Originally Published IVD Technology January 2002
Building blocks for the point-of-care boom
Selecting and implementing the right components is the essential first step toward market success for POC devices.
Ian Macfarlane and Fred Davis
 |
|
Figure
1. Key components of current-generation POC testing devices include
user interfaces, cartridges, detection mechanisms, and communications
and control elements. Shown here is a digital rendering of the SensiDx
POC analyzer by Ambri (Chatswood, NSW, Australia), which is currently
under development.
|
Although IVD testing is predominantly conducted in central laboratories, many manufacturers are increasingly developing point-of-care (POC) test systems to extend the reach of and, in some cases, to replace lab-based systems.
POC
testing is a multibillion-dollar market and one of the most rapidly growing
IVD testing segments. Two key factors are driving the growth of this market.
First, the need of clinicians for quick test results in settings where patient
treatment decisions benefit from immediate diagnostic findings, such as hospital
emergency rooms and physician offices. Second, the development of technologies
that are capable of delivering the operational simplicity required by POC
systems.
Manufacturers
that are looking to develop successful POC products face three key challenges.
First, they must select and implement features and benefits that will cost-effectively
match the needs of end-users. Second, they must develop the core technologies
necessary to create a functional product. And finally, they must ensure that
they have selected appropriate market applications for the core technology.
This
article examines four building blocks of POC technology that are critical
to market success: user interfaces, cartridges, detection mechanisms, and
control and communications elements (see Figure 1). These subsystems play
important roles in the POC instrument's interactions with end-users (see
Figure 2). Effective design of these subsystems contributes directly to a
POC instrument's success in fulfilling the varied needs of end-users,
including rapid results, high accuracy and precision, simple and safe test
operation, minimal quality control and maintenance requirements, and minimum
cost per result.
In this article, each of these subsystems is reviewed with reference to three products that are currently on the market, the ABL 77 by Radiometer (Brønshøj, Denmark), the i-Stat by Abbott Laboratories (Abbott Park, IL), and the Micromat II by Bio-Rad (Hercules, CA); and a product still under development, the SensiDx by Ambri (Chatswood, NSW, Australia; see Figure 1).
The
User Interface
) |
|
Figure
2. Primary user interactions with a point-of-care system (left) and
key subsystems with which the interaction takes place (right). The subsystems
discussed in this article are in the four upper-right-hand boxes, while
the additional systems not covered are in the "supporting subsystems"
box.
(click to enlarge) |
Second
only to the device's core technology, the user interface is a critical
factor in differentiating products. The interface is literally the “face”
of the product, and comprises all of the components that users see and interact
with. The simplicity and intuitive feel of the interface are major factors
in determining market acceptance. In current-generation POC devices, the user
interface typically consists of several elements, including the display, the
keypad or other data-entry device, a bar code reader, a printer, sample access
and delivery components (usually including a test cartridge), and waste-removal
features.
Because
POC devices are often used in environments in which the operator's attention
is divided between patient care and diagnostic tasks, it is important that
all of the user-interface elements contribute to expediting the sample-in,
result-out process. A well-designed interface should require minimal interaction,
while guiding the operator through the process and, where possible, tolerating
minor data-entry errors. The minimum number of steps required to perform a
test using such a device would include entry of the operator, patient, and
test identifiers; and loading of the sample and cartridge. After the completion
of those steps, the device should output the test results without further
operator intervention. Many POC device manufacturers also include submenus
for quality control and maintenance functions that should be available to
the operator when necessary.
Displays.
The display is the primary means of providing control and status feedback
to the user. Confirmation of data entry, operational instructions, and output
data presentation must be clear and intuitive. The display also helps ensure
that devices are used as intended, providing reliable, valid results and minimizing
the need for operator decisions.
 |
|
Figure
3. The Bio-Rad Micromat II is a CLIA-waived HbA1c analyzer. Utilizing
a 10-µl fingerstick sample, the healthcare provider can generate
the HbA1c result within 5 minutes, allowing dialogue with the patient
during the visit.
|
To
maintain their small size, truly portable devices such as the i-Stat and Micromat
II typically feature a minimal display and keypad entry (see Figure 3). Portable
and benchtop devices with larger footprints, such as the ABL 77 and the SensiDx,
have displays of sufficient size to deliver image and graphic clarity, maximizing
visual data communication. In addition, these larger displays allow information
to be legible from a distance, enabling visual prompts to signal test completion
or ready status.
In some instances, the display is the only means of conveying test results
to users. The base model Micromat II, for example, does not include a printer,
but simply presents a result on the display. Such an electronic display of
a single-value, non-critical result may be all that is needed during a physician
office visit, when the physician simply needs to discuss the result with the
patient and can easily write the result in the patient's record.
Many
hospital-based POC devices provide a selection of quality assurance features
such as calibration, service scheduling, password access, and self-testing
capabilities. The inclusion of such features leads to more-complicated menu
interfaces and data interactions with the user. To support these options,
configurable displays and the availability of a numeric keypad are generally
essential. Other factors that can contribute to a design requirement for a
flexible and configurable display include the needs to offer multiple language
character sets, to configure date and time displays according to local convention,
and to comply with reporting standards that differ among various target markets.
 |
|
Figure
4. The i-Stat1 handheld analyzer and single-use cartridge provide a
variety of critical-care assays, including electrolytes, general chemistries,
blood gases, and hematology. Different cartridges provide specific parameters
or a range of measured parameters.
|
Keypads.
Some POC instruments are used as stand-alone devices without any connection
to hospital information systems (HIS) or other data entry and tracking systems.
The stand-alone interface may therefore need to support keypad entry of sufficient
patient demographics to satisfy requirements for hospital test records.
Overall,
the POC market is currently dominated by devices with membrane-type keypads,
as seen on the i-Stat and the SensiDx (see Figure 4). Such membrane keypads
are cost-effective and universally accepted, but they have the disadvantage
of adding to the space requirements of a device. On the ABL 77, the keypad
footprint has been eliminated through the use of a touch screen with overlays
that enable the instrument's data-entry buttons to be relabeled during
use. Manufacturers seeking to adopt touch screen technologies must be careful
in their selection, however, as some technologies may be ineffective in environments
in which high levels of fluid contamination can result in operator lockout.
Despite being less flexible, quality membrane keypads are generally more reliable
in such environments.
Bar
Coding Systems. Bar codes are used to transfer data such as test identification,
calibration, system instructions, and patient and user identification. Instead
of the tedious process of data entry through a keypad, two or three sweeps
past a bar code reader can allow all of the necessary input information to
be loaded. Bar codes therefore help to deliver speedy results and simplicity
of operation, and are found in the ABL 77, the recently released i-Stat1,
and the SensiDx (see Figure 5). A range of other technologies, such as magnetic
strips or smart chips, can also be used to fulfill similar rapid data-entry
functions. While the decision to include a bar code system in a product may
be fairly straightforward, it pays to have someone with bar coding experience
on the development team in order to select the best solution and avoid pitfalls.
 |
|
Figure
5. The Radiometer ABL 77 is a blood analyzer intended for use in emergency
rooms and other critical-care environments. It takes whole blood samples
and measures pH, blood gases, electrolytes, and hematocrit. From the
measured parameters, it can infer a range of other derived parameters,
and its measuring cycle time is about 70 seconds.
|
Many POC systems utilize cartridges that employ bar codes to identify test type, batch, and lot information. Depending on the complexity of these bar codes, calibration information may also be provided. In some cases, the POC system must access an external database to determine the actual calibration parameters associated with each lot number. In other cases, the bar codes may contain the actual calibration parameters, either on the cartridge itself or on a separate batch calibration bar code that is read at the start of each new box of tests.
Patient
identification bar codes can be read from a sample tube or chart, while operator
identification can be obtained from a bar code that is affixed to an employee
badge. Many products include an operator lockout function in which operator
IDs can be entered into the analyzer and only authorized persons will be able
to run a test.
Bar Code Type and Reader Selection. The most important considerations in bar code selection are how much data needs to be represented by each bar code and how large an area is available for the bar code label. Such factors will determine whether one- or two-dimensional bar codes are appropriate and contribute to the selection of the proper symbology.
The next decision involves reader type. Assuming that one of the more standard formats is used, there are many reader models from which to choose. For some of the high-density, proprietary formats that are applied to cartridges, however, there is no choice: the reader will be whatever the symbology provider supplies. Manufacturers must also select a core technology type that is best suited to such parameters as cost and performance. Performance considerations include the robustness of the data, consequences of an erroneous read, and the possibility of automatic error detection. It is not possible to provide hard data for error rates, since many factors must be considered, including the use environment and the quality of the label. One environmental problem is condensation that can form on the labels of reagent cartridges after they have been removed from refrigerated storage. The simplest solution to this situation is to develop nonrefrigerated cartridges. Alternatively, the manufacturer may also consider the use of some form of condensation management, such as heating.
Printers. In situations in which the device has several test options or produces critical results, the manual transcription of data is more error prone and less desirable. Printed reports provide proof of testing for billing and for maintaining results and records in patient charts. Miniature thermal printers that print results onto stickers or small pieces of paper for insertion into patient records are a convenient and cost-effective choice. These types of printers are found in the i-Stat and the ABL 77.
More-complicated
data may require full-page printed reports that can be accommodated by a small-footprint,
benchtop printer. Integrated printers have the potential disadvantage that
they may malfunction or jam, resulting in inconvenience, lost result records,
or wasted tests. It is usually a better choice to supply the printer as a
separate module selected from one of the many commercially available, low-cost,
plain-paper printers. Such an approach also ensures that those who do not
want the print capability are not paying for it.
Printing
can also be achieved through POC system connectivity to an HIS and associated
networked printers.
Alternative
User-Interface Technologies. Clinicians' desire to have portable
instruments that occupy a small footprint means that traditional personal
computer interfaces with monitor, keyboard, and mouse are undesirable for
most POC applications. Contamination and cleaning requirements also make it
difficult to use devices such as trackballs and nonmembrane keypads.
One solution could be pen-based personal digital assistant computing devices that have been used peripherally for patient data entry in consulting and waiting room environments for several years. For POCs, though, reliable text entry presents challenges, and devices that require the operator to use a writing tool could prove cumbersome. In the future, however, there may be increased use of this approach.
In addition, voice activation technologies such as those used in mobile phones are starting to emerge as candidates for use in POC devices. Adoption of such technologies could have significant advantages, minimizing the potential for contamination by eliminating the need for physical contact with the instrument, and freeing the users' hands for other concurrent activities. The future may see voice-activated interfaces mature to a clinically acceptable level.
Another interesting area is the ongoing development of expert systems that may expand the diagnostic capabilities of POC devices. First Medical (Mountain View, CA) has recently licensed such a system from Duke University (Durham, NC). Called the Advanced Reperfusion Algorithm, the system promises to provide physicians with expert-system assistance in interpreting the results of cardiac marker tests.
Choosing the Right Interface. Ideally, a manufacturer will choose the form of the user interface through a process of market research and testing. End-users and purchasing decision makers should be included in this research.
Product concepts may be reviewed by specialized human-interface labs that offer experts in the analysis of intuitive flow and consistent user experience. Concepts can be selected and refined through two simultaneous channels: cognitive walk-through sessions with representative end-users, and sample tests with purchasing officers of major market targets.
The cognitive walk-through process is particularly useful in guiding the development of the user interface. The process involves selecting pairs of typical end-users across the range of skill levels expected to use the system, such as nurses, doctors, and supervisors. Using a prototype system, the pairs of users receive a training session of a length that is equal to what is expected in the end-use situation. They are then given a preplanned set of tasks to complete. In order for the developers to get effective feedback from this process, two things must be emphasized: the participants are testing the intelligence of the designers (not vice versa), and they must verbalize to one another their thinking as they work out how to perform tasks and their reactions to how the information is presented.
The
Lowly but Ubiquitous Cartridge
POC cartridges usually house the reaction cell, store all test reagents, and accommodate the sample while the test is being conducted. They are designed to protect both the user and the POC reader from biohazard contamination. Cartridges also help to ensure adherence to the test protocol, and reduce the potential for error and the need for sophisticated user training. With cartridge-based devices, operators can concentrate on making clinical decisions and on patient management, rather than on the intricacies of performing the test.
To minimize the cost of consumables, the cartridge should be as simple as possible. In fact, designers typically elect to add functionality and cost to the reader, rather than the cartridge. Cartridges can generally be manufactured with a range of reagents according to customer requirements. This approach enables the manufacturer to expand the test menu of an instrument simply by releasing newly approved test cartridges. When this approach is used, it is important that users be easily able to recognize which test each cartridge is for as well as which system it belongs to.
At the same time, using cartridges extends the product life cycle of certain instruments, which is attractive for both users and suppliers. Cartridges enable manufacturers to launch a POC device with a limited test menu, and to stagger their R&D investment for tests to be added after launch.
During cartridge development, the manufacturer should assess the intended use environment and the sensitivity of the test process to factors such as ambient light, temperature, humidity, and airflow. Reasonable shelf life and insensitivity to storage temperature are important design targets for cartridges. In the event that cold storage is required to extend shelf life, a minimum stabilization time before use should be targeted. Careful design can overcome tight chemistry requirements by stabilizing the cartridge temperature once it has been inserted in the reader.
 |
|
Figure
6. With the i-Stat cartridge, the user applies 2 to 3 drops of blood
to the deposition area. The cartridge is inserted into the reader and
electrically connected to the detection system.
|
Although
designed for simplicity of use, the cartridge system will most likely include
a number of smart and sometimes complex design features. For example, a unique
feature of the Micromat cartridge is the use of curved optical wells. Since
the light source in the instrument is placed at the center of rotation, minor
positional variations do not cause changes in the distance that is traveled
by the light beam through the solution to be measured. A change in light path
would otherwise result in unacceptable inaccuracies from run to run. Such
details do not add cost to the final product, but require significant engineering
effort during development.
Cartridges may also involve microfluidic circuit design, in which surface
interactions, wettability, and capillary behaviors can be significant. For
those used to dealing with the flow of larger sample volumes, resultant fluid
behaviors can seem contraintuitive. Getting microfluidics specialists involved
in design review or development can hence be very useful. Other cartridge
design issues to consider include features to prevent incorrect insertion
and retention of biohazardous waste after use.
 |
|
Figure
7. The Bio-Rad Micromat II test cartridge is a self-contained cartridge
containing sample buffer, wash buffer, and elution buffer reagent vials.
Each of the reagent vials rises and is presented to the operator at
appropriate steps in the test.
|
Another challenge in cartridge development is the need for scalability of manufacturing systems. At launch, product volumes are likely to be low to moderate. As sales volumes increase, though, scaling up production to automated assembly is likely to be necessary for capacity and cost reduction reasons. It is important that this requirement be considered from the outset and that production plans allow adequate lead time for design, implementation, and qualification of manufacturing facilities.
Dealing
with Samples. POC devices are generally designed to handle one of five
types of patient samples: whole blood; blood preparations such as plasma,
serum, or EDTA/citrate stabilized samples; urine; saliva; or expired gases.
Whole blood is preferred because it avoids the need for sample preparation,
thereby reducing the time to result. For example, of the 84 POC analyzers
that were reviewed by ECRI in 2001 (excluding portable glucose monitors),
more than half were capable of performing testing with whole blood samples
of less than 100 µl, while only seven tested urine samples.1
All
of the devices depicted in this article use whole blood samples. They generally
use less than 100 µl per test, and the trend in the POC market is toward
even smaller sample volumes that are suitable for neonatal testing. The ABL
77 and i-Stat use whole blood (~65 µl and 40–60 µl, respectively)
that must be dispensed into the cartridge (see Figure 6). The Micromat II
cartridge uses a fingerstick and 10-µl capillary to deliver samples
to the test cell; the SensiDx cartridge will enable whole blood to be accessed
directly from a closed vacutainer.
In their intended environments, each of these sample access systems has its advantages. The accuracy required in the delivery of the sample volume will also influence the design of the sample access system. To determine the right solution for their products, manufacturers must evaluate the target market and confirm the solution through trials.
Detection
Systems
 |
|
Figure
8. The SensiDx cartridge provides access to closed-tube samples, sample
temperature control, and access for electrical contacts to provide connection
to the detection system.
|
The
mechanism is an extension of the cartridge, taking the chemical reaction or
test result from the cartridge into the reader system for interpretation and
analysis.The detection system will be specific to the chemistry of the cartridge,
and products in the POC market today are based on at least 15 different detection
systems, measuring a range of chemical, electrical, and optical properties.1
Many products employ similar detection systems but achieve measurements through
different and usually proprietary chemistries.
In
the Micromat II system, the chemistry reaction produces a homogeneous solution
of glycated hemoglobin (see Figure 7). A blue light-emitting diode and narrowband
pass filter are then used to perform spectrophotometric measurements at a
fixed wavelength (430 nm). The spectrophotometer output is converted to a
concentration result using a specific algorithm and calibration data.
In the SensiDx system, the cartridge is electrically connected to the data acquisition system (see Figure 8). The sample reaction causes a change in impedance in the membrane. This change in impedance is measured by accurate bridge circuitry and converted to a concentration result, again using a programmed algorithm and calibration data.
Control
and Communications
The
brains of any POC system are provided by the control and communications subsystems,
which are responsible for controlling the measurement and detection processes
and for regulating the flow of information to and from the users.
Control.
In POC devices, the control subsystem coordinates all other subsystems.
It sequences operations, using system elements to analyze the sample according
to the test request from the menu. Controlled operations typically include
cartridge insertion, temperature control, sample control, detection, waste
control, and cartridge ejection. The control system produces a result in accordance
with the calibration information and saves the result with appropriate identification
for reporting or transmitting to an HIS. In handheld products, power management
may also be controlled.
Data
Management and Storage. POC instruments are often capable of retaining
the results of many samples for reporting or uploading to an HIS or laboratory
information system (LIS). For example, the Abacus mobile hematology POC analyzer
by Diatron (Vienna) can store up to 10,000 samples with histograms. The ABL
700–series from Radiometer can hold 2000 patient results of blood gases
or electrolytes, chemistry, or hematology tests; 1000 calibration results;
1500 QC reports; and 3000 system messages. The i-Stat1 can store up to 5000
patient results.
Communications. The communications subsystem allows the uploading of test results to an HIS or LIS database. A proposed standard for POC connectivity that was developed by the Connectivity Industry Consortium was approved by the National Committee for Clinical Laboratory Standards (NCCLS; Wayne, PA) in October 2001. This standard describes the attributes of an access point, the communication protocols between the POC device and the access point, and communications between a data concentrator and clinical information systems. As acceptance and implementation of this standard grows, vendors of POC devices will be able to seamlessly integrate test results into an institution's diagnostic information management system, regardless of where the test was performed, and in compliance with the institution's overall diagnostic testing quality policies.
The
NCCLS standard specifies the use of either RS-232 or infrared IrDA connections.
RS-232 is currently the most widely used interface, possibly due to its simplicity
and availability as a standard interface on most computers. In the ECRI survey,
among the 62 POC devices that had communications functionality, 58 used RS-232
interfaces, three used network connectors, and one (the Opti CCA by AVL) was
capable of communications via either RS-232 or infrared link.1
Although the NCCLS standard focuses on these common communications interfaces,
it also acknowledges that wireless standards may be added as future options.
In fact, it is likely that the future of POC communications may lie in the
direction of faster and more-exotic protocols such as Bluetooth or IEEE 802.11
wireless networking.
Conclusion
For
POC devices, the principal technology challenges will continue to be in test
chemistry. This core science must facilitate laboratory-quality test results
while requiring minimal training for users as well as little or no calibration
and maintenance. Quality control procedures must be simple and reliable, and
the acceptable package size for the cartridge will continue to be reduced.
The development of fully self-contained cartridges with an increasing range
of tests, smaller sample volumes, and operating costs competitive with laboratory-based
instruments will be tough challenges for many applications.
The
development of noninvasive POC instruments is also an area where diagnostic
applications are pressing technology forward. While noninvasive blood glucose
meters are beginning to come to market, in the years to come a great deal
more activity is likely to be seen in new products and the expansion of noninvasive
test menus.
Miniaturization
of the technology to enable patients to wear devices is another likely area
of technology development. It is unlikely that the POC testing marketplace
will drive technology advances in the other key subsystems. More likely, such
subsystems will leverage developments such as voice-activated user interfaces,
wireless communications, and data security once these are proven reliable
and accepted in markets with less-critical applications.
Major R&D Activities. The building blocks described in this article will all provide significant challenges to R&D teams working on products for the broad range of POC testing environments. The major technology challenges and intellectual property development will be focused on the cartridge, reader, and chemistry performing the core science.
Embarking on a POC system development program invariably requires electronic, mechanical, software, and fluid-engineering input. Effectively managing interactions between elements such as hardware and software development, chemistry and assay development, optics, communications technologies, disposable design and molding, and manufacturing systems is critical to success. Though a seemingly endless list, the need to consider such varied disciplines contributes to the challenge and excitement of designing an exceptional POC system.
The innovative products that are evolving from this activity will provide benefits to the healthcare system and patients in many areas. Physicians advising patients on the management of diabetes, for example, will obtain real-time data to more effectively help patients during consultation. In the emergency or critical-care environments, POC testing will on one hand provide data that expedite release of healthy patients, freeing up valuable bed space. Alternatively, it will help ensure that the right care is applied as quickly as possible to those in need.
REFERENCE
1. "Point-of-Care Analyzers, Clinical Laboratory," Healthcare Product Comparison System (Plymouth Meeting, PA: ECRI, 2001), 1–60.
Ian Macfarlane, PhD, is manager of point-of-care systems development and Fred Davis, PhD, is director of instrument design and development at Invetech (Melbourne, Victoria, Australia), a contract biomedical instrument development and manufacturing firm. The authors can be reached at imf@invetech.com.au and fhd@invetech.com.au, respectively. The authors wish to thank Invetech specialists Andrew Evans, Chris Hubbard, Frank Samuhel, Rohan Smith, and Steve Cox for their contributions to this paper.
Copyright ©2002 IVD Technology
