Orders for laboratory tests are also often included in discussions of data of value for biosurveillance. The data themselves contain the test ordered, reason for the order encoded in ICD-9, patient gender, birth year, time stamp, and zip code (Ma et al., 2004). Like prescriptions, they have "face validity,'' and in fact are being collected routinely by CDC's Biosense system (Ma et al., 2004). However, there are as yet no published studies on their informational value.
Chotani and Lewis of Johns Hopkins Applied Physicians Laboratory compared orders received by health departments for influenza tests, results of those tests, and onset of influenza season (unpublished results are summarized in Wagner et al. ).They found high correlation between influenza testing and the onset of the influenza during four seasons. There was, however, no lead time for test requests compared to patterns of influenza-like illnesses ICD-9 diagnoses in physician office visit records. This result suggests that either source can be used for monitoring and the preference between them depends on whether the data can be obtained without time delay and differences in cost and effort to build the data collection system. A caveat to this research is that the data source investigated was limited to tests ordered from the state laboratory whereas influenza testing is also done by private laboratories, which may produce an earlier or stronger signal, due to increased sampling.
Was this article helpful?