Organization of Guide
Purpose and Rationale based upon Epidemiology
There are four principal reasons acute low back problems were selected as a subject for guideline development. One reason is the prevalence. Low back pain is the fifth most common reason for all physician visits in the United States (Hart et al.`95, Deyo et al`.06). About one-quarter of U.S. adults report low back pain lasting at least a whole day in the last three months (Deyo et al`.06) and 7.6% had at least one episode of severe acute low back pain in a one-year period (Carey et al.`96).
Most people report low back problems at some time in their lives, and national statistics indicate a general yearly prevalence in the U.S. population of 15-20 percent.1
Among working-age people surveyed, 50 percent admit to back symptoms each year.2
Surveys in Canada in 1998 (Cote) published in 1998 ] and Sweden in 2000 (Linton & Ryberg) found total population annual prevalence as high as 73%. Seeking the prevalence of neck and back pain in 3000 randomly selected 35-45 year old Swedish men, Linton & Ryberg also observed 17% using related sick-leave and 14% missing work without sick-leave.
Back symptoms, in fact, are the most common cause of disability for persons under age 45.4
At any given time, about 1 percent of the U.S. population is chronically disabled because of back problems, and another 1 percent temporarily disabled.1
A second reason for a guideline on assessment and treatment of acute low back problems is cost. Low back problems are expensive. Many patients have self-limited episodes of acute low back pain and do not seek medical care (Carey et al.`96). Among those who do seek medical care, rapid improvements in pain, disability, and return to work are typically seen in the first month (Pengel et al.`03). However, up to one-third of patients report persistent back pain of at least moderate intensity one year after an acute episode, and 20% report substantial activity limitations (von Korff & Saunders et al.`96). Approximately 5% of people with back pain disability account for 50-60% of the costs associated with low back pain (Spengler`87, Frymoyer`91, Linton & Ryberg`00).
Their total costs to society are difficult to calculate, but evidence indicates that both the economic and psycho-social costs are substantial. Low back problems are the second most common symptomatic reason expressed by patients for office visits to primary care physicians.5
They are the most common reason for office visits to orthopedic surgeons, neurosurgeons, and occupational medicine physicians. They rank third among the reasons for surgical procedures. About 2% of the U.S. work force is compensated for back injuries each year (Andersson`99).
Moreover, although medical costs are high, loss of time from work as well as the disability payments for work-related low back problems can together cost up to three times as much as medical treatment.6
About 2 percent of the U.S. work force has compensable back problems each year.7
Various estimates of the total annual societal cost of back pain in the United States range from $20 to $50 billion.8 These estimates have risen greatly since 1994 (Stewart et al.`03, Ricci et al.`06) In 1998, total incremental direct health care costs unrelated to days lost from work are attributable to low back pain in the U.S. were estimated at $26.3 billion (4). N
onmonetary costs of low back problems can also be substantial. The inability to function normally at work and in other daily activities has an impact on both patients and their families.
A third important reason for this guideline is the increasing evidence that many patients with activity intolerance due to low back symptoms may be receiving care that is inappropriate or at least less than optimal. Rates for hospitalization and surgery for low back problems vary substantially among regions of the United States as well as among small areas within states.9
Marked regional variations also occur in the use of diagnostic tests for assessing low back problems.9 More recent data indicate little has changed.(Lurie et al.`03)
These variations imply a lack of consensus about appropriate assessment and treatment of low back problems, suggesting that some patients may be receiving inappropriate or suboptimal care.
In addition, some patients appear to be more disabled after treatment than before, another potential indicator of sub-optimal care. Perhaps the most obvious examples involve surgery. Despite an extensive medical literature on "failed back surgery" and evidence that repeat surgical procedures for low back problems rarely lead to improved outcome, there are documented examples of patients who have had as many as 20 spine operations.9
However, surgery is not the only treatment that can lead to increased disability. Common treatment methods such as extended bed rest or extended use of high-dose opioids can prolong symptoms and further debilitate patients.
A fourth reason for the guideline is a growing body of research on low back problems, allowing a systematic evaluation of commonly used assessment and treatment methods. Although the existing literature has shortcomings, there is sufficient scientific evidence for a number of conclusions about the efficacy and safety of current assessment and treatment methods.
Scope and Organization
This Clinical Practice Guideline is intended to provide primary care clinicians with information and recommended strategies for the assessment and treatment of low back problems in adults. To develop this guideline, AHCPR convened a private-sector, multidisciplinary panel of clinicians, researchers, and a consumer representative to evaluate the scientific evidence in the medical literature, draw conclusions, and make recommendations.
In determining the scope of the guideline, the panel focused on information needed for primary care assessment and treatment of adults with low back problems. "Back problems" were defined as activity intolerance due to back-related symptoms and "acute" as limitations of less than 3 months' duration. Back symptoms include pain in the back as well as back-related leg pain (sciatica). The panel agreed that the guideline should provide information on initial detection of underlying serious conditions (such as fracture, tumor, infection, or cauda equina syndrome) that could be causing low back problems, but that treatment of these conditions is outside the scope of the guideline.
The panel agreed further that the assessment and treatment of patients who have chronic low back problems (with symptoms lasting over 3 months) may be quite different than for patients with acute problems. Patients who become disabled due to chronic low back problems represent less than 5 percent of those with low back problems, but they account for up to 60 percent of the societal costs for this disorder.6
To a much greater extent than acute problems, chronic low back problems are influenced by complex psychological, behavioral, socioeconomic, demographic, legal, and occupational factors, many of which are not easily controlled.8
For these reasons, the panel decided that chronic low back problems are beyond the scope of a guideline on acute problems. The recommendations included in the guideline may not apply to persons younger than 18 years since diagnostic and treatment considerations for this group are often different than for adults.
Evaluation of Evidence. The panel agreed that this guideline on low back problems should be anchored to published scientific evidence, and that such evidence should take priority over panel opinion in making guideline recommendations. In looking at a proposed assessment or treatment method, the panel considered: (1) efficacy, (2) potential harms, and (3) costs.
The panel considered randomized controlled trials
(RCTs) that focused on patient-oriented clinical outcome measures such as symptom relief or improved level of functioning to be the acceptable method for establishing the efficacy of treatment methods. Evidence about efficacy of assessment methods was considered adequate if results of the diagnostic test studied were compared to an independent reference standard in a way that allowed calculation of standard test parameters, such as the test's true-positive rate (sensitivity) and true-negative rate (specificity).
The panel agreed to give the greatest weight to scientific research evidence that met the above criteria. When such strong scientific evidence was not available, the panel labeled the evidence as weak and indirect and used the combined expert opinion and clinical judgment of panel members for interpretation. In all cases, the guideline explicitly states the type of evidence used by the panel as the basis for recommendations. The scale used for labeling the evidence is at the end of this chapter.
. The panel found in 1994 that studies of interventions aimed at preventing low back problems or their risk factors present conflicting findings and explain only a small portion of back complaints. Few prevention studies had been well designed, and most have been conducted in workplace settings focusing on injury claims or have used interventions that could not easily be carried out by primary care providers. When information from these studies was applicable to primary care, however, it was included under specific areas of assessment or treatment in the guidelines. (See Primary Prevention Trials
The panel agreed that a methodological problem commonly associated with studies of the prevention of back problems is lack of precision in specifying the goal(s) of the preventive intervention. Researchers often fail to establish whether the goal is to prevent the first episode of low back symptoms, activity limitations, recurrent episodes, injury claims, time lost from work, chronic disability, and/or medical care utilization and costs. In addition, some authors have suggested that efforts to prevent first or recurrent episodes of low back symptoms at work may be futile, and that research should focus instead on preventing long-term disability that results in high-cost disability claims.6
The Pengel et al.`03
review of 15 studies of variable methodological quality found rapid improvements the rule. Within one month pain (mean reduction 58% of initial scores), disability (58%), and return to work (82% of those initially off work) improved. Further improvement was apparent until about three months. Thereafter levels for pain, disability, and return to work remained almost constant. Within 12 months 73% of patients had at least one recurrence. They concluded that acute low back patients' pain and disability usually improve rapidly within weeks. None the less, pain and disability are typically ongoing, and recurrences are common.
So far other than conditioning for primary prevention
sparked by Gundewall et al.`93, and secondary prevention (Conditioning
) and treatment with activity paradigm supported by reliable science for secondary prevention (Activity Paradigm
), our prevention impact is not impressive according to prospective prevention trials and the prospective cohort studies (See Prevention attempts
Organization and Clinical Categories ______________________________________
Chapter 2 of this guideline focuses on the initial assessment of the patient with activity limitations
due to low back symptoms, and Chapter 2 addresses initial treatment methods for these patients. The assessment and treatment methods considered in these chapters can typically be managed by the primary care clinician. Up to 90 percent of patients with acute low back problems recover within 1 month from activity limitations due to symptoms.6
Chapter 3 addresses diagnostic and treatment considerations for the small percentage of patients who still have substantial symptoms or limitations after 1 month. Many of these diagnostic and therapeutic methods can be managed by the primary care clinician; others will require consultation with a specialist.
The panel recognized that different clinical disciplines use a variety of diagnostic labels that implicitly suggest a cause for low back symptoms. However, these labels are often unreliable for categorizing causes of low back problems. Even after an extensive workup, only about 15 percent of patients can be given a definitive diagnosis.16
Since the many diagnostic labels currently used to describe low back problems may confuse patients and clinicians, the panel considered it more useful to classify a patient's low back problem into one of three descriptive clinical categories based on medical history and physical examination findings:
- Potentially serious spinal condition: spinal tumor, infection, fracture, or cauda equina syndrome suggested by findings from medical history or physical examination ("Red Flags
- Sciatica: back-related lower limb symptoms suggesting nerve root compromise.
- Nonspecific (neurological negative) back symptoms: symptoms occurring primarily in the back that suggest neither nerve root compromise nor a serious underlying condition.
In the panel's opinion, clinicians would have enough information to make appropriate decisions about initial assessment and treatment, as well as some hints about prognosis, after correctly classifying patients with low back problems into one of the above three categories. The panel used this classification scheme in making guideline recommendations about assessment and treatment methods.
Methodology for Guideline Development
The general theory and principles underlying development of clinical practice guidelines are presented in an Institute of Medicine report;17
other reports published by AHCPR provide specific information on the clinical guideline development process.18
These materials provided a starting point for developing the Clinical Practice Guideline on low back problems. AHCPR provided the general parameters for guideline development. The panel, aided by the methodologists and consultants, then independently determined the specific methodology for the project, directed the literature review
, and developed the guideline findings and recommendations.
Formation of the Panel and Staff__________________________________________
AHCPR initiated formation of the panel and appointed its chairperson and members. Important considerations in the choice of panel members were: (1) demonstrated knowledge about low back problems, (2) representation of major clinical disciplines involved in back care, and (3) geographic diversity. Nominations were solicited through a Federal Register announcement and from professional and consumer organizations and persons interested in the care of patients with low back problems.
More than 200 individuals were nominated. AHCPR selected 23 representing the fields of biomechanical and spine research, chiropractic care, emergency medicine, family medicine, internal medicine, neurology, neurosurgery, occupational health nursing, occupational medicine, occupational therapy, orthopedics, osteopathic medicine, physical and rehabilitation medicine, physical therapy, psychology, rheumatology, and radiology.
The panel also included a consumer representative who had experienced low back problems, but did not work in the health care field. Several consultants with expertise in spine research, clinical care of low back problems, clinical epidemiology, and health economics were appointed to the panel. Two methodologists with experience in developing clinical practice guidelines were assigned to the panel by AHCPR. Both methodologists were physicians with MPH degrees, one an emergency medicine physician and one an internist. The methodologists aided the panel in determining the scope of the literature search and the criteria to be used for selecting articles for panel review.
The panel chair formed a research and support staff that included two physicians: a spine-fellowship-trained orthopedic surgeon and an occupational-medicine-trained physician with an MPH degree. National Library of Medicine representatives aided the staff in retrieving literature. The staff screened articles and constructed evidence tables for articles according to panel review criteria. These evidence tables and the original articles were presented to the panel for review and interpretation. The panel used this information as the basis for its guideline findings and recommendations.
Public Comment and Peer Review________________________________________
An open forum was held early in the guideline development process to give interested individuals, organizations, and agencies the opportunity to present written or verbal testimony. Later in the process, drafts of the guideline were sent out for peer and pilot review. AHCPR selected peer and pilot reviewers from those who had expressed interest in the guideline, participated in the open forum, or were nominated by professional organizations or panel members.
Over 100 peer reviewers were selected based on their expertise in the care of low back problems. They were asked to evaluate the comprehensiveness of the literature review as well as the panel's findings and recommendations. The pilot reviewers who were selected represented a cross-section of primary care settings including private and group practices, health maintenance organizations, and occupational medicine clinics. They were asked to evaluate the practical applicability of the guideline in their own practice settings by using examples published in the Quick Reference Guide for Clinicians and by soliciting feedback from patients given the Consumer Version. The panel used comments from peer and pilot reviewers to guide final revisions of the guideline.
Literature Search ________________________________________________________
The panel initiated a comprehensive literature search of topics deemed applicable to low back problems. The Quebec Task Force on Spinal Disorders had previously published an evidence-based guideline on low back problems, based upon an exhaustive literature search through 1984.19
The bibliography from their report was the starting point in the literature search for this AHCPR guideline.
The literature search of articles published after 1984 was performed through the National Library of Medicine. Abstracts of 10,317 articles which met the search criteria were each independently evaluated by the orthopaedic surgeon and occupational medicine physician on the research staff. If either reviewer thought an article might be useful, the entire article was retrieved. A total of 3,918 articles (38 percent of all abstracts evaluated) was obtained for further evaluation.
Additional articles came from panel members, from the open forum process, and from unsolicited sources. All articles were entered in a comprehensive bibliography, classified by topic, and screened methodologically to determine if they contained information that might be useful to the panel.
Evaluation of Efficacy ___________________________________________________
In evaluating efficacy of assessment and treatment methods, the panel decided to focus on how each method affected clinical outcomes important to patients and society. Examples of such outcomes are symptoms, level of physical functioning, patient satisfaction, and morbidity and mortality (as complications of the assessment or treatment method). The panel dealt with costs, another outcome of interest to patients and society, as a separate issue. Cost was not considered when evaluating efficacy.
The panel used a standard methodology to identify and evaluate the best scientific evidence available on the efficacy of each assessment and treatment method, while focusing on clinical outcomes. This process included a systematic evaluation of each study's quality and its clinical applicability to patients with low back problems. The panel used this information to screen all articles, using minimum article selection criteria for efficacy. Articles meeting these minimum criteria were prioritized (giving priority to articles of higher quality and clinical applicability), and data from the higher priority articles were abstracted onto evidence tables.
The panel then reviewed the available data from both evidence tables and original articles to decide how much weight to give each study in developing the "findings and recommendations" statements for this guideline. The greatest weight was given to studies of high quality that evaluated adults with low back problems, although few such studies were found.
For most topics, the quality and clinical applicability of studies reviewed were limited. Inclusion and exclusion criteria for subjects were often either incompletely described or so broad that they allowed for wide variations in age, symptoms, symptom duration, examination findings, prior treatments, and other potentially confounding factors. Studies often inadequately described the baseline demographic and clinical characteristics of subjects. Many studies did not distinguish acute from chronic patients; others failed to either describe or control for factors known to cause significant variation in outcome (such as prior back surgery). Certain studies lacked appropriate statistical analysis or included too few subjects to attain adequate statistical power.
Evaluation of Potential Harms and Costs __________________________________
Evaluating Harms. Since back problems are rarely life-threatening, the panel paid special attention to potential harms (side effects or complications) of assessment and treatment methods. Controlled trials evaluating treatment and assessment methods, however, seldom included enough subjects to detect rare but potentially serious complications. This information was found only in large case series or case reports. On the other hand, controlled trials of oral medications often included extensive information on side effects. Thus, accurate comparison of the relative risks of side effects and complications of different assessment and treatment methods was not possible.
A lack of published evidence about harms related to specific treatment or assessment methods does not mean that potential harms do not exist. In many instances, the side effects and complications of assessment and treatment methods have never been extensively studied or comprehensively reported. In addition, articles evaluating newer treatment and assessment methods are often written by advocates of these methods, who may tend to deemphasize the harms.
The panel felt it was important for both clinicians and patients to have a sense of potential harms relative to the potential benefits of these methods. Therefore, the panel considered information about potential harms from a variety of sources, including case series, case reports, cross- sectional surveys, clinical trials, and in some instances studies of patients who did not have low back problems. Finally, if no specific information was available from any of these sources, the panel generally considered whether the method was invasive or carried the potential for an allergic reaction.
Evaluating Costs. Both clinicians and patients need to consider relative costs of assessment and treatment methods before making informed decisions about care. Costs vary greatly, however, and the cost data on assessment and treatment methods for low back problems are limited. The unit cost of a service may vary within and between geographical regions. The aggregate cost of services also varies depending on the frequency and duration of services for the individual patient. Although costs of various medical services have generally increased in recent years, they have done so at inconsistent rates. Given these variations, the panel decided to make broad statements about whether methods appeared to be of low, moderate, or high cost, graded according to the following system (based on 1993 dollars):
1. Low cost: under $200.
2. Moderate cost: $200 to $1,000.
3. High cost: over $1,000.
This grading system provides no more than a rough comparison of costs, and the panel recognized that the divisions between cost categories are somewhat arbitrary. For example, some Americans may not consider a $199 expense that comes directly out of pocket to be "low cost."
Developing the Guideline Recommendations ______________________________
To develop recommendations for each assessment and treatment method, the panel considered: (1) the quality and amount of evidence for efficacy, (2) the strength of the effect found for the method, (3) the consistency of findings between studies, (4) the clinical applicability of the evidence to adult patients with low back problems, and (5) any evidence on harms or costs. For each assessment and treatment method the panel then sought to answer the following questions:
1. What is the likelihood that this assessment or treatment method will:
- Benefit the patient?
- Harm the patient?
2. Does the likelihood and magnitude of potential benefit outweigh the likelihood and magnitude of potential harm enough to justify the cost for this method?
The development of "findings and recommendations" statements required the collective judgment of the panel in interpreting the available evidence. The panel rated the amount and quality of evidence supporting each guideline statement using the scale in Figure 1 below.
Figure 1. Strength of Evidence: Panel ratings available guideline statement supporting evidence
A = Strong research-based evidence (multiple relevant and high-quality
B = Moderate research-based evidence (one relevant, high-quality
scientific study or multiple adequate scientific studies*).
C = Limited research-based evidence at least one adequate scientific
study* in patients with back symptoms).
D = Panel interpretation of information that did not meet inclusion
criteria as research-based evidence (not RCT or specific to back symptoms)
* Met minimal formal criteria for scientific methodology and relevance to population and specific method addressed in guideline statement.
The guideline Panel's "findings and recommendations" statements represent the panel's assessment of a method's strength of clinical effect to achieve the intended assessment or treatment goals after accessing the potential benefit relative to potential harms and costs. The strength of effect has no bearing on A B C D above about level of scientific support for each statement that may deal with the relative effectiveness.
A B C D is not an indication that modality with B level evidence is more effective than one with C level of evidence - it only indicates the quality of data supporting the statement.
This rating system (A, B, C, or D) is assessment of the reliable literature upon which to give:
YELLOW LIGHT or
according to evidence of benefit relative to potential harms and cost:
Recommendations for: If the available A,B,C evidence indicates that potential BENEFITS outweigh potential harms.
Options: If the available A,B,C,D evidence of potential benefits is WEAK or EQUIVICAL (inconsistency in some studies) but potential harms and costs appear small.
Recommendations against: If the available A,B,C evidence indicates either a lack of benefit or that POTENTIAL HARMS and costs outweigh potential benefits.
SRC Spinemate - only used or reproduced with written permission from SResourceC@aol.com