| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
General Clinical Chemistry |
1 Department of Clinical Chemistry and Laboratory Medicine, Robert-Bosch-Hospital, 70376 Stuttgart, Germany.
2 Institute of Anesthesiology and Intensive Care Medicine and
3 Institute of Clinical Chemistry, University Hospital Mannheim, University of Heidelberg, D-68167 Mannheim, Germany.
4 Institute of Clinical Chemistry and Laboratory Medicine, Clinic Nuernberg, D-90419 Nuernberg, Germany
aAddress correspondence to this author at: Clinic Nuernberg, Institute of Clinical Chemistry and Laboratory Medicine, Prof.-Ernst-Nathan-Strasse 1, D-90419 Nuernberg, Germany. Fax 49-911-3985710; e-mail thomas.bertsch{at}klinikum-nuernberg.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: We estimated blood loss by multiplying the number and volumes of sampling tubes collected from 2654 adult inpatients. We compared the number of tests per patient for all inpatients and intensive care unit patients during the first period and again in the same time period 1 year later when cumulative blood-loss volumes were being reported to physicians and educational information had been given to decrease blood loss from laboratory tests.
Results: For 95% of the patients, blood loss during hospitalization was <196 mL. The largest proportion of the blood samples was used for clinical chemical tests (median, 45%), followed by hematologic (median, 26%) and coagulation (median, 17%) tests. In the surgical and cardiovascular surgical intensive care units, however, blood gas analyses accounted for 19–34% (medians) of the use. For 5% of the patients, all undergoing intensive care, blood loss was >200 mL and for 0.7% was >600 mL during their hospital stay. Such high blood losses were observed in patients with long-term ventilation, coagulation disorders, and repeated surgery. The largest median blood loss was in patients undergoing cardiovascular surgery (median, 201 mL). The mean number of tests was 44 per inpatient before cumulative blood loss was being reported and 46 when it was being reported.
Conclusions: Blood loss from laboratory diagnostic testing is not likely to pose a problem for most hospitalized patients. Blood loss is greater in intensive care patients and after cardiovascular surgical procedures. Reporting of the cumulative individual blood loss did not decrease blood loss for laboratory testing.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Increased blood loss for laboratory testing seems to be primarily attributable to the increased number of analytes measured, an increased frequency of testing, and easier collection with arterial and venous indwelling catheters. Patients with indwelling catheters are usually critically ill and require repetitive testing. Moreover, a volume of blood (flush) is discarded (9)(10).
Published studies were performed before the advent of the small sample volume analyzers and smaller blood collection containers now in use. We wished to obtain current data to determine for which patient groups blood loss was greatest and to gauge the effect on blood loss of informing the physician in charge in a simple, ongoing manner, of the diagnostic-related blood loss of each patient.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We estimated blood loss associated with laboratory testing data for 2654 patients for whom the hospital stay was started within a 2-month period until the stay was finished. Most phlebotomies for patients outside intensive care units (ICUs) were performed by medical students in their last year of clinical training, whereas most phlebotomies for patients in ICUs were performed by experienced nurses from arterial and venous lines. The investigated patient samples are categorized and summarized in Table 1
. For each patient, the blood loss was estimated cumulatively by adding the volume of the tubes used for blood collection with the aid of a computer program (Medat). This software was used to calculate the cumulative diagnostic blood loss of each patient. Results were sent together with the laboratory results to the attending physicians. The volumes of the collection tubes (Sarstedt) are shown in Table 2
. Routine chemistry tests (including hormones, tumor markers, therapeutic drugs, and serology) were from a menu of 105 tests and were performed with serum.
|
|
The frequency of secondary tubes was <1%. Analyzing instruments were as follows:
A complete blood count was defined as one test. Dead space is defined as the remaining sample volume after performance of one determination.
It should be noted that the estimation of blood loss by adding the volumes of the tubes used for blood collection does not account for the fact that not all containers were filled to nominal capacity. Consequently, the numbers indicated in the following tables are upper limits and have been rounded to take into account this lack of accuracy. However, because the participating hospital has established a fully centralized laboratory testing process and all intraoperative determinations and laboratory tests of intensive care patients were also performed by the central laboratory, the estimated data on blood loss are relatively complete except for the above-mentioned restrictions. No point-of-care testing was performed on the ward, on the ICU, or in the operating rooms except for 
4000 capillary blood glucose tests/month. This equates to 400 mL of whole blood and to a mean blood loss of 0.15 mL/patient. Therefore, the amount of blood loss attributable to point-of-care testing can be neglected, but only under the conditions of our hospital.
The laboratory is equipped and organized such that sample aliquoting is reduced to a minimum and the analyses are performed in the primary tube whenever possible. Sample transport from the ICUs and operating rooms is via a pneumatic tube sample transport system. For selected departments, the blood loss for patients was assessed both during treatment on the ward and during intensive care, including intraoperative analyses. The amount of testing was at the discretion of the ordering physicians except for patients enrolled in studies and ICU patients. No groups of tests were mandated. All tests were selectively ordered.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. The differences between median and mean clearly verify an asymmetrical distribution of measured values. The maximum values show deviations from the corresponding medians of up to 120-fold. For 5% of the patients tested, the total blood loss exceeded 196 mL. Our study group included patients receiving special treatments (daily blood loss in parentheses): seven patients with aneurysms (14–67 mL), one patient with autologous stem cell transplantation (5 mL), one patient with resection of liver metastasis (119 mL), and two patients with simultaneous ovariectomy and hysterectomy (122–126 mL).
|
Larger collection volumes were found in particular with patients undergoing cardiovascular surgery (Table 4
). Note that 98% of all cardiac surgical cases received intensive care.
|
Gynecologic and obstetric patients were not investigated separately because the numbers of tests per patient were similar: 22 and 20 tests per patient, respectively. The number of abnormal results were six (gynecology) and three (obstetrics) per patient. The percentage of abnormal results varied between 45% (cardiovascular surgery) and 15% (obstetrics).
The upper percentile limits for the total blood loss of the various departments (Table 4
) demonstrate the interdepartmental differences for diagnostic-related blood loss. Results for nephrology and gastroenterology, which were responsible for all patients of internal medicine with unexplained medical problems, were nearly identical. Similarly, there was practically no difference between these two departments in daily blood loss, number of blood drawings, and number of determinations. The daily blood losses were nearly identical for all four departments of internal medicine, but hospitalization time, total blood loss, number of blood drawings, and number of laboratory tests were lower in oncology and cardiology.
Looking at the total blood loss of the patients according to department, it is striking to see a similarity between visceral surgery and gynecology/obstetrics. However, the department of gynecology/obstetrics performed substantially fewer clinical chemical analyses over almost the same period of time (Table 5
).
|
As expected, the values for patients undergoing cardiovascular surgery were all markedly higher. The high number blood drawings among these patients (66) were attributable to repeated tests and were 9 times higher than the number of blood drawings for visceral surgery.
The major portion of total blood loss was attributable to clinical chemical analyses (36–51%), followed by hematologic (20–30%) and coagulation analyses (13–21%), without significant differences among the hospital departments.
Laboratory testing for intensive care treatment is considerably different from testing for treatment on a ward, as shown in Table 5
. It should be noted that for cardiovascular surgery, a high percentage of the total blood loss (34%) is required for blood gas analyses in the ICU. The data for the 42 ICU patients in the visceral surgery and gynecology departments showed similar results except that 19% were attributable to blood gas analyses.
Assuming that a total blood loss during an average hospital stay of <200 mL is usually not critical, we evaluated separately 131 patients from whom larger blood volumes had been collected (Fig. 1
). As can be seen clearly from the data, that the test covered two subgroups. A smaller group (n = 18; data points within the boxed area in Fig. 1
) showed a linear increase in total blood loss during the intensive care treatment. For these patients, we found a linear correlation between length of stay and blood loss (Spearman correlation coefficient, 0.678). In the larger group, we found no clear correlation between blood loss and duration of the hospital stay.
|
Twenty-two patients of the entire group had a laboratory-related blood loss >600 mL during their hospital treatment. All of these patients were on long-term ventilation and had venous and/or arterial lines. These patients had a special blood-drawing protocol: at the beginning of an 8-h shift, a cell count, electrolyte and creatinine determinations, and depending on the clinical situation, acid-base assessment were ordered. Furthermore, when the ventilation scheme was changed, acid-base studies were again ordered.
We determined the number of tests per ICU patient within the first study period and in the same time period 1 year later after physicians had been informed about the individual cumulative blood loss. Before and after this intervention, physicians ordered, respectively, 44 and 46 tests per patient concerning all inpatients. For ICU patients, they ordered 93 and 85 tests per patient, respectively.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Only 5% of the 2654 patients included in the study had a laboratory-related blood loss >196 mL. The fact that the median is only two analyses per milliliter of serum demonstrates that the sample volume required for clinical chemical tests can be further reduced. Assuming that mostly primary blood collection tubes are used for analyses, that analyzer dead volume is 100 µL, and that the mean testing volume is 10–15 µL, the calculated sample volume for three determinations amounts to 130 and/or 250 µL of serum (12). According to the results of Dale and Pruett (5), 45 times (2- to 108-fold) more than the necessary amount of blood is taken. A comparison of our data with published results (4)(9)(10)(13) reveals a much lower diagnostic blood loss in the current study (18 mL/day vs 37–63 mL/day).
For 95% of all samples, the number of clinical-chemical determinations per milliliter of serum was five analyses or fewer. For each of these, 4.7 mL of blood was drawn. A major reduction in the diagnostic-related blood loss could be achieved by use of 2-mL sample containers for repetitive clinical-chemical tests and reducing the sample volume for blood gas analyses to 1 mL.
In 22 patients with long-term ventilation and venous and/or arterial catheters, laboratory-test-related blood loss was >600 mL. Among these patients, four were is intensive care for 70–80 days. The other 18 patients show in a shorter time a linear increase in diagnostic blood loss in a critical clinical situation. The decisive factor for the blood loss was not the number of tests but the number of sample drawings caused by a rigid blood drawing scheme so that repetitive sample drawings were performed routinely irrespective of the patient’s clinical condition. On the basis of the total number of patients treated in the year of the study in the hospital, 142 patients with this high blood loss would be expected. This special patient group requires further measures to reduce the collection volumes.
In addition to the group of intensive care patients listed here, there are also other risk groups: e.g., low-birthweight newborns or patients with already existing anemia. According to Hicks (7), the increase in elderly patients could be another reason to reduce laboratory-related blood loss. Finally, the reduction of excess amounts of blood used for laboratory testing lowers the infection risk associated with administration of blood products.
A reduction in required blood volumes is in particular necessary for the above-mentioned risk groups. As shown in the investigation by Foulke and Harlow (14), the use of special collection containers in pediatric units led to a 50% reduction in the mean daily blood loss, but according to the results of Smoller et al. (15), the use of smaller collection containers did not affect the requesting behavior for laboratory tests that are not medically indicated. This statement compares to our experiences. The introduction of special catheters that reduce the blood loss up to 30% (9) was not realized because of the higher costs of these catheters.
Other methods for reduction of diagnostic blood loss include the use of plasma instead of serum, reduction of distribution into secondary tubes (12), and the use of whole-blood analyzers(6). A reduction in the number of repetitive tests and a change in requesting behavior would be the most effective method, but this is laborious to achieve because of the habits of the medical staff on the ward (7). In our study, reporting of the individual cumulative blood loss may have had an influence on the requesting behavior because the tests per ICU patient decreased. This reduction may be achieved not only by reporting the daily cumulative blood loss on the daily laboratory report but also by discussion in the wards.
In summary, in our study diagnostic blood loss did not pose a serious problem except for patients who underwent heart surgery and some patients who underwent visceral surgery. In all other departments, critical blood loss caused by laboratory testing was seen in only isolated cases under intensive care. To inform the attending physicians properly, our daily laboratory report always indicated the current cumulative blood loss of each individual patient. We close with the following recommendation "... but we shouldn’t draw more blood than we need. It can’t be good for the patient and it means biohazardous waste for the laboratory to dispose" (16).
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  D. Aronson, W. Markiewicz, and H. Hammerman Nosocomial anaemia: reply Eur. Heart J., October 2, 2007; 28(20): 2551 - 2552. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
