Do NSAIDs Work for Treating Sciatica?

Weber H, et al. "The Natural Course of Acute Sciatica with Nerve Root Symptoms in a Double-Blind Placebo-Controlled Trial" Spine 1993; 18(11):1433-1438

< Investigation Results | The Study | Discussion >

The main focus of this double-blind placebo-controled trial was to investigate the effectiveness (efficacy) of a nonsteroidal anti-inflammatory (Feldene) as a treatment for patients suffering back and leg pain (sciatica) that was thought to be induced by disc herniation. It also followed the cohort (214 patients) for a 12 month period, in order to assess the natural recovery pattern of true radicular pain (sciatica).

Unfortunately, there was absolutely NO confirmation of the diagnosis, i.e., no MRI, CT, or EMG! So, although obvious spinal conditions (Rheumatoid arthritis, Ankylosing spondylitis, definite surgical indications) were excluded, it could only be 'assumed' that disc herniation was the underlying pain-generator. There may well have been other causes (as openly admitted by the authors) to the patients sciatic pain, such as stenosis spondylolisthesis, tumors, etc.

The bottom line of this investigation was that the NSAID Feldene, which is a very powerful anti-inflammatory, was useless in reducing the pain of sciatica during the first four weeks of care; however it was effective at irritating the stomachs of 12% of group!

In the long-term (one year) about 30% of the cohort continued to suffer significant pain, limitations in leisure activity and restrictions at work. Even more noteworthy is the fact that only 54% of the cohort subjectively 'completely recovery' from their back and leg pain! The 54% number, although a little high, fell in line with other investigations into the prognosis of disc herniation-associated sciatica (15,16,17,19,77).

The Study :

During a period of three years, 214 radicular pain suffering patients were gathered together for an investigation into the efficacy of an NSAID (Feldene) on true radicular pain (sciatica) that was confined to the S1 or L5 root level. Most of these patient's pain was moderate to severe in nature because only 2% of them required hospitalization. Great care was used to select ONLY patients who presented with root-related L5 or S1 sciatica pain.

The following conditions were excluded for this investigation: cauda equina syndrome, progressive paresis (worsening muscle weakness), Rheumatoid Arthritis, Ankylosing Spondylitis, and Psychiatric Illness. Unfortunately no MRI's, CT's, or EMG's were done on the group; therefore, Stenosis and/or Spondylolisthesis were probably complicating some of these cases. (Bad design!)

Treatment consisted of strict bed rest for one week, followed by slow "mobilization" over the next few weeks. No physical therapy was allowed for the first 4 weeks. All patients were randomly given either a strong anti-inflammatory (Piroxicam aka: Feldene) OR a 'placebo pill' that looked exactly like the anti-inflammatory; the cohort took the pills for four weeks and then assessed via questionnaire.

Pain in the back and leg was measured by the VAS (visual analog scale 0 - 100 where 100 is severe pain).   The degree of functional-disability was measured by a modified 17 question Roland - Morris disability questionnaire.

VAS scores were matched to descriptive words in order to define a successful outcome:

Free from pain corresponds to VAS 5 (less then 10)    
Mild pain corresponds to VAS 20 (10 to 30)
Moderate pain corresponds to VAS 45 (30 to 60)
Severe pain corresponds to VAS 80 (60 to 100)

The Results:

Results:

The patient participation rate was adequate (26) although not impressive: 88% of the patients responded to the mailed questionnaire at 3 months and 82% responded at 1 year. (see result chart)

At the 4 week mark, which was the full investigation period for the NSAID, there was no difference in pain relief scores (VAS) between the two groups. Ironically, the patients who were given the real NSAID (feldene) medication, not only had slightly higher pain scores, they also suffered nearly twice as much stomach-upset as the placebo group!

Poor Prognosis:   the only factor associated with a poor outcome was a past history of sciatica, i.e., if the patient had a past history of sciatica, their chances for full recovery were diminished when compared to those patient who had never had sciatica before.

Discussion:

The most sobering results from this investigation, aside from the conclusion that the NSAID Feldene is completely useless for the treatment of radicular pain, was the finding that people often do NOT fully recover from the syndrome of true sciatica (aka: radicular pain, radiculopathy).

The obvious down-fall of this investigation, as an outcome study, was the fact that there was no confirmation of the diagnosis by MRI or CT. However, great care was taken to only included patients with obvious root-related sciatic pain into the investigation. You can rest assured that the overwhelming majority of this cohort was suffering from disc herniation-associated sciatica.

The authors also hinted that a portion of the cohort came from work-related injuries. Workers' compensation patients, as well as other litigation type cases, often can skew the data within investigations, for these types of patients do NOT always tell the 'whole truth' when it comes to assessing their level of improvement (for fear of damaging their cases). The latter 'Workers' compensation phenomenon' may explain why Weber's data is a little out-of-line with the other investigations into the natural history of sciatica. in regards to the patient's reported recovery from sciatica. (15,16,17,19,77)

Study Quotes:

"About 60% of all patients were back at work by 4 weeks.   At one year 7.5% of the patients were still off work.

"There is hardly any indication for the use of nonsteroidal anti-inflammatory drugs in acute cases of sciatica."

* "That 30% of the patients after 1 year had back pain, reduced capacity in work, and restriction of leisure activity underlines the seriousness of the illness as in also seen from a medico-social point of view."

REFERENCES:

15) Atlas SJ, et al. "Surgical & nonsurgical management of sciatica secondary to a lumbar disc herniation: Five year outcomes from the Maine Lumbar Spine Study." Spine - 2001; 26(10):1179-1187

16) Henrik Weber, '1982 Volvo Award in Clinical Science' "Lumbar Disc Herniation: A controlled, Prospective Study with Ten Years of Observation." Spine - 1983; 8(2):131-140

17) Nykvist F, et al. "A prospective 5-year follow-up study of 276 patients hospitalized because of suspected lumbar disc herniation" Int. Disabil. Studies - 1989; 11(2):61-67

19) Weber H, et al. "The Natural Course of Acute Sciatica with Nerve Root Symptoms in a Double-Blind Placebo-Controlled Trial" Spine 1993; 18(11):1433-1438

26) Nachemson AL, LaRocca H. "Editorial

77) Saal & Saal, "Nonoperative Treatment of Herniated Lumbar IVD with Radiculopathy: An outcome Study." Spine 1989; 14(4):431-437