Printer Friendly

Evolving issues related to bedside glucose testing.

Evolving issues related to bedside glucose testing

As the practice of medicine becomes more efficient for the individual patient, an increasing amount of testing will be done outside the laboratory. Glucose testing was the first blood test to leave the confines of the laboratory for the bedside. Many other tess can potentially be performed outside the laboratory as well.

During the 1980s, bedside capillary glucose measurement became standard practice in hospitals, clinics, emergency rooms, nursing homes, and doctors' offices. Reagent strips impregnated with glucose oxidase have been available since the 1960s, and portable reflectance meter, built on microchip technology, have been marketed since 1970.

Bedside glucose meters provide accurate results when used by well-trained personnel. Results can, however, be skewed by many sources of error; in fact, one study showed that inadequately trained individuals err 20 per cent of the time. [1]

The American Diabetes Association, jgkjjunction with the Centers for Disease Control and the Food and Drug Administration, convened a Consensus Development Conference in 1986 to evaluate training procedures for bedside glucose testing (BGT) meters and other aspects of the devices, including their general impact on patient care, both at home and in the hospital. One recommendation that emerged from the conference was to implement a complete quality control program for all users. The Health Care Financing Administration (HCFA) Health Standards and Quality Regional Letter 83-27 (sent to Medicare laboratory inspectors) specified QC training and proficiency xasting required when BGT was performed by nonlaboratory personnel in hospitals or longterm-care facilities.

The Joint Commission on Accreditation of Healthcare organizations (JCAHO) has approved new standards on decentralized laboratory testing for its 1990 manual. "If diagnostic clinical laboratory testing for the organization's patients is done within the organization outside a central laboratory," the manual states, "personnel performing the tests must . . . demonstrate satisfactory levels of competence, . . . [have] current written policies and procedures [that] are readily available, ... . [have] quality control checks conducted on each procedure each day, . . . [resolve] identified problems, . . . and [maintain] appropriate quality control and test records." [2]

The challenge to the laboratory in the '90s will be to insure the quality of testing performed outside as well as inside the lab. This article will address the responsibilities the lab must assume willingly or be delegated to do by the hospital administration.

* Responsibility. The laboratory can assume responsibility for BGT in two basic ways. First, it can establish a laboratory-based program. Second, it can work with one or more other departments in a laboratory-directed or laboratory-monitored program.

In the laboratory-based program described by our laboratory in 1985, [3] BGT is performed by the laboratory phlebotomy team under the supervision of a medical technologist. A similar laboratory-based program was recently outlined by a group from the Mayo Clinic. [4' In a laboratory-directed program, testing is performed by nurses or other personnel; the laboratory monitors the operations and assures quality control and quality assurance. [5]

* Equipment and method. Although many meters are on the market, three manufacturers provide most of the meters for hospital-based BGT: the Glucometer series (Ames, Division of Miles Pefk exkriew?Alkhart, Ind.), the Accu-Chek series (Boehringer-Mannheim, Indianapolis, Ind.), and the Glucoscan series (life-scan, Mountain View, Calif.). To date, only one strip-and-meter combination on the market--One Touch, also by Lifescan--does not require blotting to remove excess blood from the reagent strip. The same company supplies individually packaged strips. The other two companies provide reagent strips in vials.

The staff of each facility must decide which features of each meter/strip combination best meet their requirements--and then choose only one. Because meters are calibrated in different ways to correlate with either plasma or whole blood glucose levels, it is critically important that the same type of meter be used throughout the facility.

Each institution must be fully aware of the differences in measured glucose levels between the laboratory method and the bedside method. The laboratory method uses venous serum or plasma, whereas in the bedside testing method, capillary whole blood simulates a plasma level, since the erythrocytes do not penetrate the reagent strip matrix. Table I shows that bedside methods correlate well with laboratory methods (R values) but regression equations show significant differences between methods. Method validation. According to most published studies, results of bedside and lab methods of glucose testing vary by approximately 10 per cent. Each hospital laboratory should validate the BGT method being used in its own institution. The most accurate way is to compare the test with the laboratory method for Stat serum or plasma glucose levels on at least 25 patients whose glucose levels range from <50 mg/dl to >400 mg/dl. This is not a significant burden for the laboratory because for most diabetic patients who would undergo BGT, physicians would have requested other chemistry laboratory tests, such as electrolytes, as well. Monitoring the timing of BGT and phlebotomy is critical in relationship to meals, insulin injections, or the infusion of intravenous solutions containing glucose or destrose.

Physicians, especially endocrinologists, have expressed concern about the differences in glucose levels obtained with venous blood and capillary blood glucose methods. Studies have shown that these differences are trivial during fasting, but after consumption of a glucose load, the capillary glucose concentration exceeds the venous concentration by 15 to 20 per cent. [1] Our experience shows that the differences are method dependent (Table II). The primary difference between serum or plasma glucose and bedside glucose testing methods is the manner in which each strip/meter combination has been calibrated; that is, whether the strip/meter has been calibrated for whole blood or serum/plasma, over what glucose range, and at what hematocrit level. The differences between the major manufacturers of strips and meters with respect to hematocrit are shown in Figure I. [6]

Glucose levels established as standards for diagnosing diabetes from the National diabetes Data Group were obtained by laboratory methods. [1,7] These standards indicate that BGT methods average 10 mg/dl more than venous values, while glucose levels of capillary whole blood (e.g., YSI from Yellow Springs Instrument, Yellow Springs, Ohio) are, on average, 15 mg/dl less than those of venous blood. [7]

* Result reporting. Whoever reports BGT results--whether the laboratory or another hospital department--should use a standardized report format. The results should be easily located in a patient's chart in a format that will enable the physician to scan for glucose results at equivalent times of day. If BGT is performed by a computerized laboratory, the results can be displayed in the cumulative report format. [3] If bedside glucose testing is performed by nursing personnel, a specific format can be used to document results [5] at a chart location other than with the nurses' notes. The report format must clearly distinguish BGT results from those of glucose tests performed within the laboratory.

It may be helpful for the laboaratory to prepare an equivalence chart to illustrate the differences between laboratory-generated glucose results and those produced by BGT. Table IIi provides an example of such a chart for the three strip/meter combinations described above.

Each institutions must also determine the standard procedure that will be followed when very low values (<50 mg/dl) or very high glucose levels (>400 mg/dl) are detected using BGT. The department of internal medicine at our community hospital decided that further laboratory testing beyond the BGT was not routinely required. If values were very low, glucose would be administered; if they were very high, insulin would be administered and BGT repeated until an acceptable level was established. [3] In other institutions, unusually high or low values are routinely followed by venipuncture to obtain serum or plasma glucose results by the laboratory methods. [4]

* Charging for BGT. If BGT is performed by laboratory staff, charges to patients are handled like other lab orders and charges. The current procedure terminology (CPT) code for Glucose: Blood Stick Test, code 82948, should be used to distinguish BGT from serum or plasma glucose, which is code 82947. If BGT is done by nurses or another department, care must be taken to make sure all procedures performed are duly charged and the results are well documented for insurance auditing purposes.

* Training operators. Training of all employees performing BGT should include four steps:

1. Test performance instructions. Phlebotomists and nurses are oriented to patient care and learn best with visual instructions. [8]

2. Simulations or practice sessions, to be performed by each employee until proficiency is achieved.

3. A procedure manual developed with the trainee in mind and including visual aids. The videotapes develped by many manufacturers for training new operators are much easier to follow than written instructions.

4. Verbal or written quizzes administered to each employee learning the technique. The questions are designed to ascertain their comprehension of common sources of error and their consequences and clinical significance.

* Sources of error. Bedside glucose testing results can be skewed by approximately eight different problems.

1. Inadequate specimen. No deleterious effect has been associated with applying an excess of blood to the strips that require blotting. Nevertheless, obtaining a correct reading from a Glucoscan strip, for example, requires the application of at least 50 [mu]l of whole blood (Table IV). Similar minimum requirements are present with the One Touch, which requires no blotting. With that device, however, excess blood was deleterious; it contaminated the optics and interfered with the reading. [4] Each reagent strip and meter combination should be evaluated to determine the minimum or maximum amount of blood that will provide accurate test results. Operators should be familiar with visualizing these relative amounts, since it is impractical to actually measure the specimen at the time of service.

2. Inaccurate calibration. Calibration of the instruments used for bedside glucose testing is critical to proper operation. Each instrument is calibrated differently. Some meters are calibrated to specific lot numbers, vials of reagent strips, or both. With such instruments, using strips differing from those appropriate for the specific calibration used will lead to erroneous results.

3. Deteriorated enzyme in strips. The reagent strips contain glucose oxidase, an enzyme that deteriorates upon prolonged exposure to light, air, or excess heat. If an operator is using vials of reagent strips and leaves the cap off a vial overnight, the enzymes on the strips remaining in the vial will have deteriorated and should not be used for patient testing.

4. Expired strips. The strips are marked with an expiration date that must be respected. Monitoring these dates is particularly important when meters and reagent strips are kept at each nursing unit and not used regularly.

5. Inadequate validation. If the comparison of results between BGT and the laboratory method is not known, physicians may make incorrect decisions based on BGT test results, especially when the standards of practice were developed using the laboratory method. An example is decision points for glucose levels to change insulin dosages.

6. Insufficient instrument maintenance. This is a significant source of error, especially when departments other than the laboratory are responsible for BGT. The optics must be cleaned daily to prevent a film of dried blood from obstructing the optical reading source. In fact, obtaining the same glucose result two or more times for consecutive patients suggests that the optics have been obstructed. This can happen whenever the reagent strip has not been adequately blotted to remove all blood that has not been absorbed by the reagent strip.

7. Bad timing. Human failure is a primary source of error in BGT, especially when the operators do not regularly use such instruments. The major error, incorrect timing--a problem with visually read reagent strips in particular been alleviated by built-in timers with audible signals in the second- and third-generation meters. But such timers decrease this source of error only when they are used correctly.

8. Wrong materials. Failure to blot the reagent strips with the materials recommended by the manufacturer is another source of human error. Accu-Chek II suggests cotton balls for blotting, but does not provide them. Glucometer II recommends using facial tissues but does not specify whether hospital-type wipes, such as Kimwipes (Kimberly-Clark, Roswell, Ga.), may be used. The Glucoscan provides the blotting paper recommended for use with its reagent strip. This source of error is eliminated in the One Touch meter/strip combination, which does not require blotting.

* QC and QA. Since laboratories are accustomed to implementing quality control, QC activities tend to be well established in facilities whose labs are fully responsible for BGT or direct a BGT program administered by other departments. In most hospital situations currently, the BGT program is administered through the nursing department and does not necessarily include a routine QC program at this time.

Meter manufacturers have done an excellent job of developing QC programs that can be easily adapted by nursing personnel. The Ames Glucometer, for example, has a computerized program with data that can be downloaded from the meter to a personal computer (PC) or dot matrix printer for storage or printing.

Quality assurance programs include regular testing of heparinized venous blood with the BGT method and comparing results with those of the laboratory method for plasma glucose on the same specimen. [3] Taking this step insures that the BGT method equivalence to the laboratory method is regularly monitored. Another example of a QA program is to compare the BGT method between operators or instruments using the same heparinized venous blood specimen. [5] Such a program, when followed routinely, insures that all instruments and operators are measuring essentially the same glucose concentration.

A QA program emphasizing BGT utilization established at our hospital evaluates two aspects: the number of patients on whom BGT is performed simultaneously with the laboratory venous blood glucose method, and the number of times BGT has shown normal levels for 48 hours. Both monitors would detect overutilization of glucose testing and, if it were found to any great extent, indicate a lack of trust in the BGT method. According to our monitors over six months, venous serum glucose testing was ordered and performed for less than 10 per cent of patients who had BGT, and less than 5 per cent of patients had normal glucose levels for more than 48 hours. We interpret the data to indicate that our physicians have confidence in the BGT performed by our phlebotomy team and that overutilization is not a problem in our hospital.

The QC program must specify the alternatives available to operators if QC fails. Are extra batteries or instruments available if an instrument fails QC? Are extra reagents of the same lot number available in case an operator has left the vial cover off for a prolonged period of time? Who is responsible for troubleshooting problems with meters or reagent strips? Who is responsible for regularly cleaning the optics of meters? Who has the authority to withdraw the privilege of performing the test if an operator consistently fails QC?

* Patient self-testing. Each institution must develop a policy regarding patient self-testing. If a patient who has been doing home BGT is admitted to the hospital for management of a diabetic crisis, should this patient be entitled to do his or her own BGT while hospitalized? Does the answer to that question change depending on whether the patient's hospital bill is covered by insurance?

After experiencing problems with patient self-testing at our hospital, we developed a policy. We permit patients to do their own testing after their results have favorably compared with laboratory-based BGT for the first two days of hospitalization. This step insurers that the patient's meter is properly calibrated, that the optics are reading properly, and that a patient's loss of visual acuity or an inability to perform the BGT properly has not contributed to the individual's current diabetic crisis. Once BGT proficiency has been established, the attending physician decides whether the patient or the laboratory will perform any future testing. Most physicians prefer to have the laboratory continue to perform bedside glucose testing on hospitalized diabetics so that they will have the documentation necessary to justify the patient's continued hospitalization to the payer.

* Keeping step with technology. In a laboratory-directed BGT program in which nursing or another department performs the testing, who determines whether and when to upgrade meters to a new technology? In our laboratory-based program, it has been difficult to upgrade technology because this would require rrtraining the entire phlebotomy team, consisting of approximately 50 individuals, including full-time and part-time employees. If our hospital's nursing staff performed BGT, more than 500 people would have to be retrained, since diabetic patients can be admitted to any medical service or nursing unit on any shift.

To justify making a change, the cost: benefit analysis of a new meter must be convincing. At our hospital, we postponed upgrading to the One Touch (no-blot) bedside glucose testing method until an inexpensive disposable plastic optic cover was developed. The One Touch is now feasible to use by dedicating a separate optic cover for each patient because the same meter is used for many patients. This makes the One Touch cost-effective without risking a patient's contact with another patient's blood.

* Other applications. It is necessary not only to measure glucose levels rapidly so that insulin requirements can be assessed and possibly modified, but also to measure partial thromboplastin time (PTT) rapidly to assess and possibly modify heparin dosages for inpatients who require anticoagulant therapy. For outpatients, prothrombin time (PT) is measured to assess and perhaps modify dosages of warfarin (Coumadin); rapid measurement of PT would facilitate that process. Miniature instruments of roughly the same size as BGT meters are available to measure PT alone (Coumatrak by Biotrack, Mountain View, Calif., distributed by Du Pont Medical Products Dept., Wilmington, Del.) or to measure both PT and PTT (512 Coagulation Monitor developed by Biotrack, distributed by Ciba Corning Diagnostics Corp., Medfield, Mass.).

Intraoperative, postoperative, or postnatal measurement of hemoglobin or hematocrit by a rapid bedside method would facilitate monitoring of blood loss. Such an instrument would be useful in obstetric office practice to detect anemia. The hand-held instrument, slightly larger than BGT meters, measures hematocrit by changes in electrical conductivity, in that the erythrocytes act as resistors. The meter calculates hemoglobin levels based on a fixed mean corpuscular hemoglobin concentration (MCHC). The Stat-Crit is available from Wampole Laboratories, Cranbury, N.J., and U.S. Surgical Corp., Hospital

W wxa s Division, Norwalk, Conn.

The cost per test for the coagulation instrument and for the hemoglobin and hematocrit instrument is substantially greater than the cost of performing the tests with current lab instruments. Similarly, BGT is more expensive than testing for glucose using the Astra-8 (Beckman Instruments, Brea, Calif.), the CX-3 (Beckman), or virtually any other chemistry lab instrument that measures glucose and other analytes.

BGT increased the productivity of our phlebotomy team without adding staff. The void left in the chemistry lab was filled by implementing new tests. [3]

Other investigators have reported x exx a rapid turnaround time of whole blood testing helps shorten the length of patient stay. [9,10] If so, differences in cost are justified for the hospital even if not for the laboratory.

In conclusion, BGT is being performed by nurses in most hospitals. Regulatory agency guidelines from the JCAHO and CLIA establish the laboratory as the department responsible for insuring accurate alternate-site testing results. Technology is available to enable additional analyses to be tested using whole blood techniques at the bedside. The hospital laboratory management staff must be prepared to evaluate and implement this testing.

z1] Weiner, C.P.; Faustich, M.; Burns, J.; et al. The relationship between capillary and venous glucose concentration during pregnancy. Am. J. Obstet. Gynecol. 155: 61-64, 1986.

[2] "Accreditation Manual for Hospitals," p. 157. Chicago, Joint Commission on Accreditation of Healthcare Organizations, 1990.

[3] James, K.; Latham, D.; Marrero, J.; et al. Capillary blood glucose measured by the laboratory phlebotomy team. Lab. Med. 16: 417-421, 1985. Additional material presented in teleconference sponsored by the American Society of Clinical Pathologists, February 1986.

[4] Burritt, M.F.; Klee, G.G.; Hanson, E.; et al. Evaluation of the One Touch glucose meter for use by a phlebotomy team. Lab. Med. 21: 512-516, 1990.

[5] Bain, O.F.; Brown, K.D.; Sacher, R.A.; et al. A hospital-wide blind control program for bedside glucose meters. Arch. Pathol. Lab. Med. 113: 1370-1375, 1989.

[6] Poon, R., and Hinberg, I. Effect of hematocrit on three blood glucose monitoring systems. Clin. Chem. 34(abstr): 1314, 1988.

[7] National Diabetes Data Group Standards. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 28: 1039-1057, 1979.

[8] Hilton, B.A. Nurses' performance and interpretation of urine testing and capillary blood glucose monitoring measures. J. Adv. Nurs. 7: 509-521, 1982.

[9] Kost, G.J. The impact of whole-blood testing on response time. Arch. Pathol. Lab. Med. 112: 921-922, 1990.

[10] Trundle, D.S., and Weizenecker, R.A. Capillary glucose testing: A cost-saving bedside system. Lab. Manage. 24: 59-62, May 1986.

Dr. James is associate director of clinical laboratory services; MacPhail is chief technologist, support services; and Davis is support services technician, Central DuPage Hospital, Winfield, Ill.
COPYRIGHT 1990 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:James, Karen; MacPhail, Gertrude; Davis, Robert
Publication:Medical Laboratory Observer
Date:Dec 1, 1990
Previous Article:Some wins, some losses in final budget agreement.
Next Article:Developing a laboratory chemical hygiene plan.

Related Articles
Protocols for bedside testing.
Protocols for bedside testing.
Quality management of bedside glucose testing.
Blood glucose monitoring: is blood glucose monitoring for many of those people with type 2 diabetes really necessary in the management of the...
Diabetes testing supplies: a how-to selection guide; High-quality blood glucose testing applied appropriately is critical. What to look for.
Answering your questions; finger-stick fasting sample in GTT.
Confirmation of critical POCT glucose.
The effects of two differing techniques on the accuracy of reagent strip blood glucose testing.
Basics of home glucose monitoring.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters