J Manipulative Physiol Ther 2007 (Nov); 30 (9): 617–683 ~ FULL TEXT
André E. Bussières, DC, Cynthia Peterson, DC, RN, MMedEd, John A.M. Taylor, DC
Université du Québec à Trois-Rivières,
Purpose Imaging technology can improve patient outcomes by allowing greater precision in diagnosing and treating patients. However, there is evidence that overuse, underuse, and misuse of imaging services occur. The purpose of this project was to develop evidence-based diagnostic imaging practice guidelines for musculoskeletal complaints for use by doctors of chiropractic and other primary health care professionals.
Methods An electronic search of the English and French language literature (phase 1) was conducted on several databases. Cross references, and references provided by clinicians, were also used. Independent assessment of the quality of the citations used to support recommendations in the guidelines was performed using the QUADAS, the AGREE, and the SPREAD evaluation tools. A first draft of a diagnostic imaging practice guideline was produced, using the European Commission's Referral Guidelines for Imaging document as a template. Results were sent to 12 chiropractic specialists for a first external review. A modified Delphi process, including 149 international experts, was used to generate consensus on recommendations for diagnostic imaging studies. The reliability of proposed recommendations was further tested on field chiropractors and on a group of specialists both in chiropractic and in medicine in both Canada and the United States. All recommendations were graded according to the strength of the evidence.
Results The research procedure resulted in the recommendations for diagnostic imaging guidelines of adult extremity and spine disorders supported by more than 685 primary and secondary citations. High levels of agreement among Delphi panelists were reached for all proposed recommendations. Comments received by specialists were generally very favorable and reflected high levels of agreement with the proposed recommendations, perceived ease of use of guidelines, and implementation feasibility.
Conclusions These evidence-based diagnostic imaging practice guidelines are intended to assist chiropractors and other primary care providers in decision making on the appropriate use of diagnostic imaging for specific clinical presentations. In all cases, the guidelines are intended to be used in conjunction with sound clinical judgment and experience. Application of these guidelines should help avoid unnecessary radiographs, increase examination precision, and decrease health care costs without compromising the quality of care. All guidelines are documents to be refined and modified regularly with new information and experience.
Key Indexing Terms: Diagnostic Imaging, Radiology, Diagnostic X-Ray, Radiography, Practice Guideline, Musculoskeletal System, Chiropractic
From the FULL TEXT Article:
Approximately 150 musculoskeletal diseases and syndromes associated with pain and loss of physical function commonly affect children and adults in all regions of the world. The most common conditions include osteoporosis, arthritis, spinal disorders (including low back pain [LBP]), and the consequences of severe trauma. The total cost of musculoskeletal disease in the USS in the year 2000 has been calculated at US$254 billion. In developing countries, the cost of musculoskeletal injury care is estimated at US$100 billion, a figure nearly twice that of total foreign aid for these nations. [1, 2] Diagnostic imaging is the most rapidly evolving specialty within health care. [3, 4] Imaging technology can improve patient outcomes by allowing greater precision in diagnosing and treating patients. Despite the advent of new and advanced techniques, conventional radiography remains the cornerstone of diagnostic imaging for patients presenting with regional musculoskeletal pain, especially for acute fractures of the extremities or for suspected neoplasia. [5–7] However, evidence of overuse, underuse, and misuse of imaging services has been emphasized in the literature. [8–13] In the US alone, more than $500 million are spent each year on lumbar spine radiography  and in Ontario, Canada, the Ontario Health Insurance Plan spends more than $16 million annually on physician-requested imaging for LBP.15 Such costs are unlikely to outweigh the effect of small reported increases in patient satisfaction, especially when considering potential risks of ionizing radiation exposure and lack of demonstrable benefits to patients. [16–18] It should be noted that this significant rise in expenditure has recently motivated the US Congress to reduce payments for such imaging services.  Several factors contributing to escalating imaging costs have been identified: the aging population; the practice of defensive medicine; overuse of diagnostic imaging; self-referral abuses; duplicative studies; consumer demand; and advanced technology. Research has shown many of these costs could be decreased if the referring physician community had a solid understanding of the appropriate ordering of diagnostic imaging studies and the risks related to radiation exposure. 
There is an urgent need throughout the health care professions to develop practice guidelines. Brown  summarized the characteristics of quality care as (a) patient centered, (b) scientifically based, (c) population outcomes based, (d) refined through quality improvement and benchmarking, (e) individualized to each patient, and (f) compatible with system policies and resources. Government agencies, professional accreditation agencies, and public institutional policies are now demanding that all health professions develop practice guidelines that may be used to evaluate what are deemed reasonable and appropriate indications for care. In addition, recent recommendations of the Institute for Alternative Futures for the chiropractic profession include accelerating research activity, striving for high standards of practice, engaging in consumer-driven health models, engaging integration and collaboration with mainstream health care, achieving greater professional unity, and meaningfully contributing to the public health agenda.  Evidence-based diagnostic imaging practice guidelines facilitate the implementation of such recommendations. [23, 24]
The purposes of guidelines are to:25
describe appropriate care based on the best available scientific evidence and broad consensus;
reduce inappropriate variation in practice;
provide a more rational basis for referral;
provide a focus for continuing education;
promote efficient use of resources;
act as focus for quality control, including audit;
highlight shortcomings of existing literature and suggest appropriate future research.
These guidelines have been created to improve patient care by detailing the appropriate information gathering and decision-making processes involved in the diagnostic imaging of musculoskeletal care.
The initial impetus for this project occurred in 2003 when the Council on Chiropractic Education in Canada (CCEC) requested a revision of the diagnostic imaging guidelines used at the chiropractic teaching clinic of l'Université du Québec à Trois-Rivières (UQTR). CCEC recommended incorporating protocols that were centered on the clinical presentation of patients. We first identified the need for an extensive literature review to propose recommendations on the use of imaging studies for musculoskeletal disorders. The review began with a search for the relative risks and benefits of imaging, clinical indications for imaging of the spine and extremities, knowledge of existing guidelines, and consideration of potential barriers to dissemination and implementation strategies of any such guidelines.
The purposes of guidelines are to: 
Does ionizing radiation from radiography carry a potential risk to patients?
What is the frequency of conventional radiography use?
What are the associated costs associated with the use of conventional radiography?
What factors influence practitioners in determining the need for radiographs in adults?
What is the evidence supporting the recommendations of various national and international practice guidelines for radiography utilization for musculoskeletal conditions for adults?
What is the clinical usefulness of spine radiography in biomechanical analysis, including spinal curvatures and scoliosis, especially with respect to the effects on spinal degeneration?
What are the indicators of potentially serious pathology involving the musculoskeletal system (red flags) in adults?
What is the prevalence of congenital anomalies that are considered contra-indications for high velocity, low amplitude manual spinal adjustment or manipulation? Furthermore, what is the associated complication rate of these maneuvers in patients with such anomalies?
What are the clinical indicators leading to optimal investigation and management of osteoporosis?
What are the appropriate clinical findings indicating radiographic evaluation of the extremities in adults?
Sensitivity, specificity, reliability, validity, predictive values, and likelihood ratios were considered when available and appropriate. Two hundred and fifty-seven citations were retrieved and synthesized by the principal investigator (AB) and presented in a French language narrative review to UQTR clinicians in June 2003. Relevant findings were also presented at a research conference. 
This narrative review helped identify the need for development of diagnostic imaging guidelines for musculoskeletal disorders. Although several answers to the above 10 questions are integrated within the guidelines, the routine use of conventional radiography and the consequent risks of low-dose ionizing radiation exposure remain significantly contentious and will be further addressed in a separate article. Proponents of the routine use of conventional radiography offer the following justifications for that practice: 
To screen for and prevent rare possible complications associated with spinal manipulative therapy;
For nonclinical motives, including patient preference and satisfaction; medicolegal concerns; political influence and administrative factors;
For postural and biomechanical analysis of spinal disorders both before and during the provision of manual therapy;
To provide patients with a more accurate prognosis;
Because they consider the effects of ionizing radiation exposure on human health to be negligible.
The proposed diagnostic imaging practice guidelines represent a shift to a more evidence-based approach and are supported by more than 680 primary and secondary citations. When available, meta-analyses and systematic reviews, randomized clinical trials (RCTs), nonrandomized studies, cohort and case-control studies, nonexperimental studies, and existing high-quality clinical guidelines were used to propose recommendations for various patient presentations (specific conditions) and comments supporting those recommendations.
The guidelines pertain only to adult patients and are divided into 3 parts. Part I specifically addresses lower extremity disorders; part II addresses upper extremity disorders; and part III addresses spinal disorders. All 3 parts are condition specific (symptom-based), a feature that should make these guidelines practical, user-friendly, and efficacious for practitioners.  Because accurate history and examination are crucial initial steps in the diagnosis of any musculoskeletal disorder, specific clinical criteria are emphasized for most every condition. The initial triage of patients with LBP is a constant recommendation of various clinical guidelines and we believe it to be equally applicable to the initial management of other regional painful conditions. [29, 30] One major objective of the initial triage is to determine the presence of clinical indicators (red flags) for serious pathologies requiring diagnostic imaging, appropriate referral, or urgent surgical intervention. In the absence of such clinical indicators, simple decisions based on separation into articular vs nonarticular disorders and on the duration of the patient complaint (acute, subacute, persistent/chronic) are favored by most recent expert consensus statements. [30, 31]
Clinical presentations are divided accordingly throughout the guidelines. Recommendations for radiography (indicated, not initially indicated, not routinely indicated) appear for every clinical presentation throughout the guidelines. Because professional self-regulation and knowledge about the clinical usefulness of specialized diagnostic modalities are necessary for all primary health care professionals, recommendations for further imaging studies are therefore discussed for each condition. [32, 33] In some instances, laboratory assessment is recommended.
Recent evidence suggests that implementation strategies of imaging guidelines can assist in achieving a sustained reduction in the number of radiographic examinations obtained of the cervical spine, lumbar spine, and knees.  Application of practical, highly sensitive, and reliable decision rules for cervical spine, shoulder, elbow, wrist, knee, ankle, and foot trauma has a potential to reduce the use of radiography by a factor of 16% to 50% in emergency departments and sports injury clinics alone. [35–46] Although the authors are uncertain of precise chiropractic utilization rates in these cases, it is likely that utilization rates in chiropractic parallel those in general medical practice because a substantial portion of chiropractic patients present with spine and extremity disorders or have a history of cervical spine trauma.  Although targeting a different population (generally not as acutely injured), application of decision rules in private practice should also help reduce the use of radiographic studies. For example, in acute LBP in adults, the age criterion proposed in earlier guidelines (>50 years of age) as well as the tendency to order radiographs in patients with symptom duration of less than 4 weeks has low specificity and substantially increased utilization. [12, 48] One authority now recommends obtaining lumbar spine radiographs in patients older than 65 years rather than 50 and waiting up to 7 weeks to obtain radiographs in patients with an untreated first episode of LBP. 
The aim in conducting an evidence review is to facilitate the integration of the best available evidence with clinical expertise and the values and beliefs of patients. Although these proposed guidelines are more restrictive and likely more cost-efficient, practitioners are provided with the necessary information to allow for significant flexibility in decision making. As with all guidelines, clinical judgment and clinician experience must be used in all cases. It must also be emphasized that clinical practice guidelines (CPGs) are intended to assist, not replace, clinicians in their decision-making processes. Guidelines aid in the application of selective diagnostic tools and therapeutic approaches for specific clinical presentations while considering patient preference. 
Statement of Purpose
The purpose of this project was to develop evidence-based diagnostic imaging practice guidelines for adult musculoskeletal complaints for use by chiropractors and other primary health care professionals. The guidelines aim to describe appropriate care based on the best available scientific evidence and broad consensus, reduce inappropriate variation in practice, provide a more rational basis for referral, provide a focus for continuing education, promote efficient use of resources, act as a focus for quality control (including audit), highlight shortcomings of existing literature, and suggest appropriate future research. The project consists of 9 phases; a flow diagram demonstrating the placement of each phase is provided in Figure 1.
Phase 1: Literature Search
The literature search focused on the appropriate use of conventional radiography and specialized (advanced) imaging for adult musculoskeletal disorders of the spine and extremities. Studies conducted in any setting were included in the review, including general practice, accident and emergency, and hospital clinics. The selection criteria were applied to the abstracts of the publications retrieved by the search strategy described below. The publications were retrieved and read if the abstracts provided insufficient information to enable selection. The review also considered biomechanics, degenerative changes and spinal curvatures (and their relationship to patient symptoms), indicators of potentially serious pathology (red flags), as well as congenital anomalies, trauma, scoliosis, and osteoporosis. Any study that compared clinical examination and/or assessment of patient history with an acceptable gold standard for the evaluation of patients with musculoskeletal pain was included. A review of various national and international chiropractic and medical guidelines discussing the use of imaging studies for musculoskeletal disorders was also incorporated in the literature review.
Phase 1 consisted of an electronic search in English and French language literature. Two literature updates were conducted using a similar strategy. This included MEDLINE (Pub Med) (1966–2003; –2005; -August 2006), the Cochrane Register (-January 2004, 2006), the National Research Register (-January 2004, 2006), the National Guideline Clearinghouse (-January 2004, 2006), the internet “Google” search engine, as well as cross references and references provided by clinicians. The search strategy also included other databases such as EMBASE to identify relevant studies for several of the disorders. A combination of subject headings (MeSH headings) and keywords were searched (specific disorders of the adult hip, knee, ankle and foot, shoulder, elbow, wrist and hand, pain, extremity, lumbar spine, thoracic spine, cervical spine, low back pain, thoracic pain, neck pain, whiplash, trauma, diagnostic, radiograph, x-ray, imaging, US, CT, MRI, MRA, bone scan, nuclear medicine, musculoskeletal, chiropractic, and medical guideline).
Specific disorders of the extremities and spine were subdivided into the following categories:
Adult hip pain: acute hip injury, strain, tendinitis or tendinosis, piriformis syndrome, trochanteric and iliopsoas bursitis, osteoporotic fractures, septic arthritis, chronic pain, congenital/developmental abnormalities, osteoarthritis (DJD or OA), inflammatory arthritis, osteonecrosis, tumors and metastatic lesions, stress (fatigue or insufficiency) fractures, physical examination.
Adult knee pain: acute knee injury, osteoarthritis (DJD or OA), inflammatory arthritis, bursitis, tendinitis, strain, tendinosis, anterior knee pain, internal joint derangement, physical examination.
Adult ankle and foot pain: acute ankle and foot injury, chronic ankle and tarsal pain, impingement syndromes, peroneal tendinosis, lateral premalleolar bursitis, tarsal tunnel syndrome, chronic foot pain, hindfoot-heel pain, plantar fasciitis, calcaneal enthesosphyte (spur), sinus tarsi syndrome, midfoot pain, flat foot with posterior tibial tendon dysfunction/rupture, navicular tuberosity pain and tenderness, complex regional pain syndrome, forefoot pain, metatarsal bursitis, Morton's neuroma, stress (fatigue or insufficiency) fracture, osteonecrosis (avascular necrosis), hallux rigidus and hallux valgus, sesamoiditis, physical examination.
Adult shoulder pain: localized pain, acute injury, glenohumeral joint (trauma, instability, arthritis), rotator cuff, adhesive capsulitis, osteoarthritis, acromioclavicular joint, physical examination.
Adult elbow pain: acute injury, lateral and medial epicondylitis, forearm (or wrist) pain and trauma, physical examination.
Adult wrist and hand pain: acute injury/localized trauma (wrist, hand, finger), forearm/wrist pain, tenosynovitis, DeQuervain's disease, trigger finger, carpal tunnel syndrome (CTS), osteoarthritis, arthropathy (inflammatory, crystal, infection), rheumatoid arthritis, physical examination.
Adult spinal trauma/injury: lumbar, thoracolumbar, thoracic, cervical spine, whiplash, chest wall (thorax), pelvis, sacrum, coccyx, coccydynia.
Adult lumbar spine disorders: acute, subacute, and chronic LBP, nontraumatic, uncomplicated, sciatica, radicular pain, stenosis (lateral, central), spondylolithesis, inflammatory spondyloarthropathy, fracture, neoplasia, infection, structural deformity, surgical fusion, laboratory examination, cauda equina syndrome, abdominal aortic (aneurysm, dissection, condition, risk factors), complication, epidemiology, physical examination.
Adult thoracic spine disorders: acute, subacute, chronic, osteoporosis (collapse, compression, fracture, risk factors, evaluation, monitoring), scoliosis, chest wall pain, aortic dissection, complication, epidemiology, physical examination.
Adult cervical spine disorders: acute, subacute, chronic, nontraumatic, radicular, neck and arm pain, cervical myelopathy, radiculomyelopathy, ligament laxity, atlantoaxial instability, congenital anomaly, cervical artery dissection/condition (vertebral/carotid artery, transient ischemic attack [TIA], stroke, risk factors, past trauma), complication, epidemiology, physical examination.
When available, systematic reviews and meta-analyses, RCTs, nonrandomized studies, observational studies (cohort and case-controlled studies), nonexperimental (nonanalytic) studies, and existing high-quality clinical guidelines took precedence over case studies, case series, or opinion for inclusion in the study. Articles were excluded from the recommendations if they were not available in English or French or could not be translated, if they were published before 1980 (except for some rare classic papers), and if they specifically focused on thermography, electrodiagnostic studies, intervention, nonmusculoskeletal disorders, or on children. Details of the literature review (2003–2006) are available upon request.
Phase 2: Independent Literature Assessment
Early phases of this guideline project used the European Commission classification (2001)  to classify grade of recommendations (Table 1). However, several new instruments have been proposed to evaluate scientific literature. Weaknesses identified in the grading system initially proposed by the US Agency for Healthcare Research and Quality led to the development of new tools to assess available literature. 
With the objective of independently assessing the level of evidence of citations listed in the guidelines and considering the recently developed assessment tools, 5 literature reviewers (licensed practitioners, 4 with specialized training in research methods) were invited to review the citations of the spine and extremity imaging guidelines. A number of citations were assessed by at least 2 evaluators to compare the level of agreement between evaluators. In cases of disagreement, the research team reached a decision and classified the citation.
The methodological quality of all studies that were included in the review was assessed by means of a methodological quality assessment list. Participants were asked to complete an evidence extraction table (Excel spreadsheet), provided at the outset of the review, with a list of all citations included in the spinal and extremity disorders guidelines. Requested information included summaries of study design; study objective in brief; type of analysis; methods (including population studied, and follow-up period); study results/evidence provided; scores for the scoring instruments (QUADAS, AGREE, and SPREAD); significant biases.
Three methodological scoring instruments were used, as described below.
The QUADAS instrument  of the Health Technology Assessment Diagnostic Evaluation is an evidence-based assessment tool to evaluate the quality of primary studies of diagnostic accuracy. The tool, derived from 3 systematic reviews, does not include a quality score. Instead, it is structured as a list of 14 questions that should be answered “Yes,” “No,” or “unclear” (Figure 2). Although some minor modifications have been proposed, [54, 55] QUADAS appears useful for highlighting the strengths and weaknesses of existing diagnostic accuracy studies. Effects of a negative response to item 3 (quality of the reference standard), item 5 (verification bias), items 10 and 11 (blinded interpretation of index test and reference standard), and item 14 (withdrawals) are assumed to be particularly important and thus are identified as “critical.”
Therefore, the following scoring system was used:
- High quality study = all 5 critical items + at least 3 other “yes” items = 8 yes or more
- Good quality = 1 critical item missing but > 8 yes
- Fair quality = 2 critical items missing but > 8 yes
- Poor quality = >2 critical items missing and/or <7 yes
Return to RADIOLOGY