Medical Acupuncture
A Journal For Physicians By Physicians

Volume 13 / Number 1
"Aurum Nostrum Non Est Aurum Vulgi"

Medical Acupuncture For Non-Radicular
Low Back Pain: A Preliminary Investigation Of A Treatment Protocol And Outcome Measures
Michael Fredericson, MD
John Giusto, MD

ABSTRACT
Background Randomized controlled trials to date do not clearly indicate whether acupuncture is effective in the treatment of low back pain.
Objectives To find a more sensitive measure for evaluating the effectiveness of medical acupuncture, and to determine a reproducible treatment protocol for non-radicular low back pain.
Design, Setting, and Patients Case series at Stanford University Physical Medicine Clinic, including 5 patients with sub-acute, non-radicular low back pain, all of whom had not improved after at least 6 weeks of standard physical therapy.
Intervention Six treatments of medical acupuncture supplementing continued physical therapy.
Main Outcome Measures Visual analog scale, Oswestry Questionnaire, Short Form-36 Questionnaire Global Health Survey (SF-36), and the Patient-Generated Index (PGI) for low back pain.
Results In a relatively high-functioning patient population, the Oswestry Questionnaire did not reflect clinical improvements, whereas the SF-36 and visual analog scale, in combination with the newer PGI, appeared to be more sensitive to treatment effects.
Conclusions This study suggests a combination of outcome measures more sensitive for determining the usefulness of medical acupuncture for non-radicular low back pain. It also defines a modular treatment protocol for a reproducible approach to the range of conditions contributing to low back pain in an outpatient clinic setting.
KEY WORDS
Acupuncture, Low Back Pain, Outcome Assessment, Visual Analog Scale, Oswestry Questionnaire, Short Form-36 Questionnaire Global Health Survey (SF-36), Patient-Generated Index (PGI)

INTRODUCTION
Non-radicular low back pain does not reliably respond to epidural steroids,1 intradural steroids,2 or facet injections,3 and is rarely considered an indication for surgery. Although there is interest in the use of intradiscal therapy (electrothermal heating of the annulus) for discogenic pain confirmed by discography, the results of this treatment are still preliminary, the treatment is invasive, and long-term consequences are unknown. Investigation is necessary into other treatments for non-radicular low back pain, whether from disc, facet joint, ligamentous, or soft tissue pathology.

Despite the historical use of acupuncture to treat low back pain and the increasing interest in its applicability in the United States, acupuncture for the common clinical entity of non-radicular low back pain is not well studied in Western medical literature. A recent review4 of 11 randomized controlled trials did not clearly indicate acupuncture's effectiveness in the treatment of low back pain. Some of the problems in the studies included limited information on the training of acupuncturists, a lack of detailed description of the points needled or the frequency and duration of acupuncture treatment, and an inability to define which types of low back pain best respond to acupuncture.

The purposes of our study were to gain experience in the use of a modular treatment protocol and to determine which outcome measures might be sensitive to the effects of this treatment in non-radicular low back pain.

METHODS
The study was approved by the Stanford University Investigations Review Board, and all subjects signed informed consent prior to beginning the study.

Inclusion Criteria
Eligibility criteria included low back pain for at least 3 months, but less than 1 year, radiating no further than the buttocks, 10th rib, or
posterior axillary line. Subjects had to be aged 18-65 years. In addi-
tion, they must have had a trial of at least 6 weeks of physical therapy that did not alleviate the pain.

Exclusion Criteria
Subjects were not eligible for the following reasons:
1. Previous treatment with acupuncture for any reason
2. Previous low back surgery
3. Compression fractures of the spine
4. Spondylolisthesis greater than grade I
5. Sacroiliac sclerosis
6. Documented moderate/severe spinal stenosis
7. Moderate/severe hip or knee osteoarthritis
8. Certain medical conditions, e.g., cancer, AIDS, systemic infection, osteomyelitis, diabetes, rheumatologic disease (including fibro-
myalgia), reflex sympathetic dystrophy/complex, regional pain syn-
drome, myelopathy, endometriosis, operable fibroids, or a diagnosed psychiatric disorder
9. Certain medications, e.g., currently taking daily narcotics, prednisone within 6 months, antipsychotics, antidepressants, benzodiazepines, or warfarin
10. Specific radiologic/diagnostic findings (see below)
11. Pending litigation
12. Pregnancy

Work-up Algorithm
A detailed work-up algorithm, not reproduced here, was used to rule out diagnoses under exclusion criteria. The application of the work-up was prompted by certain complaints in the history (e.g., primary hip pain warranted plain x-ray films of the hip to rule out osteoarthritis), or certain findings on the physical examination (e.g., pain with sacroiliac or hip joint compression or a positive Patrick's test result warranted plain x-ray films of the pelvis to rule out disorders of the hip or sacroiliac joints).

Subjects
All patients were referred to the Stanford Physical Medicine and Rehabilitation Department for the evaluation and management of low back pain. Five patients were entered in the treatment protocol, all of whom met the inclusion criteria. Three patients who did not meet the inclusion criteria (2 due to pain longer than 1 year, and 1 due to complaints of pain radiating to the foot and a positive straight leg raise)
were evaluated as a separate treatment group. All patients had not improved after at least 6 weeks of standard physical therapy consisting of spine stabilization exercises and low back instruction in proper posture, body mechanics, and lifting techniques to minimize stress on lower back. The baseline characteristics of the patients are presented in Table 1. Acupuncture treatments were provided without charge and patients continued with weekly physical therapy. All patients were able to complete the protocol and none were denied physical therapy for any reason.

TREATMENT
A modular treatment protocol was used that allowed a reproducible approach to the range of conditions presenting as non-radicular low back pain in the outpatient clinical setting. Treatments were given 1 time weekly for 6 weeks. Point location, the size and type of needle, method of skin preparation, needle insertion, and depth and direction of insertion are all described by Helms.5 Needles were left in place for 20 minutes and were removed from cephalad to caudad. Patients were allowed to continue over-the-counter medications during the study, but no other medications were used.

Week 1
o General Tonification: KI 3, SP 6, LU 7, LI 4, LI 11, ST 36,
and CV 4 or CV 6.
o Relaxation: GB 21, LI 18, ST 6, Tai Yang, and GV 20.
Week 2
o Treatment protocol was the same as in week 1.
Week 3
o Local Spine Points. An interspinal point (GV) at the interspace of the most involved level(s) of the problem was treated. This was determined by imaging studies, physical examination, and patient history. Paraspinal points were needled bilaterally at the most involved level, and at 1 level above. These points were located in a muscle cleavage plane approximately 2 cm lateral to the inferior border of the spinous process at the chosen level. The needles were carefully directed in a medial and cephalad direction (both at a 60°-angle to the skin).
o Trigger Points (as described by Travell and Simons6). Up to 8 of the points were chosen from among the following: inner and outer paraspinal muscles (in addition to local spine points), iliolumbar
ligament, quadratus lumborum, gluteus medius, gluteus minimus, tensor fascia latae, iliotibial tract, piriformis, internal/external obliques, and iliopsoas muscles. These sites were chosen if they reproduced
greater than 4/10 pain with mild-to-moderate pressure.
o Distant Points. BL 10 and BL 40 were used to release spasm and stiffness of the paraspinal muscles;7 SI 3 and BL 62 if trigger points were more active along Tai Yang channel; TH 5 and GB 41 if trigger points were more active along Shao Yang channel. Three of the following
Yin points were chosen, depending on constitutional deficiencies:
KI 3, KI 7, KI 10, SP 4, SP 6, LR 3, or LR 8.
o Electrical Stimulation. Electrical stimulation at 2 Hz was added to the paraspinal points. The stimulation intensity was increased until the patient felt a tapping sensation that was not painful. After 20 minutes, the electrical stimulator was turned down, removed, and turned off. All needles were then removed.
Week 4
o The same points were used as in week 3; by then, up to 12 trigger points were chosen.
Week 5
o Treatment was the same as in week 4, but before the electrical stimulator was attached, a specialized needle technique was applied. The points chosen lie in the muscle cleavage plane just inferior and slightly medial to the lateral tip of the transverse process at the level(s) of the main spine pathology. Needles were inserted in the fascial plane at a 45°-angle to the skin and directed toward the spine to a maximum depth of 3 cm (if bone was struck, the needle was withdrawn ~2 mm). The technique consisted of twirling the needle in a clockwise, then counterclockwise rotation (360° each). The end point occurred when the patient felt a deep sensation spreading out from the point and beginning to move down the back (the needle may have to be redirected several times to obtain this sensation). The frequency of rotation was 1 clockwise/counterclockwise movement per second.
Week 6
o The treatment was the same as for weeks 4 or 5, depending on the degree of effectiveness for the patient.
Component Adjustments and Indications
Electrical Stimulation. After week 3, if the patient felt that electrical stimulation worsened his/her condition, the intensity was lowered in
future treatments to a barely noticeable sensation. If that was still not tolerable, it was discontinued.
Relaxation. After its initial use in the first 2 treatments, relaxation was repeated if the patient requested this component. If the patient was tender to mild pressure on 2 or more of the points and had a history of headaches or pain/tightness in the shoulders, neck, or jaw, this component was recommended. However, the patient could still choose whether to have the treatment repeated.
Outcome Measures
The outcome measures were (1) the visual analog scale, VAS (for current pain, average pain in the last 2 weeks, pain on the best day in the last 2 weeks, and pain on the worst day in the last 2 weeks), (2) the Oswestry Disability Questionnaire, a popular low back pain research questionnaire,8 (3) the Short Form-36 Questionnaire Global Health Survey (SF-36), an equally well-known global health survey,9 and (4) the Patient-Generated Index for low back pain (PGI), a newer patient-generated scale for low back pain more sensitive to patient concerns.10

The PGI is not as familiar and is considered a quality-of-life scale; it measures the gap between one's hopes and expectations, and one's actual experience. There are 3 stages to the PGI. In stage 1, patients list the 5 most important areas or activities in their lives that are affected by their back pain (<5 areas may be listed). General areas and activities mentioned by other patients were provided in this study to encourage patients to include anything important to them, including physical limitations, specific aspects of pain, medication use, persistent or troublesome emotions or moods, challenges to independence or self-image, interference with personal or professional goals, interference with work roles or leisure activities, economic stress, and relationship problems. In stage 2, patients rated each item from stage 1 on a scale of 0-100, with 0 being the "worst they could imagine" and 100 being "exactly as they would like to be." An additional item labeled "all other aspects of your life not mentioned above" (in their personal list) was included and rated. In stage 3, patients distributed 60 points among the items on their list to weight them in terms of importance. It was not permitted to rate any items 0, or 1 item a 60. The final score was obtained by multiplying each item's percentage rating (0%-100%) by its weighting (0-60) and totaling all items, which yields a maximum score of 60. In this study, the final score was divided by 60 to express patient expectations realized on a 0-100 scale. An index of change was included to represent the percentage of the final score that could potentially come from new items. It was calculated as the sum of the weights of newly mentioned items, divided by 60.

The above outcome measures were recorded at initial presentation (within 1 week of the 1st treatment), at short-term follow-up (1-2 weeks after the last treatment, which was 6-8 weeks after the first treatment), and at long-term follow-up (3-4 months after the last treatment).

RESULTS
Results are presented as percentage changes in Table 2. All scales have been adjusted to 0-100 scales. For all scales other than the VAS, 0 represents the poorest and 100 represents the best outcome. For the VAS, negative change implies decreased pain and, therefore, improvement. In general, some improvements were noted in the group that met entrance criteria, while the group that did not meet these criteria did not show improvement with the exception of the VAS scales for subject 8.

For the short-term follow-up, the greatest differences were in the VAS best (actually a 237% increase in pain), the PGI (127%), and SF-36 subscales social functioning (61%), bodily pain (154%), and vitality (118%).

For the long-term follow-up, the greatest differences were in the PGI (505%) and SF-36 subscales social functioning (302%), bodily pain (148%), and vitality (235%). The VAS showed improvement in all areas at the long-term follow-up.

A change in the items listed in the PGI (index of change) was noted at short- and long-term follow-ups in the group that met entrance criteria, and was not noted in the group that did not meet these criteria.

DISCUSSION
This investigation was undertaken as a preliminary study for the use of medical acupuncture in sub-acute low back pain. The purpose was to gain experience in defining an appropriate study population, using a modular treatment protocol, and choosing sensitive outcome measures. No definitive conclusions can be drawn from such a small sample size and without the use of a control group. Nonetheless, our experience suggests that the study population should be defined to exclude conditions that are unlikely to respond to the short treatment protocols likely to be part of a larger randomized trial. Ninety percent of all low back pain resolves by 3 months, making it difficult to prove a treatment effect. Low back pain of longer than 1 year is by definition chronic and unlikely to respond to a short, focused treatment course. All patients in this study had experienced low back pain for more than 3 months but less than 1 year. The other inclusion/exclusion criteria and the work-up algorithm reflected theoretical concerns and our previous clinical experience.

The modular treatment protocol was designed to allow for an eclectic mix of acupuncture inputs most likely to be effective in the range of conditions that could be contributing to low back pain in this study cohort: degenerative lumbar disc or facet disease, myofascial pain with active and latent trigger points, iliolumbar ligament syndrome,
blocked Qi in meridians that cross the low back (Principal, Curious, and Distinct meridians), or local stagnant Qi. The treatment protocol components were ordered to capitalize on any general acupuncture effects, and then to escalate the intensity of the input while minimizing the chance of treatment aggravation. Our experience suggests that a modular protocol is usable and in an appropriately defined cohort, it allows for a reproducible but minimally-constrained series of treatments representative of a focused medical acupuncture approach.

Our experience also suggests that the Oswestry Questionnaire is minimally sensitive to any effects of acupuncture in this study population. Possible reasons for this include a relatively high functioning population with high initial scores (range 54-80/100, data not shown), fixed categories that may not cover a patient's most important or most improved problems, and an equal weighting of categories that dilutes the effect of changes of greater importance to the patient. Several SF-36 subscales and the total score appeared sensitive to changes possibly associated with the acupuncture treatments. However, the SF-36 subscales have the same shortcomings as the Oswestry Questionnaire; they can vary greatly with slight changes in responses because they are based on few questions with limited choices. The social functioning and bodily pain subscales each are calculated from 2 questions: 1 with 5 choices, the other with 6; the vitality subscale is calculated from 4 questions, each with 6 choices. The difficulties noted with the Oswestry and SF-36 outcome measures apply to all standardized fixed-category instruments, and indeed motivated the decision to include the PGI.
While the PGI appears sensitive to possible acupuncture treatment effects, it cannot by itself differentiate between improvement in the original condition or a re-evaluation of that condition. The index of change was meant to provide some assessment of this latter possibility.

The VAS actually showed an increase in average pain on the best day in the last 2 weeks at the short-term follow-up. Perhaps an explanation for this is that functional activity was increased at this time. We believe that by combining the acupuncture and physical therapy as described in this study, patients were able to increase their exercise tolerance, allowing them to achieve greater levels of spine stabilization and generalized fitness. This would be consistent with the improved PGI and SF-36 subscale scores at short-term follow-up, and the general improvement in all VAS categories at long-term follow-up.

CONCLUSION
This study suggests a combination of outcome measures that are more sensitive for determining the utility of medical acupuncture for subacute, non-radicular low back pain. A condition-specific low back pain questionnaire other than the Oswestry should be sought that is sensitive to the possible effects of acupuncture in this patient population. Its inclusion with the SF-36, PGI, and VAS would appear to be the most meaningful way to assess outcomes from physician and patient perspectives in non-radicular low back pain of intermediate duration. Further studies would also benefit from the modular treatment protocol described in this study for a reproducible approach to the range of conditions presenting as non-radicular low back pain in the outpatient clinical setting.

REFERENCES
1. Spaccarelli KC. Lumbar and caudal epidural corticosteroid injections. Mayo Clin Proc. 1996;71:160-178.
2. Simmons JW, McMillin JN, Emery SF, Kimmich SJ. Intradiscal steroids: a prospective double-blind clinical trial. Spine. 1992;17(6 suppl):S172-S175.
3. Dreyer SJ, Dreyfuss PH. Low back pain and the zygapophysial (facet) joints. Arch Phys Med Rehabil. 1996;77:290-300.
4. van Tulder MW, Cherkin DC, Berman B, Lao L, Koes BW. The effectiveness of acupuncture in the management of acute and chronic low back pain: a systematic review within the framework of the Cochrane Collaboration Back Review Group. Spine. 1999;24:1113-1123.
5. Helms JM. Acupuncture Energetics: A Clinical Approach for Physicians. Berkeley, Calif: Medical Acupuncture Publishers; 1995:281-286.
6. Travell JG, Simons DG. Myofascial Pain and Dysfunction: The Trigger Point Manual. Vol 2. Baltimore, Md: Williams & Wilkins; 1992:chap 3-12.
7. Seem M. A New American Acupuncture. Boulder, Colo: Blue Poppy Press; 1993:94.
8. Fairbank JCT, Couper J, Davies JB, O'Brien JP. The Oswestry low back pain disability questionnaire. Physiotherapy. 1980;66:271-273.
9. McHorney CA, Ware JE Jr, Raczek AE. The MOS 36-Item Short-Form Health Survey (SF-36), II: psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care. 1993;31:247-263.
10. Ruta DA, Garratt AM, Leng M, Russell IT, MacDonald LM. A new approach to the measurement of quality of life: the Patient-Generated Index. Med Care. 1994;32:1109-1126.

AUTHORS' INFORMATION
Dr Michael Fredericson is Assistant Professor and Director of Sports Medicine Clinics in the Division of Physical Medicine and Rehabilitation, Stanford University School of Medicine, Stanford, California. Dr Fredericson practices Medical Acupuncture at the Stanford Complementary Medicine Clinic.
Michael Fredericson, MD*
Stanford University Medical Center
Division of Physical Medicine and Rehabilitation
Department of Orthopaedics and Sports Medicine
300 Pasteur Drive, Edwards Building R-105B
Stanford, CA 94305
Phone: 650-723-1410 o Fax: 650-498-7546
E-mail: mfred2@leland.stanford.edu

Dr John Guisto is a Physiatrist in Durham, North Carolina, integrating Medical Acupuncture, manual medicine, and physiatry in treating musculoskeletal and pain problems in his orthopaedic practice. Dr Giusto is a
preceptor in the UCLA Medical Acupuncture course.
John Giusto, MD
Triangle Orthopedic Associates, P.A.
120 William Penn Plaza
Durham, NC 27704
Phone: 919-317-1081

*Address correspondence to: Michael Fredericson, MD, Assistant Professor, Stanford University School of Medicine, Division of Physical Medicine and Rehabilitation, 300 Pasteur Drive, R-105B, Stanford, CA 94305. Phone: 650-723-1410, Fax: 650-498-7546, E-mail: mfred2@leland.stanford.edu.

Back to Top | Table of Contents | On-line Journal Index