Home Improving turn-around times in low-throughput distributed hematology laboratory settings with the CellaVision® DC-1 instrument
Article Open Access

Improving turn-around times in low-throughput distributed hematology laboratory settings with the CellaVision® DC-1 instrument

  • Cheri Mayes , Tracey Gwilliam and Etienne R. Mahe ORCID logo EMAIL logo
Published/Copyright: November 7, 2023
Download Article (PDF)
Become an author with De Gruyter Brill
Submit Manuscript Author Information
Journal of Laboratory Medicine
From the journal Journal of Laboratory Medicine Volume 49 Issue 2

Abstract

Objectives

Digital pathology is becoming standard in the delivery of timely, high-quality clinical services, inclusive of morphological assessment in laboratory hematology. While many digital hematology systems are designed with high-throughput in mind, CellaVision® has recently developed a low-throughput instrument, the CellaVision® DC-1. The utility of the CellaVision® DC-1 was tested in a distributed laboratory system, with a focus on turn-around times (TATs).

Methods

We evaluated the TATs of a CellaVision® DC-1 workflow, with specimens originating in a small spoke-laboratory referring materials to a central hub-laboratory. Our spoke-laboratories perform on-site complete blood counts (CBC’s) and manual peripheral blood smears (PBS’s), with complex cases referred for review to the hub-laboratory. Baseline TATs were collected, followed by prospective evaluation of 21 cases analyzed using the CellaVision® DC-1, with digital review by spoke-laboratory staff in concert with remote review by hub-laboratory staff. The TATs for the same 21 cases by standard manual assessment were compared.

Results

Improvement in the distribution of TATs using the CellaVision® DC-1 was noted relative to the retrospective spoke-laboratory data (Mann–Whitney U=26, p<0.0001) and the parallel manual PBS review (Wilcoxon W=190, p<0.0001). The CellaVision® DC-1 permitted a significant reduction in case-assessment times (Wilcoxon W=105, p=0.0001). No significant diagnostic discrepancies were identified during the testing timeframe.

Conclusions

We describe a real-world assessment of the CellaVision® DC-1 analyzer in a distributed (hub-and-spoke) laboratory network, linking low-volume laboratories to high-throughput sites. Our evaluation highlights significant improvements in case TATs with a CellaVision® DC-1 assisted digital pathology workflow.

Keywords: digital hematology; distributed laboratory network; turn-around-time

Introduction

Digital pathology is becoming increasingly important in routine clinical laboratory practice and many advances in digital pathology on the whole have been facilitated by advances in digital hematology. Indeed, digitized hematology images featured in one of the earliest telepathology experiments undertaken in North America [1]. Furthermore, in parallel to many of the ground-breaking technological advances central to digital anatomical pathology, numerous advances initially supporting digital hematology were being made [2], [3], [4], [5], [6].

While a number of digital hematology instruments are currently available, the CellaVision® suite is an industry leader in the area [2]. CellaVision® instruments employ a combination of high power microscopy and efficient optical processing, as well as artificial neural network-based computational analysis, to rapidly identify and pre-classify the cellular components of peripheral blood smears (PBS). Cells are automatically pre-classified (i.e., assigned a preliminary cell identification) and counted; the user is then able to review and alter or adjust the CellaVision® pre-classification. The CellaVision® algorithms have demonstrated excellent pre-classification accuracy relative to manual assessment [7], 8]; the CellaVision® differential counting functions have proven accurate and reproducible [8], [9], [10], [11]; and studies of blood film analysis turn-around times (TATs) strongly support the use of systems such as CellaVision® in routine clinical practice [12]. CellaVision® also provides capabilities to perform slide reviews remotely by using CellaVision® Server and Remote Review Software.

Many laboratory networks employ a hub-and-spoke service delivery model. Alberta Precision Laboratories, for example, provides clinical laboratory testing and referral services for a geographical catchment area of over 80,000 km2, including both large urban centers (“hub-laboratories”) as well as small and remote communities (“spoke-laboratories”). As with most comparable laboratory systems [13], much of our test volume derives from hematology, both in hub- and spoke-laboratories. In some instances, expert morphologic assessment of PBS is required, thereby necessitating the transport of materials from spoke to hub. In our experience, this process may be subject to commensurate TAT delays.

The new low-throughput CellaVision® DC-1 was developed in order to offer a semi-automated advantage to low test-volume laboratories [14]. In addition to the potential advantages that such a system might have for reducing hands-on time, this system has the theoretical advantage of interpretation of a blood film produced in a small community-based spoke-laboratory by large-volume experts in a hub-laboratory [14]. Using Remote Review capabilities, experts available in a hub-laboratory can review digital images of the slide remotely, thereby obviating the need to physically transport the slides from spoke to hub laboratories. The latter might permit reductions in test TAT if expert review in a hub-laboratory is required.

Materials and methods

This work was undertaken after formal review and approval by the Calgary Health Research Ethics Board (REB17-2133).

Part I: historical review of spoke-laboratory data

We began by performing a baseline TAT evaluation of PBS from our pre-selected spoke-laboratory (High River, Alberta, Canada). This process involved a retrospective assessment of all PBS evaluation time-stamps logged at the spoke-laboratory in the months from March to June, 2017, spanning backward to the time stamp of CBC request. Of note, CBC requests did not arise from specimen receipt by the hub-laboratory (the Diagnostic and Scientific Center, Calgary, Alberta, Canada) and the span-back period was restricted to the March–June window.

Since a priori knowledge of the expected CellaVision® DC-1 TAT distribution in the hub-and-spoke context was unavailable, we used the above data to derive an estimate of the sample size required for reasonable testing of a proposed TAT improvement workflow using the CellaVision® DC-1. For ease of calculation, sample size estimation was made by way of up-scaling the sample size computation for the Student’s t-test under the assumption of normality [15].

Part II: comparison of CellaVision® DC-1 workflow to standard-of-care

Figure 1 provides a schematic summary of the study protocol specific to Part II.

Figure 1: Study schematic, specific to Part II.
Figure 1:

Study schematic, specific to Part II.

Cases reflexed (by automated analyzer flagging) for PBS were scanned using the CellaVision® DC-1 and analyzed using CellaVision® Peripheral Blood Application software for cell pre-classification. Both the CellaVision® DC-1 analyzer and related software were provided for the evaluation by CellaVision® AB (Lund, Sweden). Pre-classifications were then reviewed by the spoke-laboratory technical staff, and in consultation with hub-laboratory pathologists using the CellaVision® Remote Review Software. The time-stamps of receipt and completion for each process step were logged, with particular attention paid to the time-stamps from initial evaluation of the PBS in the spoke-laboratory to the time-stamp of evaluation in the hub-laboratory. In parallel, the generated slides were assessed using normal protocols by way of a manual assessment by the spoke-laboratory technical staff, followed by transport of slides to the hub-laboratory for pathologist review.

Part III: CellaVision® software classification vs. manual classification

We undertook an exercise of TAT evaluation from the perspective of the CellaVision® classification analysis workflow. For this, the period of time required for evaluation of a given slide using the CellaVision® Peripheral Blood Application (after automated scanning) was compared with matched manual evaluation of glass slides under a microscope. This exercise was undertaken across the spectrum of technologist expertise (including junior and senior technologists), over several work-time periods (addressing potential performance differences based on time of day) and at several laboratory sites (accounting for potential patient differences and case-type differences between our hub-laboratories). A total of 14 timed exercises, each with a range of 20–50 slides per session were undertaken.

Results

Part I

The TAT data for CBC-only, local PBS review only and central PBS review are presented in Figure 2. These data highlight the extreme discrepancy in median TAT when hub-laboratory referral is required (Figure 1, referral PBS) relative to spoke-laboratory review only (Figure 1, local PBS). It should also be noted that hub-laboratory referral TATs appeared relatively consistent, irrespective of STAT or routine levels of urgency.

Figure 2: Historical TAT data (originating from the spoke-laboratory) for CBC, local (spoke-laboratory) PBS and central (hub-laboratory) PBS review for STAT and routine priority cases.
Figure 2:

Historical TAT data (originating from the spoke-laboratory) for CBC, local (spoke-laboratory) PBS and central (hub-laboratory) PBS review for STAT and routine priority cases.

Part II

For the purposes of sample size calculation, assuming a desired reduction of median TAT to 120 min on average, we estimated that approximately 30 blood films evaluated using the CellaVision® DC-1 workflow would be sufficient (with a desired power of at least 0.95).

We evaluated the CellaVision® DC-1 prospectively on 32 consecutive cases flagged for hub-laboratory PBS review subsequent to CBC evaluation by the spoke-laboratory. In accordance with spoke-laboratory standard operating procedures, cases flagged for review included cases flagged by the local automated CBC analyzer and/or cases for which clinical history or other clinical parameters indicated a need for PBS review. Seven of the 32 cases were later rejected owing to incomplete time-stamp data or irregularities of collection/analysis or PBS indication. Figure 3 highlights the relative frequencies of analyzer flags, noting that possible lymphocyte abnormalities were the most common analyzer indications for a PBS.

Figure 3: Analyzer flags (possible abnormalities identified by analyzer), relative frequency. RBC, red blood cell.
Figure 3:

Analyzer flags (possible abnormalities identified by analyzer), relative frequency. RBC, red blood cell.

We compared the distributions of TATs before and after the implementation of the CellaVision® DC-1 workflow from the point of referral from hub-to-spoke laboratory. A total of 21 cases assessed on the CellaVision® DC-1 could be analyzed (as some additional cases were rejected owing to cancellation of review during transport to the hub-laboratory). Figure 4 (upper panel) compares the distribution of PBS referral TATs from baseline (i.e., pre-automation data from Part I) with those logged after implementation of the CellaVision® DC-1 workflow; a significant difference is highlighted by the Mann–Whitney U test (U=26, p<0.0001; baseline median 1794 min vs. CellaVision® DC-1 assisted workflow median of 82 min).

Figure 4: TAT distributions compared: upper: CellaVision® workflow compared to baseline (historical) workflow TATs; lower: CellaVision® workflow compared to matched manual workflow TAT.
Figure 4:

TAT distributions compared: upper: CellaVision® workflow compared to baseline (historical) workflow TATs; lower: CellaVision® workflow compared to matched manual workflow TAT.

We also used the logged time-stamps of case review request at the hub-laboratory and compared the TAT to pathologist review with CellaVision® with the TAT for physical slide review of the same slides. Figure 4 (lower panel) highlights the significantly different distribution of pathologist TAT, with a markedly improved median TAT when the CellaVision® DC-1 workflow was employed (W=190, p<0.0001; manual pathologist PBS review median TAT 1447 min vs. CellaVision® DC-1 assisted pathologist PBS review median TAT 82 min). The noted outlier CellaVision® DC-1 assisted pathology review TATs related to overnight requests for review that were received and acknowledged the following day.

Part III

We undertook an exercise of TAT evaluation from the perspective of the CellaVision® classification analysis workflow. As highlighted in Figure 5, the CellaVision® software workflow demonstrated a superior TAT relative to a manual classification (mean 1.92 vs. 4.05 min; Wilcoxon matched-pairs signed rank test p=0.0001).

Figure 5: Comparing CellaVision® classification to manual slide review and classification.
Figure 5:

Comparing CellaVision® classification to manual slide review and classification.

Discussion

We undertook a real-world evaluation of the CellaVision® DC-1 semi-automated cell morphology assessment system. This study was aimed to assess the impact that the CellaVision® DC-1 might have on improving the TATs in a distributed laboratory network setting. By selecting a spoke-laboratory with relatively infrequent PBS volumes, but at a notable geographic distance from the nearest hub-laboratory, we aimed to establish how a CellaVision® DC-1-assisted workflow might improve TATs for PBS review. We also undertook a rigorous evaluation of the TAT differences of CellaVision® software classification and manual microscope-based classification.

We noted significant improvements in TATs in all measured aspects, including time to central laboratory and time to pathologist PBS review. In addition, we noted a significant difference in TAT in favor of CellaVision® software-based classification as compared to manual microscopic slide review. In addition, the resounding majority of the feedback received from DC-1 users was positive, including strong recommendations in favor of high image quality and usability.

It bears noting that our study sample size is small, and indeed smaller than our a priori desired sample size. The latter could not be met in light of post-study data review that resulted in the rejection of several cases that were inappropriately accrued to the study. Nevertheless, we feel that our limited data (and the resulting statistical analyses) are compelling. We are also unable to comment on the direct impact that improved TAT might have on the clinical outcomes of the patients whose specimens were included in the study. However, we are confident that general improvements in TAT will serve to improve the quality of laboratory performance.

Conclusions

Our study provides strong real-world support for the DC-1 automated cell morphology assessment system. In addition to highlighting the TAT advantages of the CellaVision® system, our study highlights the specific benefit of the CellaVision® DC-1 platform to link larger referral laboratories with smaller low-throughput laboratories that could be outfitted with a CellaVision® DC-1 instrument.

Corresponding author: Etienne R. Mahe, Department of Pathology and Laboratory Medicine, University of Calgary, Calgary, Alberta, Canada, Phone: 403 944 5290, E-mail:

Funding source: Calgary Laboratory Services

Award Identifier / Grant number: Unassigned

Acknowledgments

The authors acknowledge Alberta Precision Laboratories (formerly Calgary Laboratory Services) for their support of this work.

  1. Research ethics: This work was undertaken after formal review and approval by the Calgary Health Research Ethics Board (REB17-2133).

  2. Informed consent: Research ethics board waiver of patient consent was granted, as this study did not include direct interaction with patients.

  3. Author contributions: The authors have accepted responsibility for the entire content of this manuscript and approved its submission. The study was conceived by EM. Logistics related to the study activities were led by TG. Laboratory activities were overseen by CM. Statistical analyses and manuscript writing were performed by EM.

  4. Competing interests: The authors state no conflicts of interest relevant to the contents of the manuscript.

  5. Research funding: Both the CellaVision® DC-1 analyzer and all related software were provided by CellaVision AB, which was also involved in the study design. The authors received no financial compensation for executing the study.

  6. Data availability: Data are available for use by external parties, provided that any such external parties receive research ethics approval from our regional research ethics board for use thereof.

References

1. Murphy, RL, Bird, KT. Telediagnosis: a new community health resource. Observations on the feasibility of telediagnosis based on 1000 patient transactions. Am J Public Health 1974;64:113–9. https://doi.org/10.2105/ajph.64.2.113.Search in Google Scholar PubMed PubMed Central

2. Kratz, A, Lee, S, Zini, G, Riedl, JA, Hur, M, Machin, S, et al.. Digital morphology analyzers in hematology: ICSH review and recommendations. Int J Lab Hematol 2019;41:437–7. https://doi.org/10.1111/ijlh.13042.Search in Google Scholar PubMed

3. Weinstein, RS, Holcomb, MJ, Krupinski, EA. Invention and early history of telepathology (1985–2000). J Pathol Inf 2019;10:1. https://doi.org/10.4103/jpi.jpi_71_18.Search in Google Scholar PubMed PubMed Central

4. Fanghraeus, C, Söderlund, P, Sjöstedt, L, Segersten, R. to CellaVision AB. Device for optical analysis of specimens. US 6,014,210. CI 356-244.000, 1997.Search in Google Scholar

5. Pettersson, M, Rosenqvist, A. assignors to CellaVision AB. Image content autofocus for microscopy using a noise-insensitive focus filter. US 60/071,176. Cl 382-255:1997.Search in Google Scholar

6. Pettersson, M, Rosenqvist, A, Heyden, A, Almers, M, to CellaVision AB. Feature-free registration of dissimilar images using a robust similarity metric. US6268611B1; 2001.Search in Google Scholar

7. Cornet, E, Perol, JP, Troussard, X. Performance evaluation and relevance of the CellaVisionTM DM96 system in routine analysis and in patients with malignant hematological diseases. Int J Lab Hematol 2008;30:536–42. https://doi.org/10.1111/j.1751-553x.2007.00996.x.Search in Google Scholar

8. Lee, LH, Mansoor, A, Wood, B, Nelson, H, Higa, D, Naugler, C. Performance of CellaVision DM96 in leukocyte classification. J Pathol Inform 2013;4:14. https://doi.org/10.4103/2153-3539.114205.Search in Google Scholar PubMed PubMed Central

9. Horn, CL, Mansoor, A, Wood, B, Nelson, H, Higa, D, Lee, LH, et al.. Performance of the CellaVision DM96 system for detecting red blood cell morphologic abnormalities. J Pathol Inform 2015;6:11. https://doi.org/10.4103/2153-3539.151922.Search in Google Scholar PubMed PubMed Central

10. Gao, Y, Mansoor, A, Wood, B, Nelson, H, Higa, D, Naugler, C. Platelet count estimation using the CellaVision DM96 system. J Pathol Inform 2013;4:16. https://doi.org/10.4103/2153-3539.114207.Search in Google Scholar PubMed PubMed Central

11. Mahe, ER, Higa, D, Naugler, C, Mansoor, A, Shabani-Rad, MT. Accuracy of the CellaVision DM96 platform for reticulocyte counting. J Pathol Inform 2014;5:17. https://doi.org/10.4103/2153-3539.133127.Search in Google Scholar PubMed PubMed Central

12. Kratz, A, Bengtsson, HI, Casey, JE, Keefe, JM, Beatrice, GH, Grzybek, DY, et al.. Performance evaluation of the CellaVision DM96 system: WBC differentials by automated digital image analysis supported by an artificial neural network. Am J Clin Pathol 2005;124:770–81. https://doi.org/10.1309/xmb9k0j41lhlatay.Search in Google Scholar

13. Horton, S, Fleming, KA, Kuti, M, Looi, LM, Pai, SA, Sayed, S, et al.. The top 25 laboratory tests by volume and revenue in five different countries. Am J Clin Pathol 2019;151:446–51. https://doi.org/10.1093/ajcp/aqy165.Search in Google Scholar PubMed

14. CellaVision® DC-1 Concept Brochure.Search in Google Scholar

15. Nikitin, YY. Asymptotic relative efficiency in testing. Encyclopedia of Mathematics. Available from: http://encyclopediaofmath.org/index.php?title=Asymptotic_relative_efficiency_in_testing&oldid=38458.Search in Google Scholar
Received: 2023-06-18
Accepted: 2023-09-27
Published Online: 2023-11-07
Published in Print: 2025-04-28

© 2023 the author(s), published by De Gruyter, Berlin/Boston

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Frontmatter
  2. Editorial
  3. Can a digital smear review be helpful in the routine haematology laboratory?
  4. Original Articles
  5. The impact of mutational burden, spliceosome and epigenetic regulator mutations on transfusion dependency in dysplastic neoplasms
  6. Improving turn-around times in low-throughput distributed hematology laboratory settings with the CellaVision® DC-1 instrument
  7. Hematology instruments don’t speak the same language: a comparison study between flagging messages of sysmex XN-1000 and alinity H
  8. Platelet clump assessment using the Cellavision peripherical blood application – do we need manual microscopy?
  9. Short Communication
  10. Development of a peripheral blood morphology proficiency assessment program using the CellaVision® Proficiency Software
  11. Images From the Medical Laboratory
  12. Giant granules in white blood cells
https://doi.org/10.1515/labmed-2023-0073
Keywords for this article
digital hematology; distributed laboratory network; turn-around-time
Creative Commons
BY 4.0

Articles in the same Issue

  1. Frontmatter
  2. Editorial
  3. Can a digital smear review be helpful in the routine haematology laboratory?
  4. Original Articles
  5. The impact of mutational burden, spliceosome and epigenetic regulator mutations on transfusion dependency in dysplastic neoplasms
  6. Improving turn-around times in low-throughput distributed hematology laboratory settings with the CellaVision® DC-1 instrument
  7. Hematology instruments don’t speak the same language: a comparison study between flagging messages of sysmex XN-1000 and alinity H
  8. Platelet clump assessment using the Cellavision peripherical blood application – do we need manual microscopy?
  9. Short Communication
  10. Development of a peripheral blood morphology proficiency assessment program using the CellaVision® Proficiency Software
  11. Images From the Medical Laboratory
  12. Giant granules in white blood cells
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Have an idea on how to improve our website?
Please write us.
© 2025 De Gruyter Brill
Downloaded on 7.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/labmed-2023-0073/html?lang=en&srsltid=AfmBOoq3l2v3YqzaYJJm8P5TvOSntJGTbZSWNK2TV90fBcFqAnc-getm
Scroll to top button