VARIATIONS IN THE FLOW OF CEREBROSPINAL FLUID THROUGH SPINAL NEEDLES

Br. J. Anaesth. (1987), 59, 1465-1471

VARIATIONS IN THE FLOW OF CEREBROSPINAL FLUID THROUGH SPINAL NEEDLES S. P. GERRISH AND J. E. PEACOCK

SUMMARY The dimensions of currently available spinal needles (0.7mm (22-gauge) to 0.45mm (26gauge) external diameter) were measured and compared with the International Standard (ISO 7864-1984 E). The in vitro performance of the needles was assessed by comparing times to appearance of cerebrospinal fluid and flow rates through the needles. There were wide variations in the measurements and performance of the different needles.

British Standard Wire Gauge (BSWG) as its designated measurement. We have used metric units, and given the old British Standard size in parentheses. MATERIALS AND METHODS

Fourteen different types of spinal needle, with external diameters in the range 0.7-0.45 mm (2226-gauge), which are currently available in the U.K., were assessed (table I). The internal and external diameters of 10 needles of each type were measured before they were divided into two groups of five to measure CSF flow at either high or low pressure. All needles were of a standard 90-mm length. TABLE I. List of needles studied. O = opaque hub; C • 1 clear hub

S. P. GERJtISH,* M.B. B.CH., F.F.A.R.C.S.; J. E. PEACOCK.t M.B. CH.B., F.F.A.R.C.S. ; Department of Anaesthetics, Northern

Manufacturer

Genera] Hospital, Herrics Road, Sheffield S5 7AU. Accepted for Publication: May 7, 1987. Present addresses: *Dcpartment of Anaesthetics, Royal Hallamshire Hospital, Glossop Road, Sheffield S10 2JF. tUniversity Department of Anaesthetics, Sheffield University Medical School, Beech Hill Road, Sheffield S10 2JF.

Steriseal BD Monoject Spinocan Vygon Everett

External diameters (mm) 0.7 (O) 0.7 (C) 0.7 (O)

0.5 (O) 0.5 (O&C) 0.5 (O)

0.7(0 0.7(0

0.5(0 0.5(0

0.7 (O)

0.45 (O) 0.45 ( O

Downloaded from http://bja.oxfordjournals.org/ at University of Michigan on June 25, 2015

The use of intrathecal drugs to produce either analgesia or anaesthesia has been limited, in part, by the problem of post-spinal headache. The headache, which may be disabling, has discouraged the use of "spinal anaesthesia", especially in younger patients in whom the incidence is said to be greater. However, the introduction of fine gauge needles has decreased the risk to less than 1 % with 0.5-mm (25-gauge) and 0.45-mm (26-gauge) needles (Bonica, 1970). One problem in using fine gauge needles (especially for the inexperienced operator) is visualization of the free-flow of cerebrospinal fluid (CSF) to confirm the position of the needle tip in the subarachnoid space. The only previous study which examined the in vitro performance of spinal needles (Messahel, Robinson and Mathews, 1983) found considerable differences in flow through two types of 0.5-mm (25-gauge) needle. We have related the dimensions of some currently available spinal needles (0.7-, 0.5- and 0.45-mm; 22-, 25- and 26-gauge) to CSF flow at "high" and "low" pressures (equivalent to the sitting and lying positions for lumbar puncture). We have used the term "time to appearance of CSF" in an attempt to quantify how long the operator may expect to wait, under ideal conditions, before 'seeing a fluid meniscus within the needle hub. In 1984 an International Standard for hypodermic needles was published and defined specifications in terms of metric measurements. This replaced the British Standard of 1976 which used

1466

BRITISH JOURNAL OF ANAESTHESIA

appearance times. These results were then used to calculate the mean flows for the individual needles. Five results were obtained for each type Measurements were made at the tip (across the bevel) using a measuring microscope incorpor- of needle. Analysis of variance was used to compare data ating a micrometer screw-gauge. The internal diameter was taken as the difference between two from different groups of needles. Tukey's Honreadings made in a plane perpendicular to the estly Significant Difference test was used to detect length of the needle and tangential to the curves of which pairs of needles within the groups were the the inner wall of the needle. The external diameter cause of any significant difference. The Mannwas measured in a similar way, using the outer Whitney U test was used to compare individual pairs of needle samples. Differences were consurface of the bevel. sidered statistically significant when P < 0.05. Two-tailed P values were obtained for the CSF flow An in vitro model was used to measure CSF Mann-Whitney U test. flow. A fluid filled reservoir was connected by large bore tubing and a three-way tap to a selfRESULTS sealing bung for needle insertion. A simple liquid manometer was attached to the third port of the The results of the viscosity measurements perthree-way tap. The pressure in the system was formed on the different fluids are listed in table altered by adjusting the height of the reservoir so II. that the reading on the manometer was either 12 or 50 cm of CSF, and thus equivalent to the lying Needle dimensions There were major differences in dimensions or the sitting position. The system was filled with an artificial CSF between the needle types (fig. 1). The mean which had a chemical content and viscosity similar external diameter of the 0.7-mm Spinocan needle to that of real CSF and physiological temperature was significantly smaller than that of the Vygon (37 °C) was maintained by passing the connecting and BD needles (P < 0.05). The mean internal diameters of the 0.7-mm tubing through a water bath. Viscosities were measured using a Carrl-med controlled stress needles showed much larger differences between needle types. The mean internal diameter of the rheometer. The time to appearance of CSF was measured Vygon needles was significantly larger than all by inserting the needle through the bung and the other 0.7-mm needles (P < 0.01) and the starting a stop-watch as the stylet was withdrawn. Steriseal internal diameter was larger than the The same two observers were used throughout Monoject, Everett (P < 0.05) and Spinocan (P < the study to look down into the hub, or through 0.01) needles. The results for the 0.5-mm needles showed that the hub when it was transparent, until the CSF was first seen—at which time the watch was the Spinocan needles were of significantly greater stopped. This first evidence of CSF within the external diameter than all the other needles (P < hub—whether viewed through a clear hub from 0.05) except for the Monoject. The diameter of above or looking into an opaque hub—was defined the Monoject needles tended to be larger than the as the "time to appearance of CSF". Five results were obtained for each needle type. TABLB II. Different fluid viscosities at different shear stress After the appearance time had been recorded, the CSF was collected and weighed continuously Viscosity (mPa s"1) on an electronic balance. Readings were taken at Artificial 1-min intervals for 5 min for 0.7-mm (22-gauge) Shear1 stress CSF CSF Saline needles and 8 min for 0.45-mm and 0.5-mm (26- (Pa" ) and 25-gauge) needles. Flow volumes at 1-min 0.82 0.84 0.76 0.88 intervals were obtained for each needle by sub1.69 0.84 0.81 0.76 tracting consecutive results. This produced four 4.21 0.78 0.81 0.76 8.43 0.78 0.80 0.76 results for the larger, and seven for the smaller, 10.11 0.76 0.78 0.80 needles because we discarded the result for the 12.64 0.78 0.79 0.76 first 1 min to avoid differences attributable to Dimensions

Downloaded from http://bja.oxfordjournals.org/ at University of Michigan on June 25, 2015

CSF FLOW THROUGH SPINAL NEEDLES ISO range of tolerance for external diameters.

ISO minimum internal diameters. *

1467

*

0.45 4 0.5 mm

0.7 mm

0.45 mm

0.7 mm

0.5 mm

•—** '

*

Vygon 0.7mm

'

B.D. 0.7mm

1

"**~1

'

*

'

1

+

•

Steriseal 0.7mm Monoject 0.7mm Everett 0.7mm

"-*- 1 •*—'

Spinocan 0.7mm

Vygon 0.5mm Spinocan 0.5mm

•—o>

Steriseal 0.5mm Monoject 0.5mm B.D. 0.5mm (clear (opaque hubs) B.D. 0.45mm Steriseal 0.45mm

0.24

0.28

0.32

0.36

0.40

'

°

«

•—'

a44

0.48

a52

0.56

0.60

0.64

0.68

0.72

0.76

Needle diameters (mm) FIG. 1. Internal (A) and external (O) needle diameters (mean and range) of the different needles tested with ISO specifications.

remaining needles, but the differences did not reach statistical significance. The external diameters of the 0.45-mm needles did not differ significantly. The internal diameters of the 0.5-mm and 0.45mm needles are considered together, because the ISO specifications for their internal diameters are identical. The Vygon 0.5-mm needle had a significantly larger mean internal diameter than all the other needles in the group (P < 0.01). The Spinocan needles were of significantly larger diameter than the remaining needles (P < 0.05, except for the Steriseal 0.45-mm needle: P < 0.01).

pressure than the larger needles with opaque hubs, such as Monoject, Everett and Steriseal 0.7mm needles (P < 0.001 at the lower pressure). Using pooled data, needles with clear hubs all had significantly shorter appearance times than needles of the same specified diameter with opaque hubs (P < 0.001). A direct comparison of the two forms of the BD 0.5-mm needles showed a significant difference between the two needles (P < 0.001).

CSF flow The results are summarized in figure 3. There were no significant differences between the flows of individual needles of the same type at the same Time to appearance of CSF pressure. Each needle type had a significantly The results are summarized in figure 2. Delay greater flow at the higher pressure (P < 0.001). in appearance times occurred mainly in the The Vygon 0.5-mm needle had significantly faster smaller diameter needles at the lower pressure. All flows than all the other 0.5-mm needles at both the needles had a significantly shorter appearance pressures (P < 0.01) and the Spinocan 0.5-mm time at the higher pressure (P < 0.001, except had faster flows than all the remaining 0.5-mm the Spinocan: P < 0.01). Some smaller diameter needles at both pressures (P < 0.01). The Vygon needles with clear hubs, for example the Vygon 0.7-mm needle also had greater flows at both 0.5-mm, had shorter appearance times at the same pressures than the other 0.7-mm needles (P <

Downloaded from http://bja.oxfordjournals.org/ at University of Michigan on June 25, 2015

*•

•—°-"

BRITISH JOURNAL OF ANAESTHESIA

1468

'

°

'

Steriseal 0.45mm ( o )

Monojed 0.5mm (o)

Steriseal 0.5mm (o)

B.O. 0.5mm (o) •<>• —»*

* M l

-°-1

B.D. 0.5mm (c)

B.D. 0.45mm (c)

—o- 1

Steriseal 0.7mm (o)

-°—'

Everett 0.7mm (ol

Monojed 0.7mm (o)

Spinocan 0.5mm (c) B.D. 0.7mm -o-

*

tt

Vygon 0.5mm (c)

Spinocan 0.7mm (cl

Vygon 0.7mm (c)

10

15

20

30

40

65

60

70

Time to appearance of CSF (s)

FIG. 2. Time to appearance of CSF (mean and range) measured at 50 cm CSF (A) and 12 cm CSF (CO pressure for the different needle types.

Vygon 0.7mm '

»

»

•

Steriseal 0.7mm

°

'

Monoject 0.7mm

*—'

H

—

Everett 0.7mm

Spinocan 0.7mm •—O-"

•-*»-«

B.D. 0.7mm

*

Vygon 0.5mm Spinocan 0,5mm

Steriseal 0.5mm Monojed 0.5mm B.D. 0.5mm (clear) B.D. 0.5mm (opaque)

>

B.D. 0.45am Steriseal 0.45mra

"•

O.5

1.0

1.5

2.0

3.5

2.5

4.0

4.5

Flow (ml s~1) at 50 cm CSF pressure

[ O

i O.1

i O.2

l 0.3

l O.4

l O.5

l 0.6

i O.7

i 0.8

• 0.9

l 1.0

l 1.1

Flow (ml s"1) at 12 cm CSF pressure

FIG. 3. Flow related to needle types (mean and range) at 12 cm CSF (O) and 50 cm CSF (A) pressure.

Downloaded from http://bja.oxfordjournals.org/ at University of Michigan on June 25, 2015

•

*-*• | - ° *

CSF FLOW THROUGH SPINAL NEEDLES

1469

0.01), but the 0.7-mm Spinocan had significantly slower flows than all the other 0.7-mm needles at both pressures (P < 0.01).

TABLE III. Comparison of external diameter measurements using ISO and SWG standards, and ISO tolerances for internal and external diameters Internationa] Standard: nominal external diameter (mm)

Equivalent standard wire gauge (SWG) Metric conversion of SWG (mm) ISO tolerances (mm) External diameter Minimum Maximum Minimum internal diameter (mm)

0.45

0.5

0.6

0.7

0.9

26

25

23

22

20

0.457

0.508

0.610

0.711

0.914

0.44 0.47

0.50 0.53

0.62 0.65

0.70 0.73

0.86 0.92

0.24

0.24

0.32

0.39

0.56

Downloaded from http://bja.oxfordjournals.org/ at University of Michigan on June 25, 2015

ments which may produce a doubling of the error of each individual measurement. Hence an accuracy of 0.005 mm in each individual measurement could produce an error of 0.01 mm for any needle diameter. All the mean external diameters DISCUSSION measured were within 0.01 mm of the quoted We have compared the dimensions and in vitro standard tolerance. However, the mean diameter flow rates of different spinal needles using artificial of the Spinocan 0.5-mm needles was larger than CSF and a system that would mimic the clinical the tolerances allowed by the International Standard. Messahel, Robinson and Mathews (1983) situation. The viscosity measurements (table II) showed showed that the Spinocan 0.5-mm needles were that, while saline is a Newtonian fluid, the two larger than the BD 0.5-mm needles, which is forms of CSF are not. Although the artificial consistent with the results that we have obtained. preparation is not identical to real CSF, its The mean external diameter of the Spinocan 0.7mm needle was smaller than the tolerance quoted viscosity is closer to it than is saline. Variation in the diameters of needles from in the International Standard. Although individdifferent manufacturers was documented by Mes- ual needles from the Monoject 0.7-mm and 0.5sahel, Robinson and Mathews (1983) when they mm and the BD 0.45-mm samples had measurecompared the BD and Spinocan 25-gauge needles. ments greater than the permitted tolerance, the They suggested that the differences were the results were not statistically significant and the result of the manufacturers using different stand- mean values of the needle samples all lay within ards and they thought that an International the range of tolerance. Similarly, the Steriseal 0.7Standard should be produced to avoid this mm sample had a measurement smaller than the problem. An International Standard is now quoted tolerance, but the mean result lay within available (ISO 7864-1984 E) and defines the the range of tolerance. specifications in terms of metric equivalents of the The differences in external diameters demonpreviously used gauge sizes. The differences strated in the present study can be explained in between the last British Standard (BS 5081 1976) one of two ways. Either the manufacturer is and the International Standard are shown in table working to an outdated standard now that the III. International Standard is available, or the manuAny conclusions drawn from our results must facturers' specifications or tolerances do not take into account two additional factors: the match those of the International Standard. permitted tolerance within the standard used and Most needles are supplied with the needle size the accuracy of our measurements. The diameters marked on the outer packaging. This may be in were obtained by subtraction of two measure- metric or BSWG units. On the basis of our results

1470

may be to aspirate gently as advised by Slattery, Rosen and Rees (1980). The additional advantage of both these methods is that the colour of the fluid can be identified to help avoid the possibility of i.v. injection. The shorter appearance times with the clear hub needles is effectively a perceptual phenomenon, as can be seen from the comparison of CSF flow in the two types of 0.5mm BD needles, where there was no significant difference (P > 0.05). The use of such needles may aid the inexperienced operator and avoid unnecessary multiple attempts at dural puncture. The differences in CSF flow that are documented would appear to be related to the internal diameter, because the order of ranking for the internal diameters is similar to that for the rates of CSF flow (table IV). However, simple calculation (assuming laminar flow) using the product of measuredflowand the fourth power of the internal diameter, which should be constant for the same fluid and pressure gradient, showed that this was not the complete picture. Messahel, Robinson and Mathews (1983) found considerable differences in the finish of the internal surface of the two needle types they examined and suggested that this may have a marked effect on fluid flow. Although we did not perform longitudinal sections in this study, we found, on microscopic examination of the needles for measurement purposes, that some needles had a relatively smooth finish and others had a considerably rougher finish on the internal surface. Although the major factor in CSF flow is the internal diameter of the needle, the quality of finish of die internal surface may also be important, since it may result in non-laminar flow. This may explain the differences in flow between certain needles, for example 0.7-mm BD and Steriseal needles, where the ranking for flow and internal diameter is reversed. In conclusion, the study confirms and extends the previous work of Messahel, Robinson and Mathews (1983) and shows the considerable differences that exist between the various needles available. In the hands of an experienced anaesthetist, it would probably make very little difference which needle was used, but for the novice the help given by using the sitting position and being aware of the potential differences in needle performance may help decrease the morbidity associated with dural puncture.

Downloaded from http://bja.oxfordjournals.org/ at University of Michigan on June 25, 2015

it would seem possible that an operator could use a needle larger or smaller than that consistent with the International Standard. This may be the result of different manufacturing standards or individual needle variation. Since spinal headaches are related to needle size, this may produce increased morbidity, since the needle would be larger than anticipated. The differences in internal needle diameter between the different needle types are of less consequence when related to the International Standard, because only a minimum diameter is quoted and not a range, as for external diameter. All the needles measured exceeded this value. Although related to the external diameter of the needle, the internal diameter is also related to the thickness of the needle wall. The manufacturer must balance the increased flow with a thinner wall needle against a more rigid needle with a thicker wall. Only one manufacturer seems to have moved towards creating a thin-wall needle. Vygon needles had significantly larger internal diameters than other needles of the same specified external diameter. The 0.5-mm Spinocan also had a larger internal diameter, but this would seem to be related to its larger external diameter. Each needle was used only once because reinsertion of the stylet may produce metal shards. In addition, examination of used samples revealed the presence of crystalline material within the needle and this might have produced erroneous results with re-use. The material was thought to be a deposit of substances such as protein from the artificial CSF used in the study. It was most noticeable after the liquid had dried out, but we felt that differences would occur even after a single use. In consequence, repeated measurements were not made on individual needles and tests could not be applied to look for variation in appearance times within an individual needle. By weighing the CSF at 1-min intervals, we obtained a series of CSF flow rates for each needle and these were used to assess differences in CSF flows within needle types. The results showed the prolonged period of time (up to 66 s) which may elapse before a fluid meniscus is seen when small diameter needles are used. This was particularly noticeable in needles with opaque hubs at the lower pressure. Clear hub needles allowed a shorter appearance time and may be useful for more rapid assessment of successful dural puncture. An alternative method

BRITISH JOURNAL OF ANAESTHESIA

CSF FLOW THROUGH SPINAL NEEDLES

1471

TABLE IV. Rankings and measurements for internal diameters and flows at 12 cm CSF pressure for all needles

Flow at 12 cm CSF pressure

Internal diameter Needle

Size (mm)

(ml s"1)

Rank

1 2 3 4 5 6 7 8 9 10 11 12 13

Vygon 0.7 Steriseal 0.7 BD0.7 Monoject 0.7 Everett 0.7 Spinocan 0.7 Vygon 0.5 Spinocan 0.5 Steriseal 0.5 BD0.5 BD 0.45 Monoject 0.5 Steriseal 0.45

0.492 0.448 0.440 0.420 0.420 0.412 0.344 0.316 0.288 0.288 0.288 0.284 0.272

0.986 0.588 0.679 0.597 0.591 0.501 0.226 0.148 0.122 0.116 0.082 0.103 0.084

1 5 2 3 4 6 7 8 9 10 13 11 12

British Standards Institution (1976). Sterile hypodermic syringes and needles for single use. British Standard B.S.:5081. International Standards Institution (1984). Sterile hypodermic needles for single use. International Standard ISO 78641984(E). Messahel, F. M., Robinson, J. S., and Mathews, E. T. (1983). Factors affecting cerebrospinal fluid flow in two spinal REFERENCES needles. Br. J. Anaesih., 55, 169. Slattery, P. J., Rosen, M., and Rees, G. A. D. (1980). An aid Bonica, J. J. (1970). Obstetrical Anaesthesia: Current Concepts and Practise, p. 175. Baltimore: The Williams and Wilkins to identification of the subarachnoid space with a 25G Co. needle. Anaesthesia, 35, 391.

ACKNOWLEDGEMENTS We are grateful to Mr T. Moore for the preparation of the artificial CSF, to Dr D. Northcliffe for measuring the viscosities and to Dr J. Alderson and Professor W. Nimmo for their help and advice.

Downloaded from http://bja.oxfordjournals.org/ at University of Michigan on June 25, 2015

Rank