Portal de Periódicos da CAPES

Evaluating Laboratory Performance With the Six Sigma Scale

2000; American Medical Association; Volume: 124; Issue: 12 Linguagem: Inglês

10.5858/2000-124-1748a-elpwts

1543-2165

Stephen J. Sarewitz,

Statistical Methods in Clinical Trials

To the Editor.—I read with interest the article by Nevalainen and colleagues1 on clinical laboratory quality. Several points come to mind.The economic environment in which most clinical laboratories currently function is not friendly. Laboratories operate under stringent budgets. Personnel are worked to the limit and have little time to think creatively about problems and improving systems. Managers are overloaded with responsibilities. Revenue per unit of work continues to shrink (eg, the 2001 Medicare budget calls for further reductions in payments to laboratories). Meanwhile, overhead attributable to regulatory requirements keeps increasing. Clinical laboratories may not have the resources to make the significant investments in system re-engineering that will be required to achieve significant quality improvement—the orders-of-magnitude improvement that might reduce the defect rate of clinical laboratories to something close to that of other industries.Furthermore, quality improvement does not always lead to reduced costs to the laboratory. For example, consider a new test that is more expensive than the old one but carries 1/100th the false-positive rate. This new test may decrease costly unnecessary treatments, but (except in a fully integrated health care plan) the laboratory will not see the savings. Laboratories would not be encouraged to provide such a new test unless third parties provide reasonable reimbursement.The process that starts with ordering a test and ends with getting an accurate report in the hands of the clinician is complex. I'm not sure that it is reasonable to expect laboratories to achieve the same defect rate as activities such as baggage handling or producing a restaurant bill.Finally, some of the functions listed in Tables 1 and 2 are not necessarily under the laboratory's direct control, for example, missing information on requests, timing of therapeutic drug monitoring specimens (when laboratory personnel do not perform phlebotomy, a practice that is becoming more common), and patients without wristbands.