In the face of the COVID-19 pandemic, efficient outpatient management is paramount to minimise wait times and reduce exposure to respiratory illnesses, particularly in patients with cancer receiving chemotherapy. The complete blood cell count (CBC) with white cell count (WCC) differential (CBC-D) is an essential laboratory test used to screen cancer patients prior to chemotherapy. Patients receiving cytotoxic chemotherapy are at risk for developing neutropenia; the absolute neutrophil count (ANC) can be used to determine if a patient is neutropenic.1 Neutropenia is generally defined as an ANC <1.5×109/L–2.0×109/L, with severe neutropenia defined as <0.5 x 109/L.2
WCC differential analysis is performed by automated haematology platforms in the clinical laboratory. These platforms have been found to provide accurate and precise ANC results.3 However, automated results often require manual review if the analysis is associated with instrument flags that indicate the potential for inaccurate results. Review and confirmation of flagged results by laboratory technologists is time consuming and increases test result turnaround time (TAT), thereby prolongs onset of chemotherapy.4–7 Given that chemotherapy administration requires an adequate ANC, we examined the reliability of the automated ANC in the presence of instrument flags in out patients with solid tumour and haematological malignancies. We aimed to confirm if discrete ordering of a ‘CBC with ANC only’ test could be used to replace the traditional CBC with full WCC differential (CBC-D) and shorten patient wait times.
We performed a retrospective analysis of 1648 sequential CBC with WCC differential results flagged for manual peripheral blood smear review performed for patients with cancer. Results across five outpatient locations were analysed during October 2018. The cases were separated based on malignancy type, solid tumour (n = 1078) or haematological malignancies (n=570). The accuracy of automated ANC results (Sysmex XN) was assessed by comparison to the manual differential using Rumke statistics.8 Manual review of automated differential results was performed by experienced clinical laboratory technologists and consisted of either a peripheral blood smear review with confirmation of automated differential results or a full 100 cell manual differential.
Differential review flags associated with automated CBC-D results are summarised in table 1. Multiple flags were frequently observed in the same sample. As shown in figure 1, the automated differential was confirmed by peripheral blood smear review in 327 of 1078 (30%) samples from patients with solid tumour malignancies. A full manual differential was performed in the remaining 751 of 1078 cases (70%). In contrast, the automated differential was confirmed by peripheral blood smear review in 108 of 570 (19%) samples from patients with haematological malignancies, and a full manual differential was performed in the remaining 462 of 570 cases (81%). The correlation between manual and automated ANC results was strong for patients with solid tumour and those with haematological malignancies with a R2 of 0.99 (slope=0.89) and a R2 of 0.96 (slope=0.91), respectively, despite instrument flags (figure 2). Subset analysis of haematological malignancies showed a similarly strong correlation between automated and manual ANC results for samples from patients with leukaemia, lymphoma, plasma cell dyscrasia, bone marrow transplant (BMT), myelodysplastic syndrome (MDS), myeloproliferative neoplasm (MPN) and non-malignant haematological disorders (table 2).
Table 1.
Distribution of differential flags requiring manual peripheral blood smear review
| Instrument flags | No of flagged instances | |
|---|---|---|
| Solid tumour malignancies, % | Haematological malignancies, % | |
| Nucleated red blood cells (NRBC) | 512 (33) | 256 (20) |
| Blasts/abnormal lymphoid | 386 (25) | 297 (23) |
| Monocytosis (≥1.5×103/μL or ≥30%) | 106 (7) | 167 (13) |
| Immature granulocytes | 189 (12) | 137 (11) |
| WCC abnormal scattergram | 123 (8) | 112 (9) |
| Atypical lymphoid | 127 (8) | 51 (4) |
| Previous sample with blasts | 65 (4) | 167 (13) |
| Eosinophilia (≥20%) | 35 (2) | 10 (0.8) |
| Absolute lymphocytosis (≥5×103/μL) | 25 (2) | 93 (7) |
| Basophilia (≥1×103/μL or ≥5%) | 4 (0.3) | 9 (0.7) |
| New patient with WCC >11×109/L | 1 (0.1) | 1 (0.1) |
| Total flags | 1573 | 1300 |
WCC, white cell count.
Figure 1.
Study design to evaluate automated ANC performance in flagged CBC-D samples. ANC, absolute neutrophil count; CBC-D, complete blood cell count differential.
Figure 2.
Correlation between manual and flagged automated ANC results in patients with solid tumour malignancies (top) and in patients with haematological malignancies (bottom). ANC, absolute neutrophil count.
Table 2.
ANC correlation in samples from patients with haematological malignancies
| Haematological malignancy type | Total cases, % | ANC confirmed by smear review, % | Manual differential performed, % | Automated and manual differential correlation characteristics | |
|---|---|---|---|---|---|
| R2 | Slope | ||||
| Lymphoma | 196 (34) | 35 (6) | 161 (28) | 0.97 | 0.85 |
| Leukaemia | 189 (33) | 18 (3) | 171 (30) | 0.95 | 0.99 |
| Plasma cell dyscrasia* | 59 (10) | 21 (3) | 38 (6) | 0.97 | 0.84 |
| BMT | 56 (10) | 15 (3) | 41 (7) | 0.97 | 0.85 |
| MDS | 38 (6) | 13 (2) | 25 (4) | 0.99 | 0.89 |
| MPN | 25 (4) | 6 (1) | 19 (3) | 0.99 | 0.94 |
| Non-malignant haematological disorders | 7 (1) | 0 (0) | 7 (100) | 0.98 | 0.88 |
Plasma cell dyscrasias include multiple myeloma and plasmacytoma.
ANC, absolute neutrophil count; BMT, bone marrow transplant; MDS, myelodysplastic syndrome; MPN, myeloproliferative neoplas
Discordant automated and manual neutrophil results were identified in a total of 247 cases; 102 from patients with solid tumour malignancies and the remaining 145 from patients with haematological malignancies. The majority of these cases (n=217), had an automated ANC >1.0×109/L, and the difference in ANC results was not clinically significant. In general, automated ANC results were slightly lower than manual results. Thirty samples, 10 from patients with solid tumours and 20 from patients with haematological malignancies, were found to have discordant ANC results with an automated ANC of ≤1.0x 109L (figure 3).
Figure 3.
Differences between flagged automated and manual ANC results for samples with an automated ANC ≤1000/μL in patients with solid tumour (n=10) and in haematological (n=20) malignancies. ANC, absolute neutrophil count.
This study investigated the feasibility of reporting a CBC with ANC only in cancer patients with solid tumour or haematological malignancies. The data support creating a distinct test order for ‘CBC with ANC only,’ when the ANC is required for the initiation of chemotherapy, discharge or other intervention. Our results confirm previous reports describing the validity of the automated ANC in a general patient population3,5–7 and that the automated ANC results are consistently marginally lower than the manual count.9 Additionally, this study provides new data on patients with haematological malignancies, subclassifying haematological malignancy by type (leukaemia, lymphoma, MDS, MPN and plasma cell dyscrasias), as well as including patients with recent BMT. Based on our review of discordant manual and automated ANC results, as well as the clinical definition of neutropenia, we suggest manual review of samples with flagged automated differential results, when the ANC is <1.0×109/L.
A ‘CBC with ANC only’ order has the potential to decrease the manual review of CBC with differentials in patients with solid tumour and haematological malignancies by 91% and 75%, respectively (table 3). If clinicians transition from ordering a conventional CBC with differential to the ‘CBC with ANC only’ test option patient wait times can be reduced and laboratory and clinical resources conserved. In our institution, the TAT for a CBC with automated ANC is 10–15 min compared with 30–60 min for a manual ANC. A reduction in manual peripheral blood smear reviews from 1648 to 247 would represent substantial laboratory labour savings. Additional institutional savings are expected based on more efficient management of patient appointments and discharge.
Table 3.
Proposed reduction in automated ANC review of flagged CBC-D samples
| Solid tumour malignancies (N) | Haematological malignancies (N) | |
|---|---|---|
| Current state: manual review of all flagged samples | 1078 | 570 |
| Proposed manual review for flagged samples with automated ANC ≤1000/μL | 102 | 145 |
| Anticipated reduction in manual review for automated ANC in flagged samples | 91% | 75% |
ANC, absolute neutrophil count; CBC-D, complete blood cell count differential.
The need for more efficient patient management is highlighted during the COVID-19 pandemic. Reduction in staffing and the necessity for social distancing in the clinical laboratory as well as patient care areas has led healthcare institutions to reconsider established workflows. A ‘CBC with ANC only’ test offers an opportunity to preserve resources without adversely impacting patient care.
Funding
This study was funded in part by National Cancer Institute (Cancer Center Support, grant P30 CA008748).
Footnotes
Competing interests None declared.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; internally peer reviewed.
This article is made freely available for use in accordance with BMJ’s website terms and conditions for the duration of the covid-19 pandemic or until otherwise determined by BMJ. You may use, download and print the article for any lawful, non-commercial purpose (including text and data mining) provided that all copyright notices and trade marks are retained.
REFERENCES
- 1.Yang B-B, Savin MA, Green M. Prevention of chemotherapy-induced neutropenia with pegfilgrastim: pharmacokinetics and patient outcomes. Chemotherapy 2012;58:387–98. [DOI] [PubMed] [Google Scholar]
- 2.McPherson RA, Pincus MR, Henry JB. Henry’s clinical diagnosis and management by laboratory methods. Philadelphia: Saunders Elsevier, 2011. [Google Scholar]
- 3.Amundsen EK, Urdal P, Hagve T-A, et al. Absolute neutrophil counts from automated hematology instruments are accurate and precise even at very low levels. Am J Clin Pathol 2012;137:862–9. [DOI] [PubMed] [Google Scholar]
- 4.Jatoi A, Jaromin R, Jennings L, et al. Using the absolute neutrophil count as a stand-alone test in a hematology/oncology clinic: an abbreviated test can be preferable. Clin Lab Manage Rev 1998;12:256–60. [PubMed] [Google Scholar]
- 5.Sireci AN, Herlitz L, Lee K, et al. Validation and implementation of an algorithm for reporting the automated absolute neutrophil count from selected flagged specimens. Am J Clin Pathol 2010;134:720–5. [DOI] [PubMed] [Google Scholar]
- 6.Friis-Hansen L, Saelsen L, Abildstrøm SZ, et al. An algorithm for applying flagged Sysmex XE-2100 absolute neutrophil counts in clinical practice. Eur J Haematol 2008;81:140–53. [DOI] [PubMed] [Google Scholar]
- 7.Stephens L, Hintz-Prunty W, Bengtsson H, et al. Impact of integrating Rumke statistics to assist with choosing between automated hematology analyzer differentials vs manual differentials. JALM 2017. [DOI] [PubMed] [Google Scholar]
- 8.Rumke C The statistically expected variability in differential counting. In: Koepke J, ed. Cap conference. Aspen, CO: College of American Pathologists, 1977: 1978. 39–45. [Google Scholar]
- 9.Ruzicka K, Veitl M, Thalhammer-Scherrer R, et al. The new hematology analyzer Sysmex XE-2100: performance evaluation of a novel white blood cell differential technology. Arch Pathol Lab Med 2001;125:391–6. [DOI] [PubMed] [Google Scholar]