HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Correspondence to:

SPECIAL ARTICLES: Carmel Armon and Randolph W. Evans Addendum to assessment: Prevention of post–lumbar puncture headaches: Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology Neurology 2005; 65: 510-512 [Abstract] [Full text] [PDF]

Correspondence: Submit a response to this article

Correspondence published:

Addendum to assessment: Prevention of post–lumbar puncture headaches: Report of the Therapeutics and

Phillip L. Pearl, William M. McClintock, Dept. Neurology, Children's National Medical Ctr, Washington, DC   (15 November 2005)

Reply to Pearl et al

Carmel Armon, and Randolph W. Evans. M.D. for the Therapeutics and Technology Assessment Subcommittee   (15 November 2005)

Addendum to assessment: Prevention of post–lumbar puncture headaches: Report of the Therapeutics and 15 November 2005
Phillip L. Pearl, Children’s National Medical Center Dept. of Neurology; 111 Michigan Ave. NW, Suite W4-800; Washington, DC 20010-2970, William M. McClintock, Dept. Neurology, Children's National Medical Ctr, Washington, DC Send Correspondence to journal: Re: Addendum to assessment: Prevention of post–lumbar puncture headaches: Report of the Therapeutics and ppearl{at}cnmc.org Phillip L. Pearl, et al.	We read with interest the report of the Therapeutics and Technology Assessment ((TTA) Subcommittee of the AAN providing an addendum to the assessment on the prevention of post-lumbar puncture headaches (PLPHAs) following diagnostic LPs. [1] We were surprised that a consensus was reached for a Level A recommendation. Only one class I article was cited. [2] In the Strupp et al study, 306 patients were allocated randomly to the “atraumatic” versus “traumatic” needle, yet 230 are evaluated, representing a drop-out rate of 25%. Of the 76 drop-outs, 25 did not return the evaluation sheet. Since the Subcommittee’s recommendation includes a mandate to enact widespread educational strategies to impact neurologic practice, we think an intention-to-treat analysis that incorporates all randomized patients would be important. When further assessing the Strupp et al [2], the control event rate of 24% versus experimental event rate of 12% translates to a relative risk reduction of 50% and absolute risk reduction of 12% (NNT=8.3). However, the confidence interval calculates to 0.12 +/- 0.1, (i.e., an absolute risk reduction of 0.02 – 0.22). The new conclusion: “now also one study providing Class I evidence in a patient population undergoing diagnostic LPs with a 22-guage needle support the use of an atraumatic spinal needle to reduce the frequency of PLPHA" is unlikely. Additionally, the studies leading to the addendum do not provide data higher than Class IV evidence addressing other relevant primary outcomes (e.g., occurrence of back pain, technical variables between the two procedures). We question the new conclusion that supports the use of an atraumatic spinal needle to reduce the frequency of PLPHA, and the recommendations to develop and disseminate standardized training materials for practitioners, and to track acceptance and implementation within the neurologic community. References 1. Armon C and Evans RW: Addendum to assessment: Prevention of post-lumbar puncture headaches: Report of the TTAS of the AAN. Neurology 2005; 65:508 -512. 2. Strupp M, Schueler O, Straube A, Von Stuckard-Barre S, Brandt T. “Atraumatic” Sprotte needle reduces the incidence of post-lumbar puncture headaches. Neurology 2001; 57:2310-2312. The authors report no conflicts of interest.
Reply to Pearl et al 15 November 2005
Carmel Armon, From the Division of Neurology, Baystate Medical Center 759 Chestnut Street, Springfield, MA 01199-001, and Randolph W. Evans. M.D. for the Therapeutics and Technology Assessment Subcommittee Send Correspondence to journal: Re: Reply to Pearl et al Carmel.Armon{at}bhs.org Carmel Armon, et al.	We thank Drs. Pearl and McClintock for their comments and appreciate this opportunity to respond. Pearl and McClintock state that 306 patients were randomized. The authors of the original article [2] state that 51 patients did not meet the inclusion criteria and were not randomized. Therefore, 255 subjects were randomized. Of the randomized patients, 25 did not return the evaluation sheet - 12 from the traumatic group and 13 from the atraumatic group. The 230 patients who returned the evaluation sheets were equally divided and the results for them are reported using intention-to-treat analysis. A ten percent drop- out rate is acceptable and does not change the evidence class. We agree with Pearl and McClintock that the number needed to treat to prevent one PLPHA by using a 22G atraumatic needle rather than a 22G traumatic needle, based on the article in question [2] is 8, and indicated this in the Discussion. We agree also that a reduction from 24% to 12% represents a relative risk reduction of 50% and an absolute risk reduction of 12%. We elected to present the information only in terms of numbers needed to treat, because we consider that number to be easier to understand than RRR or ARR. Pearl and McClintock state that the confidence interval on the absolute risk reduction of 12% is 2% to 22%, without providing the basis for their statement. If they are correct, the best estimate for the absolute risk reduction remains 12%; eight is the most likely estimate of the "number needed to treat," and the likelihood that the true number is greater than 8 is the same as the likelihood that it is smaller than 8. The original assessment and the addendum discussed technical aspects within the Discussion. The presence of a learning curve to the use of a new technology is not considered, in general, a "primary outcome" of the use of that technology. Pearl and McClintock suggest also that the occurrence of back pain should be considered an additional relevant primary outcome. We point out that there can be only one primary outcome, while recognizing the need to be aware of secondary outcomes that might detract from the value of benefits measured within the primary outcome. The decision for a level A recommendation was based on available evidence for post-LP headache. It took into consideration not only the new (2001) class I evidence in diagnostic LPs, but also the class I evidence in spinal anesthesia studies, all pointing to the reduction in the frequency of PLPHA by using non-cutting needles. The recommendation to develop training materials and track acceptance are consistent with the mission of the AAN to serve its constituent members as they, in turn, serve patients with neurological diseases. We stand by our original conclusions and recommendations. Considering Drs. Pearl and McClintock’s affiliation, we point out that our recommendations pertain to the adult population, in which there are high quality data and make no comments about the pediatric population. The authors have no conflicts of interest to report.