|
|
| |||||||||||||||||||||||||||||||
Background Chronic pain in the cervical zygapophyseal joints is a common problem after a whiplash injury. Treatment with intraarticular injections of corticosteroid preparations has been advocated, but the value of this approach has not been established. We compared the efficacy of a depot injection of a corticosteroid preparation with the efficacy of an injection of a local anesthetic agent in patients with painful cervical zygapophyseal joints.
Methods Sixteen men and 25 women with pain in one or more cervical zygapophyseal joints after automobile accidents (mean age, 43 years; median duration of pain, 39 months) were randomly assigned to receive a 1-ml intraarticular injection of either bupivacaine (0.5 percent) or betamethasone (5.7 mg) under double-blind conditions. The patients were followed by means of regular telephone contact and clinic visits until they reported a return to a level of pain equivalent to 50 percent of the preinjection level. The time from treatment to a 50 percent return of pain was compared in the two groups with the use of a survival analysis.
Results Less than half the patients reported relief of pain for more than one week, and less than one in five patients reported relief for more than one month, irrespective of the treatment received. The median time to a return of 50 percent of the preinjection level of pain was 3 days in the 21 patients in the corticosteroid group and 3.5 days in the 20 patients in the local-anesthetic group (P = 0.42).
Conclusions Intraarticular injection of betamethasone is not effective therapy for pain in the cervical zygapophyseal joints after a whiplash injury. .
We conducted a controlled trial to determine the efficacy of intraarticular corticosteroid injections for the treatment of patients with pain in the cervical zygapophyseal joints. The hypothesis tested was that a depot corticosteroid preparation injected into the painful joint would relieve the pain for a longer period than would an injection of a local anesthetic agent.
Methods
The study was conducted at a university unit operating as a tertiary referral center at a regional public hospital. The unit accepted referrals of patients over the age of 18 years with neck pain attributed to a motor vehicle accident and lasting for more than three months. The study protocol was approved by the ethics committees of the university and the area health service, and written informed consent was obtained from each patient enrolled in the study. On referral, each patient underwent a full clinical evaluation. At the initial interview, the severity of pain was assessed on the basis of the score on a 100-mm visual-analogue scale of pain, from 0 (no pain) to 100 (worst possible pain); the total word count and score for the present intensity of pain on the McGill Pain Questionnaire7; and the score on a checklist of psychological symptoms, which uses a scale of 0 to 4 to score such symptoms as somatization, interpersonal distress, anxiety, and depression8.
Diagnostic Phase
The diagnostic criteria for cervical zygapophyseal joint pain, which were detailed and validated in a previous study,9 are based on controlled experiments in subjects without pain that showed that the anesthetic effects of bupivacaine outlast those of lidocaine10,11.
A putatively painful cervical zygapophyseal joint was selected for initial study on the basis of the distribution of pain12,13. Patients were randomly selected to receive either 2 percent lidocaine or 0.5 percent bupivacaine and were not told which agent was to be administered. The joint was anesthetized by blocking the medial branches of the dorsal rami of the cervical nerves with the use of a previously validated technique9,14. For the C2-3 joint, the third occipital nerve was blocked. The joints from C3-4 to C6-7 were anesthetized by blocking the medial branches above and below the joint15. The response to the injection was assessed by an independent observer who was unaware of the agent that had been injected. Pain relief was rated as "complete," "definite," "partial," or "none." Complete relief of pain was defined as the abolition of all the usual pain in a discrete anatomical region; definite relief as substantial and gratifying relief of pain, with some residual discomfort from an additional source of pain or trauma from the procedure; and partial relief as some diminution of pain but no more than might be explained by spontaneous variation. Pain relief was considered to have occurred only if the response was complete or definite.
If a block of the first joint failed to relieve the patient's pain, additional joints were blocked either until the pain was relieved or until all joints that might have been the source of pain had been tested. Once a block of a joint succeeded in relieving the pain, the initial study was terminated, and the patient returned after two weeks for the second block of that joint, with the use of the second agent.
For enrollment in the study, patients had to have relief of pain on both occasions when the joint was blocked and had to have a longer period of relief with bupivacaine than with lidocaine. We also enrolled six patients who had inordinately prolonged responses to lidocaine or to both agents. A prolonged response is consistent with the literature on anesthetic agents16 and was therefore considered an insufficient reason for excluding a patient from the study9. Patients with any other response were excluded.
Therapeutic Trial
A single cervical zygapophyseal joint was identified as the sole source of neck pain after a median of three blocks in 27 of the 42 patients who enrolled in the study. Of the 15 patients with two or more painful joints, 8 had bilateral joint pain at a single level; the other 7 patients had either less severe pain in a second joint or another, unidentified source of pain. Only one joint was treated in each patient. All treated joints were the primary source of pain in the neck, and discrete and readily identifiable symptoms could be attributed to them. Patients were enrolled in the trial at least one week after their usual pain had returned after the diagnostic blocks.
Tables of random numbers were used to assign the patients to treatment with a single, 1.0-ml intraarticular injection of either bupivacaine (0.5 percent) or betamethasone (5.7 mg). A local anesthetic agent was selected as the control for ethical and methodologic reasons. An intraarticular injection of an anesthetic agent is a legitimate diagnostic procedure, unlike injection of a placebo, which is an invasive procedure with attendant risks and x-ray exposure for no benefit to the patient. Furthermore, the use of a local anesthetic agent provides a more stringent control.
With the use of fluoroscopy and an image intensifier, the painful joint was anesthetized by medial-branch blocks with 0.5 percent bupivacaine, so that the subsequent intraarticular injections were painless. A lateral approach was used to direct a 25-gauge, 10-cm spinal needle onto the articular pillar adjacent to the target joint. The needle was then carefully guided just into the joint space, which was detected by the loss of bony resistance. Correct placement of the needle was confirmed by arthrography after injection of 0.05 to 0.2 ml of contrast medium (Figure 1). If the needle could not be inserted into the joint space, a second attempt was made on another day. The needle was always introduced into the joint by the same operator, who did not know which agent was to be injected.
|
Follow-up
During the course of the trial, patients were permitted to use their usual analgesic and physical therapies, but surgical or neurolytic procedures were proscribed. Physical activity was not restricted. All follow-up examinations were performed by the same observer, who did not know which treatment had been given. Each patient was contacted by telephone 1, 4, 8, 16, and 20 weeks after the injection and was examined in the unit at 2 weeks and 12 weeks. The visual-analogue scale and McGill Pain Questionnaire were administered only during the visits to the unit. During each telephone call and unit visit, the patients were asked to describe their symptoms and to state the percentage of preinjection pain being experienced. The end point of the study was a return of pain to 50 percent of the level experienced before the injection. Patients were encouraged to arrange an immediate visit to the unit when their pain reached the 50 percent threshold.
Statistical Analysis
The principal outcome measure was the time from the injection to the return of pain to 50 percent of the preinjection level. We used the log-rank test to compare the Kaplan-Meier curves for the two treatment groups, with a failure of treatment defined as the return of pain to the 50 percent threshold. The characteristics of the two groups at base line were compared with Student's two-tailed t-test for parametric data and with the Mann-Whitney U test for nonparametric data.
Results
Intraarticular injections were successfully performed on the first attempt in 38 of the 42 patients and on the second attempt in 3. One patient withdrew from the study after a single, unsuccessful attempt and was not included in the analysis. Seven patients received less than the full 1.0-ml injection because of excessive resistance in the joint, but no patient received less than 0.7 ml. The only symptoms that might be considered side effects of the treatment were transient facial flushing in two patients and temporary exacerbation of the usual pain when any analgesic effect of the injections had worn off.
The treatment and control groups were well matched, with no significant differences in age; duration of pain; or base-line scores on the visual-analogue scale, pain questionnaire, or symptom checklist (Table 1). No patients were lost to follow-up.
|
In over half the patients, neither treatment provided relief of pain for more than a week, and less than 20 percent of the patients had any substantial relief after one month (Figure 2). Only 15 patients completed a second set of pain assessments before their pain had returned to base-line levels. Their verbal assessments of the severity of pain, expressed as a percentage of base-line pain, correlated closely with the change in the pain-intensity score on the McGill Pain Questionnaire (r = 0.89, P<0.001), the total word count (r = 0.87, P<0.001), and the score on the visual-analogue scale (r = 0.64, P<0.001). Consequently, the patients' verbal assessments were taken to be a reliable measure of pain and formed the basis for further analysis.
|
|
Our study provides little reassurance for clinicians using corticosteroid injections to treat patients with pain in the cervical zygapophyseal joint after whiplash injuries, since the corticosteroid was no more beneficial than a local anesthetic agent. In all studies of corticosteroid therapy, the patients were only presumed to have pain in the cervical zygapophyseal joint2,3,4,5,6. In contrast, we used a rigorous diagnostic protocol to select patients in whom the cervical zygapophyseal joint was the source of pain9. Such patients are most likely to have pain in the joint and are therefore most likely to have a response to a targeted treatment. Thus, the negative result of our study cannot be attributed to improper patient selection.
All injections were performed by experienced physicians. The correct position of the needle was confirmed to ensure that the injections were correctly placed. We used a lateral rather than a posterior approach, and instead of obtaining a formal arthrogram, we injected only enough contrast medium to confirm intraarticular placement. There is no reason to suppose that these modifications altered the efficacy of the injection.
A return of pain to 50 percent of the preinjection level was selected as a threshold below which the benefit of treatment would have been unequivocal but some residual discomfort would also have been possible. If 100 percent relief of pain had been the outcome measure, the results might have been biased in favor of a poor result despite some benefit of treatment. Our outcome measure compares favorably with the measures used in previous studies, in which the degree of pain relief was often poorly defined.
We enrolled only patients who had been referred to a specialized unit because of chronic and intractable pain in the cervical zygapophyseal joint after whiplash injuries, and the results of our study should not be extrapolated to the treatment of patients with cervical zygapophyseal joint pain from other causes. Whether spontaneous pain in the cervical zygapophyseal joint responds to intraarticular steroid injections is unknown.
Some patients had substantial and long-lasting relief of pain after injection with either agent. These responses may have been due to a placebo effect. However, all the patients had long histories of ineffective treatments and, if anything, should have had an attenuation of any placebo effect17. Alternatively, the patients who derived a benefit from either treatment may have had a condition that was improved by the stretching of the joint capsule during intraarticular injection, irrespective of what was injected.
In summary, intraarticular injections of betamethasone are not effective in patients with pain in the cervical zygapophyseal joint after a whiplash injury. There would seem to be little justification for pursuing such therapy, with the attendant risks of infection, exposure to radiation, side effects of the corticosteroid, and damage to nearby structures during insertion of the needle.
Supported by a grant from the Motor Accidents Authority of New South Wales.
Source Information
From the Cervical Spine Research Unit, Faculty of Medicine, University of Newcastle, Callaghan (L.B., S.M.L., B.J.W., N.B.), and the Mater Misericordiae Hospital, Waratah (L.B., S.M.L., N.B.) -- both in New South Wales, Australia.
Address reprint requests to Dr. Barnsley at the Cervical Spine Research Unit, Faculty of Medicine, University of Newcastle, University Drive, Callaghan, NSW 2308, Australia.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
