ABL77
Studies
Evaluation and use of the Radiometer ABL77 blood gas analyzer
By: François Braconnier*, Annie Dupeyrat, P. Odelut.
From: SPECTRA BIOLOGIE – VOL. 22 – No. 133 – August-September 2003
Summary: The Radiometer ABL77 blood gas analyzer was evaluated in our laboratory to assess its correlation with the current analyzers in use (Radiometer ABL725) and to investigate its ease of use in a point-of-care setting. The ABL 77 performance demonstrated that it could be used as an excellent analyzer for point-of-care services.
I Introduction
The object of the Radiometer ABL77 evaluation was to compare its performance with the current analyzers used in the laboratory (ABL725) with a view to of implementing this type of instrument in point-of-care settings (emergency wards, intensive care units, etc.)
II Materials and methods
1. Materials
The ABL77 is a stand-alone analyzer that includes a touch screen, a sensor cassette, a floppy drive and a calibration solution pack (Cal Pack). Small, light weight and with a handle, the analyzer transports easily. Although it is usually operated from an electrical outlet, its internal battery allows operation at point-of-care locations or during electrical failure for up to 25 patient tests. The calibration solution (Cal Pack) can be found on the back of the analyzer along with the standard connections: barcode reader, RS232 serial port, Ethernet port RJ45. The reference analyzer (Radiometer ABL725) was operated according to the recommendations of the manufacturer under well-defined quality control conditions.
2. Reagents and consumables
2.1. Sensor Cassette Sci
Each cassette bears a label with the expiration date (100 days at constant temperature), a lot number and connectors for data transmission to the ABL77. The cassette is stable at room temperature and has a lifetime of 15 or 30 days (30 days for 50 tests) from its installation on the analyzer. Three types of cassettes are available, depending of the selected panel:
- Full panel: pH, pCO2, pO2, cNa+, cK+, cCa2+, cCl– and hematocrit
- BGH: pH, pCO2, pO2 and hematocrit
- Lytes: cNa+, cK+, cCa2+, cCl–
and four different capacities (depending on the number of tests: 50, 100, 150 and 300).
The cassette features a flap for positioning the inlet probe for sampling (syringe or capillary), a window for observation of the sample progress through integrated sensors and a pump tubing for connection to the peristaltic pump.
2.2. Cal Pack (Calibration solution pack for ABL77)
This standard cartridge for all sensor cassettes has a 90-day shelf life from the date of production and a lifetime of 30 days from installation on the analyzer. This cartridge allows measuring of pH, pCO2, pO2, Na+, K+, iCa2+, cCl– and hematocrit. It does not need any preparation or handling and contains all the necessary items for calibration, rinse and waste collection:
- a 650-mL pouch of calibration solution 1 (Cal1) for 360 cycles of calibration and/or rinsing
- a 350-mL pouch of calibration solution 2 (Cal2) for 180 cycles
- a 1.2-L pouch for waste collection
The exact calibration values of the parameters are included in the attached barcode. After use, the Cal Pack has to be disposed of as biohazardous material, according to regulations.
2.3. Quality control
The four-level Radiometer quality control QUALICHECK4+ (QC4+) was used. These colored solutions contain a biological buffer in an aqueous solution with germicide added. The different levels are balanced with carbon dioxide and oxygen to give predefined pH, pCO2 and pO2 values. When a QC solution is injected, the ABL77 recognizes the QC level.
2.4. Patient samples
Arterial patient samples were taken with the Radiometer PICO70 sampler.
III Evaluation protocol
This protocol covers four studies.
- Repeatability study
1.1. on each level of control (QC4+), each level was run 20 times successively, each measurement being performed with a new ampoule,
1.2. on patient arterial blood samples, these samples being “tonometered” in ambient air with temperature control as follows:
Arterial blood from two patients was sampled with three PICO70 syringes from Radiometer. Blood from one of the patients was then pooled in a precipitating tube in ambient air for 45 minutes. All the blood was then transferred to a syringe and 20 consecutive measurements were performed on the ABL77.
- Reproducibility study: one run of each level of quality control solution every day for 20 days.
- A correlation study performed on 103 patient samples. The distribution of values was done according to SFBC Guidelines for pH, pCO2 and pO2 (1). The accuracy was evaluated by linear regression performed on each parameter (Y = ABL77; X = ABL725).
- Correlation test reliability study
In order to evaluate the inaccuracy caused by ambient air during consecutive sampling on two different analyzers, a series of 50 samples was run, one after the other, on the ABL725. The same protocol was applied to the ABL77.
IV Results
1. Repeatability
Repeatability on QC4+ quality control solutions
Table I presents the coefficient of variation (CV) obtained for each evaluated parameter for the four QC levels studied. Analysis of the results shows that the obtained CV percentages are all excellent or lower than 2.14 % except for pO2 level 1 (mean 50 mmHg), which shows a CV of 4.22 %. This result is affected by two pCO2 assays performed at the beginning of the evaluation (52 mmHg and 54 mmHg). If these two values are excluded, CV is 2.31 %.
TABLE I – Repeatability on QC4+ quality control solutions
|
|
pH |
pCO2 |
pO2 |
cNa+ |
cK+ |
cCa2+ |
cCl– |
|
Level 1 |
0.05% |
1.05% |
4.22% |
0.14% |
0.46% |
0.68% |
0.67% |
|
Level 2 |
0.00% |
1.30% |
0.85% |
0.00% |
0.00% |
0.83% |
0.39% |
|
Level 3 |
0.06% |
2.14% |
0.81% |
0.00% |
0.00% |
1.62% |
0.28% |
|
Level 4 |
0.05% |
0.82% |
1.06% |
0.20% |
0.00% |
0.32% |
0.00% |
Repeatability on “tonometered” arterial blood samples
Tables II and III summarize the results obtained on two patient samples according to modalities described in protocol (Section III 1, 1-2).
TABLE II – Patient No. 1
|
|
pH |
pCO2 |
pO2 |
cNa+ |
cK+ |
cCa2+ |
cCl– |
|
Values |
20 |
20 |
20 |
20 |
20 |
20 |
20 |
|
Min. |
7.52 |
16 |
186 |
136 |
4.7 |
1.09 |
108 |
|
Max. |
7.55 |
18 |
206 |
140 |
4.8 |
1.12 |
113 |
|
Mean |
7.54 |
17 |
196 |
137 |
4.8 |
1.10 |
109 |
|
SD |
0.01 |
0.73 |
6.29 |
0.81 |
0.0 |
0.01 |
1.04 |
|
CV% |
0.14% |
4.39% |
3.21% |
0.59% |
0.86% |
0.68% |
0.95% |
TABLE III – Patient No. 2
|
|
pH |
pCO2 |
pO2 |
cNa+ |
cK+ |
cCa2+ |
cCl– |
|
Values |
20 |
20 |
20 |
20 |
20 |
20 |
20 |
|
Min. |
7.52 |
13 |
185 |
137 |
4.8 |
1.04 |
109 |
|
Max. |
7.61 |
15 |
201 |
138 |
4.9 |
1.08 |
113 |
|
Mean |
7.59 |
14 |
192 |
138 |
4.8 |
1.06 |
111 |
|
SD |
0.01 |
0.70 |
5.31 |
0.31 |
0.0 |
0.01 |
1.30 |
|
CV% |
0.15% |
5.04% |
2.76% |
0.22% |
1.01% |
1.40% |
1.17% |
The obtained coefficients of variation are excellent for all parameters (< 1.40 %) except for pCO2 (4.39 % and 5.04 % respectively, due to low values of pCO2) and for pO2 (3.21 % and 2.76 % respectively for a mean pO2 value of 190 mmHg).
2. Reproducibility on QC4+ quality control solutions
Table IV groups the coefficients of variation obtained for the four levels of control during the 20 days of the reproducibility study. Results are excellent for pH, Na+, K+ and Ca2+ (< 1.96 %). For pO2, results are affected by extreme pO2 values (mean at 50 mmHg and extreme at 46 mmHg and 55 mmHg).
TABLE IV – Reproducibility on QC4+ quality control solutions
|
|
pH |
pCO2 |
pO2 |
cNa+ |
cK+ |
cCa2+ |
|
Level 1 |
0.10 % |
2.21 % |
5.65 % |
0.36 % |
0.91 % |
1.37 % |
|
Level 2 |
0.08 % |
1.44 % |
1.90 % |
0.44 % |
0.00 % |
1.18 % |
|
Level 3 |
0.07 % |
1.91 % |
2.07 % |
0.46 % |
1.12 % |
1.44 % |
|
Level 4 |
0.10 % |
2.65 % |
3.71 % |
0.40 % |
0.75 % |
1.96 % |
3. Correlation study on patient samples
Table V groups the totality of the results. The few deviating points have not been suppressed. The difference which can be observed on the correlation line for pO2 could be due to the fact that the reference analyzer (ABL725) uses an electrode based on an anode/cathode couple of silver/platinum whereas the ABL77 uses a gold/platinum couple. As the membranes also have different compositions, these sensors do not provide the exact same electrical response.
TABLE V – Correlation studies on patient samples
|
pH |
Y = 1.117 X – 0.8563 |
R2 = 0.944 |
|
pCO2 |
Y = 0.9551 X + 0.0258 |
R2 = 0.9616 |
|
pO2 |
Y = 0.9286 X + 7.5307 |
R2 = 0.9845 |
|
Na+ |
Y = 0.9475 X + 7.6673 |
R2 = 0.8338 |
|
K+ |
Y = 1.0052 X – 0.0602 |
R2 = 0.9097 |
|
Ca2+ |
Y = 0.9941 X + 0.0105 |
R2 = 0.9416 |
4. Correlation test reliability study with ABL77 and ABL725
According to the protocol described in Section III, 4, Tables VI and VII show the calculated difference (2nd value – 1st value) for each analyzer and each studied parameter, expressed in the parameter unit. The obtained results for pH, Na+, K+ and Ca2+ were unremarkable. For pCO2 and pO2, the maximum variations observed in this study are almost identical for both analyzers for pCO2 (2.7 mmHg on ABL725 and 2 mmHg on ABL77) and for pO2 (18 mmHg on ABL725 and 17 mmHg on ABL77). This last value was only observed on one patient and on a pO2 value higher than 100 mmHg.
These findings differ little from published tracking standards by (1), illustrating the difficulty of comparing blood gas analyzers due to possible air contamination during consecutive runs of a given sample. Only the use of two identical samples performed at the same time and on the same patient could decrease this influence, and this would be ethically questionable.
TABLE VI – Calculated differences on ABL725
|
|
DpH |
DpCO2 |
DpO2 |
DNa+ |
DK+ |
DCa2+ |
|
Min. |
–0.02 |
–1.00 |
–31.00 |
–1.00 |
–0.10 |
–0.01 |
|
Max. |
0.01 |
2.70 |
18.00 |
1.00 |
0.10 |
0.03 |
|
Mean |
0.00 |
0.48 |
–1.75 |
0.09 |
–0.02 |
0.0 |
TABLE VII – Calculated differences on ABL77
|
|
DpH |
DpCO2 |
DpO2 |
DNa+ |
DK+ |
DCa2+ |
|
Min. |
–0.02 |
–1.00 |
–20.00 |
–1.00 |
–0.10 |
–0.04 |
|
Max. |
0.01 |
2.00 |
17.00 |
1.00 |
0.10 |
0.04 |
|
Mean |
0.00 |
0.20 |
–1.78 |
0.15 |
–0.01 |
0.0 |
V Practicability
During the evaluation period, the ABL77 analyzer performed without any trouble. A four-hour staff training was sufficient for the personnel to learn basic operation and how to perform blood gas measurements. An obvious advantage is the short turnaround time (1 min 20 sec) plus the fact that the analyzer is maintenance free (except for the replacement of the thermal paper). The inlet may appear fragile and the orifice can get obstructed if the necessary cleaning is not carried out after each sample has passed, as recommended by the manufacturer.
The measurement begins once the inlet flap has been closed. To save time the operator may close the flap before cleaning the probe, which can cause blood spillage on the sensor cassette. Despite these restrictions, the probe did not fail during the evaluation period and there was no clogging or build-up observed. It should be remembered that the probe is part of the sensor cassette and will therefore be replaced along with the cassette within a maximum of 15 days after its installation on the analyzer. Replacing the Cal Pack solutions and the sensor cassette with its probe does not need any comment.
The analyzer was found to be user-friendly. It may be mentioned that the on-screen keyboard does not respect the usual sequence (AZERTY-keyboard as featured in the ABL700 Series). The general information appearing on the various screens is very close to that in the ABL700 Series. The possibility of entering patient demographics with a barcode reader and a connection to the LIS (either direct or through the HIS) will greatly improve data entry, especially when the analyzer is used in a care unit, since the onboard software makes it possible to obtain a true bi-directionality of the analyzer.
The battery allows utilization during a power failure or in the case of patient transport (ambulance). A maximum of 25 tests can be obtained from a fully charged battery, which seems to be sufficient for any situation. When battery operated, the system automatically reduces its power consumption by shutting down in case of inactivity. The restart corresponds to a system reboot and takes 45 seconds. According to the manufacturer, mobile transport possibilities are being explored (ambulance) and tested at the moment. The ABL77 has also been tested at altitudes of up to 1,500 meters.
The ABL77 is well suited for point-of-care testing. A number of recent articles (2 to 9) describe the limitations in point-of-care testing and the importance of commitment from the lab team (lab managers and lab assistants) concerning user training, quality control, biochemical validation, medical record management or cost management (economical issues have not been addressed in the present protocol).
VI Conclusions
This evaluation made it possible to validate several analytical aspects of the ABL77. The simplicity of operation allows utilization by care unit personnel. Performance could be enhanced by an IT connection with RADIANCE software. The portability of this analyzer and its internal battery broaden the applications of the analyzer (e.g. pressurization).
VII References
- Vassault A, Grafmeyer D, De Graeve J, Cohen R, Beaudonnet A, Bienvenu J. Analyses de biologie médicale : spécifications et normes d’acceptabilité à l’usage de la validation de techniques. Ann Biol Clin 1999; 57: 685-95.
- Gouget B, Laforge M, Feuillu A. Les systèmes de gestion à distance des analyseurs de pH, gaz du sang et analytes. Spectra Biologie 2002; 21, 122: 28-36.
- Pollet J, Sarran A, Daumain P, Collin J-P, Fonteneau J, Gouget B. Maîtrise biologique dans l’externalisation des analyses et télégestion. Revue Française des laboratoires 1997; 292: 76-79.
- Magny E, Launay J-M. Expérience de biologie délocalisée à l’hôpital Lariboisière. Revue Française des laboratoires 1999; 312: 56-58.
- Thebault H, Mollard J-F. Gestion à distance de l’analyse délocalisée : contrôle de qualité automatique. Revue Française des laboratoires 1999; 312: 66-71.
- Borgard J-P, Goguelin A, Duvezin F, Rota M. Biologie au chevet du patient : implantation dans un centre hospitalier général. Le rôle du biologiste. Revue Française des laboratoires 1999; 312: 170-73.
- Gruson A, Verchain S, Desenclos P, Bouchet D, Huddlestone P. Biologie délocalisée : problématique et évaluation d’une solution informatisée. Revue Française des laboratoires 1999; 312: 174-76.
- Duchassaing D. Gestion à distance d’un analyseur de gaz du sang, Revue Française des laboratoires 1999; 312: 179-80.
- Magny E, Beaudeux J-L, Launay J.-M. Gazométrie et electrolytes sanguins délocalisés : exemple d’organisation et évaluation économique. Ann Biol Clin 2003; 61: 344-51.
