Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Does Leg position alter CSF opening pressure during lumbar puncture?

Three Part Question

In [patients undergoing lumbar puncture and CSF opening pressure recordings] does [lower limb position—flexed versus extended] effect [opening pressure measurements]?

Clinical Scenario

A child with suspected benign raised intracranial pressure requires a lumbar puncture and opening pressure readings. As you prepare for the procedure the consultant on call asks you to make sure the child’s legs are straightened out before you measure the opening pressure. You wonder whether measuring CSF pressure with the lower limbs in the flexed position truly does falsely elevates the reading, and whether you should take the reading with the lower limbs in the extended position?

Search Strategy

Medline using NHS Evidence 1950–14 June 2013
(((exp SPINAL PUNCTURE/) OR ((lumbar AND puncture).ti,ab)) AND ((exp LEG/) OR ((leg AND position$).ti,ab) OR ((leg AND posture).ti,ab))) OR (exp SPINAL PUNCTURE/ AND PATIENT POSITIONING/mt [Methods])

Embase using NHS Evidence 1980–14 June 2013
(((exp LEG/ OR ((leg AND position$).ti,ab) OR ((leg AND posture).ti,ab)) AND (((lumbar AND puncture).ti,ab) OR (exp LUMBAR PUNCTURE/))) OR (exp lumbar puncture AND exp patient positioning/)

The Cochrane Library : Issue 6 of 12, June 2013
MeSH descriptor: [Spinal Puncture] explode all trees AND (MeSH descriptor: [Posture] explode all trees OR leg position, ti, ab, kw)—24 records 0 unique results.

Search Outcome

Five relevant articles were found. The citations in these articles revealed no further relevant results

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Avery et al.
53 children age >1-<19yrs (mean age 11.7 years; 60% male) undergoing LP in inpatient or outpatient setting. Single center prospective cohort studyOpening pressures in flexed and extended lower limb position; with difference between the two measured.Mean opening pressures in flexed position = 25.1 +/- 9.2cmH20; mean pressures in extended position = 24.4 +/- 8.4cm H20 (Wilcoxon sign rank test p < 0.03) Mean difference between extended minus flexed was -0.6cm +/- 2.2cm H20 (range of +6.5cmH20 to -7cmH20.Small patient numbers, single centre study.
Sithinamsuwan et al
83 patients. Inclusion criteria: > 14 years, LP indicated for diagnostic/therapeutic procedures, and good cooperation from the patient. Exclusion criteria: pregnancy, contraindication for LP, increased intra-abdominal or intra-thoracic pressure, and a marked fall in CSF pressure during measurement.Single centre prospective cohort study.Comparison of CSF opening pressure in the flexed position with that in the relaxed postionMean range of flexed opening pressures: 17.854cm H20 (3.0-45.0) and relaxed opening pressures 16.052cm H20 (2.5-38.0). Mean difference and standard error between flexed and relaxed opening pressures was 1.802 +/- 0.27mmH20 [95% CI: 1.266-2.330], paired t-test, p <0.001.Adult study, single centre.
Abbrescia et al
39 adult patients (aged 18 – 84 years, mean 39 years; 66% male) attending an urban emergency department an determined by the attending physician to require an LP during their evaluations. Exclusion criteria: contraindication to LP, haemodynamic instability, pregnancy, bleeding diathesis, & inability to assume positions being studied.Prospective crossover study: Patients randomly assigned to one of two sequences for three successive measurements of CSF pressure. Sequence A: flexed, extended, flexed. Sequence B: extended, flexed, extended.Changes in CSF opening pressures obtained when moving the lower extremities from flexed to extended, and from extended to flexed positions. Change from 1st to 3rd measurement also obtained to see if serial testing itself affected CSF pressures.CSF pressure changes little b/w position 1 & 3 (0.2cm H20 (1.7% difference) with a 95% CI = 0.9 - - 0.6cm H20). Corresponding median difference was 0%, suggesting that pressure changes little from simply performing repeated measurements. Changing from flexion to extension decreased pressure by a mean of 0.9cm H20 (2.5%) 95% CI = 2.1 - -0.1cmH20. Corresponding median difference was 0.5cmH20 (1.8%). Changing from extension to flexion increased CSF pressure by a mean of 1.1 cm H20 (6.1%) with a 95% CI = 0.2 – 2cm H20). This is a statisitically significant difference, but 2 cm H20, the maximum effect within 95% CI’s is not likely to be clinically important. The corresponding median difference was 1cm H20 (4.5%). Within 95% confidence limits, the maximum possible mean change in pressure measurements resulting from any change in lower extremity positions was 2.1cm H20. Small convenience sample of patients coming through ED.
Dinsmore et al,
12 adult neurosurgical patients requiring lumbar subarachnoid drains under anaesthetic & hyperventilated to maintain EtCO2 at 4kPa; all had undergone elective craniectomy for clipping of cerebral aneurysm or for microvascaular decompression of trigeminal nerve; (age and sex of patients not given).Single center prospective cohort study.CSF pressures & central venous pressures measured in 3 positions: (1)fully flexed (full flexion of thoracolumbar spine & both the hips and neck fully flexed), (2) flexed (hips flexed at 90 degrees to the thoracolumbar spine, but with thoracolumbar spine & neck in neutral position) & (3)straight (thoracolumbar spine, hips and neck all in the neutral position).Small & statistically insignificant change in CSF pressure when moving from straight (mean 10.3 +/_ 1.5 mmHg)to flexed position (mean 11.4 +/- 1.2mmHg); larger & statistically significant increase in CSF pressure when moving from flexed (mean 11.4 +/- 1.2mmHg) to fully flexed position (16.7 +/- 1.1mmHg) (increase of 5.3mmHg; p < 0.0001).Small size adult study/ ‘Fully flexed’ position not often employed in paediatrics
Watanabe et al
42 women aged 18-60 y undergoing gynaecological surgery under spinal anaesthesiaSingle Centre prospective cohort studyCSF pressure measured prior to spinal anaesthesia injection in right lateral decubitus position, with legs non-full-flexed to fully-flexed.In women <40 years mean increase in pressure 8 (SD 4) cmH20 moving from partially to fully flexed position In woman >40 years increase in pressure 6 (SD 6), (P <0.05 for each group)Primary purpose of study to investigate effect of opening pressure on level of anaesthesia achieved. All participants premedicated prior to LP LP performed L2/L3


LP to obtain a sample of CSF is a commonly performed procedure in paediatric practice. Measurement of CSF opening pressure is a key part of the procedure, providing important diagnostic information. Children are frequently unwilling participants, requiring holding within the flexed lateral recumbent position at the time of the procedure. Moving a child to the extended position may pose a technical challenge, particularly in the absence of sedation. Recommendations for the performance of LP often suggest measurement of CSF opening pressure with the legs in the extended position, due to concerns that pressure may be falsely elevated due to raised intra-abdominal pressure.

We were able to find only five studies that compared the measurement of CSF opening pressure with legs flexed and extended (Watanabe et al, 1991; Dinsmore et al, 1998; Abbrescia et al, 2001; Sithinamsuwan et al, 2008; Avery et al, 2010), with only one of these studies performed in children (Avery et al, 2010). All five studies were prospectively performed in single centres, and included a total of 229 subjects. In four of these studies a small increase in CSF pressure was found with the legs in a flexed compared to extended position, with the mean differences in pressure less than 2 cm H2O. In most of the studies identified, some patients had increased pressure in the extended compared to the flexed positions, contrary to group trends.

Although the total differences in pressure were small, the differences may have a diagnostic effect depending on the cut-off value chosen to define a raised pressure; for example, one study used a cut-off value of 18 cm H2O to define a high pressure, although other studies suggest normal values for CSF opening pressure in a flexed lateral decubitus position in children of 10–28 cm H2O (Ellis, 1994), with other authors concluding that opening pressures greater than 20 cm H2O should not necessarily be considered abnormal (Lee and Vedanarayanan, 2011).

Clinical Bottom Line

CSF opening pressure measurement with legs in a flexed compared to extended position usually results in a very small increase in measurement that is unlikely to be of clinical significance. Altering leg position from straight to flexed for an individual may either increase or decrease opening pressure, and so if serial recordings are required a consistent leg position should be adopted.


  1. Avery RA, Mistry RD, Shah SS, et al. Patient position during lumbar puncture has no meaningful effect on cerebrospinal fluid opening pressure in children. J Child Neurol. 2010;25:616-9.
  2. Sithinamsuwan P, Sithinamsuwan N, Tejavanija S et al. The effect of whole body position on lumbar cerebrospinal fluid opening pressure. Cerebrospinal Fluid Res 2008;5:11.
  3. Abbrescia KL, Brabson TA, Dalsey WC, et al. The effect of lower-extremity position on cerebrospinal fluid pressures. Acad Emerg Med 2001;8:8-12.
  4. Dinsmore J, Bacon RC, Hollway TE. The effect of increasing degrees of spinal flexion on cerebrospinal fluid pressure. Anaesthesia. 1998;53:431-4.
  5. Watanabe S, Yamaguchi H, Ishizawa Y. Level of spinal anesthesia can be predicted by the cerebrospinal fluid pressure difference between full-flexed and non-full-flexed lateral position. Anesth Analg. 1991;73:391-3.
  6. Ellis R Lumbar cerebrospinal fluid opening pressure measured in a flexed lateral decubitus position in children. Pediatrics. 1994;93:622-3.
  7. Lee MW, Vedanarayanan VV. Cerebrospinal fluid opening pressure in children: experience in a controlled setting. Pediatric neurology. 2011;45:238-40.
  8. Haslam R. Neurologic Evaluation. Nelson Textbook of Pediatrics. 2004.