The Evidence Mapping and Methodological Quality of Clinical Practice Guidelines of Diagnosis and Management for Spinal Pain: A Cross-Sectional Survey

Objectives: To systematically review clinical practice guidelines (CPGs) for treating spinal pain, to assess the quality of different CPGs, and nally to provide an evidence map for the specic explication of research trends and gaps. Method: We searched CPGs in PubMed, Embase, Web of Science, Guidelines International Network (GIN), National Health and Medical Research Council (NHMRC), Scottish Intercollegiate Guidelines Network (SIGN) and the National Institute for Health and Care Excellence (NICE). We extracted basic information, recommendations, methodological quality, and reporting quality of the CPGs. Four researchers independently evaluated the quality of the CPGs according to the Appraisal of Guidelines Research and Evaluation (AGREE II) and Reporting Items for Practice Guidelines in Healthcare (RIGHT). Results: We included 21 CPGs into our study. According to the AGREE II checklist, we found that the mean score was relatively high in four domains, namely, scope and purpose (78.39), stakeholder involvement (63.04), clarity of presentation (72.04), and rigor of development (61.25). However, the mean score of two domains (editorial independence and application) was relatively low. Among the seven domains of the RIGHT checklist, one eld (basic information) had the highest reporting rate (86.61%), while another eld (funding, declaration, and management of interest) obtained the lowest reporting rate (53.00%). Conclusion: Our study provided evidence mapping, which is a good tool to reduce research waste and facilitate the process of knowledge transfer. We found the mean score of the application of included CPGs was the lowest and most of CPGs didn’t consider patient preferences. Therefore, guideline makers should concentrate on patient preference and application in future guidelines. The results of our study can also be used to optimize the implementation of these recommendations and to improve the development of reliable CPGs for treating spinal pain. methodological quality of recent spinal pain CPGs. Our study offered a comprehensive assessment of methodological quality of recent spinal pain CPGs. 3. We found the mean score of the application of included CPGs was the lowest and most of CPGs didn’t consider patient preferences. The results of methodological quality could help guideline developers identify the limitations of recent guidelines and then promote improvement in the development of reliable CPGs for treating spinal pain


Introduction
Spinal pain includes cervical spine pain, low back pain and sciatica. [1][2][3] It is one of the leading causes of disability and loss of human labor capacity worldwide. [4][5][6][7][8][9][10] Approximately 50-85% of the population has back pain for some period over their lifetimes. [11] , [12] The incidence of neck pain was reported to be approximately 20-40% in adolescents and to increase as age increases. [13,14] Approximately 2-14% of the population suffers from sciatica. [15,16] The annual cost of treating spinal pain ranks third behind diabetes and heart disease treatment. [17] Therefore, spinal pain is a serious public health problem that creates many problems in people's daily lives.
Clinical practice guidelines (CPGs) for treating spinal pain can help improve the quality of care. Care following CPG recommendations can result in better outcomes and lower costs. [18,19] However, various organizations give con icting CPGs and recommendations. [20] Furthermore, CPGs have potential issues; for example, several CPGs include voluminous documents that are not user-friendly, [21] and there are shortcomings in their quality [22,23], which may limit clinical use. [24] Some contemporary spinal pain CPGs are of poor quality and have the same problems as mentioned above. [24] Therefore, we conducted a systematic evaluation and investigation of the quality of different spinal pain CPGs.
Evidence mapping, an emerging rapid and accurate review method, includes systematic search and speci c characterization of existing studies on topics of interest. [25] Evidence mapping aims to identify the gap between different studies to open avenues for future research. [25] In our study, the evidence mapping method was used to visually present different CPGs regarding diagnosis, management strategies and con ict recommendations of spinal pain care.
The Appraisal of Guidelines Research and Evaluation (AGREE II) and Reporting Items for Practice Guidelines in Healthcare (RIGHT) are commonly used checklists to assess the quality of CPG. The international RIGHT Working Group followed an existing framework for developing guidelines and the EQUATOR (Enhancing the Quality and Transparency of Health Research) Network approach to build the RIGHT checklist.
[26] The RIGHT checklist can not only help developers in reporting guidelines but also help health care practitioners understand and implement guidelines.
[26] AGREE II was developed to address the issue of variability in guideline quality. It is a tool that assesses the methodological rigour and transparency in which a guideline is developed.
In present study, the aim is to systematically review CPGs for treating spinal pain, to assess the quality of different CPGs, and nally to provide an evidence map for the speci c explication of research trends and gaps.

Search strategy and selection criteria
We de ned CPGs as being identi ed by authors and being consistent with the de nition of the Institute of Medicine. [27] We searched for CPGs in PubMed, Embase and Web of Science using medical subject headings (MeSH) and keywords. We also searched some online guideline websites: Guidelines

Study process
Teams of paired reviewers, trained in trial and systematic review methods, screened abstracts and full texts for eligibility and abstracted data from eligible studies independently and in duplicate, using pilot-tested, standardized forms, together with detailed instructions. The reviewers resolved disagreements through discussion or, if needed, adjudication by a third reviewer.

Classifying recommendations
Two authors independently reviewed and classi ed the recommendations. We classi ed recommendations into four groups, namely, 'recommended to do', 'could be considered to do', 'do not recommend to do' or 'uncertain/unclear', to accommodate the various terminologies used in CPGs. For a particular topic, if there were no recommendations provided by the CPG, or if it was out of scope of the CPG and was not included, it was not classi ed.

Narrative summary
Two authors initially developed a narrative summary including classifying recommendations into different theme areas, and these were then reviewed and re ned by all other authors. Our author group included academic and practicing physiotherapists, MSK pain researchers, an indicator development researcher, an emergency care physician, a senior medical o cer in emergency medicine and a pain medicine physician.
Assessment of guideline quality AGREE II and RIGHT were used to assess the quality of CPGs. The RIGHT checklist can help report guidelines. AGREE II was developed to assess the variability in guideline quality. Four independent reviewers were trained to perform CPG appraisals, and they independently reviewed and scored each eligible CPG.
[26] Four reviewers independently assessed the quality of each CPGs by using "Yes", "No", and "Partial" for each item according to the RIGHT checklist.
Finally, we combine AGREE II and RIGHT to de ne the quality of CPGs. We de ned the quality of CPGs with both AGREE II scores > 60% and RIGHT scores > 20 as high quality. We de ned the quality of CPGs with AGREE II scores between 30% and 60% or RIGHT scores between 10 and 20 as middle quality. We de ned the quality of CPGs with AGREE II scores < 30% or RIGHT scores <10 as low quality.

Statistical analysis
For each CPG, we calculated the AGREE II score for each domain and overall scores as a percentage of the maximum possible score and standardized range.
Then, we calculated the mean and standard deviation (SD) for six main domains.
The number of RIGHT checklist items reported in each CPG was presented to assess the reporting quality data.
Summarizing and grading the quality of CPGs After completing the AGREE II score and RIGHT score, we summarized two scores simultaneously to visualize the quality of each CPG. The darkness of the bubbles represented the quality of CPGs (green: high; yellow: middle; red: low).

Patient and public involvement
Patients and/or the public were not involved in this research.

Quality
The highest mean score was 78.39 ± 13.34 for scope and purpose, and the lowest mean score was 40.67 ± 22.61 for applicability. Generally, nine CPGs had overall scores higher than 60%; these CPGs were deemed as strongly recommended for clinical care [34-36, 38, 39, 41, 42, 49, 50], ten CPGs were de ned as recommended with modi cations [31-33, 40, 44-48, 51] , and three CPGs were deemed as not recommended [37,43,47]. The mean and SD of AGREE II scores for each domain and overall scores of various guidelines are shown in Table S1 and Figure 2A. The brief overall scores of CPGs are shown in Table S2.
The RIGHT checklist contains 22 requirements organized into 7 sections, with a total of 35 items. According to the RIGHT checklist, the CPGs with the largest number of reported items was NICE, DHA and SIGN (RIGHT score = 34), followed by OPTIMa (RIGHT score = 31) and Canada Guideline (RIGHT score = 30). Among the seven domains of the RIGHT checklist, eld one (basic information) had the highest reporting rate (86.16%), and eld six (funding, declaration and management of interest) obtained the lowest reporting rate (53.00%). The numbers of reported items are shown in Table 2. The mean compliance rates of each item are shown in Figure 2B. AGREE II and RIGHT were combined to calculate the quality of the CPGs ( Figure 3). Figure 4 summarizes the various strengths of recommendations of CPGs for neck pain and low back pain patients. Table 5 summarizes the various recommendations of assessment, diagnosis, and management of spinal pain CPGs.

Patient-centered care
Most CPGs suggested that clinicians should provide patient-centered care that can lead to better outcomes. Patient-centered care included care methods such as providing individualized care [31][32][33], providing information for patients in a reassuring manner, [34] and shared decision-making [35].
Exclude serious pathology/red ag conditions and classify the stage of neck pain Five CPGs offered recommendations regarding screening patients for potentially serious or structural pathology or for 'red ag' conditions at the initial assessment. [32,33,[35][36][37] For example, clinicians should determine if there is a fracture, malignancy, infection, in ammatory cause of pain or another pathology reason.
Two CPGs mentioned that classi cation of neck pain grades should be conducted at the initial assessment. [32,35] Three CPGs stated that one should classify neck pain as acute, subacute or chronic. [31,33,36] Then, management should be provided based on the stage of neck pain.

Conduct physical examination and assess psychosocial factors
Four CPGs recommended that clinicians or physical therapist conduct comprehensive physical examinations to determine patients' baseline status. [32,33,36,37] The aim of the physical examination is to help diagnose and classify neck pain disorders. Three CPGs advised assessment of psychosocial factors, which are important factors for prognosis. [32,35,36] Evaluate patient progress Three CPGs recommended evaluating patient progress and using validated outcome measures. [32,33,35] Outcome measures included numeric pain rating scale and the patient-speci c functional scale [32], utilizing easily reproducible activity limitation and participation restriction measures [33] and self-rated recovery questions. [35] Provide education or information Six CPGs strongly recommended providing education or information for patient about their condition and management methods. [31][32][33][34][35][36] Education is a part of rst-line, essential treatment. Common education consisted of providing structured and individualized education [34][35][36] such as advice, training, supervision, instruction, prognosis, encouragement, reassurance and pain management [31][32][33]. However, one CPG indicated that there was insu cient evidence to prove the effect of educational interventions for patients with neck pain. [37] Patients should do physical activity or exercise All CPGs recommended that patients to perform physical activity or exercise. [31][32][33][34][35][36][37] Exercise (e.g., strengthening, endurance and exibility exercises) was recommended for the treatment of acute and chronic neck pain for both short-and long-term bene ts. [31,32,36,37] One CPG showed that clinicians could use exercise therapy for management of neck pain. [35] Two CPGs mentioned that exercise should/could combined with manual therapy. [34,37] Manual Therapy All CPGs recommended that patients receive manual therapy to relieve pain. Manual therapy (manipulation and/or mobilization) combined with other modalities was recommended for treating both acute and chronic neck pain. [31][32][33][34][35][36][37] One CPG recommended manipulation alone for patients with acute neck pain and combination therapy for chronic neck pain. [37] However, for massage, the recommendations of different CPGs differed. Two CPGs recommended massage combined with other treatments to treat chronic neck pain. [31,37] One CPG stated that clinicians could consider massage to be a potential treatment method. [32] Two CPGs did not recommend massage to treat neck pain. [34,35] Cervical collar One CPG suggested that a cervical collar could be considered only if primarily advised treatments are ineffective. [32] Another CPG stated that short-term use of cervical collar can be considered for patients with acute neck pain with radiating pain. [33] One CPG did not recommend the use of cervical collar, [35] and one CPG indicated that there was insu cient evidence to give a recommendation [37].

Pharmacotherapy
Some CPGs recommended nonsteroidal anti-in ammatory drugs (NSAIDs), [34,36] paracetamol, [37] tramadol [34] or muscle relaxants[36] after careful consideration of adverse events, contraindications, and patient preference. Furthermore, the duration should be short. However, one CPG did not recommend the use of paracetamol and recommended that clinicians consider the use of muscle relaxants. [35] For opioids, clinicians must carefully consider before using and only use these drugs in the short to medium term when other therapies have been insu cient. [ 39-41, 45, 48, 50].

Provide education and exercise therapy
Most CPGs recommended patient education[36, 39-46, 48-50] (e.g., explaining the expected course, advising to remain active and providing information about self-management options) and exercise therapy[36, 38-46, 48-50] as the rst-line choice for the management of acute, subacute and chronic low back pain.

Manual Therapy
Three CPGs recommended manual therapy for acute [38], subacute[38], and chronic low back pain. One guideline indicated manual therapy could be considered for short-term pain relief for chronic LBP.
[36] In addition, some guidelines suggested that manual therapy could be applied if self-care and medication therapy did not provide signi cant improvements [39] or that it could be offered as a part of multimodal management [40,42,49,50].
One CPG recommended massage for treating acute and subacute low back pain[38], while two guidelines indicated that massage was not recommended for acute low back pain but could be considered for chronic low back pain [43,44]. One guideline stated that there was insu cient evidence to show the effect of manual therapy including massage. [48] Pharmacotherapy For the management of chronic low back pain, nonpharmacologic therapy should be the primary treatment. For those who have had an inadequate response to nonpharmacologic therapy, pharmacologic treatment should be considered. For pharmacologic treatment, NSAIDs should be treated as a rst-line therapy [38,39,45,49] and should be given in the lowest effective dose and for the shortest period [45,50]. One CPG stated that NSAIDs should only be offered in addition to usual care for acute LBP [42].
Two CPGs recommended paracetamol as the rst choice and NSAIDs as the second choice of pharmacotherapy for LBP management. [39,48] In contrast, many CPGs did not recommend paracetamol to treat LBP [40,42,43,45,49,50].
Most CPGs recommended the use of opioids only if NSAIDs were ineffective, not tolerated, or contraindicated [40,50] or if common care failed and if the known risks and realistic bene ts were discussed with patients[36, 38, 41].

Epidural steroid injection
Most CPGs had a consensus opinion about epidural steroid injection. They were all against epidural steroid injection for patients without radiculopathy[48-50], long-term reduction of radicular low back pain [49], or spinal stenosis [49]. For patients with acute or subacute severe radicular pain [40,50] or for very shortterm reduction of radicular low back pain [49], an epidural steroid injection can be considered.

Surgery
For patients with the following situations, surgery can be considered: 1. severe and disabling pain persisting after nonsurgical treatment [42,47,48]; 2. radicular symptoms, symptoms of neurogenic claudication, or severe stenosis [47,48]; and 3. nerve root compression proved by MRI or CT [47]. In addition, NICE did not recommend disc replacement or spinal fusion for low back pain [50].

Sciatica
Two CPGs did not recommend routine use of imaging for patients with sciatica [42,50], while one CPG recommended MRI for patients with history and physical examination ndings consistent with sciatica [51]. Exercise and manual therapy can be considered to treat sciatica, but only as a part of multimodal treatment [42,50]. NICE recommended antidepressants and gabapentin as initial treatment for neuropathic pain [50]. Epidural steroid injection was recommended to relieve severe radicular pain [50,51]. For surgery, one guideline recommended discectomy, which can provide more effective pain relief than other care for patients with sciatica whose symptoms warrant surgery treatment [51]. However, another CPG suggested that clinicians consider spinal decompression for patients with sciatica when nonsurgical treatment is ineffective in improving pain or function and their radiological ndings are consistent with sciatic symptoms [50].

Discussion
A total of 21 CPGs were identi ed in our international review of CPGs on the management of spinal pain. According to the results of guideline quality assessment, it was obvious that the quality of different CPGs varied considerably. In a recent study, we identi ed nine relatively high-quality CPGs [34-36, 38, 39, 41, 42, 49, 50]. According to a standardized grade system, the several organizations summarized their recommendations for spinal pain care. We found that the recommendations and the strength of the recommendations varied widely among the CPGs, which may lead to confusion on the part of clinicians and patients or may even lead to inappropriate management regimes. Therefore, in our study, we tried to systematically evaluate and summarize the recommendations from various CPGs to provide a simple and clear consensus of current management of spinal pain for clinicians, researchers, healthcare managers, and policymakers. Our results could also help improve the quality of clinical care for spinal pain. [53,54] Our ndings were consistent with those of previous reviews. [55][56][57][58][59][60] Advice, education, and exercise were recommended as rst-line treatments for spinal pain.
For neck pain, Cohen et al. recommended that clinicians should conduct a thorough history and physical examination to distinguish neuropathic pain from mechanical neck pain and provide nonpharmacological alternative treatments before pharmacological treatments.
[56] The authors also showed that there was moderate and weak evidence to support NSAIDs for acute and chronic neck pain and muscle relaxants for subacute neck pain associated with muscle spasm.
[56] These results were similar to our ndings. [34,36] For sciatica, Jensen et al. indicated that most patients improved after conservative treatments such as exercise, manual therapy, and self-management and that imaging should not be routinely used. [60] Though surgery may speed up recovery, the effect was similar to conservative care over the long term.
[60] Surgery can be considered if symptoms do not improve after 6-8 weeks of conservative treatment.
[60] Epidural corticosteroid injections only reduce short-term pain and disability caused by sciatica, but not back pain.
[61] For acute low back pain, there was high-quality evidence to support the use of pharmacological therapies (e.g., NSAIDs, muscle relaxants and paracetamol) and exercise, moderate-quality evidence to support use of acupuncture, and low-quality evidence to support manual therapy. [4] For persistent low back pain, there was high-quality evidence to support the use of exercise and antidepressants, moderate-quality evidence to support acupuncture, TENS, psychotherapy and opioids, low-quality evidence to support NSAIDS and massage and very low-quality evidence to support paracetamol, muscle relaxants and manual therapy. [4] However, that evidence was only based on short-term pain outcomes and did not weigh against the risk of harm. Therefore, the results of that review may provide different opinions from those of the present review.

Imaging
In our study, we found that most CPGs consistently recommended against the routine use of imaging. The problem of indiscriminate imaging is serious, possibly contributing to the waste of medical resources and unnecessary exposure to radiation. In previous surveys, approximately one-quarter of patients with recent-onset low back pain undergo imaging.[62, 63] The abuse of imaging was worse in emergency departments.[64-66] Therefore, both patients and clinicians should know the indications of imaging for spinal pain as we concluded above.

Physical activity
In our study, most CPGs recommended physical activity/exercise for managing spinal pain. However, we needed to know the different patterns of physical activity between work and leisure.
[67] Previous studies showed that physical behavior at leisure time can reduce the risk of NP and LBP.[68-71] Furthermore, Sitthipornvorakul et al. pointed out that increased both work and leisure physical activity can reduce risk of spinal pain. [72] However, previous studies indicated that the reduced risk of spinal pain was associated with increased sitting at work in working populations. [73][74][75][76] Therefore, for the different populations, the recommendation should be varied. For the white-collar population, it was recommended to increase physical activity both in work and leisure time, while the blue-collar population, it was recommended that they should limit work intensity and increase rest time at work as well as perform physical activity during leisure time. Physical activity was highly recommended; however the ideal duration, intensity and methods of training remain unclear.
[77] Therefore, future studies should focus on the ideal duration, intensity and methods of physical activity.

Manual Therapy
Cervical manipulation and mobilization produced similar immediate-term or short-term effects.
[78] Previous studies showed that combinations of manual therapy and other common therapies (e.g., education and exercise) were more effective for pain relief and function rehabilitation. [79][80][81][82] Similarly, in our study, all CPGs recommended manual therapy combined with other treatments for treating both acute and chronic neck pain. [31][32][33][34][35][36][37] [42,50]. In general, acupuncture can be considered only for short-term pain relief, but does not have a long-term bene t for spinal pain patients.

Pharmacotherapy
Previous studies have shown the effect of NSAIDs [92,93] and muscle relaxants [92,94] for the alleviation of pain and disability in patients with acute, subacute, and chronic low back pain and antidepressants only for chronic low back pain[95, 96]. Short-term opioid therapy has analgesic bene ts for low back pain; however, adverse effects must be considered. Systemic corticosteroids, anticonvulsants and antidepressants showed statistically signi cant effects in reducing sciatica pain. [99][100][101][102] There was no su cient evidence to show the effects of paracetamol, benzodiazepines, or opioids for treating sciatica. [100,103,104] In addition, previous systematic reviews pointed out that three commonly used medicines (NSAIDs, paracetamol and opioids) did not provide clinically important effects for spinal pain.[2, 105, 106] Therefore, new analgesics or new pharmacotherapy regimes for spinal pain are needed.
CPGs of low methodological quality are still being developed and published. [107,108] Previous studies mentioned that recent guidelines failed to clearly show the selection criteria of the literature, to adequately describe the strengths and limitations of included literature and to speci cally describe the methods used to formulate recommendations. [109] Wong et al. stated that future high-quality guidelines should focus on providing clear implementation strategies and the applicability to speci c populations. [110] Our results also indicated that the mean score of the application was the lowest. Besides, some CPGs included in our study seem to be too complex and long to get points. Readers may end up "giving up" reading everything and may miss something "important". Therefore, guideline makers should concentrate more on how to convey the information quickly and easily and the application of the clinical guidelines.
Evidence-based decision-making was proved to have favorable health and economic outcomes. [111] However, current real-world practices differ substantially from recommendations in evidence-based guidelines. Patient preferences are becoming increasingly important in healthcare policy decision-making. [112,113] However, most included guidelines did not consider the views and preferences of patients when designing guidelines. Previous studies pointed out that consideration of patient preferences could increase adherence to treatment, satisfaction with treatment, and health outcomes. [114,115] Therefore, future guideline development should consider patient preferences as an important factor.

Limitations
First, due to the variations of reporting and expression of different CPGs, there may be some inappropriate and inaccurate interpretations of our research because there is heterogeneity in the way that CPGs are conceptualized and how their evidence and recommendations are presented. This issue is inherent.
Second, for various reasons, we may fail to identify all relevant and standard-compliant documents. Third, different CPGs used different grading systems that were based on different coding systems to classify the quality of evidence and strength of recommendations. Therefore, it is hard to summarize and unify the quality of evidence for each recommendation and conduct further studies.

Conclusions
To improve the management of spinal pain, we summarized CPG recommendations to identify consensuses of management regimes. However, the recommendations of some CPGs were not very clear and speci c, and they even provided inconsistent recommendations. Evidence mapping is a good tool to reduce research waste and facilitate the process of knowledge transfer. The results of our study can be used to optimize the implementation of these recommendations and to promote improvement in the development of reliable CPGs on spinal pain. We found the mean score of the application of included CPGs was the lowest and most of CPGs didn't consider patient preferences. Therefore, guideline makers should concentrate on patient preference and application in future guidelines. Once these weaknesses are resolved, a more effective and accurate management regime that will lead to reduced costs of spinal pain can be established.   Figure 1 Flow diagram of study identi cation, screening, eligibility assessment, and inclusion. Figure 2