College of American Pathologists
Printable Version

  Q and A




August 2008

Question Q. Is it necessary to analyze the fasting glucose before giving a patient glucola for a glucose tolerance test, or is it acceptable to draw the fasting glucose and give the glucola right away?

A. The main indication for performing an oral glucose tolerance test, or OGTT, is to detect and diagnose gestational diabetes mellitus.1 Though not recommended by the American Diabetes Association (ADA) for routine clinical use to diagnose diabetes in nonpregnant individuals (fasting plasma glucose, or FPG, is the preferred method), the OGTT may be used for further evaluation of patients when diabetes is strongly suspected despite a normal FPG.1 The latter recommendation is based on observations that the OGTT is more sensitive than FPG to diagnose diabetes. The ADA prefers FPG because the OGTT is poorly reproducible and difficult to perform in practice. The World Health Organization recommends either FPG or an OGTT for diagnosis.2

As alluded to above, the OGTT is affected by a number of factors, ranging from the form of glucose and rate of ingestion to posture and time of day, that result in poor reproducibility.3 Moreover, approximately 20 percent of OGTTs fall into the nondiagnostic category. Unless results are grossly abnormal initially, the OGTT should be performed on two separate occasions.1

It is not clear from the question whether the OGTT is being used in pregnant or nonpregnant patients. For the latter, the patient should have had FPG measured on a prior occasion (if ADA guidelines are followed). If the FPG is =126 mg/dL (7.0 mmol/mol) on two separate days, the patient has diabetes and an OGTT is unnecessary. If the recent FPG was <126 mg/dL (7.0 mmol/mol), the glucola can be administered immediately after drawing the fasting sample when performing an OGTT.

The diagnosis of gestational dia­betes mellitus, or GDM, is a little different. Initial screening is usually accomplished by performing a one-hour glucose loading test at 24 to 28 weeks of gestation.1,2 Patients do not have to fast for this test. The patient is given a 50-g oral glucose load regardless of the time of the last meal. Blood is drawn one hour after completion of the glucose drink. If the plasma glucose at one hour exceeds the chosen threshold, a diagnostic OGTT should be performed on a different day. This can be either a 100-g three-hour glucose tolerance test or a 75-g two-hour OGTT. Two different thresholds are used for the 50-g glucose load test. A glucose threshold =140 mg/dL (7.8 mmol/mol) identifies GDM with 80 percent sensitivity, while the sensitivity is increased to 90 percent if 130 mg/dL (7.2 mmol/ mol) is selected as the cutoff.

Neither the ADA nor the World Health Organization has specific recommendations for measuring fasting glucose before administering glucola. As mentioned, the OGTT is unnecessary if the FPG is diagnostic of diabetes. Though there is no evidence that it is harmful to give glucola to someone with increased fasting glucose, it seems prudent to avoid doing so. Some do capillary finger-stick glucose measurements before the OGTT, and if the glucose concentration is high, do not perform an OGTT. For GDM, some do not perform an OGTT if the result of the 50-g challenge test is above a certain value, commonly 180 to 200 mg/dL (10.0 to 11.1 mmol/mol). None of these practices is evidence based. Findings from the Hyperglycemia and Adverse Pregnancy Outcome study—a large, multinational prospective investigation in which more than 23,000 patients were recruited—are expected to be reported later this year. The results of this study are likely to lead to revised diagnostic criteria for GDM.


  1. American Diabetes Association. Standards of medical care in diabetes—2008. Diabetes Care. 2008;31(Suppl 1):S12–54.
  2. Sacks DB, Bruns DE, Goldstein DE, et al. Guidelines and recommendations for laboratory analysis in the diagnosis and management of diabetes mellitus. Clin Chem. 2002;48:436–472.
  3. Sacks DB. Carbohydrates. In: Burtis CA, Ashwood ER, and Bruns DE, eds. Tietz Textbook of Clinical Chemistry and Molecular Diagnostics. St. Louis, Mo.: Elsevier Saunders;2006:837–902.

David B. Sacks, MBChB, FRCPath
Associate Professor of Pathology
Harvard Medical School
Brigham and Women’s Hospital
Member, CAP Chemistry Resource Committee

Question Q. How long should you wait after a unit of blood has been transfused before drawing a complete blood count, or doing other lab work, to ensure accurate test results?

A. Optimum timing of post-transfusion phlebotomy is critical for ensuring meaningful laboratory testing results, and medical judgment is required in making this determination. Several factors must be considered, including the type and amount of blood product given, purpose of the test (that is, the question it is intended to answer), and clinical setting.

In general, it is best to perform phlebotomy when the patient’s circulatory system is in homeostasis. A patient who is bleeding or undergoing blood product transfusion, or both, is not in a steady state. Whenever possible, samples for laboratory testing should be postponed until bleeding has stopped and transfusion is complete. One obvious exception to this rule, however, would be the setting of massive transfusion, during which monitoring certain laboratory values, such as cell counts and coagulation parameters, is essential to guide ongoing therapy. Variables such as patient blood volume, cardiac output, renal function, and volume of blood products transfused affect how quickly homeostasis is achieved following transfusion.

For the evaluation of post-transfusion increments in hemoglobin, hematocrit, and platelet counts, a practical approach is to draw blood samples within 10 to 60 minutes after completing transfusion, as this time interval is aimed at measuring peak recovery.1 Results determined from blood samples drawn later than 60 minutes post-transfusion are increasingly affected by confounding conditions, such as splenic sequestration, sepsis, and consumption.1,2 If the intent is to determine the extent of such confounding processes on red cell and platelet counts, one should combine a 10-minute post-transfusion sample with sequential samples drawn at one hour and 24 hours post-transfusion.

Alterations in chemistry test results following transfusion are not usually a concern in the low-volume transfusion setting. However, assay results may be affected for varying periods following transfusion of large amounts of blood products, as seen in massive transfusion, red cell, or plasma exchange—particularly if the recipient has impaired hepatic or renal function. Banked storage of red cells results in elevated plasma levels of hemoglobin, potassium, LDH, and iron in the blood unit that may, particularly in the metabolically impaired patient, be reflected in the post-transfusion laboratory values. In addition, citrate anticoagulant present in blood products may result in transient hypocalcemia in the recipient.3 Therefore, following large-volume transfusions or exchanges, waiting 12 to 24 hours before drawing samples for chemistry assays will provide results that are more reflective of the patient’s underlying metabolic state.


  1. Choo Y. The HLA system in transfusion medicine. In: McCullough J, ed. Transfusion Medicine. New York, NY: ­McGraw–Hill Book Co;1998:401.
  2. Legler TJ, Fischer I, Dittman J, et al. Frequency and causes of refractoriness in multiply transfused patients. Ann Hematol. 1997;74:185–189.
  3. Brecher ME, ed. Technical Manual. 15th ed. Bethesda, Md.:AABB;2005;649–650.

Rita A. Reik, MD
Pathology Consultants of South Broward
Medical Director/Transfusion Medicine Services
Memorial Healthcare System
Hollywood, Fla.

Question Q. Are there accepted policies regarding the performance of repeated manual differential white blood cell counts for patients with persistent low total leukocyte counts?

A. Practice varies widely on this topic. The CAP Q-Probes group recently published the results of its study on the rate of manual peripheral blood smear review (including scans and differential counts) at 263 institutions in 45 states and four foreign countries.1 A total of 95,141 complete blood count determinations were studied. The study found that the percentage of CBCs with a manual review, scan, or differential performed ranged from a 10th percentile of 9.9 percent to a 90th percentile of 50 percent, with a median of 26.7 percent. In other words, 10 percent of laboratories did a manual review, scan, or differential on less than 9.9 percent of their CBCs; another 10 percent did more than 50 percent; and the median group did reviews on approximately 25 percent of their CBCs. The most frequent prompts resulting in reviews were hematology analyzer flags, including WBC flag (36.7 percent), immature cell flag (25.5 percent), red blood cell flag (13 percent), and platelet flag (5.9 percent). Physician request accounted for only 3.7 percent of reviews. When the prompts were set at higher levels, the number of reviews decreased, as would be expected. In addition, the number of reviews was lower if the laboratory had a policy of limiting reviews of repeat specimens within certain time frames or if the reviews were done only for abnormal RBC parameters (and not other hematologic abnormalities). Only 41 laboratories reported having a time interval limit, and of those 41, the most common interval limit was 24 hours. When all reviews were considered, they were thought to have contributed new information in about one-third of cases.

The International Society for Laboratory Hematology has suggested criteria for action following automated CBC and WBC differential analysis.2 A group of 20 hematology laboratorians from six countries and 17 laboratories reached consensus on rules that should be used in various circumstances, which it then tested on 1,000 samples, with 200 repeated to test the delta rules. The group developed criteria to determine if the use of these rules revealed a positive smear finding. Analysis of this method showed a true positive rate of 11.2 percent, true negative rate of 67.3 percent, false-positive rate of 18.6 percent, and false-negative rate of 2.9 percent. The group indicated that it hopes individual laboratories will consider applying these rules in their own institutions, and it has provided a protocol to facilitate this validation.

As for the question of repeated low total leukocyte counts, the group recommends a slide review with the first low result (<4.0) and then a repeat slide review if the delta failed and within three days. In addition, it recommends a manual differential and slide review in circumstances in which the analyzer is unable to give a differential or the differential is incomplete. Therefore, the group would recommend a slide review (or manual differential) with the first low total leukocyte count, but would only repeat it in the first three days if the delta check failed.


  1. Novis DA, Walsh M, Wilkinson D, et al. Laboratory productivity and the rate of manual peripheral blood smear review: a College of American Pathologists Q-Probes study of 95,141 complete blood count determinations performed in 263 institutions. Arch Pathol Lab Med. 2006; 130: 596– 601.
  2. Barnes PW, McFadden SL, Machin SJ, et al. The international consensus group for hematology review: suggested criteria for action following automated CBC and WBC differential analysis. Lab Hematol. 2005;11:83–90.

Katherine A. Galagan, MD
Chief of Pathology and
Director of Clinical Laboratories
Virginia Mason Medical Center
Seattle, Wash.

Question Q. Our laboratory in Missouri is located inside a critical access hospital, where we have a busy blood bank. A question has arisen about the regulations for allowing nurse practitioners or physician assistants, or both, to write orders for crossmatches and infusions of blood products. What are the recommended policies/pro­cedures for proceeding with orders written in these situations?

A. According to standard in the 25th edition of the AABB Standards for Blood Banks and Transfusion Services, “a physician or other authorized health professional shall order blood, components, tests, and derivatives.” The governing body of each hospital has the authority to define the functions and activities of nurse practitioners and physician assistants in accordance with state and federal laws. The scope of practice can include writing orders for transfusion amid consultation with the supervising physician. In general, the supervising physician must be available for consultation and assistance whenever physician assistants or nurse practitioners are employed. The extent of supervision, such as the amount of on-site supervision and the distance of the physician from the hospital, is delineated by the licensing boards in each state.

Waivers to the standard supervision requirements may be available for critical access hospitals to ensure adequate access to health care services for the community. Federal regulations pertaining to critical access hospitals can be found in title 42 of the Code of Federal Regulations, part 485. Your state licensing board will be able to provide local requirements. Additional information and assistance may also be available from the American Academy of Physician Assistants and the American Aca­demy of Nurse Practitioners.

Lesley A. Kresie, MD
Medical Director, Laboratory Services
Carter BloodCare
Bedford, Tex.