Changes in Automated Complete Blood Cell Count and Differential Leukocyte Count Results Induced by Storage of Blood at Room Temperature Open Access

Clinical laboratory professionals are familiar with the scenario whereby a blood specimen is collected on a Friday but not delivered to the laboratory for processing until the following Monday or even later. When such a specimen arrives at the laboratory, the staff needs to decide (1) whether to accept or reject it; (2) if accepted, whether to perform all of the ordered tests or only those deemed appropriate based on the age of the specimen; and (3) what comments, if any, should be appended to the reported results regarding their reliability or unreliability. Such decision making requires laboratorians to be familiar with changes known to occur in blood specimens during storage. Manufacturers of automated analyzers and published literature often cite that blood specimens, kept at either room temperature or at 4°C (refrigerated) for up to 24 hours, generally yield reliable results for complete blood cell count (CBC) and automated differential leukocyte count (differential).1–4 However, specific information concerning the suitability or unsuitability of specimens older than 1 day for various laboratory tests, including the CBC and the automated differential, is scarce, particularly in the recent literature. This study delineates changes that occur in various parameters of automated CBC and differential during storage of blood at room temperature for several days.

A total of 40 K₃ (tripotassium ethylenediamine-tetraacetate) EDTA-anticoagulated blood specimens were processed through the Coulter Gen.S (Beckman Coulter, Miami, Fla) for CBC and differential once daily, specimen volume permitting, for 3 to 7 days. Specimens were selected from the routine laboratory workload over several days to represent a range of normal and abnormal CBC and differential values, as indicated in Table 1. Among the 40 specimens, 11 were leukopenic (white cell count, <4.0 × 10³/μL), 16 had leukocytosis (white cell count, >11.0 × 10³/μL), 15 were thrombocytopenic (platelet count, <140 × 10³/μL), and 11 had thrombocytosis (platelet count, >400 × 10³/μL). All but 4 specimens were anemic, with hemoglobin levels less than 12.5 g/dL. Specimens were kept at room temperature throughout the study. The Coulter Gen.S uses impedance technology to generate a 10-part CBC and volume (impedance), conductivity, and light scatter measurements, commonly referred to as VCS technology, to generate a 5-part differential.1 Software versions 1C and 2A were used during this evaluation. Calibration and quality control of the analyzer were performed using S-cal and 3 levels of 5C control (both from Beckman Coulter), respectively, according to the manufacturer's instructions.1 The 10-part CBC consisted of white blood cell count (WBC), red blood cell count (RBC), hemoglobin, hematocrit, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), red cell distribution width (RDW), platelet count, and mean platelet volume (MPV). The 5-part differential consisted of relative percentages and absolute numbers of neutrophils, lymphocytes, monocytes, eosinophils, and basophils.

The CBC and differential results were tabulated using a personal computer with Excel software. Percent change from the day 0 value for each parameter for each subsequent day of the study period was calculated. For percentages and numbers of eosinophils, basophils, and monocytes, absolute difference rather than percent change from the value on day 0 was calculated because some of the values obtained initially and/or on subsequent days were 0. Ultimately, mean percent change or difference and SD were determined, also using Excel software, for each parameter for each day. Adobe Illustrator, version 8.0, was used to prepare the figures.

Changes observed in the CBC results of blood specimens stored at room temperature are summarized in Table 2 and illustrated in Figure 1. For each CBC parameter, mean percent change from the value on day 0 is plotted along with its 95% confidence interval (CI) (±2 SDs) against time in days after collection of blood (Figure 1).

The mean percent change in the WBC was less than 1 for the first 3 days but fluctuated between −5.3 and +6.3 from days 4 to 7. The 95% CIs ranged between −12.2% and +13.8% for the first 3 days and between −30% and +42% from days 4 to 7.

The RBC, hemoglobin, and MCH were relatively stable, each with a mean percent change of equal to or less than 1.1 and a combined 95% CI in the range of −3.7% to +5.1%.

The MCV, hematocrit, and RDW each increased over time, revealing maximum mean percent changes on day 7 of 10.4, 10.3, and 15.8, respectively. The respective 95% CIs for MCV, hematocrit, and RDW ranged from −1.8% to +14.8%, −2.8% to +16.5%, and −5.4 % to +25.7% throughout the 7-day period.

The MCHC trended downward over time, with a maximum mean percent change of −9.4 on day 7 and 95% CIs of −13.9% to +3.2% for the 7-day period.

The mean percent change in the platelet count fluctuated between −0.9 and +4.3 for the first 4 days and ranged from +6.1 to +6.6 for the subsequent 3 days. Similarly, the 95% CIs ranged from −23.8% to +30.2% for the initial 4 days and from −42.5% to +54.8% for the subsequent 3 days. The platelet count of 2 of the 40 specimens dropped progressively (up to 37% of the value on day 0) from day 3 onward, and on blood smear review was found to be associated with platelet clumping. Only a slight increase in mean percent change and either no change or minimal decrease in SD was noted on each day when the platelet count data of these 2 specimens were excluded from final calculations (PLT_a in Table 2 and Figure 1).

The MPV initially increased steadily, reaching a maximum mean percent change of +12.5 on day 5. On days 6 and 7, the mean percent changes were +6.3 and +8.2, respectively. The 95% CIs for MPV ranged between −12.0% to +37.1% throughout the 7-day period.

When changes observed in the CBC parameters were categorized based on whether the initial (day 0) value was low (below the lowest normal value), normal (within the normal range), or high (above the highest normal value), the number of specimens analyzed per day decreased in each group to between 4 and 16, but certain trends became evident among the groups. Leukopenic and thrombocytopenic specimens revealed the greatest day-to-day mean percent changes and the widest day-to-day SDs compared with the changes observed for specimens with normal or high initial respective counts (Table 3). Furthermore, the platelet counts of thrombocytopenic specimens tended to increase over time, whereas the opposite was true for specimens with high initial count (Figure 2, B). Similarly, the WBC of leukopenic specimens tended to increase from day 5 onward, whereas the opposite was true, although minimally so, from day 4 onward for specimens with high initial counts (Figure 2, A). Changes observed in the remaining CBC parameters (data not shown) did not show any particular relationship to the levels of initial values.

Changes observed in the automated differential results of blood specimens stored at room temperature are shown in Table 4 and Figure 3.

A minimal initial decrease in the percentage of neutrophils on days 1 and 2 and in the number of neutrophils on day 1 was followed by a variable degree of increase in both on subsequent days. A maximum mean percent change in percentage of neutrophils of +21.4 was noted on day 4 and in number of neutrophils of +24.5 on day 6. The 95% CIs during the 7-day period ranged from −69.9% to +90.8% for the percentage of neutrophils and from −119.2% to +168.1% for the number of neutrophils. The widest 95% CI was noted on day 6 for both the percentage and number of neutrophils.

The percentage and number of lymphocytes tended to increase over time, and the mean percent change fluctuated between +29.1 and +97.8 for the former and between +3.3 and +81.3 for the latter. The respective CIs for the percentage and number of lymphocytes throughout the 7-day period ranged from −782.3% to +978% and −207.5% to +355.2%.

The percentage and number of monocytes tended to decrease over time. The mean percent change fluctuated between −85.4 and +0.4 for the former, and the mean absolute difference fluctuated between −0.6 and −0.1 for the latter. The respective 95% CIs for the percentage and number of monocytes throughout the 7-day period ranged from −406.7% to +407.5% and −2.9 to +2.8, respectively.

The mean absolute difference from the value on day 0 for the percentage and number of eosinophils increased steadily with time, with a maximum difference of +7.9 for the former and +0.6 for the latter, both on day 7. The respective 95% CIs for the percentage and number of eosinophils ranged from −25.6 to +40.4 and −1.1 to +2.0.

Basophils represented the smallest fraction (0.0% to 1.7%) of all WBC and revealed minimal or no change with time in both the percentage and number. The maximum mean absolute difference from the value on day 0 was 0.2 on day 4 for the percentage of basophils. The mean absolute difference from the value on day 0 for the number of basophils remained unchanged at 0 throughout the study. The respective 95% CIs for the mean difference in the percentage and number of basophils ranged from −1.4 to +1.5 and −0.4 to +0.3.

Storage of blood at room temperature for up to 7 days caused changes in CBC and automated differential results. However, the changes observed in some CBC parameters may be considered clinically insignificant.

The maximum change of 4% in RBC, 3.4% in hemoglobin, and 5.1% in MCH carries little, if any, clinical significance. Moreover, the narrow 95% CIs for each of these 3 parameters clearly demonstrate their stability at room temperature for at least up to 7 days. The MCV, and consequently the hematocrit and RDW, increased over time, at an apparently faster rate during the first 4 days when compared with that during the subsequent 3 days. The increase in MCV reflects red blood cell swelling at room temperature. The degree of change in MCV, hematocrit, and RDW observed from day 2 onward may be considered less than desirable, particularly when the results are borderline normal or abnormal. A continuous increase in hematocrit with a relatively stable hemoglobin level led to a steady decline in MCHC over time. A potential change of 5% or greater in MCHC, as noted from day 2 onward, is also less than desirable. In general, the WBC appeared stable for the first 3 days. However, the degree of change in the WBC of specimens with a normal or high count observed throughout the study may be considered acceptable for clinical utility. In contrast, the changes in the WBC of leukopenic specimens, as observed from day 5 onward in particular, are less than desirable. Like the WBC, the platelet count in general appeared stable for the first 4 days. However, the degree of change in the platelet count observed throughout the study in specimens with the platelet count within or above the normal range may be considered acceptable for clinical utility. A less than desirable degree of change was observed only in thrombocytopenic specimens, particularly from day 5 onward. The MPV behaved like the MCV for the initial 5 days, showing a continuous increase, with a relatively faster rate of change during the first 2 days when compared with that during the subsequent 3 days. A mean percent change in the MPV of 10 or greater, as noted from day 2 onward, is also less than desirable.

Our observations of changes in CBC results induced by storage of blood at room temperature agree with those of Cohle and associates5 published 2 decades ago. This agreement is noted despite the difference in the total number (12 to 40 vs 10) and types of specimens (normal plus abnormal vs only normal) used in the 2 studies. The automated analyzer used by Cohle et al5 to obtain CBC results was Coulter S-Plus, also an impedance technology–based cell counter. These findings suggest that clinically usable results may be obtained for some CBC parameters from blood specimens stored at room temperature for up to 7 days. Specifically, the RBC, hemoglobin, and MCH were stable for the maximum study period of 7 days. The WBC and platelet count were fairly stable for 4 days or even longer (up to 7 days) if the counts were within or above the normal range. Relatively wide fluctuations (large SDs) in the WBC and platelet count of specimens with counts below the normal range are understandable and attributable, at least partly, to lower degree of precision known to be associated with low counts when compared with normal or high counts by automated or manual methods. Additionally, the small number of specimens from day 4 onward could have contributed to wider fluctuations noted during the same period for specimens with counts below or within the normal range. A progressive increase in MPV, MCV, hematocrit, and RDW and a decrease in MCHC from day 2 onward make the results of these parameters suspect in terms of clinical utility, especially in cases where the results of these parameters are borderline normal or abnormal.

Storage of blood at room temperature caused changes in virtually all of the parameters of automated differential. Beginning on day 1, 2, or 3, the relative percentages and absolute numbers of neutrophils, lymphocytes, and eosinophils tended to increase, whereas the opposite was true for the relative percentage and absolute number of monocytes. The increase in the percentage of neutrophils began on day 3, number of neutrophils on day 2, percentage and number of lymphocytes on day 1, percentage of eosinophils on day 1, and number of eosinophils on day 3. The percentage and number of monocytes trended downward from day 1 onward. The relative percentage and absolute number of basophils, although difficult to reliably evaluate because they represented a very small fraction, if any, of total leukocytes, trended either downward or showed no appreciable change, particularly in the absolute number throughout the study period. These changes in white blood cell differential are most likely the sum effect of the loss of individual cell characteristics specifically measured by the analyzer and the cellular degeneration that is known to occur as the cell ages.

An increase in lymphocytes and a decrease in monocytes after 24 and 48 hours of storage of blood at room temperature have also been reported previously by Warner and Reardon,2 who analyzed 25 specimens from healthy subjects using Coulter STKS, which, like the Coulter Gen.S, uses VCS technology for the differential count. Compared with an appreciable decrease in neutrophils noted at the same period by these investigators, we observed only a small or rather negligible decrease in percentage of neutrophils on days 1 and 2 and in number of neutrophils on day 1, beyond which both the percentage and number of neutrophils increased, although to a variable degree, for the remainder of the study period. Unlike our observations of an increase in the percentage and number of eosinophils and a decrease in the percentage and number of basophils, Warner and Reardon2 noted a decrease in the percentage of eosinophils and an increase in the percentage of basophils after 24 and 48 hours of storage at room temperature. Differences in the sample size and sample type between the 2 studies may at least partly account for the different findings. The observations on changes in neutrophils, lymphocytes, and monocytes after 1 day of storage at room temperature are also corroborated by the published reports of Wood et al,6 who analyzed 21 specimens from healthy volunteers and 113 patient specimens using Cell Dyn-3500, which performs differential analysis by measuring laser light scattering properties of cells at various angles.

To the best of our knowledge, there are no other published reports in the English language pertaining to the study of changes in the automated differential induced by prolonged storage of blood at room temperature. Based on the findings of this study and the cited published reports, it can be surmised that even after only 24 hours a specimen may yield unreliable differential results when using the VCS technology. Storage of blood at 4°C has been shown to prolong the specimen stability by several days for many CBC parameters3,4,6,7 and by a couple of days for some differential parameters.2,6 Refrigeration is not a solution, however, for specimens that arrive at the laboratory after being stored at room temperature for a few days.

Although it is preferable to perform a CBC and differential on a blood specimen as soon as possible after collection, our study suggests that clinically reliable results may be obtained for some CBC parameters, but not for the differential, from specimens older than 1 day, when analyzed on the Coulter Gen.S. Specifically, hemoglobin, RBC, and MCH are stable for at least 7 days, WBC for at least 3 days, and platelets for at least 4 days of storage at room temperature. Based on these findings, laboratories using VCS technology for CBC and differential may decide to accept specimens older than 24 hours for analysis and report only the results of the stable CBC parameters. Under these circumstances, a comment such as “unable to obtain reliable results due to the age of the specimen” or “specimen too old to obtain reliable results” could replace the results of unstable parameters. Alternatively, one may choose to report the results of all CBC parameters, but with a comment such as “specimen __ days old, results may be affected” appended to the result(s) of one or more unstable parameters.

We are grateful to the hematology laboratory staff for technical help, Medical Media Graphics staff for preparing the illustrations, and John O'Connor, PhD, of Department of Health Policy for helpful hints in data analysis.