how emergency physicians approach low back pain: choosing costly options

Pergamon The Journal of Emergency Medicine, Vol 13, No 2, pp 143-150, 1995

Copyright 0 1995 Elsevier Science Ltd Printed in the USA. AlJ rights reserved

07364679/95 $9.50 + .oo

0736-4679( 94 )00134-O

Original Contributions

HOW EMERGENCY PHYSICIANS APPROACH LOW BACK PAIN: CHOOSING COSTLY OPTIONS

Kenneth C. Elam, MD, rwH,*t Daniel C. Cherkin, PhD,*$ and Richard A. Deyo, MD, MPH*§

*Department of Health Services, University of Washington, Seattle, Washington tEmergency Department of St. Joseph Hospital, Tacoma, Washington

*Center for Health Studies, Group Health Cooperative of Puget Sound, Seattle, Washington §Health Services Research and Development Field Program, Seattle Veterans Affairs Medical Center, Seattle, Washington

Reprint Address: Kenneth Elam, MD, MPH, Emergency Department, St. Joseph Hospital, 1718 South I Street, Tacoma, WA 98401

0 Abstract-To determine ways in which emergency physicians approach the diagnosis and treatment of the common presenting complaint of low back pain, responses of emergency physicians to a questionnaire dealing with three hypothetical patients with different types of low back pain were taken from a stratified national random sample of eight medical specialties. For severe acute (with and without sciatica) or chronk! low back pain, physicians were asked which tests and consultants they would use in pursuit of the diagnosis, and which treatments and specialty referrals they would recommend in each of the three scenarios. For diagnosis in the acute cases (pain less than 1 week), up to 22% of emergency physicians recommended computed tomography (CT scan) and 36% recommended magnetic resonance imaging (MRI). Specialist consultation would be sought for 61% of the acute sciatica patients, 32% of the acute nonsciatica patients, and 47Vo of the chronic patients. In approaching treatment, over 75% of emergency physicians would advise bedrest for an average of 3.5 to 4.5 days. Between 16% and 40% suggested physical therapy for the acute patients. Referrals to surgical speciaiists (orthopedist or neurosurgeon) were highest (81%) for acute sciatica, compared with 52% for chonic low back pain, and 41% for acute nonsciatic low back pain. In conclusion, given that most cases of acute low back pain resolve with minimal intervention, diagnostic imaging, Iabo- ratory testing, and early specialist consultation favored by many emergency physicians would add iittle except expense to understanding its etiology. For treatment, emergency physician recommendations for bedrest were longer than

necessary and, for physical therapy, of no proven benefit. The tendency to refer acute cases to orthopedists and neurosurgeons could set the stage for unnecessary operative intervention. Simpler, less costly, and more reassuring approaches to emergency department patients with low back pain should be considered.

Cl Keywords-low back pain; emergency physicians

INTRODUCTION

As the second leading symptom resulting in a physician visit, low back pain contributes significantly to the cost of American health care as well as to lost worker productivity (1,2). Rare is the shift during which emergency physicians do not encounter this condition, one that may be met with a lack of enthu- siasm. Although geographical variation in the treatment of low back pain has been established (3), little is known about how this complaint is dealt with in the emergency department. A national survey of physicians in specialties likely to see patients with low back pain was conducted to better understand which diagnostic tests are considered useful and what treatments would likely be ordered (4). The data pertain- ing specifically to emergency physicians provided a useful look at how the specialty behaves when con- fronted with patients complaining of low back pain.

Original Contributions is coordinated by John A. Mcrrx, MD, of Carolinas Medical Center, Charlotte, North Carolina

RECEIVED: 27 October 1993; FINAL SUBMISSION RECEIVED: 4 April 1994; ACCEPTED: 22 April 1994

143

144 K. C. Elam et al.

METHODS

Questionnaires were designed to collect information about how physicians use tests, treatments, and hospitalization for patients with low back pain. Physi- cians were asked to indicate how they would evaluate and treat patients described in three case scenarios. Those parts of the survey applicable to emergency physicians (with questions about hospitalization omitted) are shown in Appendices A through C.

The first of the three vignettes described a case of acute back pain without sciatica. The patient was a 2%year-old woman unable to work for 5 days due to a week of back pain; she was neurologically normal with only local tenderness on examination, and had normal lumbar spine x-rays. A case of acute back pain with sciatica was described in a 35year-old male with 4 days of pain radiating to the posterior calf and lateral foot. His examination showed sensory deficits in this distribution, decreased ankle reflex, no motor weakness, and straight leg raise limited to 45 degrees in the affected leg. Plain x-rays were negative. A patient with chronic low back pain was described as a 50-year-old with 3 years of intermittent excruciating pain poorly responsive to medical and chiropractic treatments. Her severe pain did not radiate, examination showed only limited lumbar flexion, and x-rays showed osteophytes but no disc-space narrowing. Tests from which to choose were: contrast myelogram or computed tomography (CT scan)/myelogram, CT scan without contrast, magnetic resonance imaging (MRI), diskogram, electromyography/nerve conduction, bone scan, laboratory tests (for example, urinalysis, sedimentation rate), psychological evaluation (for example, MMPI), and specialist consultation (the details of when and where the consultation were to take place was not specified by the survey). Treatment options included: bedrest, traction, spinal manipulation, trigger point or epidural steroid injections, use of various medications (narcotic analgesics, NSAIDS, muscle relaxants), referrals to orthopedist, neurosurgeon, neurologist, or rehabilitation specialist, and physical therapy. For both survey questions, there was an option of “other” that al- lowed respondents to specify test or treatment choices not listed. From the survey, it was not possible to link respondents who ordered one test to the ordering of other tests or the choice of a particular treatment.

Physician Sampling

From a random sampling of American Medical Asso- ciation and American Osteopathic Associaton data-

bases, approximately 300 physicians were selected from the specialties of family or general practice, general internal medicine, general practice osteopa- thy, emergency medicine, rheumatology, physical medicine and rehabilitation, and neurology. Samples of about 350 neurosurgeons and 400 orthopedists were also obtained.

Survey Distribution

Up to three mailings of the 7-page questionnaire were carried out, beginning in early 1991. Of the 2,456 physicians surveyed who treated back pain, there was a 43% response rate. A similar percentage of emergency physicians responded- 114 of 283 surveyed or 40%. Surveys were mailed with a cover letter of en- dorsement from the relevant professional societies, including the American College of Emergency Physi- cians .

Demographics

As outlined in Table 1, the 114 emergency physician respondents were more likely to practice in the west and south, they averaged 40 years of age, and 74% were Board Certified in any specialty. The majority, 69%, worked in cities (or their suburbs) with a population of over 100,000. Fifty-nine percent typically saw 5-20 patients per month for low back syndromes or sciatica, and 35% had more than 21 such visits per month.

Table 1. Demographic Information For Emergency Phydcians

Mean age: 40.4 years

Female: 14%

Region of Practice: Northeast 14% South 30% Midwest 23% West 33%

Board certified (any specialty): 74% Number of visits in a typical month for low back pain syndromes or sciatica:

None 0 1 to4 5.6% 5to20 59.3% 21+ 35.2%

The population of the community in which physicians practiced: Less than 25,000 5.6% 25,000-100,000 25.9% City over 100,000 or its suburbs 66.5%

Emergency Physicians and Low Back Pain 145

RESULTS

Approach to Diagnosis

Imaging studies (CT scan, MRI, myelogram) were ordered most commonly for the patient scenario sug- gestive of acute sciatica (Figure 1). Over one-third of emergency physicians requested MRI, and about one-fourth would order a CT scan for this type of patient, described as having only 4 days of pain. One- third of physicians also desired laboratory tests such as sedimentation rate and urinalysis for patients with acute (less than 1 week) nonsciatic low back pain. For the chronic scenario (3 years of intermittent back pain), 47% of emergency physicians wanted these tests. The desire to seek specialist consultation as part of the evaluation for acute conditions was consider- able. At 60%, consultation for acute low back pain with sciatica was twice as likely as for acute low back pain alone. For chronic low back pain, nearly one- half of emergency physicians indicated they would seek consultation.

Approach to Treatment

Figures 2 and 3 summarizes the recommendations for bedrest, physical therapy, and specialist referral. For acute cases, bedrest was advised by at least three- quarters of physicians; for the chronic case example,

TESTS & CONSULTS:

Contrast CTMyelogrm

CT without Contrast

MRI

Lab Tests

Specialist Consults.

over half suggested it. Duration of recommended bedrest ranged from 1 to 14 days, with a mean of 4.5 days for acute sciatica and 3.5 days for the other two conditions. Although specific modalities of physical therapy in the questionnaire were not specified, it was suggested by about 40% of respondents for acute and chronic nonsciatic low back pain and by 16% for acute sciatica.

Medication (acetaminophen, anti-inflammatory, narcotic analgesic, or muscle relaxant) was suggested for all types of low back pain by a large majority of respondents for the acute scenarios (Figure 2). In the case of chronic low back pain, muscle relaxants and narcotic analgesics were less likely to be recommended. With regard to specialist referral for treatment, surgical referrals predominated over nonsurgi- cal for all clinical vignettes (Figure 3). Most striking was that 80% of emergency physicians would refer patients to an orthopedist or neurosurgeon, while only 2 or 3% would refer them to rehabilitation and neurology specialists in the instance of acute sciatica. Surgical referrals were preferred almost four to one for acute nonsciatic low back pain and almost two to one for chronic low back pain.

DISCUSSION

The inclination by up to one-third of emergency physicians to order resource-consuming tests, and for

0% 10% 20% 30% 40% 50% 60% 70% Percent of Respondents

Acute with Scietice m Acute no Scietica

m Chronic LBP

Figure 1. Tests and consultations emergency physicians would order for low back pain.

146 K. C, Elam et al.

TREATMENTS:

Acetaminophen NSAIDS

Narcotic Analgesics

Muscle Relaxants

Systemic Steroids

Redrest

Physical Therapy

100%

80%

0% 20% 40% 60% 80% 100% Percent of Respondents

Acute with Sciatica

m Chronic l5P

m Acute no Sciatica

Figure 2. Treatments emergency physlcians would recommend for low beck pain.

Percent of Respondents

81%

orthopedlstl Nauosurgeen

Neuroiogkt

Acute with Sciatka M Acute no Stdetica

Figure 3. Emergency physiclen referrals for low beck peln.


even higher percentages to seek specialty consultation for acute back pain scenarios may be excessive. The Quebec Task Force on Spinal Disorders, which pub- lished guidelines in 1987 based on a rigorous evaluation of the scientific literature, recommended tests only after 7 weeks of conservative care, except for patients with neurological deficits (5). For patients with neurological deficits as in the first scenario, the Task Force noted that obtaining plain x-rays and some kind of screen for an inflammatory process was a common practice, but one for which data were lacking. Both acute sciatic and nonsciatic low back pain are likely to resolve in a matter of weeks with minimal intervention (6).

To focus attention on newer technologies, plain radiographs had been performed on the patients described in this survey, although such films are not recommended in the radiology literature for acute, nontraumatic low back pain (7). Advanced imaging studies may actually confuse the issue for acute patients both with and without sciatica because about 20% of normal individuals without leg or back pain will show disc herniation on myelogram, CT scan (8), and MRI (9). In the absence of relevant history such as previous cancer or intravenous drug abuse or symptoms such as dysuria, laboratory tests for acute and chronic back pain are rarely helpful (5). When there is no progressive neurological syndrome accom- panying low back pain, specialist consultation, advised by 30 to 60% of emergency physicians for acute cases, seems likely to generate further testing such as nerve-conduction studies or additional imaging. Indeed, a comparison among other specialists in this survey by Cherkin and colleagues revealed a tendency for different specialties to pursue unique diagnostic pathways: neurosurgeons and neurologists favored imaging, physiatrists and neurologists requested elec- tromyograms, and rheumatologists sought laboratory tests (4). In approaching treatment, greater than three-quarters of emergency physicians advocate bedrest for acute low back pain. Although most cases of acute low back pain are not due to herniated lumber discs (5,6), bedrest has a theoretical benefit of decreasing intradiscal pressure. However, when patients roll to the side, an inevitability, intradiscal pressure rises to 75% of that observed in the standing position (10). Only one of five controlled studies of bedrest for low back pain and sciatica showed a benefit ( 11). That study, in military recruits, suffered from substantial observer bias and strictly enforced rest or activity that are not replicated in civilian life. Bedrest, Waddell states, is better regarded as a conse- quence of pain, not as a treatment. Trauma and ma- jor surgery patients are mobilized as soon as possible;

no reason exists to treat back pain patients differ- ently. Increased activity improves disc and cartilage nutrition and promotes bone and muscle strength. It also raises endorphin levels, which theoretically may reduce pain ( 11). The 3- to 4-day average recommen- dation of bedrest for acute and chronic cases may be longer than is necessary. Deyo and associates compared 2 days of bedrest to 1 week of bedrest in patients with back pain and no neurological deficits and found no differences in resolution of pain or functional recovery (12). Physical therapy, favored by about 40% of emergency physicians for acute sciatic and chronic low back pain, might be beneficial in two ways. It provides the therapist with a captive audience for purposes of education, and it adds the psychological dimension of the laying on of hands to the actual treatment. Furthermore, it is noninvasive. Yet data proving its effectiveness for acute or chronic cases are still unavailable (13). Gilbert et al. conducted a multicentered, randomized clinical trial in patients with acute low back pain comparing bedrest only (at least 4 days) with physiotherapy plus education and bedrest, physiotherapy and education only, and none of the above. No beneficial effect of any treatment was observed on several clinical outcome measures including straight leg raising, lumber flexion, activities of daily living, and pain (14). Emer- gency physicians frequently favored the use of anti- inflammatory agents, and the place of such drugs in treating low back pain seems well established (15). The role of muscle relaxants is less secure, although one recent multicenter trial suggested a modest benefit for combined analgesic and antispasmodic therapy (16). In keeping with standard recommendations (6), respondents avoided suggesting narcotics for the chronic scenario. Pain specialists, however, are beginning to recognize the appropriateness of long- term opioid analgesic use in a carefully selected group of patients with chronic pain of nonmalignant origin ( 17). The proportion (40-80% ) of emergency physicians choosing referral for treatment to orthopedists or neurosurgeons for acute low back pain, especially sciatica, appears excessive. Before possible surgical intervention should be considered, 4 to 6 weeks of conservative treatment is recommended (5). For surgery to make sense, the presence of a herniated disc on imaging must correspond to a particular pain syndrome and neurological deficit. For chronic low back pain, lumbar disk surgery provides com- plete relief of pain for one-half of the patients, but this increases to 75% if an actual bulging disk is found at operation ( 18). Surgery done for persistent pain and failure of all other treatments is less likely to be successful and may contribute to disability (6).

148 K. C. Elam et al.

Finally, the higher rate of surgery in the US vs. Eu- rope, and unexplained geographical variations in the rate of lumbar surgery in this country, suggest that surgical indications are far from consistent (6). The moderately low response rate (40%) of emergency physicians may limit the generalizability of the find- ings of this survey; that it uses a nationally represen- tative sample is a strength. Another limitation of this study may be the use of patient vignettes to predict actual physician behavior. A recent review article, however, has concluded that written simulations are probably an effective device for elucidating the deci- sion-making process (19).

CONCLUSION

For a condition as commonly encountered as low back pain, it may be time for emergency physicians (as well as other front line specialists such as inter- nists and family practitioners) to rethink common approaches. Diagnosis and treatment based on habits acquired during training must give way to practicing literature-based medicine. In the case of low back pain, what is simple is often sufficient. For the majority of patients with acute low back pain, sciatic or not, diagnostic testing is unlikely to add anything unless history and physical examination suggest ma- lignancy or specific neurological emergencies such as cauda equina syndrome. Seeking the opinion of a neurologist, physiatrist, or rheumatologist for an acute low back pain syndrome may begin a cascade of tests favored by each respective specialist.

Consultations with, or referrals to, surgical specialists may prematurely lock the patient into a path-

way of expensive imaging and possible operative pro- cedures. Absolute bedrest should rarely be advised, and even then, for a brief time. Patients should be encouraged to pursue normal daily activities as much and as soon as possible. There is no evidence that being up and about, even if in discomfort, is harm- ful, and such activity likely fosters a positive attitude toward recovery (8). Physical therapy may offer psychological, educational, and possibly physiological advantages, but its timing and duration need not be left solely to the discretion of the physical therapist when clear evidence of benefit is lacking. One large medical center in Minneapolis, recently mentioned in the national media, treats all patients with acute low back pain with only rest, ice packs, and aspirin (20). In the emergency department, reassurance plus limited medication may be the most useful interventions. Acute lumbar pain truly does resolve in most situa- tions simply with the passage of time; fewer than 2% of all patients will have low back pain that persists beyond 2 weeks (6). Although the patient should be informed that low back pain may well recur, this likelihood can be an incentive for reading educational materials and maintaining a general level of good physical conditioning. If follow-up is needed, acute low back pain may best fit into the realm of the primary care practitioner.

Acknowledgment-This work was supported in part by Grant Number HS-06344 from the Agency for Health Care Policy and Research (the Back Pain Outcome Assessment Team) and by the Health Services Research and Develop- ment Field Program, Seattle Veterans Affairs Medical Center.

REFERENCES

1. Sternbach RA. Survey of pain in the United States: The Nu- prin pain report. Clin J Pain. 1986;2:49-53.

2. Frymoyer JW, Cats-Baril WL. An overview of the incidences and costs of low back pain. Orthopaed Clin North Am. 1991; 22~263-11.

3. Deyo RA, Cherkin DC, Conrad D, Volinn E. Cost, controversy, crisis: Low back pain in the health of the public. Annu Rev Public Health. 1991;12:141-56.

9. Boden SD, Davis Do, Dina TM, Patronas NJ, Wiesel SW. Abnormal magnetic-resonance scans of the lumbar spine in asymptomatic subjects. J Bone Joint Surg. 1990,72-A&3-8.

10. Deyo RA. Bed rest for acute low back pain. In: Joyson MIV, Swezey RL, Knoplich J, Hubault A, eds. Back pain, painful syndromes and muscle spasm: Current concepts and recent advances. Park Ridge, NJ: Parthenon Publishing Group; 1990:107-16.

4. Cherkin DC, Deyo RA, Wheeler K, Ciol M. Physician variation in diagnostic testing for low back pain: Who you see is what you get. Arthritis Rheum. 1994;37:15-22.

5. Spitzer WO, LeBlanc FE, Dupuis M, et al. Scientific approach to the assessment and management of activity-related spinal diorders: A monograph for clinicians. Report of the Quebec Task Force on Spinal Disorders. Spine. 1987;12(Suppl7):S17.

11. Waddell G. Biopsychosocial analysis of low back pain. Bal- liere’s Clin Rheumatol. 1992;6:523-57.

12. Deyo RA, Diehl AK, Rosenthal M. How many days of bed rest for acute low back pain? A randomized clinical trial. N Engl J Med. 1986;315:1064-70.

13. Koes BW, Bouter LM, Beckerman H, et al. Physiotherapy exercises and back pain: A blinded review. Br Med J. 1991; 302:1572-6.

14. Gilbert JR. DW Tavlor. A Hildebrand. C Evans. Clinical trial 6. Deyo RA, Loesdr JD, Bigos SJ. Herniated lumbar-inetkerte-

bra1 disk. Ann Int Med. 1990:112:598-603. 7. Gehweiler JA, Daffner RH. Low back pain: The controversy

of radiologic evaluation. AJR. 1093;140:109-22. 8. Hitselberger WE, Witten RM. Abnormal myelograms in

symptomatic patients. J Neurosurg. 1096;28:204-6.

of commoh treatmknts’for low back pain in family practice. Br Med J. 1985;291:791-4.

15. Ingham JM, Portenoy RK. Drugs in the treatment of pain: NSAIDS and opioids. Curr Opin Anaesthesiol. 1993:6:838-44.


16. Basmajian JV. Acute back pain and spasm. A controlled multicenter trial of combined analgesic and antispasm agents. Spine. 1989;14:438-9.

17. Schofferman J. Long-term use of opiod analgesics for the treatment of chronic pain of nonmalignant origin. J Pain Sympt Management. 1993;8:279-88.

18. Spangfort EV. The lumbar disc hernation: A computer-aided analysis of 2,504 operations. Acta Orthop Stand Suppl. 1972; 142:5-95.

19. Jones TV, Gerrity MS, Earp J. Written case simulations: Do they predict physicians’ behavior? J Epidemiol. 1990;43: 805-15.

20. Staff Reporter. Strong medicine: Dose of managed competi- tion reshapes health-care system in the twin cities. The Wall Street Journal 1993;Feb 26.

APPENDIX A

National Physician Survey on Low Back Pain

Please read the folowing patient vignettes and answer each of the questions.

Clinical Vignette #1 A 35year-old auto mechanic presents with a 4-day history of severe acute low back pain with radiation to the posterior calf and Iateral foot. He has some sensory deficits in this distribution and a diminished ankle reflex, but no motor weakness. Straight leg raising is limited to 45 degrees in the affected leg. Plain x-ray of the lumbar spine is essentially normal except for postural changes sug- gestive of muscle guarding.

a. What tests would you probably order? (Check all that apply) contrast myelogram or CT/myelogram 0 psychological evaluation (for example, MMPI) computerized tomography without contrast 0 specialist consultation-Specify: magnetic resonance imaging (MRI) Cl lab tests (for example, U/A, sed rate) diskogram Cl other - Specifv: electromyography/nerve conduction (EMG-NCV) bone scan

b. What treatments would you probably order?

psychological evaluation (for example, MMPI) specialist consultation-Specify: lab tests (for example, U/A, sed rate) other: Specify:

What treatments would you probably order? (Check all that apply) bed rest-How many days? ~ spinal traction spinal manipulation trigger point injections IV medication-Specify: aspirin, acetaminophen, or NSAIDs narcotic analgesics (for example, Percodan, Tylenol No. 3) muscle relaxants (for example, Flexeril, Soma)

(Check all that app&) 0 bed rest-How many days? ____ q spinal traction 0 spinal manipulation 0 trigger point injections Cl IV medication-Specify: 0 aspirin, acetaminophen, or NSAIDs 0 narcotic analgesics (for example, Percodan,

Tylenol No. 3) 0 muscle relaxants (for example, Flexeril, Soma) 0 systemic corticosteroids q lumbar corset q exercises 0 referral to orthopedic or neruosurgeon Cl referral to neurologist El physical therapy referral 0 other-Specify:

q systemic corticosteroids 0 lumbar corset q exercises 0 referral to orthopedic or neruosurgeon 0 referral to neurologost 0 physical therapy referral 0 other-Specify:

APPENDIX B

Clinical Vignette #2 A 50-year-old homemaker presents with a 3-year history of intermittent excruciating low back pain. She has seen other medical doctors and chiropractors during this period, but was disap- pointed with the results of their care. She currently has severe low back pain but there is no radiation to the legs, and physical examination reveals a limitation of lumbar spine flexion. Plain lumbar spine films reveal osteophytes at several vertebral levels, but no disc-space narrowing.

a. What tests would you probably order? (Check all that apply)

Cl contrast myelogram or CT/myelogram 0 computerized tomography without contrast 0 magnetic resonance imaging (MRI) 0 diskogram 0 electromyography/nerve conduction

(EMG-NCV) Cl bone scan

150

APPENDIX C

Clinical Vignette #3 A B-year-old woman who runs her own catering service complains of acute severe low back pain for a week. The pain is not radiating, but is so severe she has been unable to work for the past 5 days. She is anxious to return to her usual activities, but feels immobilized by the pain at pres- ent. Physical examination reveals markedly limited anterior flexion, left paraspinous tenderness, and a normal neurologic examination. Lumbar spine films are normal.

a. What tests would you probably order? (Check all that apply)

Cl contrast myelogram or CT/myelogram 0 computerized tomography without contrast 0 magnetic resonance imaging (MRI) U diskogram 0 electromyography/nerve conduction

(EMG-NCV) El bone scan 0 psychological evaluation (for example, MMPI)

R. C Elarn et dl _._____ _._ -.-.--.- . ..-. ..-- --.- -- --

0 specialist consultation - Specifjr: .__.__ -.--._. ._ _.-.. C lab tests (for example, U/A, sed rate) 0 other-Specify: __._---__--.-.-. .__.... .“._. ._.

b. What treatments would you probably order? (Check all that apply)

•1 bed rest-How many days? ~- 0 spinal traction Cl spinal manipulation 0 trigger point injections n IV medication-Specify: c1 aspirin, acetaminophen, or NSAIDs q narcotic analgesics (for example, Percodan,

TyIenol No. 3) 0 muscle relaxants (for example, Flexeril, Soma) 0 systemic corticosteroids 0 lumbar corset El exercises CZ referral to orthopedic or neruosurgeon 0 referral to neurologost 0 physical therapy referral 0 other-Specify:

how emergency physicians approach low back pain: choosing costly options

Documents