Correspondence

DOUGLAS W. HUESTIS, M.D. , Editor

Tucson, Arizona

To the Editor Hattersley and associates described a walking

donor program for premature infants in Transfu- sion, 16:366, 1976. In a previous paper, Oberman2 discouraged t h e use of walking donors for pediatric transfusions. Hat ters ley criticized Oberman for providing no evidence that when such programs are not managed by the hospital transfusion service, their administration is difficult to maintain and presumably fraught with the hazards of nonstandard blood bank technique. Hattersley and associates then made several statements themselves for which they provide no evidence. Among their unsupported statements are:

1.

2.

3.

“In our opinion, survival was unquestionably improved by these transfusions.” “The willingness of the pediatric staff to ob- tain blood samples at frequent intervals might have been lessened under any of the non-walking donor systems described.” “The advantage of the walking donor for such patients is primarily availability along with the belief that heparinized blood better meets the needs of the acidotic infant than does citrated bank blood.”

liquid of life.”’ Phillip L. Howard, M . D . , Associate Professor, The University of Vermont, Department of Pathology, Medical Alumni Build- ing. Burlington. Vermont 05401.

References

1. Myhre, B. A,: Blood-legend laden liquid of life.

2. Oberman, H.: Special report-transfusion of the Transfusion 16:270, 1976.

neonatal patient. Transfusion 14:183, 1974.

Dr. Howard’s letter was referred to Dr. Hattersley. who replied as follows:

To the Editor: Taking our “unsupported statements” in the

order in which Dr. Howard quoted them, I can only say:

I . We obviously do not have any controlled series from which to draw statistics. The death rate among acidotic prematures in many hospitals not using a walking donor program is so very high, however, that we cannot help feeling that our transfusion program is saving lives.

2. Members of our house staff have exmessed Opinions regarding survival without the sup- very clearly the feeling that they are more willing

porting evidence of adequate controls are hardly to draw blood from a tiny baby if a walking donor “unquestionable” or acceptable as proof. If the is immediately available to replace the lost blood. authors indeed do have mortality figures for inten- 3. Availability of the walking donor is clearly a sive care nurseries where the only or the principal great advantage, as stated above. Regarding supe- variable is the method of transfusion, these data riority of heparinized blood over citrated bank should appear in the paper to support their blood-the infant whose case history we presented opinion. The same can be said regarding the effect was acidotic, with a pH of 7.12. With freshly of the pH of CPD blood versus heparinized blood. drawn C P D blood in our hands showing a pH of Equally difficult to assess is their comment re- 6.9 to 7.1, and ACD blood even a lower pH, we garding the willingness of pediatric staff to obtain feel that the higher pH of heparinized blood was samples as a reflection of the type of transfusions clearly of more help to this youngster. By no given. means all premature infants are this acidotic, and

What Hattersley and associates did establish is many will do very well on CPD blood. But for a that it is possible to set up a walking donor youngster such as this, fresh heparinized blood ap- program which can be made to work. However, peared to us to be the best-Paul G. Hattersley. they do not assess the effect of the delays in get- M . D . , Professor o f Internal Medicine ting blood transfused into the infants that result (Hematology and Oncology) and Pathology, from the necessity to postpone transfusion while University of California, Davis, California 95616. serologic tests for syphilis and hepatitis testing are [It may be important to point out that amounts of done on the proposed donor. That the use of walk- blood as small as 5 to 10 ml still fall within the de- ing donors is superior to standard blood bank finition of a “unit of blood,” and hence are subject practice is not apparent from their article. It to the FDA’s Good Manufacturing Practices, as seems that walking donors along with fresh whole specifically pointed out by the Commissioner blood continues to be part of the “legend laden (Federal Register 40, #223, 18 November 1975, p.

396 Transfusion July-Aug. 1977

Volume 17 Number 4

Volume 17 N u m b e r 4 CORRESPONDENCE 397

53533); they therefore must be subjected to all ap- propriate testing before transfusion, as required by law for any unit of blood. In addition, we must ask whether the higher pH of heparinized blood is a real consideration even in acidotic babies in the case of transfusions amounting to only a small proportion of the baby's blood volume.-Ed.]

To the Editor: We wish to raise three points regarding the

platelet/leukapheresis techniques recommended by Aisner et al. (Transfusion 16:437, 1976). First, the formula they use to determine the volume of blood processed significantly underestimates the true volume of blood entering the bowl. We found that the volume of blood processed, as calculated by the Haemonetics manufacturer's formula, was 10 to 15 per cent less than what was going through the blood pump based on predetermined flow rates. This difference was because the hematocrit of the red blood cell mass in the bowl was in- variably greater than the 77 per cent figure given by the manufacturer, and the donor's hematocrit declined with each cycle. Our work indicated that a more precise determination of the volume processed can be obtained using the following for- mula:

Volume of blood processed = A + B +

Where:

C - D - E

A = size of bowl times the number of cycles, e.g.. 225 x 6 = 1,350

B = weight of platelets and/or WBC collected +- 1.03

C = sum of the weight or volume of plasma collected in the plasma bag following each cycle

D = volume of ACD or HES used E = volume of saline used proximal to the blood

Secondly, in the calculation of the number of platelets and WBC in each 15 ml aliquot the authors assume that the number of cells they counted in their sample was representative of the number in the preceeding 15 ml. However, the data in Figures I and 2 of their paper would sug- gest that the number of cells in each ml of fluid emerging from the bowl fluctuates considerably, and that it is incorrect to extrapolate from a microliter sample to the preceeding 15 ml.

Lastly, Aisner and his colleagues suggest that one should centrifuge the platelet suspension to remove red blood cells if an ABO incompatibility exists between the donor and the recipient. A far easier step would be to avoid collecting the red blood cells in the first place (i.e., stop collection at

Pump

the appearance of red color in the exit tube) since, according to their data, approximately the same percentage of platelets would be lost (0.24 x 10" out of 1.54 x 10" or 15.6%) as with centrifuga- tion.-Richard A . Kahn, Ph. D., Scientific Direc- tor, Missouri-Illinois Regional Red Cross Blood Program. St . Louis, Missouri 63108. Timothy Coburn. MT(ASCP) SBB. Director, Technical Services, Northeast Regional Red Cross Blood Program, Boston, Massachusetts 021 15.

The foregoing letter was submitted to Doctors Aisner and Schiffer, who replied as follows:

To the Editor: The findings of Kahn and Coburn that the pack-

ing hematocrit (HCT) within the Haemonetics bowl exceeds 77 per cent agrees with the data presented by the manufacturer1 and is likely a function of flow rate. In our experiments the bowl was filled at a rate of approximately 60 ml/minute and the packing HCT was thus approximately 77 per cent. Should the predetermined flow rates be in error, then the packing HCT would be shifted slightly along the curve and would range from 76 to 83 per cent. In our experiments the volumes and efficiencies were calculated on the basis of a single cycle. We have previously determined that blood counts do not vary during the withdrawal of a single unit of blood.' We agree that the donor's hematocrit declines with the return of each cycle, and the volume of blood processed increases slightly with each cycle. Although it is somewhat less accurate, we have found that an easier for- mula for estimating the volume of blood processed and efficiencies for multiple cycles can be based on the mean of the pre- and post-donation counts.

Volume processed = bowl size/mean HCT x 0.77 x number of cycles

Efficiency = Yield/ Volume processed x mean counts

Using these formulas to calculate efficiencies for 20 recent platelet donations in ACD and 20 recent platelet and granulocyte donations with 2 per cent sodium citrate in 6 per cent Hydroxyethyl starch, we have found very close agreement with the calculations based on a single cycle as shown in Table I .

Second, the samples collected for counts and differentials were 2 ml aliquots. Calculating the total number of cells in each aliquot results in a series of bar graphs which, when connected by a continuous line, results in the graphs (Figs. I and 2) published. The reverse process, i .e., making bar graphs at various points along a continuous curve, is a method for approximating the area under that curve. We have performed similar experiments