The scope and magnitude of bedside testing by personnel without formal laboratory training have recently increased. There has been some question about the reliability of bedside test results produced by nurses, physicians, and other ward staff that are used in clinical decisions. Glucose testing is currently the most widely performed bedside test. Because of the likelihood that the use of bedside testing will increase in the future, this article discusses various issues concerning bedside glucose analysis, including the scope of such testing, staff responsibilities, training and supervision, testing frequency, work flow and result reporting, and the reliability of bedside glucose results, in three institutions. We also suggest an approach to bedside testing quality management.

With simple microbiologic and fluorescent tests, we detected two cases of classic galactosemia, confirmed by specific enzyme assays, in the first 25 000 newborn infants in British Columbia screened for this disorder. The results were equivocally abnormal for another 31 infants, and a second blood sample was requested from each, either for repeat screening or for enzyme assays. The two infants with galactosemia were in hospital with an undiagnosed acute illness and had only a trace of non-glucose reducing substances in the urine when the screening tests were done. Screening for galactosemia fits well with our established programs of screening for phenylketonuria and hypothyroidism and costs less than $1 per infant tested.

The concept of structuring complexity is described as a process in which, during medical investigation, intensity of laboratory examination is escalated as the perceived complexity of diagnostic need increases. Application of this concept to laboratory controlled testing protocols is discussed. A clinical chemistry protocol for thyroid testing of ambulatory care patients in British Columbia resulted in a 15% reduction in test use and a 12% reduction in laboratory charges to the third party paying agent over a two-year period. The need to develop medically acceptable laboratory protocols that encourage selective test utilization and reduce cost is emphasized.

We assayed more than 5000 blood spots dried on filter paper and approximately 1000 serum samples for immunoreactive trypsin, with commercial reagents (Behring and Sorin). The assay procedures were modified so that newborn screening is technically feasible. Both kits are satisfactory for serum assay, but the Sorin materials are better adapted for blood spot analysis. Immunoreactive trypsin in blood spots rapidly decreases with specimen age, but is stable in frozen serum. Values for premature infants do not differ significantly from those for full-term babies. Children with cystic fibrosis were readily distinguished from those without, up to at least one year of age.