Return to CAP Home
Printable Version

  Queries and comments


October 2001
Originally published in CAP TODAY

Albert Rabinovitch, MD, PhD

Hundreds of questions roll in each year about Laboratory Accreditation Program requirements. CAPTODAY continues to publish a sampling of those questions and their answers, written by LAP checklist commissioner Albert Rabinovitch, MD, PhD, chief medical officer-vice president of Specialty Laboratories, Santa Monica, Calif. As head of all LAP checklist activity since 1993, Dr. Rabinovitch has years of experience and insight we're pleased to share.

Q: I have a question about URN.30700, which reads, "Are manual microscopic examinations of urine sediment performed as part of routine urinalysis testing, or are there specific, documented criteria defining the circumstances under which the microscopic examination may be omitted?" According to the Health Care Financing Administration (now the Centers for Medicare and Medicaid Services), we cannot perform microscopic or reflex testing without a physician order. The physicians order "UA." Our protocol is to perform a microscopic only if protein, leukocyte esterase, nitrite, or RBC is positive on the dipstick. We perform and report a microscopic at no charge, but the question has come up from another laboratory director, whose laboratory performs only a dipstick in response to a "UA" order and does not perform a microscopic regardless of the biochemical result. I can think of one option-have the physician order "UA and micro if indicated." As you know, it might be difficult to have the physicians order this way. Any words of wisdom on this?

A: This issue of perceived discrepancies with various parts of the CAP checklists and issues of tests requested versus tests performed from a regulatory perspective comes up from time to time. I always approach these as akin to a discussion of apples and oranges, rather than conflicting issues. The thrust of this phase I question is that a "complete" (this replaces the current word "routine") UA should include microscopy, but the latter part of the question provides an escape clause more fully described in the accompanying note in the checklist, which reads: "There is evidence that in random urinalysis screening (hospital admissions, insurance physicals), urines that are yellow and clear and have negative chemical reactions have a markedly low yield on microscopic examination. Optimal service may entail protocols defining when microscopic examination of urine sediment should or should not be done." All this relates to good laboratory practice. Note that noncompliance is only a phase I and thus cannot affect your accreditation.

What your facility defines as a "UA" is up to you. If you have separate requisition entries for macro only, macro plus microscopic, etc., that is perfectly fine, with the ordering physician having a choice. Your practice of doing sediment exam when some dipstick pads are positive does not strike me as adding much clinical value and creates the impression of added tests unless that is the expressed wish of your medical staff. I suspect the regulators might not be happy that you were reporting unrequested tests (i.e. sediment microscopy) if the order was for a dipstick only. Many institutions have separate requisition requests for UA screen - dipstick only versus UA - dipstick + microscopy. The latter option could be modified by the intent of checklist item URN.30700, but you should have a document indicating that clinicians will accept a test-based algorithm for excluding the microscopic (i.e. not performing an ordered test) if the dipstick is entirely negative, or whatever criteria your physicians will accept after education from the laboratory director.

This latter circumstance is, of course, the opposite of the usual financial mischief of adding and charging for unrequested tests. As I understand it, the basic Medicare rule is that the clinical laboratory performs services that have been requested by a physician. So, it is up to each laboratory, working with its clinicians, to define requisitions to align requests with reports. The CAP Commission on Laboratory Accreditation does not address financial aspects of clinical laboratories.

Q: Question AGC.27770 in the automated/general chemistry checklist reads, "Are multiple instruments that perform the same assay checked regularly for calibration agreement and correlation of patient results at least twice a year?" This question, upgraded from phase I to phase II, asks for calibration agreement and correlation of patient results.

The calibration agreement is accomplished easily with running controls. The "correlation of patient results" needs more clarification. This question does not state that you use patient samples; it states that you correlate patient results. Would the use of survey materials that are run on the multiple instruments satisfy this criteria? Evaluation is only done on the main instrument, but if the samples are run, they can be evaluated against one another when the survey results are returned. If this material is not acceptable then I feel the question should state "correlation with patient results." Since controls, survey-validated material, and standards are all used to check the accuracy and precision of the instrument, I feel they are surrogate patient samples and are capable of satisfying this question.

A: There are indeed two elements to this question. The correlation of patient results requires direct comparisons of patient material and is not satisfied by comparing commercial controls. The selection of actual blood samples (rather than simply stabilized commercial controls with potential matrix effects) is important to directly address the issue of whether a patient sample yields the same results on all of the laboratory's instruments. Statistical agreement of commercial control materials across instruments does not guarantee comparability of patient results. We want to control the quality of the patient results, rather than controlling the control results. In other words, surrogates, while useful for daily quality control, don't fully meet the mark for this semiannual activity.