Q & A, 08/13

Editor: Frederick L. Kiechle, MD, PhD

[printfriendly]

Submit your pathology-related question for reply by appropriate medical consultants. CAP TODAY will make every effort to answer all relevant questions. However, those questions that are not of general interest may not receive a reply. For your question to be considered, you must include your name and address; this information will be omitted if your question is published in CAP TODAY.

Submit a Question [hr] [pulledquote]Q. What are considered best practices for tracking re-sult trending in the lab? We use hemoglobin running mean in our hematology department because it is built into the analyzer software. The chemistry department will have a difficult time applying moving averages without purchasing middleware.[/pulledquote]

A. Applications of averages of patient data (AOP) have been used for almost 50 years.¹ An error condition is signaled when the average of consecutive centrally distributed patient data is beyond the control limits established for the average of the patient data. The assumption underlying AOP is that the patient population is stable and a significant change in the AOP would arise from an analytical shift. Perhaps the greatest value of AOP is that it permits assessment of an analyzer during the intervals when control materials are not being run. Control materials can be analyzed at any time without the requirement to accumulate patient specimen results, and thus are especially useful at instrument startup and after maintenance and recalibration.

AOP can be used retrospectively for quality assurance (for example, comparing the means of patient data from similar analytic systems housed in the same laboratory). The use of AOP for prospective quality control is complex and has significant limitations in the acute care (hospital) environment. The error-detection capabilities of AOP depend on several factors, with the most important being the number of patient results averaged and the variances of the patient population and analytical method.² Variations of AOP have been used extensively in hematology to monitor patient red blood cell indices and, indirectly, their constituent measurements, hemoglobin and red blood cell count as well as hematocrit.3-5 In a large Pennsylvania robotic reference laboratory, the exponential smoothing of truncated groups of 60 WBC, RBC, hemoglobin, hematocrit, MCV, MCH, MCHC, RDW, platelets, neutrophils, and lymphocytes replaced the periodic analysis (once for every 60 patient specimens) of a commercial control.6 In addition to the patient moving averages, three levels of commercial controls were run at startup and then at eight-hour intervals. After implementing patient average quality control for one year, the hematology laboratory had saved $19,000 and $14,000—costs of quality control material and labor, respectively.

AOP is more suited to large reference laboratories that evaluate largely normal patients. In hospital laboratories, AOP can be shifted by changes in the proportion of patient samples originating from specific patient units or by changes in the proportions of patients with more severe illness. In hematology, for example, the averaging of a large number of specimens from a neonatal unit or hematology unit can cause the red blood cell indices to inappropriately indicate an out-of-control situation. In clinical chemistry, analysis of specimens from renal units will cause large shifts in the AOP of creatinine, glucose, and urea nitrogen. Hospital patient AOP is significantly influenced by longer-term, within-day, and within-week trends. During evenings and weekends, test volumes are reduced; this weekend and nightly testing is generally performed on more acutely ill patients. As a result, evening and weekend AOP will demonstrate higher proportions of out-of-control averages, including elevated glucose, low sodium, low protein, and low calcium averages.7 In reference laboratory testing, the patient data tend to be more centrally distributed and there is “natural randomization” of patient specimens, rendering AOP a powerful tool to guarantee acceptable analytical performance. As hematology and chemistry analyzers become more precise and accurate, investigation of outlying AOP in hospital environments will more often demonstrate changes in patient mix rather than analytic shift. Over time, the applications of traditional prospective AOP, especially in hospital environments, will wane.

1. Hoffmann RG, Waid ME. The “average of normals” method of quality control. Am J Clin Pathol. 1965;43:134–141.
2. Cembrowski GS, Chandler EP, Westgard JO. Assessment of “average of normals” quality control procedures and guidelines for implementation. Am J Clin Pathol. 1984;81(4):492–499.
3. Bull BS, Elashoff RM, Heilbron DC, et al. A study of various estimators for the derivation of quality control procedures from patient erythrocyte indices. Am J Clin Pathol. 1974;61(4):473–481.
4. Cembrowski GS, Westgard JO. Quality control of multichannel hematology analyzers: evaluation of Bull’s algorithm. Am J Clin Pathol. 1985;83(3):337–345.
5. Lunetzky ES, Cembrowski GS. Performance characteristics of Bull’s multirule algorithm for the quality control of multichannel hematology analyzers. Am J Clin Pathol. 1987;88(5):634–638.
6. Cembrowski GS, Parlapiano E, O’Bryan D, et al. Successful use of patient moving averages (PMA) as an accuracy control for multichannel hematology analyzers in a high-volume robotic clinical laboratory (abstract). Lab Hematol. 2001;7:35. Abstract 3.
7. Cembrowski GS. Thoughts on quality-control systems: a laboratorian’s perspective. Clin Chem. 1997;43(5):886–892.

George Cembrowski, MD, PhD

Director of Medical Biochemistry

Department of Laboratory

Medicine and Pathology