habits, employment status, exercise frequency, number of live-in family members, having something to do for pleasure, Center for Epidemiologic Studies Depression Scale (CESD), BS-POP, and medications used. 2.4. Drug pricing criteria An economic estimation from the perspective of a public healthcare payer in Japan was performed. Pharmaceutical pricing during the study period in Japan was used for the calculation of drug costs (Table 1). When calculating the cost of Celecox® and Calonal®, which feature several dosage forms, the dosage forms were combined to minimize the total cost.
Table 1 Official rate for Drug prices in Japan. Drugs and dosage forms
JPY
Calonal® 200 mg Calonal® 300 mg Calonal® 500 mg Loxonin® Celecox® 100 mg Celecox® 200 mg Tramacet®
7.6 8.5 15.2 15.9 68.5 105.7 211.4
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
2.5. Clinical outcomes Outcome measures included the EuroQol five-dimensions three-level (EQ-5D) for utility-based evaluation of quality of life (QOL), the Medical Outcomes Study (MOS) short-form 8-item health survey (SF-8) for comprehensive QOL measures, and the RolandeMorris Disability Questionnaire (RDQ), Japanese Orthopaedic Association (JOA) score, JOA Back Pain Evaluation Questionnaire (JOA BPEQ) consisting of 5 domains, and the visual analog scale (VAS) for disease-specific indicators of low back pain [13]. These outcomes were recorded at baseline and every month thereafter in a six-month follow-up period. 2.6. Cost-effectiveness analysis QOL and quality-adjusted life years (QALY) were calculated based on the EQ-5D questionnaire using the EQ-5D-3L index value calculator with the EuroQol group value set of Japan. Drug costs over six months and the mean acquired score of EQ-5D from baseline for six months were used to calculate the incremental costeutility ratio (ICUR ¼ drug cost/QALY). The reference willingness to pay (WTP) threshold was assumed to be JPY 5,000,000 (USD 50,000) [14,15]. Annual discount rates were not applied as this study considered follow-ups shorter than one year. 2.7. Time-dependent changes of clinical outcomes Time-dependent changes of clinical outcomes during the sixmonth observation period were compared to elucidate the differences between outcome measures (EQ-5D, VAS, RDQ, JOA score, JOA BPEQ [pain-related disorders, lumbar spine dysfunction, gait disturbance, social life disturbance, psychological disorders], SF-8 [physical component summary, FCS; mental component summary, MCS]). To standardize the units, each outcome value was normalized using a patient's mean and standard deviation, calculated from the outcome values of all observation periods for each patient. Then, the means of the difference between the normalized outcome value and the baseline for each observation period were calculated [16]. Standardized valuek ¼
Valuek Meanall SDall
Meanall ; mean value for each patient in all observation periods SDall ; standard deviation for each patient in all observation periods k ¼ 0 6 As a negative change indicates a clinical improvement in the RDQ, we reversed the sign of values in RDQ outcomes.
as median and inter-quartile range (IQR), and the categorical variables were expressed as counts and percentages. Mann Whitney U test was applied to continuous variables, and the chi-square test was applied to categorical variables. All statistical inferences were made with a two-tailed significance level of 5%. Data cleaning and analysis were conducted with R software (https://www.r-project. org/foundation/) using the “Right Management Services” package. 3. Results 3.1. Baseline demographics Among the 602 patients who met the criteria and registered for this study, 56 patients lacked in baseline data, and 72 patients dropped out before their third month of follow-up. The data of the 474 patients (79%) who followed up for more than three months (427 completed all six months of follow-up, so 47 dropped out after the third month) were used for the following analysis (Fig. 1). The patient demographics of all participants, both those included in and excluded from the analysis, are given in Table 2. Significant differences were found only in BMI (median, excluded ¼ 22.8 vs. included ¼ 23.7, p ¼ 0.046), lumbar spine dysfunction in the JOA BPEQ (median, excluded ¼ 50.0 vs. included ¼ 33.0, p ¼ 0.001), VAS for low back pain (median, excluded ¼ 50.0 vs. included ¼ 54.5, p ¼ 0.004). There were no significant differences in the major background characteristics between the included and the excluded patients (Table 2). 3.2. Mean prescribed dose of each drug The mean prescribed doses were 138.6 mg (2.3 tablets) for Loxonin®, 209.0 mg for Celecox®, 1198.5 mg for Calonal®, and 2.718 tablets (101.9 mg tramadol and 883.4 mg acetaminophen) for Tramacet®. In 32.1% (152/474) of patients, the prescribed medications were changed during the observation period. 3.3. Adverse events There were no adverse events requiring additional medical intervention. No death or hospitalization was reported during the observation period. 3.4. Cost-effectiveness analysis Total mean EQ-5D gain during the six-month (0.5-year) observation period was 0.0405, equal to the acquired QALY per 0.5 year,
2.8. Stratified analysis of cost by the median value of each factor Stratified analysis with the medians of all continuous variables (age, disease period, CES-D, live-in family members, JOA-BPEQ pain, EQ-5D, RDQ, and JOA score) and categorical variables (gender, history of spine surgery or malignant tumor, smoking habits, frequency of exercise, employment status, having something to do for pleasure) was performed to determine the trends of background characteristics with regard to drug costs, QALY, and ICUR. 2.9. Statistical analysis Last observation carried forward was used to determine the missing values of patients who dropped out four to five months after prescription based on the assumption that the CLBP values would have been stable. The continuous variables were expressed
807
Fig. 1. Study population schema.
808
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
Table 2 Participant characteristics.
Final follow-up (months) N Age (median [IQR]) Female (% [N]) Duration of disease (days) (median [IQR]) BMI (median [IQR]) History of malignant tumor (% [N]) Smoking habit Unemployment (% [N]) Exercise Rarely (% [N]) Occasionally (% [N]) Frequently (% [N]) Every day (% [N]) Live-in family members (median [IQR]) Having something to do for pleasure (% [N]) Baseline outcomes EQ-5D (median [IQR]) JOA score (median [IQR]) RDQ (median [IQR]) JOA BPEQ Pain-related disorders (median [IQR]) Lumbar spine dysfunction (median [IQR]) Gait disturbance (median [IQR]) Social life disturbance (median [IQR]) Psychological disorders (median [IQR]) VAS (median [IQR]) SF-8 PCS (median [IQR]) MCS (median [IQR]) BS-POP by physician (median [IQR]) BS-POP by patients (median [IQR]) CES-D (median [IQR]) Baseline medications Acetaminophen (% [N]) NSAIDs (% [N]) Cox-2 inhibitors (% [N]) Tramacet® (% [N]) Muscle relaxants (% [N]) Drugs for neuropathic pain (% [N]) Antidepressants (% [N]) Weak opioids (% [N]) Strong opioids (% [N])
Excluded from analysis
Included in analysis
Overall
p
Missing (%)
<3 96 73.00 [63.25, 78.00] 52.1 (50) 730.00 [300.00, 1460.00] 22.81 [21.09, 24.72] 11.5 (11) 5.2 (5) 78.1 (75)
3 474 73.00 [66.00, 78.00] 58.1 (272) 870.00 [365.00, 2190.00] 23.68 [21.46, 26.40] 6.1 (29) 9.9 (47) 73.2 (342)
570 73.00 [65.00, 78.00] 57.1 (322) 730.00 [360.00, 1825.00] 23.42 [21.38, 26.07] 7.0 (40) 9.1 (52) 74.1 (417)
0.78 0.276 0.104 0.046 0.062 0.144 0.319
2.5 1.1 6 5.4 0 0 1.2
43.7 17.2 13.8 25.3 2.00 57.3
55.7 12.5 13.4 18.4 2.00 58.6
53.7 13.3 13.5 19.5 2.00 58.4
(38) (15) (12) (22) [1.00, 3.00] (55)
(245) (55) (59) (81) [1.00, 3.00] (269)
(283) (70) (71) (103) [1.00, 3.00] (324)
0.176
7.5
0.938 0.812
12.8 2.6
0.60 [0.53, 0.77] 18.00 [16.00, 21.00] 11.00 [7.00, 15.00]
0.65 [0.58, 0.72] 19.00 [16.00, 22.00] 10.00 [6.00, 14.00]
0.65 [0.55, 0.72] 19.00 [16.00, 22.00] 10.00 [6.00, 14.00]
0.104 0.269 0.13
2.5 5.8 6.5
43.00 50.00 50.00 51.00 48.00 50.00
43.00 33.00 43.00 51.00 50.00 54.50
43.00 33.00 43.00 51.00 50.00 53.00
71.00] 66.70] 71.00] 65.00] 63.00] 70.00]
0.197 0.001 0.888 0.343 0.177 0.004
3 1.6 3.5 3.3 3.5 3.2
[14.00, [33.30, [25.00, [31.00, [39.00, [15.00,
71.00] 81.22] 71.00] 59.00] 60.25] 65.00]
[14.00, [33.00, [29.00, [32.00, [42.00, [40.00,
71.00] 58.30] 71.00] 65.00] 63.00] 71.75]
[14.00, [33.00, [29.00, [32.00, [40.25, [39.75,
36.72 [31.46, 40.11] 47.77 [43.97, 53.06] 8.50 [8.00, 9.00] 13.00 [11.00, 14.00] 15.00 [10.00, 22.00]
37.18 [29.69, 41.77] 49.37 [44.12, 54.92] 8.00 [8.00, 9.00] 13.00 [11.00, 14.00] 14.00 [10.00, 22.00]
37.10 [29.80, 41.27] 49.01 [44.12, 54.81] 8.00 [8.00, 9.00] 13.00 [11.00, 14.00] 14.00 [10.00, 22.00]
0.455 0.085 0.712 0.527 0.818
2.8 2.8 4.7 7.4 4.2
4.2 (4) 17.7 (17) 20.8 (20) 11.5 (11) 4.2 (4) 16.7 (16) 5.2 (5) 0.0 (0) 0.0 (0)
5.3 (25) 12.2 (58) 4.6 (22) 4.2 (20) 6.8 (32) 17.3 (82) 1.5 (7) 1.1 (5) 0.0 (0)
5.1 (29) 13.2 (75) 7.4 (42) 5.4 (31) 6.3 (36) 17.2 (98) 2.1 (12) 0.9 (5) 0.0 (0)
0.652 0.148 <0.001 0.004 0.342 0.881 0.02 0.312 NA
0 0 0 0 0 0 0 0 0
IQR, interquartile range; VAS, visual analog scale; JOA, Japanese Orthopaedic Association; JOA BPEQ, JOA Back Pain Evaluation Questionnaire; RDQ, RolandeMorris Disability Questionnaire; SF-8, the MOS short-form 8-item health survey; PCS, physical component summary; MCS, mental component summary; BS-POP, Brief Scale for Psychiatric Problems in Orthopaedic Patients; CES-D, Center for Epidemiologic Studies Depression Scale.
and total drug cost for the period was JPY 18,377. Thus, the ICUR (drug cost per 0.5 year over the acquired QALY per 0.5 year) was JPY 453,756 (Table 3; Fig. 2). From the reference WTP, therefore, the pharmacological management of CLBP was determined to be costeffective.
3.6. Stratified analysis of ICUR based on median values or categories of each factor Stratified analysis indicated which factors negatively affect costeffectiveness (Table 4). Negative ICUR (mean acquired EQ-5D score
3.5. Changes in clinical outcomes during the observation period Sustained improvements in outcomes were observed during the observation period except for EQ-5D, psychological disorders in the JOA BPEQ, and the MCS of SF-8. EQ-5D featured repeated improvements and aggravation, while the other two outcomes had a decreasing trend during the final two-month follow-up period, the MCS of SF-8 in particular falling below the baseline (Fig. 3).
Table 3 ICUR. Drug cost (JPY) for 6 months
Acquired QALY for 6 months
ICUR (JPY/QALY)
18,377
0.0405
453,756
ICUR, incremental costeutility ratio.
Fig. 2. Adjusted temporal change of EQ-5D.
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
809
being negative) was found in patients with a history of spine surgery or cancer, low frequency of exercise (rarely or occasionally), long disease period (870 days), low scores in lumbar spine dysfunction (<33) and gait disturbance (<43) in the JOA BPEQ at baseline, and low JOA scores (<19) at baseline. Furthermore, the ICUR of patients who were male, unemployed, older, and had lower scores in pain-related and psychological disorders on the JOA BPEQ at baseline was more than 3 times greater than that of others. Because drug costs did not differ between the stratified groups, the acquired QALY influenced the difference in ICUR. 4. Discussion In this cost-effectiveness analysis, the ICUR of pharmacological management of CLBP over a six-month follow-up period with four leading drugs was JPY 453,756 per QALY and was demonstrated to
Fig. 3. Adjusted temporal change of clinical outcomes.
Table 4 Stratified analysis of each factor.
Gender Spine surgery Cancer history Smoking habit Exercise
Unemployment Having something to do for pleasure
Age
<
Disease period (days) Live-in family members EQ-5D JOA BPEQ Pain-related disorders Lumbar spine dysfunction Gait disturbance Social life disturbance Psychological disorders RDQ JOA score CES-D SF-8 PCS MCS
Categories
Drug cost
QALY
ICUR
female male No Yes No Yes No Yes Rarely Occasionally Frequently Every day No Yes No Yes
20188.97 17207.69 16875.46 22959.12 18684.60 13658.88 17868.75 22995.74 17030.74 18555.88 19785.77 18357.10 18802.13 18133.19 19837.71 17125.90
0.061 0.016 0.061 0.023 0.046 0.038 0.042 0.025 0.002 0.060 0.125 0.039 0.095 0.027 0.083 0.019
281009.3 1289357.6 274604.1 979655.9 409770.2 362072.8 422970.0 933333.3 6967121.2 307873.2 158033.3 464760.6 196971.7 666117.3 238660.4 879844.9
Median values
Drug cost
QALY
ICUR
73
19338.68 17797.81 18320.64 18517.46 19420.87 17812.52 20184.48 17152.14
0.077 0.010 0.101 0.007 0.078 0.019 0.063 0.027
252569.8 1722845.7 181693.7 2591977.3 247964.3 917690.6 321057.7 636323.6
21381.38 16651.35 18024.10 18299.85 17925.91 18356.73 18269.01 18217.34 19146.17 17595.88 16952.28 19212.16 19809.58 17299.05 16597.64 18907.74
0.006 0.053 0.033 0.048 0.064 0.108 0.067 0.028 0.006 0.054 0.073 0.023 0.025 0.066 0.027 0.048
3348890.8 315102.1 546516.2 380069.7 282010.5 170340.5 272814.2 660665.5 3198717.8 327816.6 232573.4 851283.4 786156.4 260302.4 610273.7 391140.7
19875.46 16928.44 18149.09 18654.81
0.027 0.048 0.062 0.013
724806.7 355250.0 293757.2 1403534.6
870 < < < < < < < < < < < <
2 0.649
43 33 43 51 50 10 19 14
37.18 49.37
QALY, quality-adjusted life years; ICUR, incremental costeutility ratio; VAS, visual analog scale; JOA, Japanese Orthopaedic Association; JOA BPEQ, JOA Back Pain Evaluation Questionnaire; RDQ, RolandeMorris Disability Questionnaire; SF-8, the MOS short-form 8-item health survey; PCS, physical component summary; MCS, mental component summary; BS-POP, Brief Scale for Psychiatric Problems in Orthopaedic Patients; CES-D, Center for Epidemiologic Studies Depression Scale.
810
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
be highly cost-effective based on the reported WTP of JPY 5,000,000e6,000,000 per QALY [14,15]. This is the first costeffectiveness analysis of pharmacological management for CLBP in Japan and the first such study that was not sponsored by a pharmaceutical company in the world [7e10]. The mean acquired EQ-5D score had a positive value, but there were increases and decreases over time. This result indicates the limits of any one approach when it comes to the pharmacological treatment of CLBP [5,6]. Some studies have reported a ceiling effect in EQ-5D-3L, which was used in this study. The stratified analysis showed that patients with higher EQ-5D at baseline (>0.69) had less cost-effective CLBP management (Table 3). Recently, two further levels have been included (EQ-5D-5L), so further research using EQ-5D-5L may be able to show a greater change in EQ-5D (and thus greater cost-effectiveness). The sustained improvement in other clinical outcomes (JOA score, RDQ, pain-related disorders/ lumbar spine dysfunction/gait disturbance/social life disturbance in the JOA BPEQ, VAS, and PCS in the SF-8) except those including psychological elements (psychological disorders in the JOA BPEQ, MCS in the SF-8) support the effectiveness of the pharmacological treatment of CLBP. These results suggest that a multidisciplinary approach plays a large role among patients with poor clinical outcomes in psychological elements at baseline [4]. Stratified analysis of cost-effectiveness based on median values or categories of each factor showed interesting trends. Most notably, the ICUR of patients with a history of spine surgery or cancer, low frequency of exercise, long disease period, low scores in lumbar spine dysfunction and gait disturbance in the JOA BPEQ at baseline, and low JOA score at baseline demonstrated negative values. As to the history of spine surgery, decompressive or fusion surgery can alleviate nociceptive or neuropathic pain, so other factors might influence the ineffectiveness of pharmacological treatment of CLBP with the four leading drugs. The negative ICUR of CLBP patients with a history of cancer and older adults may be associated with frailty. History of cancer and older age are reported to relate with frailty [17,18], and an association between physical frailty and chronic pain has also been reported [19]. Other factors, including low frequency of exercise, long disease period, low scores in lumbar spine dysfunction and gait disturbance in the JOA-BPEQ at baseline, and low JOA score at baseline, may also be related to reduced muscle strength, which is one of the main causes of frailty. In patients with long disease duration, there is a possibility that all sorts of pharmacological treatments have been attempted to little or no effect, which suggests that continued discursive administration should be avoided. Finally, a higher frequency of exercise correlated with low ICUR, suggesting that the combination of exercise and pharmacological treatment has additive effects and, on the other hand, that the pharmacological treatment alone without exercise might be ineffective [3,20]. There were several limitations in this study. First, the six-month observation period was relatively short. However, under the current concept that recognizes the importance of multimodal and interdisciplinary management, a continuous approach with pharmacological treatment for a longer period is not realistic. In this study, a Markov model was not used for the estimation of longterm cost-effectiveness because the incidence of adverse events or effects on life expectancy were too low to establish a statistical model. Second, in this observation study, only the costs of the four leading CLBP-treating drugs were calculated, and the costs of gastric mucosal coating agents, agents for inhibiting gastric acid, antiemetic agents, laxative agents, and weak opioids were not included in the analysis because of the use of a wide variety of drugs and the high incidence of prescriptions for the treatment of other diseases. However, the price of the most expensive agents for inhibiting gastric acid is JPY 200 per day, the total cost for a six-
month period being about JPY 36,000. Therefore, its effect on ICUR would be limited. In conclusion, this cost-effectiveness analysis of the pharmacological management of CLBP demonstrated that it is indeed costeffective from the reference WTP. Further optimization based on patient characteristics is expected to provide evidence contributing to the sustainable development of a universal insurance system in Japan. Conflict of interest The authors declare that there is no conflict of interest relevant to the submitted work. Acknowledgements The Board of Directors and the Board of Councilors of the Japanese Society for Spine Surgery and Related Research (JSSR) deeply appreciate the following; Daiichi Sankyo Company, Ltd., Showa Yakuhin Kako Co., Ltd., and Astellas Pharma Inc. agreed with the overall research activities of JSSR and provided funds targeting all these research activities in accordance with the time when this new research project was planned. We wish to confirm that there were no known conflicts of interest associated with this publication. The authors would like to express their sincere gratitude to all supporting members at the JSSR. References [1] Hoy D, March L, Brooks P, Blyth F, Woolf A, Bain C, Williams G, Smith E, Vos T, Barendregt J, Murray C, Burstein R, Buchbinder R. The global burden of low back pain: estimates from the Global Burden of Disease 2010 study. Ann Rheum Dis 2014 Jun;73(6):968e74. [2] Dagenais S, Caro J, Haldeman S. A systematic review of low back pain cost of illness studies in the United States and internationally. Spine J 2008 JanFeb;8(1):8e20. F, Mannion AF, Pellise F, Cedraschi C. Clinical update: low back pain. [3] Balague Lancet 2007 Mar;369(9563):726e8. [4] Pincus T, Burton AK, Vogel S, Field AP. A systematic review of psychological factors as predictors of chronicity/disability in prospective cohorts of low back pain. Spine 2002 Mar;27(5):E109e20. [5] O'Connell NE, Cook CE, Wand BM, Ward SP. Clinical guidelines for low back pain: a critical review of consensus and inconsistencies across three major guidelines. Best Pract Res Clin Rheumatol 2016 Dec;30(6):968e80. [6] Foster NE, Anema JR, Cherkin D, Chou R, Cohen SP, Gross DP, Ferreira PH, Fritz JM, Koes BW, Peul W, Turner JA, Maher CG. Prevention and treatment of low back pain: evidence, challenges, and promising directions. Lancet Low Back Pain Series Working Group. Lancet 2018 Jun;391(10137):2368e83. [7] Haas M, De Abreu Lourenco R. Pharmacological management of chronic lower back pain: a review of cost effectiveness. Pharmacoeconomics 2015 Jun;33(6): 561e9. [8] Morera-Dominguez C, Ceberio-Balda F, Florez-Garca M, Masramon X, LopexGomez V. A cost-consequence analysis of pregabalin versus usual care in the symptomatic treatment of refractory low back pain: sub-analysis of observational trial data from orthopaedic surgery and rehabilitation clinics. Clin Drug Investig 2010;30(8):517e31. [9] Wielage RC, Bansal M, Wioson K, Klein R, Happich M. Cost-effectiveness of duloxetine in chronic low back pain: a Quebec societal perspective. Spine 2013 May;38(11):936e46. [10] Wielage RC, Bansal M, Acott Andrews J, Wohlreich MM, Klein RW, Happich M. The cost-effectiveness of duloxetine in chronic low back pain: a US private payer perspective. Value Health 2013 Mar-Apr;16(2):334e44. [11] Yoshida K, Sekiguchi M, Otani K, Mashiko H, Shiota H, Wakita T, Niwa Si, Kikuchi S, Konno S. A validation of study of the Brief Scale for Psychiatric problems in Orthopaedic Patients (BS-POP) for patients with chronic low back pain (verification of reliability, validity, and reproducibility. J Orthop Sci 2011 Jan;16(1):7e13. [12] Japanese Ministry of Health, Labour and Welfare. National database (NDB) open data. https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000177182. html. [13] Fukui M, Chiba K, Kawakami M, Kikuchi S, Konno S, Miyamoto M, Seichi A, Shimamura T, Shirado O, Taguchi T, Takahashi K, Takeshita K, Tani T, Toyama Y, Yonenobu K, Wada E, Tanaka T, Hirota Y. JOA back pain evaluation questionnaire (JOABPEQ)/JOA cervical myelopathy evaluation questionnaire (JOACMEQ). The report on the development of revised versions. April 16, 2007. The subcommittee of the clinical outcome committee of the Japanese
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
[14]
[15]
[16]
[17]
orthopaedic association on low back pain and cervical myelopathy evaluation. J Orthop Sci 2009 May;14(3):348e65. Shiroiwa T, Sung YK, Fukuda T, Lang HC, Bae SC, Tsutani K. International survey on willingness-to-pay (WTP) for one additional QALY gained: what is the threshold of cost effectiveness? Health Econ 2010 Apr;19(4):422e37. Bertram MY, Lauer JA, De Joncheere K, Edejer T, Hutubessy R, Kieny MP, Hill SR. Cost-effectiveness thresholds: pros and cons. Bull World Health Organ 2016 Dec;94(12):925e30. Hosono N, Takenaka S, Mukai Y, Tateishi K, Fujiwara Y, Morishita Y, Konishi H. Conventional JOA score for cervical myelopathy has a rater's bias -In comparison with JOACMEQ-. J Orthop Sci 2018 May;23(3):477e82. Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet 2013 Mar;381(9868):752e62.
811
[18] Giannotti C, Sambuceti S, Signori A, Ballestrero A, Murialdo R, Romairone E, Scabini S, Caffa I, Odetti P, Nencioni A, Monacelli F. Frailty assessment in elective gastrointestinal oncogeriatric surgery: predictors of one-year mortality and functional status. J Geriatr Oncol 2019 May;S1879e4068(18): 30344e8. [19] Nakai Y, Makizako H, Kiyama R, Tomioka K, Taniguchi Y, Kubozono T, Takenaka T, Ohishi M. Association between chronic pain and physical frailty in community-dwelling older adults. Int J Environ Res Public Health 2019 Apr 13;16(8):E1330. [20] Manniche C, Hesselsøe G, Bentzen L, Christensen I, Lundberg E. Clinical trial of intensive muscle training for chronic low back pain. Lancet 1988 Dec;2(8626e8627):1473e6.
Contents lists available at ScienceDirect
Journal of Orthopaedic Science journal homepage: http://www.elsevier.com/locate/jos
Original Article
Cost-effectiveness analysis of the pharmacological management of chronic low back pain with four leading drugs Takashi Kaito a, *, Yukihiro Matsuyama b, Toshihiko Yamashita c, Mamoru Kawakami d, Kazuhisa Takahashi e, Munehito Yoshida f, Shiro Imagama g, Seiji Ohtori e, Toshihiko Taguchi h, Hirotaka Haro i, Hiroshi Taneichi j, Masashi Yamazaki k, Gen Inoue l, Kotaro Nishida m, Hiroshi Yamada n, Daijiro Kabata o, Ayumi Shintani o, Motoki Iwasaki p, Manabu Ito q, Naohisa Miyakoshi r, Hideki Murakami s, Kazuo Yonenobu t, Tomoyuki Takura u, Joji Mochida v, The Project Committee of the Japanese Society for Spine Surgery and Related Research (JSSR)1 a
Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, Suita, Osaka, Japan Division of Orthopedic Surgery, Hamamatsu University School of Medicine, Hamamatsu, Japan Department of Orthopaedic Surgery, Sapporo Medical University School of Medicine, Sapporo, Japan d Spine Care Center, Wakayama Medical University Kihoku Hospital, Katsuragi-cho, Japan e Department of Orthopaedic Surgery, Graduate School of Medicine, Chiba University, Chiba, Japan f Sumiya Orthopaedic Hospital. Wakayama, Japan g Department of Orthopaedics/Rheumatology, Nagoya University Graduate School of Medicine, Japan h Department of Orthopaedic Surgery, Yamaguchi Rosai Hospital, Sanyoonoda, Japan i Department of Orthopaedic Surgery, University of Yamanashi, Cyuo, Japan j Department of Orthopaedic Surgery, Dokkyo Medical University, Mibumachi, Japan k Department of Orthopaedic Surgery, University of Tsukuba, Tsukuba, Japan l Department of Orthopaedic Surgery, Kitasato University School of Medicine, Sagamihara, Japan m Department of Orthopaedic Surgery, University of the Ryukyus, Faculty of Medicine, Nishihara, Japan n Department of Orthopaedic Surgery, Wakayama Medical University, Wakayama, Japan o Department of Medical Statistics, Osaka City University Graduate School of Medicine and Faculty of Medicine, Osaka, Japan p Department of Orthopaedic Surgery, Osaka Rosai Hospital, Sakai, Japan q Department of Orthopaedic Surgery, National Hospital Organization, Hokkaido Medical Center, Sapporo, Japan r Department of Orthopedic Surgery, Akita University Graduate School of Medicine, Akita, Japan s Department of Orthopaedic Surgery, Nagoya City University, Graduate School of Medical Sciences, Nagoya, Japan t Osaka Yukioka College of Health Science, Japan u Department of Healthcare Economics and Health Policy, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan v Department of Orthopaedic Surgery, Japan Medical Alliance, Ebina General Hospital, Ebina, Japan b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 May 2019 Received in revised form 6 June 2019 Accepted 7 June 2019 Available online 21 June 2019
Background: Chronic low back pain is a major health problem that has a substantial effect on people's quality of life and places a significant economic burden on healthcare systems. However, there has been little cost-effectiveness analysis of the treatments for it. Therefore, the purpose of this prospective observational study was to evaluate the cost-effectiveness of the pharmacological management of chronic low back pain. Methods: A total of 474 patients received pharmacological management for chronic low back pain using four leading drugs for 6 months at 28 institutions in Japan. Outcome measures, including EQ-5D, the Japanese Orthopaedic Association (JOA) score, the JOA back pain evaluation questionnaire (BPEQ), the RolandeMorris Disability Questionnaire, the Medical Outcomes Study SF-8, and the visual analog scale, were investigated at baseline and every one month thereafter. The incremental costeutility ratio (ICUR) was calculated as drug cost over the quality-adjusted life years. An economic estimation was performed
* Corresponding author. Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, 2-2, Yamadaoka, Suita, Osaka 565-0871, Japan. Fax: þ81 6 6879 3559. E-mail address: [email protected] (T. Kaito). 1 The Project Committee of the Japanese Society for Spine Surgery and Related Research (JSSR) consists of all the authors in this manuscript. https://doi.org/10.1016/j.jos.2019.06.004 0949-2658/© 2019 The Japanese Orthopaedic Association. Published by Elsevier B.V. All rights reserved.
806
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
from the perspective of a public healthcare payer in Japan. Stratified analysis based on patient characteristics was also performed to explore the characteristics that affect cost-effectiveness. Results: The ICUR of pharmacological management for chronic low back pain was JPY 453,756. Stratified analysis based on patient characteristics suggested that the pharmacological treatments for patients with a history of spine surgery or cancer, low frequency of exercise, long disease period, low scores in lumbar spine dysfunction and gait disturbance of the JOA BPEQ, and low JOA score at baseline were not costeffective. Conclusions: Pharmacological management for chronic low back pain is cost-effective from the reference willingness to pay. Further optimization based on patient characteristics is expected to contribute to the sustainable development of a universal insurance system in Japan. © 2019 The Japanese Orthopaedic Association. Published by Elsevier B.V. All rights reserved.
1. Introduction Chronic low back pain (CLBP) is the leading cause of disability worldwide, having a substantial effect on patients' quality of life and placing a significant economic burden on healthcare systems [1,2]. Most CLBP is nonspecific in nature and multifactorial in etiology [3]. Psychological factors have also been linked to the aggravation and chronicity of LBP [4]. Therefore, multimodal and interdisciplinary treatment approaches including pharmacological management, cognitive behavioral therapy, physical therapy, exercise therapy, acupuncture, and patient education have been demonstrated to be effective and are recommended in the guidelines [5,6]. However, pharmacological management acting on nociceptive receptors or the central nervous system remains the first line of treatment by alleviating both physical pain and psychological distress [5]. A variety of options exist for the pharmacological management of CLBP, including acetaminophen, non-steroidal anti-inflammatory drugs (NSAIDs), muscle relaxants, weak and strong opioids, and antidepressants. Although acetaminophen and NSAIDS are first-line drugs, the superiority of any one drug in terms of longterm effectiveness and side effects remains unclear. In light of a declining birth rate and an aging population in Japan, efficient allocation of limited healthcare resources based on cost-effective analyses is an urgent issue. The Japanese government introduced the concept of cost-effectiveness into the decisionmaking process for reimbursement prices of healthcare technology in 2018. However, no cost-effective analysis of pharmacological management of CLBP exists in Japan, and such analyses are scarce even in the world [7e10]. Therefore, the purpose of this study was to evaluate the costeffectiveness of the pharmacological management of CLBP using four leading drugs to provide evidence for contributing to the sustainable development of universal insurance system in Japan. 2. Materials and methods This multicenter prospective observational study was performed by the Project Committee of the Japanese Society for Spine Surgery and Related Research from January 2014 to June 2016 at 28 institutions. The study was approved by the local ethics committee, and informed consent was obtained from all participants. 2.1. Inclusion criteria and exclusion criteria Patients from 20 to 85 years of age with chronic (>3 months) low back pain were considered for inclusion. Patients were excluded if they had 1) hemorrhagic gastrointestinal diseases; 2) serious cardiac, hepatic, or renal function disorders; 3) dementia or another psychiatric disease preventing the reliable recording of
study information; 4) a history of drug or alcohol dependence; 5) a malignant tumor within the previous 5 years; or 6) Brief Scale for Psychiatric Problems in Orthopaedic Patients (BS-POP) scores indicating psychiatric problems, which is 10 points or more for the physician version and 15 points or more for the patient version [11]. 2.2. Study design Patients received pharmacological management for CLBP with one or a combination of four leading drugs for six months. The drugs were loxoprofen (Loxonin®), celecoxib (Celecox®), acetaminophen (Calonal®), and a combination drug of tramadol and acetaminophen (Tramacet®) [12]. Information about prescription contents, medical fees, clinical outcomes, and adverse events was collected each month. Use of medications for pain (e.g., neuropathic pain) other than these four drugs did not change during the observation period. Consultation fees and examination costs, including radiographs, computed tomography scans, and magnetic resonance imaging, were not recorded. Patients who followed up for more than three months were included in the analysis. 2.3. Data collection Background data collected at baseline included age, gender, duration of CLBP, body mass index (BMI), history of malignant tumors, smoking habits, employment status, exercise frequency, number of live-in family members, having something to do for pleasure, Center for Epidemiologic Studies Depression Scale (CESD), BS-POP, and medications used. 2.4. Drug pricing criteria An economic estimation from the perspective of a public healthcare payer in Japan was performed. Pharmaceutical pricing during the study period in Japan was used for the calculation of drug costs (Table 1). When calculating the cost of Celecox® and Calonal®, which feature several dosage forms, the dosage forms were combined to minimize the total cost.
Table 1 Official rate for Drug prices in Japan. Drugs and dosage forms
JPY
Calonal® 200 mg Calonal® 300 mg Calonal® 500 mg Loxonin® Celecox® 100 mg Celecox® 200 mg Tramacet®
7.6 8.5 15.2 15.9 68.5 105.7 211.4
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
2.5. Clinical outcomes Outcome measures included the EuroQol five-dimensions three-level (EQ-5D) for utility-based evaluation of quality of life (QOL), the Medical Outcomes Study (MOS) short-form 8-item health survey (SF-8) for comprehensive QOL measures, and the RolandeMorris Disability Questionnaire (RDQ), Japanese Orthopaedic Association (JOA) score, JOA Back Pain Evaluation Questionnaire (JOA BPEQ) consisting of 5 domains, and the visual analog scale (VAS) for disease-specific indicators of low back pain [13]. These outcomes were recorded at baseline and every month thereafter in a six-month follow-up period. 2.6. Cost-effectiveness analysis QOL and quality-adjusted life years (QALY) were calculated based on the EQ-5D questionnaire using the EQ-5D-3L index value calculator with the EuroQol group value set of Japan. Drug costs over six months and the mean acquired score of EQ-5D from baseline for six months were used to calculate the incremental costeutility ratio (ICUR ¼ drug cost/QALY). The reference willingness to pay (WTP) threshold was assumed to be JPY 5,000,000 (USD 50,000) [14,15]. Annual discount rates were not applied as this study considered follow-ups shorter than one year. 2.7. Time-dependent changes of clinical outcomes Time-dependent changes of clinical outcomes during the sixmonth observation period were compared to elucidate the differences between outcome measures (EQ-5D, VAS, RDQ, JOA score, JOA BPEQ [pain-related disorders, lumbar spine dysfunction, gait disturbance, social life disturbance, psychological disorders], SF-8 [physical component summary, FCS; mental component summary, MCS]). To standardize the units, each outcome value was normalized using a patient's mean and standard deviation, calculated from the outcome values of all observation periods for each patient. Then, the means of the difference between the normalized outcome value and the baseline for each observation period were calculated [16]. Standardized valuek ¼
Valuek Meanall SDall
Meanall ; mean value for each patient in all observation periods SDall ; standard deviation for each patient in all observation periods k ¼ 0 6 As a negative change indicates a clinical improvement in the RDQ, we reversed the sign of values in RDQ outcomes.
as median and inter-quartile range (IQR), and the categorical variables were expressed as counts and percentages. Mann Whitney U test was applied to continuous variables, and the chi-square test was applied to categorical variables. All statistical inferences were made with a two-tailed significance level of 5%. Data cleaning and analysis were conducted with R software (https://www.r-project. org/foundation/) using the “Right Management Services” package. 3. Results 3.1. Baseline demographics Among the 602 patients who met the criteria and registered for this study, 56 patients lacked in baseline data, and 72 patients dropped out before their third month of follow-up. The data of the 474 patients (79%) who followed up for more than three months (427 completed all six months of follow-up, so 47 dropped out after the third month) were used for the following analysis (Fig. 1). The patient demographics of all participants, both those included in and excluded from the analysis, are given in Table 2. Significant differences were found only in BMI (median, excluded ¼ 22.8 vs. included ¼ 23.7, p ¼ 0.046), lumbar spine dysfunction in the JOA BPEQ (median, excluded ¼ 50.0 vs. included ¼ 33.0, p ¼ 0.001), VAS for low back pain (median, excluded ¼ 50.0 vs. included ¼ 54.5, p ¼ 0.004). There were no significant differences in the major background characteristics between the included and the excluded patients (Table 2). 3.2. Mean prescribed dose of each drug The mean prescribed doses were 138.6 mg (2.3 tablets) for Loxonin®, 209.0 mg for Celecox®, 1198.5 mg for Calonal®, and 2.718 tablets (101.9 mg tramadol and 883.4 mg acetaminophen) for Tramacet®. In 32.1% (152/474) of patients, the prescribed medications were changed during the observation period. 3.3. Adverse events There were no adverse events requiring additional medical intervention. No death or hospitalization was reported during the observation period. 3.4. Cost-effectiveness analysis Total mean EQ-5D gain during the six-month (0.5-year) observation period was 0.0405, equal to the acquired QALY per 0.5 year,
2.8. Stratified analysis of cost by the median value of each factor Stratified analysis with the medians of all continuous variables (age, disease period, CES-D, live-in family members, JOA-BPEQ pain, EQ-5D, RDQ, and JOA score) and categorical variables (gender, history of spine surgery or malignant tumor, smoking habits, frequency of exercise, employment status, having something to do for pleasure) was performed to determine the trends of background characteristics with regard to drug costs, QALY, and ICUR. 2.9. Statistical analysis Last observation carried forward was used to determine the missing values of patients who dropped out four to five months after prescription based on the assumption that the CLBP values would have been stable. The continuous variables were expressed
807
Fig. 1. Study population schema.
808
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
Table 2 Participant characteristics.
Final follow-up (months) N Age (median [IQR]) Female (% [N]) Duration of disease (days) (median [IQR]) BMI (median [IQR]) History of malignant tumor (% [N]) Smoking habit Unemployment (% [N]) Exercise Rarely (% [N]) Occasionally (% [N]) Frequently (% [N]) Every day (% [N]) Live-in family members (median [IQR]) Having something to do for pleasure (% [N]) Baseline outcomes EQ-5D (median [IQR]) JOA score (median [IQR]) RDQ (median [IQR]) JOA BPEQ Pain-related disorders (median [IQR]) Lumbar spine dysfunction (median [IQR]) Gait disturbance (median [IQR]) Social life disturbance (median [IQR]) Psychological disorders (median [IQR]) VAS (median [IQR]) SF-8 PCS (median [IQR]) MCS (median [IQR]) BS-POP by physician (median [IQR]) BS-POP by patients (median [IQR]) CES-D (median [IQR]) Baseline medications Acetaminophen (% [N]) NSAIDs (% [N]) Cox-2 inhibitors (% [N]) Tramacet® (% [N]) Muscle relaxants (% [N]) Drugs for neuropathic pain (% [N]) Antidepressants (% [N]) Weak opioids (% [N]) Strong opioids (% [N])
Excluded from analysis
Included in analysis
Overall
p
Missing (%)
<3 96 73.00 [63.25, 78.00] 52.1 (50) 730.00 [300.00, 1460.00] 22.81 [21.09, 24.72] 11.5 (11) 5.2 (5) 78.1 (75)
3 474 73.00 [66.00, 78.00] 58.1 (272) 870.00 [365.00, 2190.00] 23.68 [21.46, 26.40] 6.1 (29) 9.9 (47) 73.2 (342)
570 73.00 [65.00, 78.00] 57.1 (322) 730.00 [360.00, 1825.00] 23.42 [21.38, 26.07] 7.0 (40) 9.1 (52) 74.1 (417)
0.78 0.276 0.104 0.046 0.062 0.144 0.319
2.5 1.1 6 5.4 0 0 1.2
43.7 17.2 13.8 25.3 2.00 57.3
55.7 12.5 13.4 18.4 2.00 58.6
53.7 13.3 13.5 19.5 2.00 58.4
(38) (15) (12) (22) [1.00, 3.00] (55)
(245) (55) (59) (81) [1.00, 3.00] (269)
(283) (70) (71) (103) [1.00, 3.00] (324)
0.176
7.5
0.938 0.812
12.8 2.6
0.60 [0.53, 0.77] 18.00 [16.00, 21.00] 11.00 [7.00, 15.00]
0.65 [0.58, 0.72] 19.00 [16.00, 22.00] 10.00 [6.00, 14.00]
0.65 [0.55, 0.72] 19.00 [16.00, 22.00] 10.00 [6.00, 14.00]
0.104 0.269 0.13
2.5 5.8 6.5
43.00 50.00 50.00 51.00 48.00 50.00
43.00 33.00 43.00 51.00 50.00 54.50
43.00 33.00 43.00 51.00 50.00 53.00
71.00] 66.70] 71.00] 65.00] 63.00] 70.00]
0.197 0.001 0.888 0.343 0.177 0.004
3 1.6 3.5 3.3 3.5 3.2
[14.00, [33.30, [25.00, [31.00, [39.00, [15.00,
71.00] 81.22] 71.00] 59.00] 60.25] 65.00]
[14.00, [33.00, [29.00, [32.00, [42.00, [40.00,
71.00] 58.30] 71.00] 65.00] 63.00] 71.75]
[14.00, [33.00, [29.00, [32.00, [40.25, [39.75,
36.72 [31.46, 40.11] 47.77 [43.97, 53.06] 8.50 [8.00, 9.00] 13.00 [11.00, 14.00] 15.00 [10.00, 22.00]
37.18 [29.69, 41.77] 49.37 [44.12, 54.92] 8.00 [8.00, 9.00] 13.00 [11.00, 14.00] 14.00 [10.00, 22.00]
37.10 [29.80, 41.27] 49.01 [44.12, 54.81] 8.00 [8.00, 9.00] 13.00 [11.00, 14.00] 14.00 [10.00, 22.00]
0.455 0.085 0.712 0.527 0.818
2.8 2.8 4.7 7.4 4.2
4.2 (4) 17.7 (17) 20.8 (20) 11.5 (11) 4.2 (4) 16.7 (16) 5.2 (5) 0.0 (0) 0.0 (0)
5.3 (25) 12.2 (58) 4.6 (22) 4.2 (20) 6.8 (32) 17.3 (82) 1.5 (7) 1.1 (5) 0.0 (0)
5.1 (29) 13.2 (75) 7.4 (42) 5.4 (31) 6.3 (36) 17.2 (98) 2.1 (12) 0.9 (5) 0.0 (0)
0.652 0.148 <0.001 0.004 0.342 0.881 0.02 0.312 NA
0 0 0 0 0 0 0 0 0
IQR, interquartile range; VAS, visual analog scale; JOA, Japanese Orthopaedic Association; JOA BPEQ, JOA Back Pain Evaluation Questionnaire; RDQ, RolandeMorris Disability Questionnaire; SF-8, the MOS short-form 8-item health survey; PCS, physical component summary; MCS, mental component summary; BS-POP, Brief Scale for Psychiatric Problems in Orthopaedic Patients; CES-D, Center for Epidemiologic Studies Depression Scale.
and total drug cost for the period was JPY 18,377. Thus, the ICUR (drug cost per 0.5 year over the acquired QALY per 0.5 year) was JPY 453,756 (Table 3; Fig. 2). From the reference WTP, therefore, the pharmacological management of CLBP was determined to be costeffective.
3.6. Stratified analysis of ICUR based on median values or categories of each factor Stratified analysis indicated which factors negatively affect costeffectiveness (Table 4). Negative ICUR (mean acquired EQ-5D score
3.5. Changes in clinical outcomes during the observation period Sustained improvements in outcomes were observed during the observation period except for EQ-5D, psychological disorders in the JOA BPEQ, and the MCS of SF-8. EQ-5D featured repeated improvements and aggravation, while the other two outcomes had a decreasing trend during the final two-month follow-up period, the MCS of SF-8 in particular falling below the baseline (Fig. 3).
Table 3 ICUR. Drug cost (JPY) for 6 months
Acquired QALY for 6 months
ICUR (JPY/QALY)
18,377
0.0405
453,756
ICUR, incremental costeutility ratio.
Fig. 2. Adjusted temporal change of EQ-5D.
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
809
being negative) was found in patients with a history of spine surgery or cancer, low frequency of exercise (rarely or occasionally), long disease period (870 days), low scores in lumbar spine dysfunction (<33) and gait disturbance (<43) in the JOA BPEQ at baseline, and low JOA scores (<19) at baseline. Furthermore, the ICUR of patients who were male, unemployed, older, and had lower scores in pain-related and psychological disorders on the JOA BPEQ at baseline was more than 3 times greater than that of others. Because drug costs did not differ between the stratified groups, the acquired QALY influenced the difference in ICUR. 4. Discussion In this cost-effectiveness analysis, the ICUR of pharmacological management of CLBP over a six-month follow-up period with four leading drugs was JPY 453,756 per QALY and was demonstrated to
Fig. 3. Adjusted temporal change of clinical outcomes.
Table 4 Stratified analysis of each factor.
Gender Spine surgery Cancer history Smoking habit Exercise
Unemployment Having something to do for pleasure
Age
<
Disease period (days) Live-in family members EQ-5D JOA BPEQ Pain-related disorders Lumbar spine dysfunction Gait disturbance Social life disturbance Psychological disorders RDQ JOA score CES-D SF-8 PCS MCS
Categories
Drug cost
QALY
ICUR
female male No Yes No Yes No Yes Rarely Occasionally Frequently Every day No Yes No Yes
20188.97 17207.69 16875.46 22959.12 18684.60 13658.88 17868.75 22995.74 17030.74 18555.88 19785.77 18357.10 18802.13 18133.19 19837.71 17125.90
0.061 0.016 0.061 0.023 0.046 0.038 0.042 0.025 0.002 0.060 0.125 0.039 0.095 0.027 0.083 0.019
281009.3 1289357.6 274604.1 979655.9 409770.2 362072.8 422970.0 933333.3 6967121.2 307873.2 158033.3 464760.6 196971.7 666117.3 238660.4 879844.9
Median values
Drug cost
QALY
ICUR
73
19338.68 17797.81 18320.64 18517.46 19420.87 17812.52 20184.48 17152.14
0.077 0.010 0.101 0.007 0.078 0.019 0.063 0.027
252569.8 1722845.7 181693.7 2591977.3 247964.3 917690.6 321057.7 636323.6
21381.38 16651.35 18024.10 18299.85 17925.91 18356.73 18269.01 18217.34 19146.17 17595.88 16952.28 19212.16 19809.58 17299.05 16597.64 18907.74
0.006 0.053 0.033 0.048 0.064 0.108 0.067 0.028 0.006 0.054 0.073 0.023 0.025 0.066 0.027 0.048
3348890.8 315102.1 546516.2 380069.7 282010.5 170340.5 272814.2 660665.5 3198717.8 327816.6 232573.4 851283.4 786156.4 260302.4 610273.7 391140.7
19875.46 16928.44 18149.09 18654.81
0.027 0.048 0.062 0.013
724806.7 355250.0 293757.2 1403534.6
870 < < < < < < < < < < < <
2 0.649
43 33 43 51 50 10 19 14
37.18 49.37
QALY, quality-adjusted life years; ICUR, incremental costeutility ratio; VAS, visual analog scale; JOA, Japanese Orthopaedic Association; JOA BPEQ, JOA Back Pain Evaluation Questionnaire; RDQ, RolandeMorris Disability Questionnaire; SF-8, the MOS short-form 8-item health survey; PCS, physical component summary; MCS, mental component summary; BS-POP, Brief Scale for Psychiatric Problems in Orthopaedic Patients; CES-D, Center for Epidemiologic Studies Depression Scale.
810
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
be highly cost-effective based on the reported WTP of JPY 5,000,000e6,000,000 per QALY [14,15]. This is the first costeffectiveness analysis of pharmacological management for CLBP in Japan and the first such study that was not sponsored by a pharmaceutical company in the world [7e10]. The mean acquired EQ-5D score had a positive value, but there were increases and decreases over time. This result indicates the limits of any one approach when it comes to the pharmacological treatment of CLBP [5,6]. Some studies have reported a ceiling effect in EQ-5D-3L, which was used in this study. The stratified analysis showed that patients with higher EQ-5D at baseline (>0.69) had less cost-effective CLBP management (Table 3). Recently, two further levels have been included (EQ-5D-5L), so further research using EQ-5D-5L may be able to show a greater change in EQ-5D (and thus greater cost-effectiveness). The sustained improvement in other clinical outcomes (JOA score, RDQ, pain-related disorders/ lumbar spine dysfunction/gait disturbance/social life disturbance in the JOA BPEQ, VAS, and PCS in the SF-8) except those including psychological elements (psychological disorders in the JOA BPEQ, MCS in the SF-8) support the effectiveness of the pharmacological treatment of CLBP. These results suggest that a multidisciplinary approach plays a large role among patients with poor clinical outcomes in psychological elements at baseline [4]. Stratified analysis of cost-effectiveness based on median values or categories of each factor showed interesting trends. Most notably, the ICUR of patients with a history of spine surgery or cancer, low frequency of exercise, long disease period, low scores in lumbar spine dysfunction and gait disturbance in the JOA BPEQ at baseline, and low JOA score at baseline demonstrated negative values. As to the history of spine surgery, decompressive or fusion surgery can alleviate nociceptive or neuropathic pain, so other factors might influence the ineffectiveness of pharmacological treatment of CLBP with the four leading drugs. The negative ICUR of CLBP patients with a history of cancer and older adults may be associated with frailty. History of cancer and older age are reported to relate with frailty [17,18], and an association between physical frailty and chronic pain has also been reported [19]. Other factors, including low frequency of exercise, long disease period, low scores in lumbar spine dysfunction and gait disturbance in the JOA-BPEQ at baseline, and low JOA score at baseline, may also be related to reduced muscle strength, which is one of the main causes of frailty. In patients with long disease duration, there is a possibility that all sorts of pharmacological treatments have been attempted to little or no effect, which suggests that continued discursive administration should be avoided. Finally, a higher frequency of exercise correlated with low ICUR, suggesting that the combination of exercise and pharmacological treatment has additive effects and, on the other hand, that the pharmacological treatment alone without exercise might be ineffective [3,20]. There were several limitations in this study. First, the six-month observation period was relatively short. However, under the current concept that recognizes the importance of multimodal and interdisciplinary management, a continuous approach with pharmacological treatment for a longer period is not realistic. In this study, a Markov model was not used for the estimation of longterm cost-effectiveness because the incidence of adverse events or effects on life expectancy were too low to establish a statistical model. Second, in this observation study, only the costs of the four leading CLBP-treating drugs were calculated, and the costs of gastric mucosal coating agents, agents for inhibiting gastric acid, antiemetic agents, laxative agents, and weak opioids were not included in the analysis because of the use of a wide variety of drugs and the high incidence of prescriptions for the treatment of other diseases. However, the price of the most expensive agents for inhibiting gastric acid is JPY 200 per day, the total cost for a six-
month period being about JPY 36,000. Therefore, its effect on ICUR would be limited. In conclusion, this cost-effectiveness analysis of the pharmacological management of CLBP demonstrated that it is indeed costeffective from the reference WTP. Further optimization based on patient characteristics is expected to provide evidence contributing to the sustainable development of a universal insurance system in Japan. Conflict of interest The authors declare that there is no conflict of interest relevant to the submitted work. Acknowledgements The Board of Directors and the Board of Councilors of the Japanese Society for Spine Surgery and Related Research (JSSR) deeply appreciate the following; Daiichi Sankyo Company, Ltd., Showa Yakuhin Kako Co., Ltd., and Astellas Pharma Inc. agreed with the overall research activities of JSSR and provided funds targeting all these research activities in accordance with the time when this new research project was planned. We wish to confirm that there were no known conflicts of interest associated with this publication. The authors would like to express their sincere gratitude to all supporting members at the JSSR. References [1] Hoy D, March L, Brooks P, Blyth F, Woolf A, Bain C, Williams G, Smith E, Vos T, Barendregt J, Murray C, Burstein R, Buchbinder R. The global burden of low back pain: estimates from the Global Burden of Disease 2010 study. Ann Rheum Dis 2014 Jun;73(6):968e74. [2] Dagenais S, Caro J, Haldeman S. A systematic review of low back pain cost of illness studies in the United States and internationally. Spine J 2008 JanFeb;8(1):8e20. F, Mannion AF, Pellise F, Cedraschi C. Clinical update: low back pain. [3] Balague Lancet 2007 Mar;369(9563):726e8. [4] Pincus T, Burton AK, Vogel S, Field AP. A systematic review of psychological factors as predictors of chronicity/disability in prospective cohorts of low back pain. Spine 2002 Mar;27(5):E109e20. [5] O'Connell NE, Cook CE, Wand BM, Ward SP. Clinical guidelines for low back pain: a critical review of consensus and inconsistencies across three major guidelines. Best Pract Res Clin Rheumatol 2016 Dec;30(6):968e80. [6] Foster NE, Anema JR, Cherkin D, Chou R, Cohen SP, Gross DP, Ferreira PH, Fritz JM, Koes BW, Peul W, Turner JA, Maher CG. Prevention and treatment of low back pain: evidence, challenges, and promising directions. Lancet Low Back Pain Series Working Group. Lancet 2018 Jun;391(10137):2368e83. [7] Haas M, De Abreu Lourenco R. Pharmacological management of chronic lower back pain: a review of cost effectiveness. Pharmacoeconomics 2015 Jun;33(6): 561e9. [8] Morera-Dominguez C, Ceberio-Balda F, Florez-Garca M, Masramon X, LopexGomez V. A cost-consequence analysis of pregabalin versus usual care in the symptomatic treatment of refractory low back pain: sub-analysis of observational trial data from orthopaedic surgery and rehabilitation clinics. Clin Drug Investig 2010;30(8):517e31. [9] Wielage RC, Bansal M, Wioson K, Klein R, Happich M. Cost-effectiveness of duloxetine in chronic low back pain: a Quebec societal perspective. Spine 2013 May;38(11):936e46. [10] Wielage RC, Bansal M, Acott Andrews J, Wohlreich MM, Klein RW, Happich M. The cost-effectiveness of duloxetine in chronic low back pain: a US private payer perspective. Value Health 2013 Mar-Apr;16(2):334e44. [11] Yoshida K, Sekiguchi M, Otani K, Mashiko H, Shiota H, Wakita T, Niwa Si, Kikuchi S, Konno S. A validation of study of the Brief Scale for Psychiatric problems in Orthopaedic Patients (BS-POP) for patients with chronic low back pain (verification of reliability, validity, and reproducibility. J Orthop Sci 2011 Jan;16(1):7e13. [12] Japanese Ministry of Health, Labour and Welfare. National database (NDB) open data. https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000177182. html. [13] Fukui M, Chiba K, Kawakami M, Kikuchi S, Konno S, Miyamoto M, Seichi A, Shimamura T, Shirado O, Taguchi T, Takahashi K, Takeshita K, Tani T, Toyama Y, Yonenobu K, Wada E, Tanaka T, Hirota Y. JOA back pain evaluation questionnaire (JOABPEQ)/JOA cervical myelopathy evaluation questionnaire (JOACMEQ). The report on the development of revised versions. April 16, 2007. The subcommittee of the clinical outcome committee of the Japanese
T. Kaito et al. / Journal of Orthopaedic Science 24 (2019) 805e811
[14]
[15]
[16]
[17]
orthopaedic association on low back pain and cervical myelopathy evaluation. J Orthop Sci 2009 May;14(3):348e65. Shiroiwa T, Sung YK, Fukuda T, Lang HC, Bae SC, Tsutani K. International survey on willingness-to-pay (WTP) for one additional QALY gained: what is the threshold of cost effectiveness? Health Econ 2010 Apr;19(4):422e37. Bertram MY, Lauer JA, De Joncheere K, Edejer T, Hutubessy R, Kieny MP, Hill SR. Cost-effectiveness thresholds: pros and cons. Bull World Health Organ 2016 Dec;94(12):925e30. Hosono N, Takenaka S, Mukai Y, Tateishi K, Fujiwara Y, Morishita Y, Konishi H. Conventional JOA score for cervical myelopathy has a rater's bias -In comparison with JOACMEQ-. J Orthop Sci 2018 May;23(3):477e82. Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet 2013 Mar;381(9868):752e62.
811
[18] Giannotti C, Sambuceti S, Signori A, Ballestrero A, Murialdo R, Romairone E, Scabini S, Caffa I, Odetti P, Nencioni A, Monacelli F. Frailty assessment in elective gastrointestinal oncogeriatric surgery: predictors of one-year mortality and functional status. J Geriatr Oncol 2019 May;S1879e4068(18): 30344e8. [19] Nakai Y, Makizako H, Kiyama R, Tomioka K, Taniguchi Y, Kubozono T, Takenaka T, Ohishi M. Association between chronic pain and physical frailty in community-dwelling older adults. Int J Environ Res Public Health 2019 Apr 13;16(8):E1330. [20] Manniche C, Hesselsøe G, Bentzen L, Christensen I, Lundberg E. Clinical trial of intensive muscle training for chronic low back pain. Lancet 1988 Dec;2(8626e8627):1473e6.