LABORATORY ERRORS - SUSPECTED, REAL OR IMAGINED?

We strive to provide rapid, accurate laboratory data by continually implementing more accurate and precise methods and by improving our quality assurance techniques. In spite of these efforts, occasional errors will occur due to the human element involved and due to the inherent limitations of current methods.

You are the ultimate quality control for the laboratories and we suggest that you scan laboratory results with these questions in mind:

  1. Is the result reasonable? (e.g., a serum potassium of 50 mEq/L in a living patient obviously is an error, probably due to a misplaced decimal point). If not, see step 4.

  2. Is the result consistent? (a) chronologically with results on previous samples from the patient? (e.g., a report of BUN of 200 mg/dl in a patient with a normal BUN the previous day); (b) internally with other results on the same sample? (e.g., Do the electrolytes balance?) If not, see step 3.

  3. If the reported result is not consistent with the clinical findings, could the inconsistency be due to:
    1. Collection technique? (e.g., wrong anticoagulant used, prolonged tourniquet application, use of a hypodermic and syringe rather than vacutainer system which is more efficient and less likely to cause hemolysis, etc.)
    2. Physiologic variation? (e.g., age, sex, pregnancy, posture, exercise, time of day, relationship to meals,etc.)
    3. Therapy? (e.g., in vivo effect of drug on the test method itself).
      U
      seful reference on drug laboratory interactions:
      Effects of Drugs on Clinical Laboratory Tests, Donald S. Young, American Association for Clinical Chemistry, 3rd Edition, 1990, ISBN 0915274531.
    4. An unsuspected disease or complication?
      Useful references:
      Effects of Disease on Clinical Laboratory Tests, Friedman and Young, AACC Press, 2nd revised edition, 1989, ASIN 0915274523
      Interpretation of Diagnostic Tests, Jacques B. Wallach, Lippincott Williams & Wilkins Publishers, 7th edition, 2000, ISBN 0781716594.
      If the unexpected result cannot be explained by any of these considerations, go to step 4.

  4. If a laboratory error is suspected after analysis of the reported data, please forward and order request marked: "Confirmation Request for (test) of (date and time).  The request must include the physician's name and pager number.
  5. Upon receipt of a confirmation request, the laboratory will:

    1. Check original data to detect transcription errors;
    2. Repeat study on original sample (if available) to detect error in original determination;
    3. Repeat determination on a new sample on the patient to detect possible error in patient or specimen identification;
    4. If steps (b) and (c) confirm the original results, the clinical findings and drug history are reviewed for possible unsuspected disease or complications and for possible drug interferences;
    5. The results of the confirmation are reported to the physician by the Clinical Pathology Resident (501-688-2820). (No charge is made to the patient for a Confirmation Request).

  6. Erroneous laboratory results (whether clerical or analytical errors) which have been entered into the patient's computer record will be replaced with the corrected results. A numeric result will appear on the patient report with the notation "Previously Reported As" followed by the old result. Text results that are replaced will display the notation "Corrected Report" with the new result underneath. "Previously Reported As" will be followed by the old text result.

Updated 2.3.10

Clin Lab Home Page

 

For web page issues you may contact the Webmaster or by phone at 603-1691.