Myth: fluids, bed rest, and caffeine are effective in preventing and treating patients with post-lumbar puncture headache.

Headache is the most common complication after lumbar puncture (LP), with reported frequency rates ranging from 6% to 36% of patients.1 August Bier (1861-1949) was the first to describe the phenomenon of post-dural puncture headache in his patients and experienced the same effect when he had the procedure performed on himself.2 Most (90%) post-LP headaches occur within 3 days of the procedure and are characteristically described as being present when the patient is in the upright position and diminished in intensity when supine. The cause of post-LP headache is uncertain. One idea is that it is possibly due to low cerebrospinal fluid (CSF) pressure as a result of CSF leakage through a dural and arachnoid tear produced by the puncture that exceeds CSF production. The continuous decrease in CSF pressure may lead to subsequent stretching of pain-sensitive structures. Another notion is that cerebral vasodilatation, in addition to traction, is responsible for headache following LP. Various treatments for this condition are thought to be effective, even though its cause is unclear. Many of these are implemented routinely in daily practice—including increased fluids, bed rest, and caffeine—despite the lack of evidence of their effectiveness. There is no evidence supporting the use of increased fluids to prevent post-LP headache.1 The only prospective study of this intervention involved oral hydration. Dieterich and Brandt performed a prospective study of 100 age-matched, randomly allocated neurologic patients and found no correlation between the incidence of post-LP headache and the amount of fluid intake.3 Half of the patients were asked to drink 1.5 L of fluids per day during the 5 days after an LP, and the other half was asked to drink 3.0 L of fluids per day for the same period. The intensity of the headache was classified into four grades according to the severity and onset of symptoms after getting up from the LP. The proportion of symptom-free individuals was 64% in both groups of patients; therefore, the incidence of post-LP headache is independent of fluid intake. Another commonly held belief is that bed rest or various body positions after LP reduce the incidence of post-LP headache compared with immediate ambulation. But Carbaat and van Crevel performed a controlled prospective study that showed that no benefit was found with 24 hours of bed rest in preventing the headache.4 A diagnostic LP was performed in 100 neurologic patients by one investigator. Half of the patients were immediately mobilized, and the other half had bed rest. To account for the possibility of improved technique by the same investigator with successive LPs, the first 25 were immediately mobilized, the next 50 were given bed rest, and the last 25 were immediately mobilized. Follow-up was for 7 days, and no significant differences were found between the two groups. Other similar studies have confirmed these findings. Oral and intravenous administration of caffeine has been recommended as a therapeutic option for post-LP headache, often as an effort to avoid using the more invasive treatment of epidural blood patching. The presumed mechanism is thought to be increased cerebral arterial vasoconstriction, resulting in decreases in cerebral blood inflow and blood volume in the brain. No well-designed, adequately powered, randomized controlled studies have been performed to prove the effectiveness of caffeine. Published information on this therapy comes from case reports or reviews that cite one study in 1975.5 The investigators in that study used a double-blind demand method to evaluate the intravenous administration of caffeine sodium benzoate in 41 patients for whom treatment with more conservative measures had failed. This study was limited in that the study size was small, it did not control for known risk factors such as sex and age, it did not include patients undergoing diagnostic LPs, and it did not investigate or correlate the quantity of daily caffeine intake before the LP was performed. In addition, the placebo arm of the study crossed over into the treatment arm. Given these limitations of the one study that is consistently cited, the evidence supporting the use of caffeine in treating post-LP headache remains weak.​weak. Table 1 Effective measures for preventing and treating post-LP headache Prevention Needle size: Halpern and Preston7 Bevel direction: Flaatten et al8 Stylet replacement: Strupp et al9 Treatment Epidural blood patch: Safa-Tisseront et al10 View it in a separate window Factors that have been shown to be associated with post-LP headache include needle size, bevel orientation, and replacement of the stylet before withdrawing the needle (box).1,7,8,9,10 Articles in the anesthesia literature have suggested that needle design is also associated, but the data in articles on diagnostic LP are conflicting and have been inadequate to assess this factor. When headache does occur, epidural blood patching had been effective in 85% to 98% of patients and is indicated for those with moderate to severe headache for more than 24 hours.2 It is performed by slowly injecting 10 to 20 mL of the patient's blood into the lumbar epidural space at the same interspace or the interspace below the previous puncture. Although it might be intuited that epidural blood patching relieves post-LP headache by tamponading the dural hole through a mass effect, the actual mechanism of action is unclear.6