| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of Pediatrics, Obstetrics and Gynecology (M.H., W.W.K.K.), Wayne State University, Detroit, Michigan
Computing and Information Technology (E.M.H.), Wayne State University, Detroit, Michigan
Neonatal Unit and Human Nutrition Research Center, University of Lyon, Lyon, FRANCE (J.-C.P.)
Neonatal Unit, University Children’s Hospital, Greifswald, GERMANY (C.F.)
Neonatal Unit, University of Liege, Liege, BELGIUM (J.R.)
Address reprint requests to: Dr. Winston Koo, Department of Pediatrics, Hutzel Hospital, 4707 St Antoine Blvd, Detroit, MI 48201. E-mail: wkoo{at}wayne.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSCONCLUSIONREFERENCES |
|---|
Methods: A set of four phantoms with total weights 1520g, 3140g, 4650g and 7490g were made with low cost and easily available materials. Each phantom was made from assembling polyethylene bottles (100 to 1000 mL) filled with either pure olive oil or electrolyte solution in different combinations, and borosilicate tubes (3 and 5 mL) and flexible polypropylene tubing filled with calcium carbonate. Triplicate measurements of each of the four phantoms were performed with three pencil beam densitometers made by the same manufacturer (Hologic Inc., Waltham, MA): two QDR 2000 (University of Liege, Liege, Belgium, and Wayne State University, Detroit, Michigan) and a QDR1500 (University Children’s Hospital, Greifswald, Germany) using infant whole body-scanning mode and analyzed with software V5.73P.
Results: DXA measured total weight, or bone, lean and fat masses, from one center were highly predictive of DXA measurements from the other centers with an adjusted r2 of 0.94 to 1.00, p < 0.001. This was the case whether the measurements from single scan or from average of triplicate scans were used in the analysis.
Conclusions: Systematic corrections, in the form of linear transformations, are possible to allow comparison of clinical data generated from different centers. Different size phantoms can be made to accommodate the varying range of weights and body composition of study subjects.
Key words: phantom, body composition, fat, lean, bone, infant
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSCONCLUSIONREFERENCES |
|---|
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSCONCLUSIONREFERENCES |
|---|
DXA Scans
Three densitometers from three centers were assessed in this study. All densitometers were from the same manufacturer (Hologic Inc., Waltham, MA): two QDR 2000 (one located at University of Liege (UL), Liege, Belgium, and the other located at Wayne State University (WSU), Detroit, MI, USA) and a QDR1500 located at the Neonatal unit of University Children’s Hospital (UCH), Greifswald, Germany. Quality control scans for each densitometer were performed daily using a manufacturer-supplied anthropomorphic spine phantom. The in vitro coefficients of variation (CV) for >1 year for the determination of bone mineral content, bone area and bone mineral density were 0.43%, 0.42% and 0.46%, respectively at UL, were 0.38%, 0.30% and 0.34% at WSU, and were 0.35%, 0.35% and 0.31%, respectively at UCH.
All densitometers were operated in the pencil beam mode, the only technique freely available for body composition studies in infants. The four phantoms were scanned in triplicate on each densitometer using infant whole body-scanning mode and analyzed with manufacturer-supplied software V5.73P. One investigator (J.-C.P.) familiar with the agreed layout of the phantoms was present at each site to insure the correct assembly and placement of the phantoms for DXA measurements. Each center used its own operator for scan acquisition and analysis. Phantoms were transported personally or shipped between centers via commercial courier.
Statistical Analysis
DXA measured total weight, lean mass, fat mass, bone mineral content, bone area and bone mineral density were used in data analysis. Percent of fat was presented as descriptive data and not analyzed further since it was calculated from fat mass and total weight. Repeated measures analysis of variance was used to determine the equivalence of the triplicate DXA measurements (within subject factor) among the four phantoms (between subject factor) and whether there was interaction between DXA measurements from different size phantoms from different instruments.
Regression analyses were performed to determine the ability of DXA measurements from UL and UCH to predict the DXA measurements of the same phantoms at WSU. Univariate analysis of variance with Helmert contrasts was used to analyze comparability of residuals from each prediction equation based on UL and UCH data respectively. The same procedures were repeated to determine the regression equation for prediction of DXA measurements of the same phantoms at the other centers from WSU DXA measurements.
The same procedures were repeated using the first of the triplicate measurements to mimic the clinical situation of generating one satisfactory scan per subject. This was done to determine whether the same relationships exist with the use of data from one or three DXA scan. All statistical tests were performed with SPSS 10.0 (SPSS Inc., Chicago, IL) for windows at an adopted significance level of 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSCONCLUSIONREFERENCES |
|---|
|
|
|
|
DISCUSSION |
|---|
Our goal for this study was to develop a set of phantoms that can be made easily and inexpensively and have sufficient flexibility for any investigator to further modify or adjust the components to create different size phantoms with varied body composition. The primary purpose for the use of these phantoms is to determine the interrelationship of the DXA measurements among different densitometers rather than the comparison of absolute accuracy of the DXA densitometers. This is consistent with the means to obtain standardized DXA measurements of the spine using instruments from different manufacturers that are known to provide different values for the same subject [9]. The design of our phantoms also satisfied the recommendations of the International DXA Standardization Committee that cross-calibration among different instruments should not be based on the use of a single phantom [9].
We have independently reported [10–12] the validity of the pencil beam DXA technique for the measurement of body composition using instruments from the same manufacturer based on animal tissue studies, and it was not our intention to reproduce the anatomically correct or exact duplication of body composition of infants, since there are great differences among infants and it would be prohibitively expensive and time consuming to achieve these goals. In any case, the physical dimensions and body composition values of our phantoms can be modified to accommodate the wide range of weights and body composition in clinical subjects, thus allowing cross comparison of any clinical studies involving small subjects.
In this study, the strongly predictive relationships of DXA measurements among the three instruments would support that data generated from different densitometers made by the same manufacturer employing the same DXA pencil beam technique and the same software are comparable. Furthermore, systematic corrections in the form of linear transformations are possible to allow comparison of clinical data generated from different studies. It is also possible that our system of phantoms can be used to determine whether these relationships remain true for data generated from the use of other DXA techniques or the use of instruments from different manufacturers.
That the intercepts of the regression equations for the prediction of DXA measurements among various centers were not significantly different from zero would support the absence of systematic difference among the densitometers tested, although the intercept for BMD prediction equation derived from UL approached significance. Even if there was a systematic difference in BMD measurements among different densitometers, it is still possible to compare data among different centers since the slope of the relationship in BMD measurements remain significantly highly correlated with an adjusted r2 of 
0.94. By way of clarification, the conversion of Celsius to Fahrenheit or vice versa would show a different intercept, but the slope would indicate that these two measurements are highly significantly related. In any case, BMD as an index of bone mass measurement is inappropriate in growing individuals such as infants [13,14]. Furthermore, since DXA bone mass measurement was validated for adults based on the mass of hydroxyapatite or other material [15] and for infants was based on the mass of carcass ash and calcium [10–12], i.e., not based on density, the potential clinical significance of any discrepancy in BMD measurements would be limited.
The same predictive ability among DXA measurements for the various components and total weight of phantoms whether using data from first or average of multiple DXA scans has major clinical implications. Thus the generation of a single good quality DXA scan, specifically without movement artifact, is likely to be adequate for clinical studies in infants. This results in reduction of radiation exposure, time and cost of clinical studies. In contrast, it is theoretically possible that a transient variability in the output X-ray source may have led to the one outlier in the triplicate measurement of one phantom. If this was the case, then the use of one DXA scan may be inadequate to determine the existence of this problem during any clinical study. The occurrence of the outlier measurement was unlikely the result of operator error since no repositioning was performed between the three scans. In any case, it is critical to maintain the instrument in optimal operating condition, remain vigilant to the assurance of uninterrupted X-ray energy output, consistent approach in data acquisition and analysis, avoidance of motion artifact and strict adherence to all aspects of quality assurance including duplicate scans on a sample of subjects in any DXA study.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSCONCLUSIONREFERENCES |
|---|
Received November 19, 2001. Accepted May 24, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSCONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. W. Koo, M. Hammami, and E. M Hockman Interchangeability of pencil-beam and fan-beam dual-energy X-ray absorptiometry measurements in piglets and infants Am. J. Clinical Nutrition, August 1, 2003; 78(2): 236 - 240. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
