Conclusions from Article about the effectiveness of Interventional spine injections
Fluoroscopically guided lumbosacral transforaminal epidural corticosteroid injections are effective in the short term, and possibly at 6 months, in treating acute/subacute lumbosacral radicular pain [13,18].
Fluoroscopically guided lumbosacral transforaminal epidural corticosteroid injections are more effective than placebo at preventing future surgeries .
One-level intra-articular cervical zygapophysial joint corticosteroid injections are not more effective than placebo in treating patients with chronic cervical zygapophysial joint pain after whiplash injury .
Corticosteroid or Sarapin injections on the cervical or lumbar medial branch nerves are not more effective than placebo in treating patients with chronic cervical or lumbar zygapophysial joint pain [33,35].
When using a technique that is superior to the technique used in the other randomized controlled trials, percutaneous radiofrequency lumbar medial branch neurotomy in patients with chronic lumbosacral zygapophysial joint pain is more effective than placebo [38,40].
Percutaneous radiofrequency cervical medial branch neurotomy is more effective than placebo in treating patients with chronic cervical zygapophysial joint pain after whiplash injury .
Sacroiliac joint corticosteroid injections are more effective than placebo at 1 month in treating patients with spondyloarthropathy and low back pain .
IDET is modestly more effective than placebo at 6 months in treating patients with less than 20% disc height loss who are diagnosed with discogenic pain via the discography criteria of concordant pain multiple times at a similar PSI with negative control discs .
Percutaneous radiofrequency neurotomy of the ramus communicans is more effective than placebo at 4 months in treating discogenic pain .
Because of the lack of PDBRPCTs or the significant methodologic flaws in the existing PDBRPCTs, no firm conclusions can be drawn about: lumbosacral epidural corticosteroid injections for the treatment of chronic radicular pain, cervical epidural corticosteroid injections for the treatment of acute or chronic radicular pain, lumbosacral or cervical zygapophysial joint corticosteroid injections for the treatment of degenerative zygapophysial joint pain, percutaneous radiofrequency cervical medial branch neurotomy for the treatment of degenerative zygapophysial joint pain, sacroiliac joint corticosteroid injections in patients without spondyloarthropathy, percutaneous radiofrequency denervation of the sacroiliac joints, or intradiscal corticosteroid injections for the treatment of discogenic pain.
If two treatments are not statistically different, what other factor could you use to determine what therapy is most appropriate for your patient?
Why is the time after onset of symptoms to treatment important to whether or not the treatment is effective?
What is the mechanism of action for each of the injected substances (corticosteroids, saline, lidocaine or similar), why would they relieve pain?
What are the possible complications of injections? What is the risk for the patient?
The current article reviews the PDBRPCTs in the interventional spine literature with strict interpretation of their results. As physicians, it is our job to understand how to properly translate this raw data into accurate and clinically relevant information. To do this, one must know how to properly evaluate medical literature.
In outcome studies, there are two potential resultsd positive (meaning that the treatment is statistically betterthan the control group) or negative (meaning that no difference was demonstrated between groups.) As with diagnostic tests, a positive study can be a true positive or a false positive, and a negative study can be a true negative or a false negative. As clinicians, it is our duty to attempt to recognize false positive and false negative studies so that the conclusions of these studies are not misinterpreted. In the interpretation of medical literature, the design of negative studies deserves closer evaluation than that of positive studies. As William Cowper said, ‘‘absence of proof is not proof of absence.’’ A negative study may be negative
because the treatment is ineffective. However, if serious flaws in study design exist, the negative results are in question.
Because proof of absence is much more difficult to demonstrate than proof of presence, negative studies should receive strict scrutiny to determine if the results are truly negative. Positive studies are different. Assuming that the randomization process has equalized baseline data between groups, and that the control group does not receive a detrimental treatment, then positive results should be accepted for what they are statistically (although not necessarily clinically) significant proof that a true difference exists between the treatment and control groups. Of course, when interpreting positive results, one must consider that the defined medical convention is to accept a 5% false positive rate (p value of .05).
With that in mind, positive results are positive. Negative results, however, require greater scrutiny to determine if the treatment is truly ineffective. This is best demonstrated with an exaggerated example. In this hypothetic example, let us evaluate if IV hydromorphone is effective in treating the pain from a mild paper cut. In designing this study, one should compare IV saline to IV hydromorphone. However, what if the control group received IV morphine instead? If negative results were obtained (ie, if there was no statistically significant difference between IV hydromorphone and IV morphine), should we conclude that IV hydromorphone is ineffective at treating the pain from a minor paper cut? Of course not a major flaw in the design of the trial explains the negative results (the control group received an active treatment instead of a true placebo). Now, what if the results were positive (ie, IV hydromorphone was shown to be significantly more effective than IV morphine)? In this case, despite the significant flaw in study design, IV hydromorphone was shown to be an effective treatment. It would be reasonable to conclude that IV hydromorphone is more effective than placebo, and even more impressively, that it is more effective than another active treatment IV morphine.
Lastly, let us examine a final example. Instead of comparing IV hydromorphone to IV saline or IV morphine, consider a study in which the control group received IV cyanide. One would expect positive results from the hydromorphone compared with the cyanide. However, because IV cyanide would be expected to be a detrimental treatment, these positive results would not truly tell us if IV hydromorphone is more effective than placebo.
In the above review of the PDBRPCTs in the interventional spine literature, none of the control groups received a treatment that can reasonably be expected to be of detriment.
Why is it important how studies were selected for this review?
How does the hypothesis proposed shape the way we interpret the results?
Is statistical and clinical significance always the same? If not, how can we tell which one is more important for our patients?
Why is it important to read the methods section of a peer-reviewed article? What information can we glean from that section?
What type of study design is the “Gold Standard?”