| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Austin Health, The University of Melbourne, Studley
Rd., Heidelberg, Melbourne, Victoria 3084, Australia.
2 Department of Epidemiology and Preventive Medicine, Monash University,
Melbourne, Australia.
Received March 23, 2004;
accepted after revision August 24, 2004.
Supported by The Royal Australian and New Zealand College of Radiologists
(RANZCR).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1. InstructionsReferences |
|---|
MATERIALS AND METHODS. Ten operators with a range of clinical biopsy experience were tested. Each operator was given two phantoms containing four targets. Each target was biopsied with both biopsy methods. Data collected included procedure time and sample quality in the biopsy specimen. Analyses were stratified by level of operator experience.
RESULTS. Median biopsy time was 23 sec with the ultrasound probe guide and 32 sec freehand. Differences between probe-guided and freehand pairs of measurements approximately followed a normal distribution. The mean time difference between probe-guided and freehand times to complete biopsy was –20 sec (95% confidence interval, –35 to –5 sec; p = 0.01). Analysis of sample quality across all operators showed no difference.
CONCLUSION. The ultrasound probe–guided technique of sonographic biopsy could be used in a complex phantom model, and there was a statistically significant time benefit with the use of probe guides compared with the freehand biopsy technique. This benefit was greatest for inexperienced operators. There was no difference in sample quality between the probe-guided and freehand techniques.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1. InstructionsReferences |
|---|
The freehand technique requires the operator to manipulate the ultrasound probe with one hand and the biopsy needle with the other. Its chief advantage is its versatility. The probe and needle can be positioned independently to achieve the best image of the lesion and a needle path free of intervening structures. To monitor the needle tip on its course to the target, the operator must maintain the needle within the plane of the ultrasound beam.
The biopsy-guide technique uses a needle guide fixed to the ultrasound probe. Its main advantage is that it keeps the needle within the plane of the sonographic image as it is advanced toward the biopsy target. However, the fixed relationship between the needle and probe reduces operator freedom in choosing a needle path. Thus, some people believe that the probe-guided technique cannot be used for more difficult biopsies.
Although many experienced radiologists use the freehand technique when performing biopsies, proponents of the biopsy-guide technique suggest that biopsy guides make the procedure of sonographically guided biopsy easier and quicker and can be used in almost all cases [1–3].
This study compares probe-guided and freehand biopsy techniques in a phantom designed to simulate focal lesions in the human liver. The following hypotheses were tested: that sample quality is improved using the probe-guided technique, that biopsy time is reduced using the probe guide, that the benefits of the probe guide are independent of operator experience, and that the probe guide can be used in a complex phantom with simulated ribs.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1. InstructionsReferences |
|---|
|
CT shows the physical design of the phantom [4] (Fig. 2). Two standard phantoms were scanned on CT to obtain average dimensions. The bowl in which the phantom was made was hemispherical in shape with flat superior and inferior surfaces. The total depth of the phantom was 8 cm. The flat circular superior surface measured 10.1 cm in diameter, and the inferior surface measured 23.2 cm in diameter. The superficial echogenic layer measured 19 mm in depth, and below this, four ribs were positioned roughly parallel to each other. The ribs measured 110 x 12 x 3 mm (length x width x depth) with the mean interspace between the ribs of 18.2 mm (range, 15.5–22.7 mm). The olives were positioned at a mean depth of 38 mm (range, 36–39.7 mm) below the surface. The olives were ovoid with average outer measurements of 17.25 x 20.84 x 18.3 mm. The inner red capsicum measured 9.5 x 15.2 x 9.8 mm, comprising 21.5% of the volume of the target.
|
One olive was positioned in each quadrant of the phantom with slight variation in the position of each olive within each quadrant and between each phantom. The distances from the olive to the periphery of the phantom varied between 23.4 and 76.3 mm. Each olive was biopsied with both techniques (Figs. 3, 4, 5, 6) to yield paired results to account for variability in difficulty among targets.
|
|
|
|
Sonography was performed on an ATL HDI 3000 unit (Philips) using a 4–2-MHz curved array. The biopsy guide was the standard fixed-angle guide supplied for use with the ATL curved array probes and consisted of a reusable plastic bracket and disposable plastic needle guide. Biopsies were performed with an 18-gauge needle (Quickcore, Cook).
Operators with a range of clinical biopsy experience were selected from the radiology department medical staff. Because this study was a single-center trial, only 10 operators were available: three novice operators who had performed fewer than 10 clinical biopsies; three operators of intermediate experience who had performed 10–50 clinical biopsies; and four experienced operators who had done more than 50 clinical biopsies. None of the operators was a frequent user of the probe guide.
Each operator was given two phantoms with four targets in each phantom. A standard typed instruction sheet was provided (Appendix 1) on which participants were asked to biopsy the central part of the olive (the red capsicum). Operators were asked to scan the phantoms to become familiar with the appearance and location of the targets. They were asked to alternate biopsy techniques in groups of four (e.g., four freehand then four probe-guided biopsies) in one phantom then to repeat the procedure in the second phantom starting with the alternate technique—for example, four probe-guided and then four freehand biopsies. Each operator performed all biopsies on both phantoms during the same session. Consecutive operators alternated which technique they used first. The time for each biopsy was measured from the time the needle (Quickcore) was handed to the operator to the time it was fired. Data collected for each biopsy included procedure time and sample quality as assessed by the weight of the red component (capsicum) and the weight of green component (olive) in the biopsy specimen.
After participating in the trial, each operator filled out a questionnaire. This addressed the use of the probe-guided and freehand biopsy techniques before and after the trial, to assess changes in their clinical use of probe guides after participation in the trial.
Statistical Methods
The design of the experiment gave rise to pairs of measurements from
freehand and probe-guided techniques of a given target from a given phantom
for a given operator. The separate sets of probe-guided and freehand
measurements followed distributions with varying degrees of skewness. The
differences between these intraoperator and intratarget pairs of measurements
were used as the outcome variable in the statistical analyses. Despite the
skewness in the individual sets of measurements, the distribution of these
differences approximately followed a normal distribution; hence, we used
t tests and analysis of variance to analyze mean differences.
Analyses were stratified by the operator's level of experience according to
the following three groups: novice, fewer than 10 clinical biopsies;
intermediate, 10–50 clinical biopsies; and experienced, more than 50
clinical biopsies.
Eight differences in the freehand versus probe-guided biopsies were observed for each operator who evaluated four targets in each of two phantoms. We made allowance for the nonindependence of these repeated measurements using robust standard errors (calculated using the information sandwich formula) for confidence intervals and hypothesis testing. A statistical software package (Stata 7.0 [2001], Stata) was used to perform all analyses.
The ratio of probe-guide time to freehand time was plotted [7] against average time (averaged over probe-guided and freehand biopsies using a geometric mean to account for skewness). A Pearson's correlation coefficient was used to summarize this analysis of whether performance (as judged by time taken to biopsy) with the probe guide was comparable to that of freehand biopsy over a range of target difficulties (with difficulty represented by average time to biopsy).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1. InstructionsReferences |
|---|
|
Despite the decreased time taken to perform biopsies with the probe-guided technique, no gains were seen when comparing the accuracy of the biopsies themselves. Analyses of sample quality across all operators showed little difference in weight of red or red plus green components. These data are shown in Tables 2 and 3.
|
|
Figure 7 shows that as average time for biopsy increased, there was no trend toward decreasing advantage to the probe guide (Pearson's correlation coefficient, r = –0.06; p = 0.61).
|
Interviews
None of the operators was a regular users of probe guides before the study.
Two operators, both from the experienced category, used probe guides
occasionally. The rest had never used probe guides before the study. After
participating in the study, nine reported improved confidence in the use of
probe guides. Four reported improved confidence in the freehand technique (all
three novice operators and one operator of intermediate experience).
Despite a shown time reduction with the probe-guided technique, few operators were influenced in subsequent clinical work by participation in the study. Three operators, one from each experience level, reported a slight shift toward the use of probe guides, having never used probe guides before the study to becoming occasional users. Reasons given for the lack of change in the remaining seven operators included lack of availability (in four operators), unfamiliarity (one), time saved was not clinically significant (two), and probe guide made needle position difficult to adjust (one).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1. InstructionsReferences |
|---|
A time saving of 20 sec with using the probe guide is insignificant relative to the time that a biopsy takes in clinical practice. However, one must be mindful of the goals of performing biopsies in patients with focal lesions in solid organs, which are to obtain an adequate sample to enable diagnosis and to perform the biopsy in a manner that is both safe and comfortable for the patient. During the time taken for a clinical biopsy, the needle is constantly being manipulated and repositioned within the target organ. We postulate that the reduced procedure time from guided biopsy will correlate with reduced needle manipulation and hence less tissue trauma and less patient discomfort. This may be greater motivation to use a probe-guided technique rather than the total time saved. Biopsy in human subjects is more complex than in phantoms, and the actual time saved may be greater, particularly if the operator is inexperienced or the target is in a difficult location. In addition, probe guides may help in planning the biopsy.
Our experiment was designed to compare the freehand and probe-guided methods of sonographic biopsy. Nyman et al. [8] partly addressed this issue in a study comparing the yield and complications of sonographically guided gun biopsies and manual biopsy techniques in percutaneous renal biopsy. The sonographically guided group was further divided into freehand and probe-guided techniques. No significant difference in sample material was found between these two subgroups. This concurs with our findings.
A number of authors have expressed opinions about methods of sonographically guided biopsy [1–3]. Esola et al. [2] described visualizing the needle and directing the needle to the lesion as being the most difficult aspects of a sonographically guided procedure. We believe that a probe guide assists with both of these difficulties. There are other factors that have been proven to aid in the visualization of the needle in sonographically guided procedures. These include the use of polymeric-coated needles [9, 10] and larger angles of insonation [11]. These factors were not specifically addressed in our study. The probe guide used in our study had a relatively low angle of insonation (30°) that was fixed. This did not appear to adversely affect needle visualization or biopsy results, although the future design of probe guides should consider the improved needle visualization that is gained from larger angles of insonation.
Some researchers believe that the probe-guided technique is cumbersome and difficult because of the fixed angle of the needle guide [1]. We showed, however, that the probe guide can be used despite the presence of "ribs." Figure 7 suggests that for more difficult lesions—that is, those that take more time to biopsy, there is as much of an advantage to using the probe guide, if not even more, as found overall. In fact, the seven longest biopsy times were all performed freehand and came from a cross section of operators.
The significance of operator experience in diagnostic accuracy of sonographically guided renal biopsy was assessed by Elvin et al. [12], who found no statistically significant difference in sample quality between operators of differing experience provided detailed instructions were given to the operators, in keeping with our results. Other studies have found the number of passes decreases as operator experience increases [13, 14].
In designing our study, we thought that a complex phantom was the best way to directly compare the probe-guided and freehand biopsy techniques, but it is difficult to know how the results will translate to human subjects. Therefore, potential limitations of our study include the design of the phantom. Although there were simulated ribs to increase the difficulty of the biopsy, in the real-life setting there are often additional structures that need to be avoided such as blood vessels in the liver. In addition, biopsies of lesions in the clinical setting are of varying difficulty. For example, the biopsy of small lesions superiorly in the right lobe of the liver may be more technically difficult than those presented in our phantom. Biopsy in clinical practice is also made more difficult because of patient motion as a result of respiration or discomfort.
Our study results have indicated factors that are likely to limit the use of currently available probe guides by radiologists. The issues identified in this study are availability, familiarity, and ability to adjust needle position while using the probe guide. We are now developing a probe guide that we hope will address these shortcomings.
Benefits for participants in this trial included increased confidence in the use of both freehand and probe-guided techniques. This benefit was greatest for inexperienced operators. We also found this complex phantom to be a useful training tool for sonographically guided biopsy and biopsy performed under CT and MRI guidance [4].
In conclusion, the probe-guided technique of sonographic biopsy was successfully used in a complex phantom model, and there was a statistically significant time benefit with the use of probe guides compared with the freehand biopsy technique. This benefit was greatest for the inexperienced operators. There was no difference in the sample quality between the probe-guided and freehand techniques.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1. InstructionsReferences |
|---|
Biopsies will be performed on an ultrasound phantom made of corn flour jelly and stuffed olives. The top layer is echogenic and contains ribs (coffee stirrers). The next layer contains stuffed olives that appear as target lesions under ultrasound.
The aim is to biopsy the central part of the olive (the red capsicum). The biopsy phantom is hemispherical in shape with a flat working surface superiorly. Biopsies can only be performed from the superior flat surface.
Each phantom contains four targets that will be labeled. The lesions will usually be placed in four quadrants of the phantom. You will be allowed time to scan the phantom to familiarize yourself with the target and the ribs. One attempt at each target with both the free-hand and probe-guided methods is permitted. A total of 16 biopsies will be performed by each person (two phantoms with four lesions each biopsied by the two biopsy techniques).
Each biopsy will be timed from when you are handed the needle to when the needle is fired. In addition, we will assess the efficacy of each biopsy by weighing the separated red and green components of the biopsy specimen.
Acknowledgments
We thank Jennifer Hollaway for clerical assistance and production of
phantoms; Sumith Nawaratne, Austin & Repatriation Medical Centre; and The
Biostatistics Consulting Service, Department of Epidemiology and Preventive
Medicine, Monash University for assistance with statistical analysis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1. InstructionsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
