< previous page	page_397	next page >
Page 397 Table 2 Continued Local Laboratory Central Laboratory Laboratory may not be operating to principles of GLP Courier reliability must be established and maintained Laboratory data audit trail may be incomplete Higher direct cost of analysis Higher up-front costs Poor notification of methodology changes Data validation required at sponsor site Manual data input at sponsor site Multiple laboratory site audits may be required used electrophoresis or radial immunodiffusion. The data was, therefore, of little value because it could not be pooled. Another common problem in Europe is that the local laboratories run their serum enzyme assays (i.e., liver transaminases and alkaline phosphatase) at different temperatures and with different reagents. These differences in testing conditions can greatly influence how much enzyme activity is measured and thus pooling the data would be impossible, short of normalizing the values (see Table 3). Local laboratories, whose primary concern is rarely the support of clinical trials, have been known to change reporting units and reference ranges midtrial, thus making it difficult to pool data, even at a single center. The inherent variability in test values, associated with the use of multiple laboratories, can Table 3 Differences in Laboratory Reference Ranges for Alkaline Phosphatase for a Single Age Range Laboratory Country Laboratory values SciCor U.S.A. 31110 IU/L Ulleval Norway 60275 IU/L ABL Holland 35140 IU/L Cortes Spain 150500 IU/L Holste Germany 0190 IU/L Zilver Holland 3095 IU/L
< previous page	page_397	next page >