Wednesday, October 26, 2011

Midazolam for Back Pain: When Data Are Not Proof

EBPM Here is a lesson in evidence-based pain management. Researchers reported that intrathecal midazolam, a benzodiazepine, was a beneficial supplement to standard analgesic therapy for recalcitrant lower-back pain. However, because of the study design and limited data this is not valid evidence of treatment effects, which raises questions about why this study was published.

Writing in the journal Pain Medicine, Jan Prochazka, MD, and colleagues from the Czech Republic describe a prospective, observational, open-label investigation of single-dose intrathecal administration of midazolam (off-label) in patients with chronic low-back pain or failed back surgery syndrome [Prochazka et al. 2011]. The primary outcome measure was pain relief determined by a patient questionnaire during subsequent visits to the pain therapy clinic. A positive analgesic-effect result was considered at least 50% pain reduction along with improved quality of life and functionality.

Between 1995 and 2010, the authors report having performed 500 intrathecal midazolam administrations: 227 in 57 male patients and 273 in 69 females. Of those, 81 administrations were for chronic low back pain and 419 for failed back surgery. Average patient age was 50 years (range 28 to 86). The single midazolam dose administered was preservative-free and ranged from 2 to 5 mg (0.02-0.06 mg/kg). Analgesic effects lasted 9.7 weeks on average, ranging from 1 week to 3 years; the most common reported duration was between 4 to 12 weeks. In 65% of patients, pain relief lasted 4 weeks or longer; however, in 13%, the treatment provided no analgesic effect at all. The incidence of side effects (drowsiness, nausea, headache, or transient worsening of complaints) was “very low,” according to the authors.

The researchers conclude that single-dose intrathecal midazolam is a useful supplement to standard analgesic therapy with opioids, non-opioids, or spinal steroids in patients with chronic low-back pain or failed back surgery syndrome who suffer recalcitrant somatic or neuropathic pain. They are careful to note that intrathecal midazolam is an off-label application of the drug, unapproved by the U.S. Food and Drug Administration.

COMMENTARY: At first glance, this study report seemed straightforward, the outcome measures were relevant, and the conclusions appeared to naturally follow from the published data. Of course, one needs to understand that the observational methodology used is a very low level of evidence; essentially, a series of case reports collected over time, assembled, and analyzed as a group.

With this type of research design and the very basic analyses in this article, there were no controls for placebo effects or confounding variables that might have distorted outcomes. Of particular concern, the authors exhibited a “trust us” stance by not including adequate data regarding participant demographics, concurrent medications or therapies, or details of responses on the outcome measures. It is difficult to assess validity with so little information and we were not even going to report on this research — until we noticed an intriguing editorial in the same journal commenting on the Prochazka et al. article.

With surprising candor, Nikolai Bogduk, MD, PhD, DSc, a Senior Editor of Pain Medicine, acknowledges that “there were disputes about this article between the referees and the authors” during the review process [Bogduk 2011]. He claims the journal welcomed the data but not the author’s conclusions.

Bogduk — who is also Professor of Pain Medicine, University of Newcastle, New South Wales, Australia — further notes that post hoc, ergo propter hoc — "after this, therefore because of this" — is a common fallacy and illusion in medicine and also is depicted in this study. While the study authors contend that they provide data demonstrating the effectiveness of their intervention, just because patients improved after treatment does NOT mean that they improved because of the treatment.

The study methodology “does not account for placebo effects or the natural course of the condition. There also may be uncontrolled confounders; eg, the intrathecal drug was a supplement to oral opioids or non-opioid analgesics,” Bogduk writes. Although, he does not actually complain about the complete lack of data in the article about co-analgesics taken by the subjects or, for that matter, the omission of specific data used in assessing pain relief, quality of life, or functionality.

Furthermore, the fact that Prochazka et al. quote others in their paper who make the same claim of midazolam efficacy does not carry any weight with Bogduk. “Doing so only refers to others who suffer the same illusion,” he states rather starkly.

He goes on to observe that, as alluded to above, while patients appeared to improve after intrathecal midazolam therapy, the data do not prove that the response was in any way due to that treatment. “The competing hypothesis is that the patients improved because they were cared for in a good manner, using a high-tech treatment. It may be that they responded to components of the treatment other than the midazolam. In other words, they responded because they had an injection, irrespective of what was injected. The history of interventional pain medicine is replete with examples of injections that work but not because of what was injected,” Bogduk says.

At this point in his editorial, Bogduk retreats a bit. He suggests the data may be of some value because of the large numbers of patients who were treated and the realistic observation that the treatment was not effective for everyone. Yet, those who responded did so for a median of 5.5 weeks, and 40% for 12 to 24 weeks. It also seemed reassuring that intrathecal midazolam appeared relatively safe.

However, Bogduk then demurs, “Relief for 6 weeks, or even 12 weeks, is not a cure, and some would wonder if it is worth administering a therapy the response to which is so short-lived.” On the other hand, he continues, “others could argue that there is precious little that can be done for patients with failed back surgery syndrome and, therefore, that it is worth knowing that intrathecal midazolam can be drawn from a ‘bag of tricks’ in the palliative management of these patients with a difficult problem.”

Again turning to the negative, Bogduk argues that only a small proportion of patients experienced long-lasting effects of the treatment, and that the role of midazolam in this was questionable. He writes, “Scientifically, we do not know if the long-term responses are due to the drug or are a placebo effect. Pragmatically, one could argue that the effects on the patient are good and worth having even if they are a placebo, but the question then becomes: ‘do we need to add the midazolam?’ We can harvest the placebo response with an injection of a less expensive agent with less risk of side effects. All that is required is to perform the injection in a convincing manner.”

“I am not against harvesting a good outcome for patients,” Bogduk concludes, “but I am hostile to wrongful claims about how and why this treatment works.” So, after dithering between condemning and defending the paper, perhaps to demonstrate fair balance, he seems to reason that, as an example of valid and enlightening research, this study is a failure. However, an important question is never answered: With such concerns and reservations, why did the journal go ahead and publish the article anyway?

As evidenced by Bogduk’s editorial, the journal was clearly conflicted about publishing the article. Unfortunately, few readers will know about the negative editorial questioning this study, so the article will likely be na├»vely accepted and cited by future authors as evidence favoring intrathecal midazolam for recalcitrant back pain conditions.

Similar troubling situations occur when journals later retract faulty articles after publication. At the time we were examining the above research and editorial, an announcement came to our attention from the journal Rheumatology [available here]. It stated….

“We regret to retract the article ‘Meta-analysis of systemic lupus erythematosus and the risk of cervical neoplasia’ because: i. Errors were identified in the publication involving several parts of the study including extraction of data, statistical analyses and interpretation of results. ii. Cervical neoplasia should have been interpreted as abnormal pap smears.”

“Unfortunately, such extensive changes cannot be addressed in a corrigendum and warrant a new manuscript. It is important to note that this is not considered to be scientific misconduct, but rather an honest error by authors. We regret any problems that this article may have caused and retract it from the literature.”

It is discouraging that the journal editors and peer reviewers did not detect in advance such fundamental flaws in the article as how data were extracted, analyzed, and interpreted. And, we always wonder in such cases how many researchers and authors already have the original article filed away and will cite it as evidence in their future papers on the subject without knowing it was erroneous.

Anyway, be on the lookout… that paper above by Prochazka and colleagues may be retracted some day.

> Bogduk N. When Data Are Not Proof [Editorial]. Pain Med. 2011(Sep);12(9):1297–1298 [
access by subscription].
> Prochazka J, Hejcl A, Prochazkova L. Intrathecal Midazolam as Supplementary Analgesia for Chronic Lumbar Pain—15 Years' Experience. Pain Med. 2011(Sep);12(9):1309-1315 [
abstract here].