Artifactual Changes in Sprague–Dawley Rat Hematologic Parameters after Storage of Samples at 3 °C and 21 °C

Abstract

Circumstances can occur that prevent timely analysis of blood samples. The purpose of this study was to characterize artifactual changes in rat hematologic parameters after storage of samples at 3 and 21 °C and to document the effects of storage on peripheral blood smear findings. EDTA-treated blood samples were collected from 12 male Sprague–Dawley rats. Samples were analyzed on an impedance hematology analyzer within 5 min after collection and then at 6, 24, 48, and 72 h after storage at 3 °C or 21 °C. Corresponding blood smears were examined microscopically. RBC count and hemoglobin concentration had not changed after 72 h at either temperature. At 3 °C, the instrument-derived hematocrit and manually measured PCV remained unchanged for 72 h. Compared with 0-h values, platelet counts and MCV at 6 h and MPV at 24 h were higher at either temperature. In general, WBC count and neutrophil and lymphocyte percentages were unchanged for at least 48 h at either temperature. Prominent blood smear findings were smudge cells, pyknotic leukocytes, echinocytes, and spheroechinocytes. Although some observed changes were within analytic variability or clinically negligible, the best practice likely is to measure hematologic parameters within 6 h after collection. In the event of delayed analysis, specimens should be stored in the refrigerator, and care must be taken not to misinterpret artifactual changes as pathologic findings.

How a blood specimen is handled prior to analysis can introduce significant preanalytical variation, producing a range of artifacts in experimental data.²⁰ Sources of preanalytical variation include blood collection technique, inconsistency in the site of blood sampling, the order in which samples are aliquoted (for example, EDTA tube followed by serum-separator tube), the time between sample collection and analysis, and storage temperature.²⁰ To limit preanalytical variation due to the time before analysis or storage time, blood specimens should be analyzed as soon as possible after collection.^1,11 However, delays in analysis occur when circumstances prevent timely analysis. For example, instrument failure, transport of specimens to a reference laboratory, and sample collection before or on weekends or holidays are all potential causes of delayed analysis. As such, identifying storage-related changes that may occur in blood specimens is important so that artifactual changes are not misinterpreted as pathologic findings.

Although many studies in the literature report effects of storage on hematologic blood variables, most involve human or dog blood,^{8-11,13,15,17,24} with little information on the effects of storage on hematologic variables in rat blood.^1,10 The variability of artifactual change that has been observed among different mammalian species has led to recommendations for the establishment of species-specific guidelines on the changes that occur due to storage.^1,8 Furthermore, delay-associated artifactual changes in hematologic parameters vary depending on the analyzer used due to the different techniques used by automated analyzers to characterize and quantify hematologic parameters.^1,10,13 When a delay in blood analysis cannot be avoided, general recommendations are that specimens be stored in the refrigerator, because in some species, this practice decreases the effects of storage on hematologic variables compared with those seen after storage at room temperature.^8,11,24 Of 2 previous studies on storage effects on rat hematologic variables, one assessed only a few hematologic blood parameters after 24 h of storage,¹⁰ whereas the other used a laser analyzer to evaluate storage effects beginning at 24 h;¹ both cited studies only assessed effects after storage in a refrigerator.

The current study used an impedance analyzer to characterize changes in rat blood hematologic parameters after storage at 3 and 21 °C, beginning after 6 h of storage and continuing through 72 h. We also evaluated corresponding blood smears for any morphologic changes due to storage.

Materials and Methods

Animals.

Male (n = 12; age, 9 wk) and female (n = 12; age, 11 wk) Sprague–Dawley rats were obtained from Harlan Laboratories (Frederick, MD). Rats were negative for Sendai virus, pneumonia virus of mice, sialodacryoadenitis virus, Kilham rat virus, H1 virus, rat minute virus, reovirus, rat theilovirus, lymphocytic choriomeningitis virus, hantavirus, mouse adenovirus, rat parvovirus, rat respiratory virus, Bordetella bronchiseptica, Clostridium piliformis, Corynebacterium kutscheri, Mycoplasma pulmonis, Salmonella spp., Streptobacillus moniliformis, Streptococcus pneumoniae, Helicobacter hepaticus, Helicobacter bilis and other Helicobacter spp., Klebsiella oxytoca, Klebsiella pneumoniae, Pasteurella multocida, Pasteurella pneumotropica, Pasteurella aeruginosa, S. aureus, and β-hemolytic Streptococcus spp. Animals were also free of endo- and ectoparasites. No pathogens were detected in sentinel rats during this study. Rats were pair-housed on autoclaved, hardwood bedding (Sani-Chips, PJ Murphy, Montville, NJ) in solid-bottom polycarbonate cages and maintained on a 12:12-h light:dark cycle at 22 ± 0.5 °C and a relative humidity of 40% to 60%. Rats were provided an autoclaved rodent diet (NIH31, Zeigler Brothers, Gardners, PA) and deionized water treated by reverse osmosis ad libitum. All procedures performed were approved by the National Institute of Environmental Health Sciences Animal Care and Use Committee. All animals were housed, cared for, and used in compliance with the Guide for the Care and Use of Laboratory Animals in an AAALAC-accredited program.¹⁴

Blood sampling method and sample handling.

Terminal blood samples were collected from all rats within 1 h on 1 d by cardiocentesis under sodium phenobarbital anesthesia and divided to completely fill two 2-mL K₃EDTA-containing tubes (Becton–Dickinson, Franklin Lakes, NJ). Samples from 1 male and 1 female rat clotted and were excluded from the study. The paired tubes then were stored at either 3 °C or 21 °C (ambient room temperature).

Blood analysis.

Within 5 min after collection, one tube from each rat was analyzed (time, 0 h) by using an automated analyzer (Hemavet 1700, Drew Scientific, Waterbury, CT) and multispecies software (Drew Scientific). After 6, 24, 48, and 72 h of storage at 3 or 21 °C, the samples were mixed gently and analyzed again. Refrigerated samples were warmed to room temperature for 30 min before analysis. In addition, blood smears stained with Wright–Geimsa stain from both the 3 and 21 °C blood samples were made at each time point. The following variables were assessed: RBC count, hemoglobin (Hgb) concentration, Hct (instrument-derived), MCV, MCHC, MCH, total WBC count, platelet count, and mean platelet volume (MPV). The automated analyzer uses the principal of electrical impedance and a patented focused flow system to measure cell counts and size. Hgb is measured spectrophotometrically by using the cyanmethemoglobin method. The parameters of RBC count, Hgb concentration, and MCV are used to calculate Hct, MCHC, and MCH. Quality control was assessed on each day of analysis, according to the manufacturer's protocols. The manually determined PCV was measured directly by using an Autocrit Ultra 3 (Becton–Dickinson). A 100-WBC differential count was performed by using the monolayer of the blood smear and a 60× oil objective; the various leukocyte populations were recorded as percentages. The number of ruptured or partially disintegrated leukocytes (smudge cells) observed per 100 WBC was recorded. The number of leukocytes with shrunken nuclei and chromatin that was condensed to a solid mass or masses (pyknotic leukocytes) observed per 100 WBC was recorded also.

Five high-power (100× oil objective) fields in the monolayer of the blood smear were examined for morphologic changes in erythrocytes; changes were categorized as previously described.²³ Briefly, the grading criteria were: occasional, present only in random fields; 1+, 1 to 5 altered RBC present in each field; 2+, an average of 6 to 15 in each field; 3+, 16 to 25 in each field; and 4+, more than 25 present in each field. The feathered edge was scanned at 10× and 40× for platelet clumps; platelet clumps within the monolayer were noted also. All of the hematologic analyses and blood smear examinations were performed by an author (DK) who is a certified medical technologist.

Statistical analysis.

To ensure that changes related to storage were not overtly different between the sexes and to minimize heterogeneity of the data due to age-, weight-, and sex-associated differences, data from male and female rats were analyzed separately. To test the effects of storage temperature (3 and 21 °C) and storage time (6, 24, 48, and 72 h), the data were analyzed by using a one-way repeated-measures design, with the levels being each temperature–time combination, including the common initial time 0 (control) group. Assumptions were tested by using the Shapiro–Wilks test for normality and Browns–Forsythe test for homogeneity of variance. For endpoints at which the assumptions were not met, Brunner and colleagues’³ nonparametric ANOVA for longitudinal data was used. For statistical comparison, the manual WBC differential percentages were used instead of the calculated differential absolute numbers, because any changes in the manual WBC differential percentages would then be unaffected by changes that may have occurred in the automated WBC count.

We first tested whether an overall difference among the temperature–time combinations was present; if so, we used contrasts to determine which temperature–time combinations differed from the control. Specifically, for a given temperature, we tested for a storage time effect by comparing hematologic values at each observed time with the initial hematologic measurements collected at time 0 (control). Similarly, to test for a temperature effect, we compared the 3 and 21 °C temperature values at each storage time point. We also tested for a trend over time within each storage temperature. The weighted κ statistic was used to test the agreement of the significance and direction of trend between the 2 sexes.²¹

Computations were carried out by using the Proc Rank and Proc Mixed protocols in SAS 9.1 (SAS Institute, Cary, NC). P values of 0.01 or less were considered statistically significant. To account for multiple comparisons, we used Bonferroni-adjusted P values. As a descriptive measure, mean percentage change from the initial control value was calculated for each temperature–time combination. Prior to analysis, extreme values identified by the outlier test of Dixon and Massey⁷ were examined; no outliers were removed for this analysis.

Results

CBC.

Effects of storage were comparable in male and female rats (κ statistic, 0.64; P ≤ 0.001); we therefore present and discuss only the results from male rats. RBC count and Hgb concentration remained unchanged for 72 h at both temperatures (Table 1). Compared with the time 0 value, Hct was unchanged by storage for 72 h at 3 °C but increased (P ≤ 0.01) after 48 h at 21 °C (Table 1). At 3 °C, PCV did not change throughout the study, whereas at 21 °C it was higher (P ≤ 0.01) at 24 h compared with the value at 0 h.

Table 1.

Hematologic data (mean [1 SD]) from 11 male Sprague–Dawley rats analyzed within 1 h (0 h) of blood collection and after storage of samples at 3 °C (refrigerator) and 21 °C (room temperature) for 6, 24, 48, and 72 h

		3 °C				21 °C
	0 h	6 h	24 h	48 h	72 h	6 h	24 h	48 h	72 h
RBC (× 1012/L)	6.20 (0.38)	6.11 (0.22)	6.09 (0.29)	6.07 (0.33)	6.07 (0.31)	6.11 (0.31)	6.13 (0.35)	6.10 (0.24)	6.05 (0.32)
Hgb (g/dL)	12.9 (0.7)	12.9 (0.5)	12.9 (0.7)	12.9 (0.6)	12.9 (0.7)	12.9 (0.6)	12.9 (0.6)	12.9 (0.6)	12.9 (0.7)
Hct (L/L)	39.5 (2.1)	40.5 (1.3)	40.6 (2.0)	40.7 (1.7)	40.8 (1.9)	40.8 (1.7)c	40.8 (2.1)	41.1 (1.2)a	41.3 (1.7)a
PCV (%)	40.5 (1.5)	40.6 (1.5)	40.6 (1.5)	40.6 (1.7)	40.9 (1.4)	40.7 (1.6)c	43.9 (1.8)ab	45.1 (1.9)ab	45.3 (1.8)ab
MCV (fL)	63.7 (2.1)	66.3 (1.5)ac	66.6 (1.8)a	67.1 (1.6)a	67.2 (1.9)a	66.8 (1.5)abc	66.5 (1.8)a	67.5 (1.6)a	68.3 (1.7)ab
MCH (pg)	20.9 (0.8)	21.1 (0.4)	21.2 (0.6)	21.3 (0.7)	21.3 (0.7)	21.1 (0.7)	21.1 (0.6)	21.2 (0.8)	21.3 (0.6)
MCHC (g/L)	32.8 (0.9)	31.9 (0.8)	31.8 (0.8)a	31.8 (0.6)	31.7 (0.8)	31.6 (0.9)d	31.7 (0.9)	31.4 (1.1)a	31.2 (0.7)a
WBC (X109/L)	7.3 (1.5)	7.2 (1.5)c	7.4 (1.5)	7.7 (1.4)	8.6 (1.3)a	7.1 (1.4)	7.6 (1.7)	7.7 (1.4)	7.6 (1.4)b
Neutrophils (%)	9.2 (3.0)	9.5 (3.9)	8.9 (2.8)	11.0 (5.0)	13.2 (4.8)	10.4 (5.2)	10.3 (3.0)	11.2 (5.2)	11.4 (6.2)
Lymphocytes (%)	87.7 (3.3)	86.3 (2.4)	89.3 (3.1)	86.6 (5.0)	85.2 (4.8)	86.1 (5.6)	87.5 (3.4)	87.2 (5.0)	88.5 (6.0)
Monocytes (%)	2.4 (1.6)	3.9 (2.8)	1.7 (1.5)	2.4 (1.2)	1.6 (1.6)	3.1 (2.4)d	1.8 (1.4)	1.5 (1.1)	0.1 (0.3)a
Eosinophils (%)	0.7 (0.9)	0.4 (0.7)	0.1 (0.3)	0 (0)	0 (0)	0.5 (0.7)	0.4 (0.5)	0.2 (0.4)	0.1 (0.3)
Platelets (×109/L)	939.9 (223.9)	1138.8 (178.4)ac	1372.1 (425.8)a	1610.0 (386.3)a	1291.2 (243.9)a	1231.0 (199.9)ac	1938.6 (471.7)ab	1948.5 (512.7)ab	1426.2 (351.0)a
MPV (fL)	5.4 (0.2)	5.5 (0.2)c	5.9 (0.4)a	6.4 (0.3)a	6.7 (0.2)a	5.1 (0.2)abc	5.9 (0.3)a	6.3 (0.5)a	6.2 (0.4)ab

^aSignificantly (P ≤ 0.01) different from value at time 0 h

^bSignificantly (P ≤ 0.01) different from value at 3 °C

^cSignificant (P ≤ 0.01) positive trend over time

^dSignificant (P ≤ 0.01) negative trend over time

Compared with the 0-h value, MCV was higher (P ≤ 0.01) after 6 h of storage at either temperature. The values plateaued or continued to increase through 72 h, with the magnitude of increase generally greater at 21 °C. Compared with the 0-h value, MCHC was lower (P ≤ 0.01) at 24 h at 3 °C and at 48 h at 21 °C. The MCHC tended to remain lower than the initial sample throughout 72 h, with a significant (P ≤ 0.01) downward trend in these values at 21 °C. MCH was not changed by storage.

Platelet counts were higher (P ≤ 0.01) after 6 h of storage at either temperature than at the respective 0-h counts. This change was consistent for 72 h, with the magnitude of the change starting at 24% and becoming as high as 115%. The magnitude of the increase in platelet count was similar between the 2 storage temperatures for most time points. At 21 °C, MPV was lower (P ≤ 0.01) after 6 h and then higher (P ≤ 0.01) after 24 h, whereas at 3 °C, MPV was higher (P ≤ 0.01) at 24 h. The value at both temperatures remained constant or continued to increase through 72 h; however, the magnitudes of the increases were greater at 3 °C than at 21 °C.

The WBC count was higher (P ≤ 0.01) after storage for 72 h at 3 °C but was unchanged after 72 h at 21 °C, compared with counts at 0 h. Neutrophil percentages were unchanged at both temperatures. Regardless of storage duration or method, lymphocyte percentages were unchanged. Compared with data at 0 h, monocytes percentages were lower (P ≤ 0.01) after 72 h at 21 °C, with no change at 3 °C. Eosinophils were observed at low numbers at time 0, and essentially no eosinophils were identified after 72 h at either temperature; however, these differences were not statistically significant.

Blood smear findings.

Compared with 0-h values, the number of smudge cells (Figure 1) counted per 100 WBCs was higher (P ≤ 0.01) after 24 and 48 h of storage at 21 °C and 3 °C, respectively (data not shown). The number of smudge cells continued to increase through 72 h, with significantly (P ≤ 0.01) higher numbers at 21 °C than at 3 °C. The mean percentage change in number of smudge cells increased over time, reaching as high as 1154%. The number of pyknotic leukocytes (Figure 2) per 100 WBC increased (P ≤ 0.01) at 24 h at both temperatures. This increase continued and was generally greater at 21 °C than at 3 °C (data not shown).

Figure 1.

Two smudge cells (arrows) on a peripheral blood smear from a Sprague–Dawley rat made 6 h after storage at 3 °C. Wright–Giemsa stain; bar, 50 μm.

Figure 2.

A pyknotic leukocyte on a peripheral blood smear from a Sprague–Dawley rat made 72 h after storage at 3 °C. Wright–Giemsa stain; bar, 50 μm.

Prominent morphologic changes in erythrocytes included crenated erythrocytes (echinocytes) and dark-staining crenated erythrocytes with loss of central pallor (spheroechinocytes; Figure 3). Echinocytes were observed by 6 h at both temperatures, with average scores of 2.5+ at 21 °C and 2.0+ at 3 °C. Spheroechinocytes were observed by 6 h at both temperatures, with average scores of 2.5+ and 2.0+ at 21 °C and 3 °C, respectively. Over time, echinocytes numbers decreased and spheroechinocyte numbers increased, with most specimens having a score of 4+ spheroechinocytes by 48 h (data not shown). Platelet aggregates were observed along the feathered edge as early as 0 h and increased in number with increasing storage times (data not shown). No significant difference was noted between the 2 temperatures; platelet aggregates seemed to be more frequent in smears from blood stored at 21 °C.

Figure 3.

Peripheral blood smears from a Sprague–Dawley rat. (A) Fresh blood smear (time, 0 h). (B) Few echinocytes (arrow) and spheroechinocytes (arrowhead) are present on a blood smear made 6 h after storage at 3 °C. (C) Numerous spheroechinocytes (arrowheads) and few echinocytes (arrow) are present on a blood smear made 48 h after storage at 3 °C. Wright–Giemsa stain; bar, 50 μm.

Discussion

The hematologic parameters of RBC count, Hgb concentration, and Hct are quantitative estimates of the circulating erythroid mass and measures of the oxygen-carrying capacity of the blood. Most impedance-based hematology analyzers, including that used in the current study, directly measure the RBC count and Hgb concentration. In contrast, Hct is a calculated variable, with the analyzer using the RBC count and another directly measured parameter, MCV, to derive the value (Hct = [MCV × RBC count] / 10). This study showed that after storage for 72 h at 3 °C (refrigerator temperature) and 21 °C (room temperature), the RBC count and Hgb concentration were unchanged, consistent with many other studies.^{1,5,8,9,11,15,17,24} Unlike the RBC count and Hgb concentration, Hct was increased at some time points, mainly due to increases in MCV.

MCV is a measurement of erythrocyte volume. In the current study, MCV increased after 6 h of storage at either temperature. This change is consistent with erythrocyte swelling and subsequent increases in MCV, which has been described in previous similar studies, including those of rats, dogs, and humans.^{9,11,15,17,24} Compared with those of other species, rat erythrocytes appear to be more sensitive to the effects of storage. For example, MCV remained unchanged for 48 h at 4 °C in humans and rabbits.^1,11,24 Increases in MCV are important to be aware of, because these artifactual changes in MCV could mask microcytosis or lead to erroneous diagnoses of macrocytosis.

Storage-related swelling of erythrocytes affected other hematologic parameters. MCHC, an estimate of the average Hgb concentration within erythrocytes, is another calculated value, requiring the Hct value for its calculation (MCHC = [Hgb concentration × 100] / Hct); MCHC therefore indirectly depends on the measured MCV value. In the current study, MCHC often decreased with storage time. In addition, the significant increases in the manually determined PCV can be attributed, at least in part, to erythrocyte swelling, given that larger erythrocytes will comprise a larger percentage of the total blood volume. Not unexpected was the finding that after storage in the refrigerator, both the PCV and Hct were unchanged throughout the 72 h, reinforcing the recommendation that with delays in analysis, specimens should be stored in the refrigerator to decrease preanalytical variability.

In the current study, platelet counts were significantly higher after storage at 6 h, with no clear differences between the 2 storage temperatures. These findings were not consistent with other studies in rats, dogs, or humans, in which platelet counts, as measured by impedance technology, were essentially unchanged or decreased over time.^9-11,15 In human studies using impedance technology, platelet counts were relatively unaltered through 96 h at both refrigerator and room temperatures, with no significant changes or a mean percentage change of only 6%.^9,11 Canine blood stored at either room or refrigerated temperatures had decreased platelet counts.^10,15,17 Another study¹⁰ reported a mean percentage change of platelet counts in rats of less than or equal to –7% after cold storage for 24 h. The reason for the increased platelet counts in the current study is unclear. Studies that reported decreased platelet counts with storage attributed the decrease to platelet aggregation. In the current study, platelet aggregation was observed to increase in the blood smears over time, but this finding was not reflected as a decrease in the platelet counts obtained from the automated analyzer. Erythrocyte microcytes or fragments may be counted as platelets when impedance technology is used for platelet and erythrocyte enumeration.^6,19 This interference seems unlikely to explain the observed increases in platelet numbers given that RBC counts remained unchanged, MCV increased, and no erythrocyte fragments (for example, schistocytes) were observed on the blood smears. Regardless of the reason for the observed increases, to maximize accuracy, platelets should be measured as soon as possible after blood collection, because counts after sample storage at both room and refrigerated temperature showed substantial increases after 6 h. MPV was decreased at 6 h with subsequent increases thereafter, probably representing initial shrinkage, with a combination of platelet swelling and aggregation contributing to the subsequent increases. Similar to platelet count, MPV should be measured soon after blood specimen collection.

The WBC count was unchanged for 72 h at 21 °C but was higher after 72 h at 3 °C. The one known rat study that reported on WBC counts after storage as measured by several different impedance counters was performed on refrigerated blood after storage for 24 h.¹⁰ The previous results were similar to ours in that WBC counts remained relatively unaltered for 24 h at refrigerated temperatures with a mean percentage change of only 4%. In humans, WBC counts were unchanged after storage for as long as 96 h at both temperatures in studies using impedance technology.^9,24 Dog studies using impedance analyzers had mixed results, which sometimes differed from what we observed in the current study.^10,15,17 For example, WBC counts having a mean percentage change of 4% or less were obtained from samples analyzed on several different impedance machines after storage at refrigerator temperatures for 24 h, whereas WBC counts measured under the same conditions on other impedance analyzers showed a mean percentage change of 25% to 58%.¹⁰ Platelet aggregates may account for storage-related increased WBC counts.¹⁵ In the current study, platelet aggregates were observed starting at 6 h of storage and in progressively greater numbers over time. However, the presence of platelet aggregates does not seem to be the only explanation for an increased WBC count, because storage at room temperature had less effect on the WBC count than did refrigeration, but similar increases in platelet aggregates occurred at both temperatures.

In general, neutrophils were unaltered for 48 h at refrigerator temperatures and for 72 h at room temperatures. These findings were accompanied by significant decreases in the monocytes after 72 h at room temperature, as well as a downward progression in eosinophil numbers with essentially no eosinophils identified by 72 h at both temperatures. Very little literature exists on the effects of storage on manual WBC differentials. One study evaluating the stability of human hematologic parameters after storage of blood compared fresh and 24-h manual differentials (reported as percentages) after storage at both refrigerator and room temperatures.²⁴ In that study, no significant changes in WBC percentages were observed after 24 h at refrigerator temperature, but significant decreases in neutrophils and increases in lymphocytes and monocytes were seen after 24 h at room temperature.²⁴ Although all leukocytes degenerate with storage, the findings of the current study suggest that rat lymphocytes and neutrophils may be more resistant to the effects of delayed analysis than are monocytes and eosinophils.

The type and degree of artifact that occurs after storage can vary depending on the species involved and the hematology analyzer that is used.^1,8,13 The basic technology used in our study is similar to that of other referenced studies but the changes we observed sometimes differed from those in previous studies. These differences may be related to machine-associated differences in impedance technology, controls for accuracy (that is, calibration specific to different species), quality control, and inherent species-specific differences in, for example, the stability of leukocytes and tendency to form platelet aggregates over time. These variations serve as reminders that results garnered from one hematology analyzer do not necessarily represent those collected by using another analyzer, even when similar technology is used.

Echinocytes, or crenated erythrocytes, and spheroechinocytes are well-known effects of prolonged blood storage and have been documented in the human literature.^2,4,12 To our knowledge, these morphologic changes have not previously been documented after storage of rat blood. In the current study, both shape changes were observed as soon as 6 h after collection. With storage, a predictable sequence of shape changes has been characterized in human blood. The normal biconcave disc shape undergoes crenation, forming an echinocyte that subsequently swells to form a crenated sphere with multiple spicules called a spheroechinocyte.⁴ Several factors seem to cause this morphologic change, including erythrocyte ATP depletion and disturbances of RBC calcium homeostasis.^2,4 Reasons for nonpathologic echinocyte formation include exposure to alkaline pH, such as may occur with exposure to glass slides or tubes.^12,18 Our study indicates that blood smears should be prepared within 1 h of specimen collection for accurate evaluation of red blood cell morphology.

Smudge cells, also known as basket cells or shadows of Gumprecht, form when a nucleated cell ruptures or is stripped of its cytoplasm during blood smear preparation, leaving a lacy nuclear chromatin that stains eosinophilic or amphophilic.^12,16,22 The presence of smudge cells suggests mechanical damage during smear preparation or increased cellular fragility, as occurs in aged blood or blood containing abnormal cells (leukemia).^12,16,22 We observed significant increases in smudge cells progressively over time in the current study. With storage, smudge cells may represent any leukocyte type, and attempts should not be made to identify the smudge cells.²² When smudge cells exceed 10% of the total WBC, the manual WBC differential count may be invalid.²² The presence of smudge cells are likely to have affected the results of the manual WBC differential count performed for the current study, explaining the changes noted in the percentages of individual leukocyte populations, given that different leukocytes deteriorate at different rates. Pyknotic leukocytes are known to form during prolonged blood sample storage.^12,22

In summary, delayed analysis of Sprague–Dawley rat blood produces artifactual changes in MCV, HCT, platelet count, and MPV. Although some of the observed changes could arguably be within acceptable analytic variability or not clinically relevant, the best practice is to measure hematologic parameters within 6 h of collection. In the event of delayed analysis, specimens should be stored in the refrigerator, and care must be taken to not misinterpret artifactual changes as pathologic findings, preferably omitting specific data for those parameters that are more sensitive to storage and substituting them with a comment.^1,11,20 In addition to the changes documented in the CBC parameters, several well-known cellular changes seen in humans (and other species) also occur in rat blood after storage. To avoid these artifactual cellular changes, blood smears should be made within 1 h of specimen collection.