How to select the ideal hematology analyzer.
Looking for the best instrument to suit your needs? You’ll save time and money by consulting the treasury of advice offered here.
AFTER MONTHS of tedious preparation, our facility had won approval for a major capital improvement project to upgrade several expensive pieces of equipment in our laboratory. One of these upgrades involved purchasing a new hematology analyzer. As we began to pursue various options available on the market, we quickly learned that all hematology analyzers are not created equal.
This article is not intended to evaluate specific hematology analyzers or to recommend one piece of equipment over another. Rather, it is intended to provide laboratory managers with guidance on selecting an instrument that will best suit their needs.
* Assessment tool. While members of our lab management staff had always considered themselves qualified to make wise purchasing decisions, we felt that we needed a more efficient way to zero in on what individual vendors could offer our facility. We knew from the start that our search for an outstanding hematology analyzer would be a real challenge. Every vendor representative offered five-part differentials, closed vial sampling, and at least 18 parameters with histograms. They all asserted that their analyzer was the best on the market. Each claimed to have won out consistently over major competitors in side-by-side comparison tests.
To help judge the individual qualities of each instrument, we developed a comprehensive questionnaire, which we gave to any vendor vying for our business. Each salesperson was asked to complete the survey and return it to us when that manufacturer’s instrument was brought into the laboratory for a trial period.
This document served several purposes. First, it helped eliminate vendors who were clearly unable to meet our needs. Second, certain responses triggered more questions to ask during the evaluation process, Finally, the survey provided important documentation to which we could refer when making our final decision. Figure 1 combines most of the questions from our original questionnaire with the later ones we asked vendors when their products were brought to us for assessment.
For the most part, we learned that vendors appreciate having an opportunity to highlight on paper the special features of their products. A few were reluctant to answer questions related to service calls, the cost of replacement parts, the number of instruments currently in service, and similar matters. Nevertheless, no one refused to comply. Any salesperson who wanted our business acceded to our requests
* Ironing out terms. After weeks of studying all the instruments we fell knowledgeable enough to prepare bid specifications for vendors who remained in the running. Among the key issues that we closely scrutinized were data management, long-term expenditures, and extended warranties–typically expensive.
Another important consideration was to obtain guaranteed prices for reagents. To calculate the reasonable yearly usage costs associated with operating each instrument, we asked every vendor to include in their bids the costs for all reagents and controls. Taking this step enabled us to obtain guaranteed reagent pricing for the next five years.
* Wise shopping. More suggestions on how to be a wise shopper follow.
Independent analysis. When possible, ask vendors to leave their analyzers in the laboratory for a few days so that your staff can assess the instrument at their convenience in the absence of an eager rep. Vendors often offered us their instruments for at least three days.
Comparison chart. Create a chart that will reflect at a glance the comparative benefits of every instrument being considered. Dissimilarities in equipment often become more apparent when they are highlighted in black and white, particularly once you have begun to establish what you want–and don’t want–in a system.
Reserving first impressions. Don’t let yourself strongly favor one instrument too early in the decision process. You may be amazed by the offers that come your way once you request additional incentives, especially as it gets closer to finalizing a deal or designing your bid specs around a specific instrument.
Doing all the research is worth every minute it takes. Good luck!
Tommy Lee Camden, M.S., MT(AMT) is laboratory director for the City of Lubbock (Tex.) Health Department.
Probing questions for vendors and evaluators
I. General information
Name of instrument
Approximate bid price
Price of maintenance contract
What kinds of discounts are available on extended warranties?
Optional attachments (autoloader, data management system, T & B lymphocytes, bar code reader, ticket printer, other).
Length of time system has been in use (excluding manufacturer’s trials)
Approximate number of instruments used and distributed in U.S.
Approximate number of instruments used and distributed worldwide
Earlier model or antecedent apparatus
How long has your company been manufacturing hematology equipment? Has it ever been marketed by another company or under a different name?
II. Confirmation of specs
What is the minimum space needed for your analyzer and reagents?
What services are required to maintain the vacuum, drainage, and voltage?
Is it necessary to control electrical interference?
Is it necessary to control the atmospheric environment?
What are the analyzer’s effects on the environment, if any (for example, acoustical disturbance, vibration disturbance, electrical or mechanical safety)?
Is your instruction book clearly written with adequate information for servicing and fault finding?
What special training is required for routine use of the system?
What special training is required for maintenance?
Does your company provide local training sessions or is training conducted elsewhere?
How many of our employees would be trained as primary operators according to the purchasing agreement?
Are training costs included in the purchase?
Are travel costs included in the training (for instance, gas mileage, airfare, food, lodging)?
What is the principle of each test performed?
Have detailed specifications of the instrument been provided?
Does the system have closed vial sampling capability?
Do the plastic shields covering the rubber stoppers on the specialized evacuated tubes pose a problem?
Describe input of samples into the system for testing.
Can program selection be modified at a user’s discretion?
Can a user select CBC with differential versus CBC without differential?
How many parameters will this instrument test?
Does this system provide a three- or five-part differential?
Is the differential interpretive or directly measured?
What is the throughput time?
What is the practical number of analyses per instrument per day?
How much time is required to start your analyzer up from a complete shutdown? Do you recommend turning the instrument off at night or over the weekend when it is not being used?
How much time is required to shut down the system each day?
Does the analyzer have a reagent warning and sensory system? Does it detect when reagents run low? Will it complete an analysis before shutting down if a low reagent is detected?
Does the system automatically clean the aspirator when used for sampling in the open vial or secondary mode? Is manual wiping required?
Do carryover or mode-to-mode programs exist within the data management system to evaluate instrument performance?
Are veterinary programs available for analyzing animal specimens?
Which parameters can be calibrated by in-house staff? Which parameters require calibration by technical field personnel? What is the recommended minimum calibration frequency for this instrument?
Are there any unique safety features on this analyzer?
Provide information on your validation methodology and the procedure we must use before putting this equipment on line and verifying linearity claims.
Is your system equipped for automated sampling? If so, is this capability standard or optional? If optional, what is the cost?
If the system does not have an autoloader, can the system be upgraded to include this feature?
When using the autoloader, how long can a user be away from the instrument once it is fully loaded (this does not mean tests per hour)?
How many specimens can be loaded onto the autoloader at one time?
Do Stat specimens create any difficulties in correctly identifying or interrupting the run in progress when using the autoloader?
Is your system equipped with a bar code reader for specimen identification? If so, which bar codes can be used for this system? Is bar coding a standard feature?
Does the position of the bar code pose any problems on the tube? Once bar coded, does the tube have to be placed in the cassette holder in any particular position so that the bar code reader can detect and correctly identify the specimen?
Are quality control materials bar coded for easy loading of assay values and automatic assignment to correct QC file?
Describe any alarm systems in place for instrument failure or disturbance.
What is the life expectancy of the tubing, laser, circuit boards, pumps, or other major components?
What is the average cost for replacing expensive items such as the circuit boards and laser?
Approximately how many service calls have been made per year for the last three years?
What is the average mean time between service calls for the last three years?
What is the name and address of the nearest authorized service facility to this laboratory?
What is your total parts inventory for your system?
How many technicians are employed at your nearest facility?
How many field service technicians are available for on-site service?
What is the hourly rate for shop labor?
What is the hourly labor rate for field service labor?
What is the average response time that can be expected for a service call to this laboratory?
What is the average delivery time for parts not found in inventory?
About how long will replacement parts be available for your system?
Explain the standard warranty coverage of your instrument.
What is your procedure for handling defective parts or poor quality of materials or workmanship?
For how many years has this particular system been in production?
If your service facility is located outside this city, explain your process for performing warranty work, service not covered under the warranty, and service work performed once the warranty has expired.
Who is responsible for transportation expenses should the unit require service out of town during the warranty period?
What is your method for calculating the transportation expense of out-of-town service?
State (in years) the useful life and expected downtime of your instrument.
What parts of your system are considered consumables and therefore not covered under your service contract? What is the annual usage of these consumables?
What type of anticoagulant does your system require?
Does the system have microsampling capability? If so, what is the dilution factor?
Can prediluted specimens of capillary blood be used? Will the system autocalculate for the dilution factor?
Do calibrators and control products come in cap-piercing vials?
What mechanism is available on the instrument to reject unsuitable specimens (hemolyzed, clotted, low volume)?
What is the specimen volume required for closed tube, open tube, and capillary specimen analysis?
List all required reagents, their functions, and the volumes used in single analysis as well as lapped analyses.
What is the unopened shelf life of each reagent?
What is the opened shelf life of each reagent?
Are reagents only supplied by your company, or are suitable reagents available from other commercial suppliers as well? If so, will using them jeopardize the warranty or the service contract?
What volume of reagents is used for priming the instrument and per run of 2, 25, and 100 specimens?
What reference materials/calibrators are required? What do they cost?
Are other suitable reference materials available for this instrument?
Are reagents prepackaged as a unit or purchased separately?
What is the monthly cost of calibrators and controls?
Do the routine quality control materials also control the quality differential? Must additional materials or procedures be purchased or followed?
Do any reagents require neutralization before discarding? If so, explain the procedure.
Are special calibrators needed and available for calibrations as opposed to day-to-day controls?
Provide detailed costs of all reagents needed to operate your system based on blood specimens per year for our institution.
VIII. Data management
Does your analyzer have a dedicated PC or data management system? Can the data management system be upgraded, and if so, how?
How large is the DMS? Detail its specifications.
Is the system capable of multitasking duties? Can you edit data while the instrument is analyzing specimens?
Does the system have standard bidirectional computer linkup capability?
Can the internal DMS be upgraded using future software revisions?
If the system can be upgraded with an autoloader, does the software program require a change? If so, how much more is the new software?
Will the system accept both alpha and numeric data entry?
How many patients with full reportable results can be stored internally in the system? When the system is full, can the results be downloaded to floppy disks for external storage?
How much optional demographic information can be assigned to a patient? Is this possible at both pre- and post-analysis?
Can the patient report be modified to the operator’s discretion? Detail the various options available for creating a typical report.
How many scatterplots and histograms can the system store?
Can the system generate a pre- and post-load list?
Are password security options available to operate the analyzer? Can they be easily printed for QA records?
If the standard system has an optional DMS attached, who is responsible for its warranty?
Does the system have flags for RBC size and color as well as WBC flags for left shift, atypical lymphs, and blasts?
Is a help screen available throughout the DMS system?
Does the DMS use an easy-to-follow window format?
Does the system have Levey-Jennings-type QC programs?
How many points or days of data are possible for each control level?
Does the system have Westgard rules within its QC program?
Does your company have an interlaboratory QA program? Can the data be transmitted via modem to your facility for analysis?
Does the system have an “X-BAR” or moving averages program? If so, how many and which parameters does the system track?
Does the moving average program selectively remove data from the calculation with Bull’s formula?
Does the system have a reproducibility program?
How many control files are available? How many lines of data can be stored in each control file?
How many lines of data can be deleted or rejected from the QC programs? Is the information totally erased or highlighted as unacceptable? Are security measures in place to prevent QC data from being erased?
When assaying controls, how are values outside the assigned ranges brought to the operator’s attention? Can the system be shut down automatically if QC values are out of range?
IX. Laboratory management staff response
How friendly is the DMS to operate?
What problems emerged during the in-house evaluation?
Did the sales rep answer all our questions without hesitation and rectify any problems that occurred as we evaluated the system?
Could our lab staff perform the required maintenance and minor repairs? How much time would it take?
How easy is the system to operate compared with the other instruments under consideration?
How great is the noise level compared with that of the other instruments under consideration?
Is the flagging criteria easy to interpret?
Is the instrument easy to calibrate? Does it hold calibration values on target?
How easy is it to set up the control libraries?
During on-site evaluation, did we have to call the troubleshooting hotline? Was it helpful?
Is there good correlation between the instrument under evaluation and the one we use now?
How well does the system separate white cell subpopulations in the cytogram?
Was significant carryover noted between specimens?
How quickly can data be retrieved from previous runs?
What distinct advantages does this system have over the other analyzers we are considering?
How does this instrument rank in relation to the other instruments under evaluation?
FLAGGING SYSTEMS FOR HEMATOLOGY ANALYZERS
Clinical laboratories routinely use hematology instruments to perform so-called
complete blood counts (CBCs) on patient specimens. About 25 different instruments are
capable of performing blood counts that include differential leukocyte counts.(1) While
the measurements performed by these instruments in general are quite
accurate and precise, the measurements may vary from one brand of
instrument to another.
Flagging is defined as “signaling or communicating a message with, or as if with,
a flag.” In the hematology laboratory a flag is the signal to the operator that the analyzed
may have a significant abnormality. Most of these automated instruments are
programmed in a variety of ways to “flag” or otherwise identify specimens that may
a report is released from the laboratory.(5-8)
among the various instruments as to which abnormalities are flagged as well as the
efficiency of the flagging procedures.
are quantitative flags, which are originally set by the manufacturer but which can be
redefined by the user if desired. But it must be recognized that reference ranges may vary
(9) Therefore, quantitative flagging must take
for both qualitative and quantitative flags to its flagging menu.
To be useful, flagging should have a low false positive rate (e.g., < 10%) coupled
potentially abnormal blood specimens. However, it is recognized that there will be
compromises between false negative and false positive flagging rates that are
which to strive, although not presently attainable.
The ultimate goal is to reduce the number of clinically non-contributory blood
third to one-half of these tedious reviews could be safely eliminated without adversely
affecting patient care. Of course, in hematology and/or oncology services, it is most
likely that most cases will have more frequent blood film reviews. As a bonus, the
shortened. But there may be a decrease in hospital revenues, which may pose a significant
financial problem. Another potential problem may be the gradual loss of morphologic
become less proficient in this important function.
In our laboratories we began the development of a flagging system by interfacing
(5) With the prototype
system in place, it was found that about one-third of the cases had no significant blood
film review abnormalities and could be reported immediately. These instruments were
later replaced by new analyzers that included acceptable flagging systems.(11) With the
acceptance by the clinical staff over several years, the laboratories were able to reduce the
rate of blood film reviews from 82% to less than 40%.
State of the Art for Flagging
Performance of the most popular hematology instrumentation reveals reasonably
good performance for quantitative abnormalities.(12-16) With time and experience, a
laboratory may safely broaden the flagging limits beyond the narrow limits of the
quantitative reference ranges. So, for example, as confidence builds that the instrument
correctly counts neutrophils in the mildly and moderately abnormal ranges, the film
review may be omitted if this is the only quantitative flag. Also, if a patient’s blood film
has been recently examined and the abnormality is still present, there is no need to repeat
the examination.(8) As an example, patients undergoing cardiopulmonary bypass
procedures frequently are thrombocytopenic and anemic in the immediate post-operative
period. There is no need to re-examine blood films on consecutive specimens if the
quantitative results are mildly abnormal and no marked changes have occurred.
Since the state-of-the-art hematology analyzers are quite acceptable insofar as
quantitative measurements are concerned, there is little to be gained by a supplementary
blood film examination. Thus a great deal of skilled technical effort can be devoted to
other more useful endeavors. The following discussion will therefore concentrate on
clinically meaningful qualitative abnormalities. In this case gender, age or even racial
differences are of lesser importance. As a result the user has much less control over the
qualitative flagging processes, which are proprietary for the most part.
The flagging performances for qualitative abnormalities have had false positive
rates of up to 30% or more and false negative rates of up to 15%. It is difficult to compare
the various instruments since many of the studies have been done with only one
instrument. These data must be interpreted with caution since many times the case mix of
the study specimens is not well defined or is variable from one study to the other;
therefore, the rates of flagging misses are subject to variation due to the makeup of the
study population rather than variation in instrument performance. Only a few studies have
been published that compare several instruments using the same set of patient
1) Flagging for Qualitative Red Cell Abnormalities
Traditionally, red cell abnormalities have been categorized by the number, size
and variability of the red cells, i.e., the red cell indices. Thus, microcytic, normocytic and
macrocytic anemias as well as polycythemias have been quite adequately identified and
categorized by hematology laboratories for many years. More recently, the determination
of red cell anisocytosis has been measured by the so-called red cell distribution width
(RDW) with some success, and additional categories of anemia have been developed.(19)
In the recent past, there have been efforts to screen for hemoglobinopathies (in addition to
thalassemias) that have shown promise of being useful in the discovery of certain
hemoglobinopathies. However, this capability has not yet been incorporated into all
Suggested flags for qualitative erythrocyte abnormalities include:
Nucleated red cells in severe anemia, metastatic
Poorly lysed red cells for hemoglobinopathies
While the quantitation of platelets has improved considerably in the recent past,
Two additional platelet parameters have been proposed. The first, the mean platelet
younger platelets are thought to be larger than normal. Thus an increase in the MPV has
often they are detected by analysis of the platelet histogram.
atelet distribution width (PDW), is an
of this measurement is still under investigation.
Studies have confirmed the accuracy and reliability of the total leukocyte count
monocyte differentiation has been problematic; this is probably of relatively minor
and variant [atypical] lymphocytosis), the instrument correlations with reference
considered to be a shortcoming of the instruments while others take the stance that there
especially surgeons and pediatricians, continue to rely on documenting these leukocyte
Variant lymphocytes for viral infections or lymphoproliferative disorders
National Committee for Clinical Laboratory Standards (NCCLS) working group with the
differential leukocyte counting standard.
4) Unexpected Changes in Hematologic Parameters
Finally, if there is a significant change (delta check) in any of the patient’s
quantitative or qualitative results even if they have occurred within the flagging limits, a
blood film should be reviewed. For example, if the platelet count drops significantly
within a relatively short period of time (hours or days), there is reason to investigate.
Platelet clumping may account for the fall, but early disseminated intravascular
coagulation might present in a similar way. Delta checking systems are not yet well
developed nor are they yet widely available(29) but should be incorporated into the
flagging procedures in the future.
Quality Assurance Procedures for Flagging
Since the flagging procedure by definition results in the examination of the blood
film either to confirm or to rule out the presence of an abnormality, there is, in fact, an
ongoing quality assurance procedure in place. A comprehensive discussion of quality
assurance strategies for automated hematology analyzers has recently been published.(30)
However, in order to ensure that no significant number of abnormalities are being missed,
a representative sampling of non-flagged specimens should also be examined and records
should be kept of such actions. This will also help to maintain the morphologic expertise
of the laboratory staff. A recent chapter on blood film review outlines a practical yet
comprehensive method for this procedure.(31)
Presently marketed hematology analyzers are capable of utilizing age and/or
gender-specific flagging. While sophisticated laboratory information systems might
include such a feature, as yet few such systems are in place. With continuing advances in
the processing of data within the laboratory, a number of interfacing systems are being
developed that may be able to help with the appropriate processing of the volume of data
generated by the automated hematology analyzers. Some systems may be able to
automatically flag specimens requiring additional study including, as appropriate, a blood
film review by senior technologists and/or the medical director of the hematology
laboratory. Some systems include the on-line evaluation of the data for validity.(32)
We have the opportunity to take advantage of the capabilities of automated
hematology analyzers. If we can properly validate the hematology data and can flag
specimens that need additional evaluation, patient care will be improved (by significantly
shortening turnaround times), the efficiency of the laboratory staff will be enhanced (by
eliminating unneeded or non-contributory blood film studies including differential
counts) and finally the costs for laboratory studies will decrease. In these days of downsizing
and cost cutting, such innovations have become more difficult to investigate,
develop and implement. Still, we should continue to strive to improve our laboratory
services as well as to decrease costs.
In order to accomplish these worthwhile goals, laboratory and clinical
professionals should continue to learn about the capabilities of the various analyzers and
encourage harmonization of the various flagging systems, including the elimination of
International Council for Standards in Haematology is to develop standards in
Filed under: Uncategorized