Low Back Pain in Adults evaluation

The following information was derived from UpToDate.com

Evaluation of low back pain in adultsAuthors:Stephanie G Wheeler, MDJoyce E Wipf, MDThomas O Staiger, MDRichard A Deyo, MD, MPHJeffrey G Jarvik, MD, MPHSection Editor:Steven J Atlas, MD, MPHDeputy Editor:Lisa Kunins, MD

Contributor DisclosuresAll topics are updated as new evidence becomes available and our peer review process is complete.Literature review current through: Oct 2020. | This topic last updated: Jun 25, 2019.

INTRODUCTIONIt is estimated that up to 84 percent of adults have low back pain at some time in their lives [1,2]. For many individuals, episodes of back pain are self-limited. Patients who continue to have back pain beyond the acute period (four weeks) have subacute back pain (lasting between 4 and 12 weeks) and may go on to develop chronic back pain (persists for ≥12 weeks) [3]. Rarely, back pain is a harbinger of serious medical illness.

This discussion will focus on an approach to the initial evaluation, including diagnostic tests, of a patient presenting with low back pain in the primary care setting. The treatment of acute, subacute, and chronic low back pain are discussed separately. (See “Treatment of acute low back pain” and “Subacute and chronic low back pain: Nonpharmacologic and pharmacologic treatment” and “Subacute and chronic low back pain: Nonsurgical interventional treatment” and “Subacute and chronic low back pain: Surgical treatment”.)

TERMINOLOGYSeveral terms are used to describe conditions related to the back, based upon radiologic findings (eg, spondylosis), physical findings (radiculopathy), and symptoms (sciatica). These terms are defined in the table (table 1).

EPIDEMIOLOGYIn 2010, back symptoms were the principal reason for 1.3 percent of office visits in the United States [4]. Spinal disorders accounted for 3.1 percent of diagnoses in outpatient clinics.

Prevalence — The prevalence of back pain has been estimated with surveys [1,5]. A 2012 systematic review estimated that the global point prevalence of activity-limiting low back pain lasting for more than one day was 12 percent and the one-month prevalence was 23 percent [6].

Other survey estimates of the prevalence of low back pain have ranged from 22 to 48 percent, depending on the population and definition [2,7-9]. For example, the 2002 National Health Interview Survey found that 26 percent of respondents reported low back pain lasting at least one day in the last three months [7].

Risk factors — Risk factors associated with back pain complaints include smoking, obesity, age, female gender, physically strenuous work, sedentary work, psychologically strenuous work, low educational attainment, Workers’ Compensation insurance, job dissatisfaction, and psychologic factors such as somatization disorder, anxiety, and depression [2,8,10-15].

ETIOLOGIESAlthough there are many etiologies of low back pain (table 2), the majority of patients seen in primary care will have nonspecific low back pain.

Nonspecific back pain — The vast majority of patients seen in primary care (>85 percent) will have nonspecific low back pain, meaning that the patient has back pain in the absence of a specific underlying condition that can be reliably identified [16-18]. Many of these patients may have musculoskeletal pain [3]. Most patients with nonspecific back pain improve within a few weeks. (See “Treatment of acute low back pain”, section on ‘Prognosis’.)

Serious systemic etiologies — Among patients who present with back pain to primary care settings, less than 1 percent will have a serious systemic etiology (cauda equina syndrome, metastatic cancer, and spinal infection) [5,19]. Almost all patients with these conditions will have risk factors or other symptoms [18].

Spinal cord or cauda equina compression – There are many causes of cauda equina syndrome, the most common being herniation of the intervertebral disc. One systematic review found that cauda equina syndrome was caused by herniation of the intervertebral disc in 22.7 percent of cases, ankylosing spondylitis in 15.9 percent, lumbar puncture in 15.9 percent, trauma in 7.6 percent, malignant tumor in 7.2 percent, benign tumor in 5.7 percent, and infection in 5.3 percent [20]. While the incidence of cord compression in patients known to have cancer varies depending on the cancer, among patients who are diagnosed with cord compression, it is the initial manifestation of malignancy in 20 percent [21]. Metastatic disease from any primary cancer can cause cord compression. (See “Clinical features and diagnosis of neoplastic epidural spinal cord compression”, section on ‘Epidemiology’ and “Clinical features and diagnosis of neoplastic epidural spinal cord compression”, section on ‘Pathophysiology’.)

Pain is usually the first symptom of cord compression, but motor (usually weakness) and sensory findings are present in the majority of patients at diagnosis. Bowel and/or bladder dysfunction are generally late findings. Early diagnosis and treatment improves outcomes. (See “Clinical features and diagnosis of neoplastic epidural spinal cord compression”, section on ‘Clinical features’ and “Treatment and prognosis of neoplastic epidural spinal cord compression”, section on ‘Importance of early detection’.)

Metastatic cancer – The bone is one of the most common sites of metastasis. A history of cancer (excluding nonmelanoma skin cancers) is the strongest risk factor for back pain from bone metastasis [22]. Among solid cancers, metastatic disease from breast, prostate, lung, thyroid, and kidney cancers account for 80 percent of skeletal metastases. Approximately 60 percent of patients with multiple myeloma have skeletal lytic lesions present at diagnosis. (See “Epidemiology, clinical presentation, and diagnosis of bone metastasis in adults”, section on ‘Epidemiology’ and “Multiple myeloma: Clinical features, laboratory manifestations, and diagnosis”, section on ‘Skeletal surveys’.)

Pain is the most common symptom. In patients with a history of cancer, sudden, severe pain raises concern for pathologic fracture. Patients may also have neurologic symptoms from either spinal cord compression or spinal instability. (See “Epidemiology, clinical presentation, and diagnosis of bone metastasis in adults”, section on ‘Clinical presentation’.)

Spinal epidural abscess –Spinal epidural abscess is a rare but serious cause of back pain. Initial symptoms (eg, fever and malaise) are often nonspecific; over time, localized back pain may be followed by radicular pain and, left untreated, neurologic deficits. Risk factors include recent spinal injection or epidural catheter placement, injection drug use, and other infections (eg, contiguous bony or soft tissue infection or bacteremia). Immunocompromised patients may also be at higher risk. Treatment of spinal epidural abscess is reviewed in detail elsewhere. (See “Spinal epidural abscess”, section on ‘Management’.)

Vertebral osteomyelitis –The majority of patients with vertebral osteomyelitis will present with back pain, which gradually increases over weeks to months [23]; fever may or may not be present. The intervertebral disc may also be become infected (discitis), and the clinical presentation (positional discomfort, pain to palpation, neurologic signs/symptoms) may vary depending upon the extent of the infection. (See “Vertebral osteomyelitis and discitis in adults”, section on ‘Clinical features’.)

The incidence of vertebral osteomyelitis generally increases with age, and men are more commonly affected than women. Many cases are thought to be health care-related or postprocedural from hematogenous spread of bacteremia. Less specific risk factors include an immunocompromised state and injection drug use. (See “Vertebral osteomyelitis and discitis in adults”, section on ‘Epidemiology’ and “Vertebral osteomyelitis and discitis in adults”, section on ‘Pathogenesis’.)

Management of vertebral osteomyelitis is discussed in detail elsewhere. (See “Vertebral osteomyelitis and discitis in adults”, section on ‘Treatment’.)

Less serious, specific etiologies — Less than 10 percent of patients who present in primary care settings with low back pain will have less serious but specific etiologies for their pain [19,24].

Vertebral compression fracture – Approximately 4 percent of patients presenting in the primary care setting with low back pain will have a vertebral compression fracture [19]. While some produce no symptoms, other patients present with acute onset of localized back pain which may be incapacitating. There may be no history of preceding trauma. Risk factors for osteoporotic fracture include advanced age and chronic glucocorticoid use (table 3). A history of an osteoporotic fracture is a risk factor for subsequent fractures, which can be mitigated by pharmacologic therapy. (See “Osteoporotic thoracolumbar vertebral compression fractures: Clinical manifestations and treatment”, section on ‘Clinical manifestations’ and “Osteoporotic fracture risk assessment” and “Overview of the management of osteoporosis in postmenopausal women”, section on ‘Post-fracture’.)

Approximately 3 to 4 percent of patients who present in primary care settings with low back pain have a symptomatic disc herniation or spinal stenosis [19].

Radiculopathy – Radiculopathy refers to symptoms or impairments related to a spinal nerve root. Damage to a spinal nerve root may result from degenerative changes in the vertebrae, disc protrusion, and other causes. The clinical presentations of lumbosacral radiculopathy vary according the level of nerve root or roots involved. Over 90 percent are L5 and S1 radiculopathies [25]. Patients present with pain, sensory loss, weakness, and/or reflex changes consistent with the nerve root involved; these are summarized in the table and discussed in more detail separately (table 4). Testing for lumbar nerve root compression is shown in a figure (figure 1). Many patients with symptoms of acute lumbosacral radiculopathy improve gradually with supportive care. (See “Acute lumbosacral radiculopathy: Pathophysiology, clinical features, and diagnosis”, section on ‘Pathophysiology and etiology’ and “Acute lumbosacral radiculopathy: Pathophysiology, clinical features, and diagnosis”, section on ‘Clinical presentations’ and “Acute lumbosacral radiculopathy: Treatment and prognosis”, section on ‘Prognosis’.)

Sciatica is a nonspecific term used to describe a variety of leg or back symptoms. Usually, sciatica refers to a sharp or burning pain radiating down from the buttock along the course of the sciatic nerve (the posterior or lateral aspect of the leg, usually to the foot or ankle) [26]. Most sciatica is attributable to radiculopathy at the L5 or S1 level from a disc disorder.

Spinal stenosis – Lumbar spinal stenosis is most often multifactorial. Spondylosis (degenerative arthritis affecting the spine) spondylolistheses, and thickening of the ligamentum flavum are the most common causes, typically affecting patients >60 years (figure 2). (See “Lumbar spinal stenosis: Pathophysiology, clinical features, and diagnosis”, section on ‘Etiologies’.)

Ambulation-induced pain localized to the calf and distal lower extremity resolving with sitting or leaning forward (“pseudoclaudication” or “neurogenic claudication”) is a hallmark of lumbar spinal stenosis. Other symptoms of lumbar spinal stenosis can include back pain and sensory loss and weakness in the legs, though many patients may present with a normal neurologic exam. Symptoms of neurogenic claudication can usually be distinguished from vascular claudication (table 5). Rare patients develop a cauda equina syndrome. Patients often have symptoms only when active. Most patients with spinal stenosis related to osteoarthritis will have stable symptoms over time. A trial of conservative, nonsurgical treatment is the initial therapy for most patients [27]. (See “Lumbar spinal stenosis: Pathophysiology, clinical features, and diagnosis”, section on ‘Clinical presentation’ and “Lumbar spinal stenosis: Treatment and prognosis”, section on ‘Prognosis’ and “Lumbar spinal stenosis: Treatment and prognosis”, section on ‘Nonsurgical treatment’.)

Other etiologies

Ankylosing spondylitis –Among patients who present in primary care settings for back pain, it is estimated that approximately 0.5 percent will have ankylosing spondylitis [19,24]. It is most commonly diagnosed in men under the age of 40 years. (See “Clinical manifestations of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults”, section on ‘Epidemiology’.)

Almost all patients report back pain, which often has characteristics suggesting an inflammatory etiology (morning stiffness, improvement with exercise, pain at night) [3]. Patients may also have extraskeletal disease manifestations (eg, uveitis). (See “Clinical manifestations of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults”, section on ‘Musculoskeletal symptoms and findings’ and “Clinical manifestations of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults”, section on ‘Extraarticular manifestations and comorbidities’.)

Osteoarthritis –Low back pain may be a symptom of osteoarthritis of the facet joints spine. Patients may also complain of hip pain, either from osteoarthritis of the hip or referred pain from the spine. Osteoarthritis most commonly presents in patients over the age of 40. Pain is typically exacerbated by activity and relieved by rest (table 6). Osteoarthritis can lead to spinal stenosis. (See “Clinical manifestations and diagnosis of osteoarthritis”, section on ‘Facet joint’.)

Scoliosis and hyperkyphosis –Back pain can be associated with scoliosis and hyperkyphosis. (See “Adolescent idiopathic scoliosis: Management and prognosis”, section on ‘Outcome’ and “Overview of hyperkyphosis in older persons”, section on ‘Other health-related consequences’.)

Psychologic distress – Psychologic distress (eg, depression or somatization) may contribute to the severity symptoms of low back pain or may be a cause of nonorganic back pain [3]. (See “Somatic symptom disorder: Epidemiology and clinical presentation”, section on ‘Clinical presentation’.)

Etiologies outside the spine –Low back pain may be a symptom of problems outside the back. Examples of other etiologies include pancreatitis, nephrolithiasis, pyelonephritis, abdominal aortic aneurysm, or herpes zoster [3,17]. Patients generally have other accompanying symptoms. (See “Clinical manifestations and diagnosis of acute pancreatitis”, section on ‘Clinical features’ and “Diagnosis and acute management of suspected nephrolithiasis in adults”, section on ‘Clinical manifestations’ and “Acute complicated urinary tract infection (including pyelonephritis) in adults”, section on ‘Clinical manifestations’ and “Clinical features and diagnosis of abdominal aortic aneurysm”, section on ‘Clinical presentations’ and “Epidemiology, clinical manifestations, and diagnosis of herpes zoster”, section on ‘Clinical manifestations’.)

There are also clinical entities that are possibly associated with low back pain symptoms:

Piriformis syndrome –The piriformis syndrome is thought by some to be a condition in which the piriformis muscle, a narrow muscle located in the buttocks, compresses or irritates the sciatic nerve [26,28,29].

Sacroiliac joint dysfunction –”Sacroiliac joint dysfunction,” a term to describe pain in the region of the sacroiliac joint believed to be due to malalignment or abnormal joint movement, is a controversial topic. Diagnosing this condition is difficult due to the absence of an agreed upon “gold standard” [30]. Tests of pelvic symmetry or sacroiliac joint movement have been shown to have low intertester reliability [31-37], and provocative maneuvers such as fluoroscopically guided injections of the sacroiliac joint have been unreliable in diagnosis and treatment [36,38,39]. The sacroiliac joint may be a referred site of pain, including from a degenerative disc at L5-S1, spinal stenosis, or osteoarthritis of the hip.

Bertolotti’s syndrome –Back pain in the setting of a transitional vertebra is known as “Bertolotti’s syndrome.” A transitional vertebra is a common finding on radiologic studies. It is a congenital anomaly with a naturally occurring articulation or bony fusion between the transverse processes of L5 and the sacrum. Estimates of prevalence of transitional vertebra range from 4 to 36 percent [40]. It remains unclear whether these individuals have a higher risk of back pain than those without such an anomaly. Generally, patients with Bertolotti’s syndrome should initially be treated similarly as patients with nonspecific back pain. Whether and when surgical intervention is appropriate remains unclear.

INITIAL EVALUATIONThe clinical evaluation of low back pain includes a history and physical to evaluate for signs or symptoms that indicate need for immediate imaging and further evaluation. For most patients with acute back pain (<4 weeks), laboratory tests and imaging are not necessary in the initial evaluation.

History — While it may not be possible to define a precise cause of low back symptoms for most patients, it is important to evaluate for evidence of specific etiologies of back pain. The history should include location, duration, and severity of the pain, details of any prior back pain, and how current symptoms compare with any previous back pain.

We also ask about constitutional symptoms (eg, unintentional weight loss, fever, or night sweats), history of malignancy, precipitants or precipitating events, therapies attempted, neurologic symptoms (eg, weakness, falls or gait instability, numbness or other sensory changes, or bowel/bladder symptoms), stability or progression of symptoms, history of recent bacterial infections (particularly bacteremia), recent history or current use of injection drugs, history or current use of corticosteroid medications, and recent history of epidural or spinal procedures.

Patients should also be evaluated for social or psychologic distress that may be contributing [3]. Potentially useful items are a history of failed previous treatments, substance use disorder, and disability compensation. Screening for depression may be helpful. (See “Screening for depression in adults”.)

Features that may suggest underlying systemic disease include history of cancer, age >50 years, unexplained weight loss, duration of pain >1 month, nighttime pain, and unresponsiveness to previous therapies [3]. Documented fevers, injection drug use, recent bacterial infection (particularly bacteremia), or recent epidural or spinal instrumentation increase the suspicion of spinal infection. (See “Spinal epidural abscess” and “Vertebral osteomyelitis and discitis in adults”.)

Physical examination — In general, the purpose of the physical examination is to identify features that suggest that further evaluation is indicated, rather than to make a primary diagnosis. The physical examination should include the following components:

Inspection of back and posture – Inspection of the patient on physical examination can reveal anatomic abnormalities such as scoliosis or hyperkyphosis (table 1). (See “Overview of hyperkyphosis in older persons”.)

Palpation/percussion of the spine – Palpation and/or percussion of the back is usually performed to assess vertebral or soft tissue tenderness. Vertebral tenderness is a sensitive, but not specific, finding for spinal infection, and may also be seen in patients with vertebral metastases and osteoporotic compression fracture [41]. (See “Vertebral osteomyelitis and discitis in adults”, section on ‘Symptoms and signs’.)

Neurologic examination – Patients should have a neurologic examination including evaluation of the reflexes, strength, sensation, and gait. (See “The detailed neurologic examination in adults”.)

For patients suspected of having a radiculopathy, neurologic testing should focus on the L5 and S1 nerve roots (table 4), since most clinically significant radiculopathies occur at these levels. (See “Acute lumbosacral radiculopathy: Pathophysiology, clinical features, and diagnosis”, section on ‘Physical examination’.)

Straight leg raising – The straight leg raise and other maneuvers can be helpful in identifying whether symptoms are radicular in nature. These are described separately. (See “Acute lumbosacral radiculopathy: Pathophysiology, clinical features, and diagnosis”, section on ‘Maneuvers’.)

Nonorganic signs (Waddell’s signs) – Patients with psychologic distress that is contributing to back pain symptoms may have associated inappropriate physical signs, also known as “Waddell’s signs” (table 7). These include patient overreaction during physical examination, superficial tenderness, straight leg raise that improves when the patient is distracted, unexplainable neurological deficits (eg, nondermatomal distribution of sensory loss, sudden giving way or jerky movements with motor exam, inconsistency in observed spontaneous activity [dressing, getting off table]), and pain elicited by axial loading (pressing down on top of head, or rotating the body at hips or shoulders) [42]. The presence of multiple Waddell’s signs suggests a psychologic component to a patient’s pain, although they do not seem to be useful for predicting the ability to return to work or success of rehabilitation [3,43,44].

Other – For patients with new or worsening urinary incontinence, we measure bladder post void residual (eg, by ultrasound) to differentiate overflow incontinence from urge and/or stress incontinence. If a patient’s history strongly suggests malignancy, we evaluate as appropriate (eg, lymph node exam, breast exam, prostate evaluation). Other physical examination components (eg, hip examination or examination for peripheral vascular disease) should be performed based on the history. (See “Clinical features and diagnosis of lower extremity peripheral artery disease”, section on ‘Physical examination’ and “Approach to the adult with unspecified hip pain”, section on ‘Diagnostic approach’.)

Laboratory studies — Most patients with acute low back pain do not require any laboratory testing. In some patients with suspected infection or malignancy, we use the erythrocyte sedimentation rate (ESR) and/or C-reactive protein (CRP) in addition to plain radiographs to determine the need for advanced imaging [22,45-47]. Because of its higher sensitivity, CRP may have similar or greater value than the ESR; however, CRP has not been similarly evaluated in the evaluation of low back pain. (See ‘Risk assessment for acute back pain’ below.)

The ESR and CRP are also used in the diagnosis of ankylosing spondylitis. (See “Diagnosis and differential diagnosis of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults”, section on ‘Laboratory testing’.)

IMAGING

Limited utility of imaging — Earlier use of imaging for low back pain without associated symptoms is not associated with improved outcomes but increases the use of invasive procedures. A 2009 systematic review and meta-analysis of six trials that compared immediate imaging (lumbosacral spine magnetic resonance imaging [MRI], computed tomography [CT], or radiography) with usual care for patients with acute and subacute low back pain, without signs or symptoms of infection or malignancy, found no significant differences in short-term (up to three months) or long-term (6 to 12 months) outcomes for measures of patient pain or function [48]. A subsequent prospective observational study in patients ≥65 years with back pain but without radiculopathy found no differences in disability at one year for patients who received early imaging (within six weeks of the index visit) compared with those who did not [49].

Additionally, imaging exams often show abnormal findings in adults without low back pain, which makes it difficult to determine which imaging findings are clinically significant. As an example, disc herniations on MRI are seen in 22 to 67 percent of asymptomatic adults and spinal stenosis in 21 percent of asymptomatic adults over age 60 [50-52]. Osteoarthritis is also often seen on imaging but correlates poorly with symptoms [53]. In a study of 188 individuals 40 to 80 years old, 60 percent of males and 67 percent of females had facet joint osteoarthritic changes on lumbar CT scans; the prevalence of radiologic facet joint osteoarthritis increased with age, but there was no correlation with low back pain [54]. In addition, in a systematic review of 33 studies of asymptomatic adults, the prevalence of degenerative lumbar spine changes (identified on CT or MRI) increased with age [55]. Facet joint degeneration, for example, was observed among 9 percent of adults in their 30s and 83 percent of those in their 80s. Disc degeneration was observed in 52 percent of adults in their 30s and 96 percent of those in their 80s. Most findings had monotonic increases with each decade of age. (See “Acute lumbosacral radiculopathy: Pathophysiology, clinical features, and diagnosis”, section on ‘Neuroimaging’ and “Lumbar spinal stenosis: Pathophysiology, clinical features, and diagnosis”, section on ‘Neuroimaging’.)

Even when the radiographic findings are consistent with clinical presentation, the magnitude of radiographic findings does not necessarily correlate with clinical severity and outcome, and clinical improvement may not correlate with resolution of the radiographic defect [56,57]. As an example, in one follow-up of a trial of 283 patients with lumbar disk herniation and sciatica who had undergone surgery, MRI at one-year follow-up showed disk herniation in 35 and 33 percent of patients with favorable and unfavorable outcomes, respectively [57].

Some findings on MRI are clinically insignificant or of uncertain significance. These include:

Annular fissures (tears) –Annular fissures, colloquially termed tears, are separations between the annular fibers of the intervertebral disc or separations of annular fibers from their attachments to the vertebral bone. Several small studies found no correlation between the presence of annular fissure and back pain [58-60]. As an example, a prospective study of asymptomatic patients found that 38 percent had evidence of annular fissures at baseline [61]. Follow-up after three years showed that annular fissures were not associated with new back pain [58].

Schmorl’s nodes –Schmorl’s nodes, representing herniation of the nucleus pulposus into the adjacent end plate, can be seen in approximately 20 percent of MRI studies in patients without back pain (image 1) [62]. Although Schmorl’s nodes are associated with degenerative changes in the lower back, they are not an independent risk factor for back pain [63].

Modic changes –Modic changes (also known as degenerative endplate changes) are of unclear clinical significance. They refer to specific signal changes in the vertebral endplate and adjacent bone marrow on a spine MRI [64]. These changes occur in 6 to 10 percent of asymptomatic adults and are common in patients with back pain, with any type of Modic change typically reported in about 20 to 40 percent of patients [65,66]. The prevalence of Modic changes increases with age and appears to be associated with degenerative disc changes. A systematic review found only a small number of treatment studies involving patients with Modic changes and concluded that it is unclear whether the presence of these changes is helpful in guiding the selection of treatment options [67]. Additionally, the type of Modic change in a single patient may progress or regress over time [68].

Modalities — The main imaging modalities to evaluate back pain are spine MRI, CT, and plain radiographs. Initial imaging is not indicated in the majority of patients with low back pain. (See ‘Indications for imaging’ below.)

Advanced imaging – For most patients with low back pain who require advanced imaging, lumbar spine MRI without contrast is generally considered the best initial examination [69,70]. It provides axial as well as sagittal views and demonstrates discs, ligaments, nerve roots, and epidural fat, as well as the shape and size of the spinal canal. MRI is more sensitive and specific than plain radiographs for the detection of spinal infection and malignancy [19]. In patients where there is a suspicion for cancer, infection, or immunosuppression, MRI is performed without and with intravenous contrast to evaluate for underlying infection or mass. Enhancement with gadolinium also allows the distinction of scar from disc in patients with prior back surgery.

In patients who require advanced imaging but cannot undergo MRI, we generally proceed with lumbar spine CT without contrast [70]. Neither Tc-99m bone scan nor myelography are considered appropriate for the initial evaluation of patients with back pain. (See ‘Infection’ below and “Patient evaluation for metallic or electrical implants, devices, or foreign bodies before magnetic resonance imaging”, section on ‘Assessing implants, devices, or foreign bodies for MRI’.)

Plain radiographs – When plain radiographs are indicated, anteroposterior and lateral views of the lumbar spine are usually adequate. Oblique and spot views substantially increases the radiation dose and adds little new diagnostic information [71]. Flexion-extension views may be helpful in patients for whom instability is a concern.

Plain radiographs are a reasonable option for imaging in patients who have risk factors for malignancy [19], and they are often combined with the erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP) for evaluation (see ‘Cancer’ below and ‘Cancer risk’ below). Plain radiographs may also be an option for in patients with osteoporosis where the primary concern is detection of a compression fracture [72].

Indications for imaging — The majority of patients with low back pain of less than four weeks duration do not require imaging [17]. Most patients who present to primary care settings will have nonspecific pain without associated symptoms and will improve rapidly. (See ‘Etiologies’ above and “Treatment of acute low back pain”, section on ‘Prognosis’.)

Approximately one-quarter of patients 18 to 50 years of age with acute low back pain who underwent imaging exams had no identifiable indication for imaging [73]. Inappropriate lumbar imaging can lead to irrelevant findings and trigger additional costly studies, unneeded treatments, and unwarranted surgical interventions [74-76].

Joint guidelines from the American College of Physicians (ACP) and the American Pain Society explicitly recommend that “clinicians should not routinely obtain imaging or other diagnostic tests in patients with nonspecific low back pain” and reserve imaging for patients with severe or progressive neurologic deficits or when serious underlying conditions are suspected on the basis of history and physical examination [17,18]. Guidelines from the National Institute for Health and Care Excellence (NICE) in the United Kingdom advise clinicians to “not routinely offer imaging in a non-specialist setting for people with low back pain with or without sciatica” [77]. The ACP provides practical advice about when imaging studies should be considered in patients with acute low back pain (table 8), and our recommendations below are consistent, with the exception of imaging for suspected vertebral compression fracture. Avoiding imaging in acute low back pain has been identified as a recommendation in the American Board of Internal Medicine’s “Choosing Wisely” campaign.

Red flags — Some guidelines suggest “red flag” symptoms, which may identify patients at risk for a more dangerous cause of back pain and represent an indication for earlier imaging exams [5,17,18,78]. There are limited data to support the use of most of the red flags as an indication for early imaging [79]. Systematic reviews of studies that used one or more of these indications for imaging found that only a history of cancer has been shown to increase the probability of finding spinal malignancy [80,81]. Systematic reviews have found that the red flags associated with the highest post-test probability of a vertebral fracture were older age, prolonged use of corticosteroids, severe trauma, and presence of contusion or abrasion [80,82].

Risk assessment for acute back pain — Among patients seen in primary care, less than one percent will have a serious systemic etiology that requires evaluation with immediate advanced imaging (algorithm 1). (See ‘Serious systemic etiologies’ above and ‘Modalities’ above and ‘Less serious, specific etiologies’ above.)

Neurologic deficits — Indications for imaging in the presence of neurologic symptoms depends upon the nature of the symptoms and the patient’s risk factors for cancer and/or an infectious etiology of back pain (algorithm 1).

●Any patient with symptoms of spinal cord or cauda equina compression or progressive and/or severe neurologic deficits should have immediate MRI for further evaluation and urgent specialist referral. Such symptoms and signs include new urinary retention, urinary incontinence from bladder overflow, new fecal incontinence, saddle anesthesia, and significant motor deficits not localized to a single nerve root. (See “Clinical features and diagnosis of neoplastic epidural spinal cord compression”, section on ‘Magnetic resonance imaging of the spine’ and “Spinal epidural abscess”, section on ‘Diagnosis’.)

●Patients with radiculopathy attributable to a single nerve root level or with stable symptoms due to spinal stenosis do not need immediate imaging unless there is a risk of metastatic cancer or moderate to high risk of infection.

•Such patients (who have a risk of metastatic cancer or in whom there is moderate to high risk of infection) should undergo immediate MRI. (See ‘Serious systemic etiologies’ above.)

•For all patients with radiculopathy attributable to a single nerve root level, we image with MRI if there is no improvement in symptoms after four to six weeks of conservative therapy. There is no indication to re-image patients with spinal stenosis symptoms if previous MRI was performed and symptoms are stable. (See “Acute lumbosacral radiculopathy: Pathophysiology, clinical features, and diagnosis”, section on ‘Evaluation and diagnosis’ and “Lumbar spinal stenosis: Pathophysiology, clinical features, and diagnosis”, section on ‘Diagnosis’ and ‘Radiculopathy or lumbar spinal stenosis’ below.)

Infection — For patients in whom there is a suspicion for spinal infection (including vertebral osteomyelitis or spinal epidural abscess), the evaluation should be guided by the degree of suspicion (algorithm 1).

Moderate to high clinical suspicion for infection –For patients with low back pain in whom there is moderate to high clinical suspicion for spinal infection, immediate MRI without and with contrast is indicated (algorithm 1) [72].

Sign and symptoms of infection may include objective fever, tenderness to palpation (vertebral osteomyelitis), and neurologic symptoms (spinal epidural abscess).

Risk factors for infection include:

•Current immunosuppression

•Current hemodialysis

•Current or recent injection drug use

•Current or recent invasive epidural/spinal procedure

•Current or recent endocarditis or bacteremia

MRI is the most sensitive imaging modality for detecting spinal infection with sensitivity of 0.96 and specificity of 0.92 [19]. For patients who are unable to obtain an MRI, a CT scan is a useful alternative to evaluate for epidural abscess, while radionuclide scans are an option to evaluate for osteomyelitis. The evaluation and diagnosis of these conditions are discussed in detail separately. (See “Spinal epidural abscess”, section on ‘Diagnosis’ and “Spinal epidural abscess” and “Vertebral osteomyelitis and discitis in adults”, section on ‘Suggested clinical approach’ and “Vertebral osteomyelitis and discitis in adults”, section on ‘Radiographic imaging’.)

Lower concern for infection –When there is a lower level of concern for the possibility of an infectious cause of back pain, it is reasonable to first evaluate patients with ESR and/or CRP. Patients with an elevated ESR and/or CRP should be evaluated with MRI (algorithm 1). (See “Spinal epidural abscess”, section on ‘Diagnosis’ and “Vertebral osteomyelitis and discitis in adults”, section on ‘Suggested clinical approach’.)

In patients with osteomyelitis or other spinal infection, the sensitivity of an elevated ESR is 0.76 to 0.95 percent and the sensitivity of an elevated CRP is 0.82 to 0.98 percent [3-5]. Infection is very unlikely in patients with an ESR <20 and no more than one risk factor for a systemic illness [22,45]. (See “Vertebral osteomyelitis and discitis in adults”, section on ‘Laboratory findings’ and “Spinal epidural abscess”, section on ‘Diagnosis’.)

Cancer — Patients with cancer or risk factors for cancer and neurologic deficits should have immediate imaging as noted above. In patients without neurologic deficits, the decision to image is based on risk. (See ‘Neurologic deficits’ above.)

Current or recent cancer –Imaging for these patients will vary depending on a variety of factors (eg, what the primary cancer is, when the patient’s most recent imaging studies were, whether or not the patient has known bone metastasis). For such patients, we consult with their oncologist to discuss the most appropriate imaging examination (algorithm 1). The diagnosis of bone metastasis is discussed separately. (See “Epidemiology, clinical presentation, and diagnosis of bone metastasis in adults”, section on ‘Overview of the diagnostic approach’.)

Moderate to high risk for cancer –In patients at moderate to high risk for cancer, we start the evaluation with plain radiographs and ESR (or CRP) (algorithm 1) [18]. Patients with a positive radiograph should have appropriate further evaluation for malignancy (eg, evaluation for primary site, other metastatic disease). Patients with a positive ESR (or CRP) but negative plain radiograph should be further evaluated with MRI. (See “Overview of the classification and management of cancers of unknown primary site”.)

What constitutes a moderate to high risk for cancer (aside from a history of cancer) is not well-defined and the approach likely differs among clinicians. Factors to consider include the patient’s age, smoking history, family history, physical examination findings, and any associated concerning symptoms (eg, recent weight loss). Vertebral metastases are often associated with localized pain and focal tenderness on examination. It is likely that patients having more than one such finding or feature are at higher risk than those with a single risk factor.

The sensitivity of plain radiographs for malignancy is 0.60 and specificity 0.95 [19]. The yield of plain radiographs is increased when combined with the ESR [22]. Cancer is very unlikely in a patients with an ESR <20 and no more than one risk factor (table 9) [22,45].

Low risk for cancer – We do not obtain immediate imaging in patients with acute back pain who are at low risk for cancer (eg, one risk factor). If the pain is persistent, we image after four to six weeks. (See ‘Cancer risk’ below.)

Compression fracture — Patients with suspected vertebral compression fracture should have plain radiographs for evaluation (algorithm 1). Features in the history that indicate an increased risk for vertebral fractures include prolonged glucocorticoid use, advanced age, significant trauma or presence of contusion or abrasion, or recent mild trauma in a patient with a known diagnosis of or risk factors for osteoporosis (table 3) [71,80,82,83]. Patients can have osteoporotic vertebral compression fractures in the absence of trauma.

Symptomatic patients with osteoporotic fractures typically describe a sudden onset of pain and have localized pain and tenderness on examination. While the American College of Physicians (ACP) guidelines suggest that imaging should be deferred in patients with suspected vertebral compression fracture, when clinical suspicion is high it is reasonable to obtain plain radiographs on initial evaluation in order to institute appropriate symptomatic and preventive therapies. The diagnosis and management of osteoporotic vertebral compression fractures is discussed in detail separately. (See “Osteoporotic thoracolumbar vertebral compression fractures: Clinical manifestations and treatment”, section on ‘Diagnosis’.)

Minor trauma — Indications for imaging in patients with trauma are discussed in detail elsewhere. (See “Evaluation of thoracic and lumbar spinal column injury”, section on ‘Decision rules for imaging thoracic or lumbar spine injury’.)

Risk assessment subacute back pain — Patients who have not improved after four to six weeks of conservative therapy and who did not receive imaging on initial evaluation are reevaluated. Patients who have developed neurologic deficits or symptoms of infection in the interim should have imaging as noted above. (See ‘Neurologic deficits’ above and ‘Infection’ above.)

Patients who present initially with low back pain of more than four to six weeks duration should undergo the initial risk assessment as presented above. (See ‘Indications for imaging’ above.)

In patients who had indications for immediate imaging and had negative findings, we do not repeat imaging in patients if symptoms are unchanged [18]. Repeat imaging is indicated in patients with new or worsening symptoms or new concerns that develop in the interim. The modality will depend upon the suspected diagnosis and the modality of the initial imaging exam.

Radiculopathy or lumbar spinal stenosis — Patients with persistent symptoms due to a lumbosacral radiculopathy or spinal stenosis who have not responded to conservative treatment and who are candidates for and interested in invasive therapies (eg, surgery or epidural injection for radiculopathy) should have an MRI for further evaluation and be referred for consideration for these therapies [18]. (See “Lumbar spinal stenosis: Pathophysiology, clinical features, and diagnosis”, section on ‘Neuroimaging’ and “Acute lumbosacral radiculopathy: Treatment and prognosis”, section on ‘Surgery’ and “Lumbar spinal stenosis: Treatment and prognosis”, section on ‘Surgical treatment’ and “Acute lumbosacral radiculopathy: Treatment and prognosis”.)

Cancer risk — In patients with low back pain who did not meet criteria for immediate imaging but who have risk factors for cancer and do not improve with conservative therapy after four to six weeks, we evaluate with plain radiographs and ESR (or CRP) [18]. Patients with a positive radiograph should have appropriate further evaluation for malignancy (eg, evaluation for primary site, other metastatic disease). Patients with a positive ESR (or CRP) but negative plain radiograph should be further evaluated with MRI. (See ‘Cancer’ above.)

Other patients

Concern for ankylosing spondylitis –Patients with persistent back pain despite four to six weeks of conservative therapy who also have signs or symptoms concerning for ankylosing spondylitis should have a plain radiograph to evaluate the sacroiliac joints [18]. These can often be well-evaluated on a lumbosacral plain radiograph. (See “Diagnosis and differential diagnosis of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults”, section on ‘Diagnostic approach’.)

Concern for osteoarthritis –Lumbosacral, pelvic, and/or hip radiographs may be considered for older patients with persistent back pain in whom there is concern for osteoarthritis of the hip referred to the back or for whom the mechanical adaptions to hip osteoarthritis are causing back pain. If a patient’s low back symptoms do not lateralize, a standing pelvis radiograph will evaluate both hips for the presence of osteoarthritis; if symptoms lateralize, a two-view hip (standing frontal and frog-leg views) is appropriate. The diagnosis and management of osteoarthritis is discussed in detail separately. (See “Clinical manifestations and diagnosis of osteoarthritis”, section on ‘Imaging’ and “Overview of the management of osteoarthritis”.)

Patients without specific concerns –In other patients where there are no concerns for a particular etiology, we generally treat with conservative therapy for another eight weeks.

Risk assessment chronic back pain — Patients who present initially with low back pain >12 weeks duration should undergo the risk assessment for acute and subacute back pain. (See ‘Risk assessment for acute back pain’ above and ‘Risk assessment subacute back pain’ above.)

In patients without specific concerns who have not received any imaging for indications noted above, if there is no improvement after 12 weeks, we generally image with a plain radiograph and consider the need for referrals for further evaluation and treatment.

Counseling patients who request imaging — Patients often expect that imaging will be ordered during their initial visit for back pain. Although it is not possible to provide a definitive physiologic diagnosis for low back pain in the majority of patients, clinicians can reassure patients without concerning history or physical examination findings that they appear to have “mechanical” or nonspecific low back pain, and that it is very unlikely that they have a serious underlying problem. Patients should be assured that improvement is to be expected and should be advised that:

●They are unlikely to have a serious underlying condition. (See ‘Etiologies’ above.)

●Incidental imaging findings, unrelated to their pain, are common, and may lead to unnecessary further tests or interventions. (See ‘Limited utility of imaging’ above.)

●Imaging is appropriate if they do not improve as expected. (See ‘Risk assessment subacute back pain’ above and ‘Risk assessment chronic back pain’ above.)

In addition, a careful physical examination with ongoing commentary (“I am checking strength for any sign of nerve injury”) may be reassuring for patients and help make it clear that the clinician is not inappropriately bypassing a further diagnostic evaluation or dismissing the symptoms.

Patients who perceive that they have received an adequate explanation for the cause of their problem are less likely to want additional diagnostic tests and more likely to be satisfied with the visit than those who do not think they have been given an adequate explanation [84]. In one randomized trial, low-risk patients who received an educational intervention rather than a plain radiograph were equally satisfied with their care and had equally good clinical outcomes [85].

SOCIETY GUIDELINE LINKSLinks to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See “Society guideline links: Lower spine disorders”.)

INFORMATION FOR PATIENTSUpToDate offers two types of patient education materials, “The Basics” and “Beyond the Basics.” The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on “patient info” and the keyword(s) of interest.)

●Basics topics (see “Patient education: Low back pain in adults (The Basics)” and “Patient education: Spinal stenosis (The Basics)” and “Patient education: Herniated disc (The Basics)” and “Patient education: Muscle strain (The Basics)” and “Patient education: Do I need an X-ray (or other test) for low back pain? (The Basics)”)

●Beyond the Basics topics (see “Patient education: Low back pain in adults (Beyond the Basics)”)

SUMMARY AND RECOMMENDATIONS

●Most patients who present with back pain to primary care settings will have nonspecific back pain. Such patients will typically improve over a few to several weeks with conservative or self-care. (See ‘Etiologies’ above.)

Less than 1 percent will have serious systemic etiologies (eg, malignancy or infection). Less than 10 percent will have less serious, specific etiologies (eg, vertebral compression fracture, radiculopathy, or spinal stenosis). (See ‘Etiologies’ above.)

●A focused history and physical examination are sufficient to evaluate most patients with back pain of less than four weeks duration. The history and physical examination should identify features that suggest that imaging and/or other evaluations are indicated. (See ‘History’ above and ‘Physical examination’ above.)

●The majority of patients with low back pain of less than four weeks duration do not require imaging. Among patients seen in primary care, less than 1 percent will require immediate advanced imaging (eg, magnetic resonance imaging [MRI] or computed tomography [CT]) (algorithm 1). (See ‘Indications for imaging’ above.)

•Any patient with symptoms of spinal cord or cauda equina compression or progressive and/or severe neurologic deficits should have immediate lumbar spine MRI for further evaluation and urgent specialist referral. Such symptoms and signs include new urinary retention, incontinence from bladder overflow, new fecal incontinence, saddle anesthesia, and significant motor deficits not localized to a single nerve root. (See ‘Neurologic deficits’ above.)

•Other patients who may require imaging on initial evaluation include those with a moderate to high suspicion for spinal infection, risk factors for metastatic cancer, and suspected vertebral compression fracture. (See ‘Risk assessment for acute back pain’ above.)

●Patients who have not improved after four to six weeks of conservative therapy and who did not receive imaging on initial evaluation are reevaluated:

•Patients with persistent symptoms due to a lumbosacral radiculopathy or spinal stenosis who are candidates for and are interested in invasive therapies (eg, surgery or epidural injection for radiculopathy) should have a lumbar spine MRI for further evaluation. (See ‘Radiculopathy or lumbar spinal stenosis’ above.)

•In patients with low back pain who have risk factors for cancer, we evaluate with erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP) and plain radiographs. (See ‘Cancer risk’ above.)

•Other patients that may need imaging include those with concerns for ankylosing spondylitis and osteoarthritis. (See ‘Other patients’ above.)

•In other patients where there are no concerns for a particular etiology, we generally treat with conservative therapy for another eight weeks.

●For patients without concerns for a particular etiology who have not improved after 12 weeks total, we generally image with a plain radiograph and consider referrals for further evaluation and treatment. (See ‘Risk assessment chronic back pain’ above.)