STRATEGIES AND CRITERIA FOR DEVELOPING NEW URINALYSIS TESTS

Urinalysis provides a non-invasive means to sequentially evaluate rena l function and disease processes. Contemporary analytical techniques a nd the rapidly expanding knowledge of cell biology are yielding new wa ys of looking at urine constituents. Application of these new analytic al techniques to screening, diagnosing and monitoring renal disease re quires much more information than is currently available about the cor relation of urine analytes with disease processes. Case definitions fo r specific renal diseases depend upon a knowledge of natural history, response to therapy and laboratory data, including biopsy. When tests are being used to detect early stages of renal damage the subjects mus t be followed for months or years before a definitive diagnosis of irr eversible disease can be established; therefore, prospective studies m ust be used to validate these tests. Analytes chosen for further study should be linked to significant renal pathophysiological processes. G old standards for evaluating the predictive value of tests results mus t be established. The influence of renal disease on the analyte should be much greater than its biological variability under non-specific st resses. The results of using the test should benefit patients, taking into account the costs of false positive results and other costs to so ciety that come from providing the test. Prospective studies needed to validate tests should be feasible and affordable. These studies could be facilitated by establishing a collaborative bank of urine samples linked to clinical data. Tests which are not used in clinical decision making are unimportant and of little value. Tests used in decision ma king should be evaluated as rigorously as the treatments that will be chosen based on the test results.