Transfusion triggers: how low can we go?

[Treating severe acute anemia due to vaginal bleeding in the Jehovah's Witness: a report of 2 cases]

For reasons of religious belief, Jehova's Witnesses do not accept blood transfusions or the infusion of blood products. In situations in which severe, life-threatening anemia develops, patient refusal to receive a transfusion can create serious ethical and legal problems. The principle of patient autonomy, which implies the freedom to accept or reject treatment, comes into conflict with the physician's obligation to safeguard the patient's life using all means possible. We report 2 cases of severe anemia in Jehova's Witnesses. One was due to menorrhagia and the other to postpartum bleeding. The physician should be aware of alternatives to infusion of blood products and know how to cope with an unexpected critical event in these patients. The measures we took were effective in our patients. In the case of menorrhagia, hormone treatment is effective when the woman wishes to preserve the ability to conceive and avoid surgery (endometrial ablation and hysterectomy). In postpartum bleeding refractory to conservative treatment, selective embolization of bleeding vessels may make it unnecessary to resort to more aggressive treatment, such as obstetric hysterectomy.

The utility of an automated electronic system to monitor and audit transfusion practice

BACKGROUND AND OBJECTIVES

Transfusion laboratories with transfusion committees have a responsibility to monitor transfusion practice and generate improvements in clinical decision-making and red cell usage. However, this can be problematic and expensive because data cannot be readily extracted from most laboratory information systems. To overcome this problem, we developed and introduced a system to electronically extract and collate extensive amounts of data from two laboratory information systems and to link it with ICD10 clinical codes in a new database using standard information technology.

MATERIALS AND METHODS

Three data files were generated from two laboratory information systems, ULTRA (version 3.2) and TM, using standard information technology scripts. These were patient pre- and post-transfusion haemoglobin, blood group and antibody screen, and cross match and transfusion data. These data together with ICD10 codes for surgical cases were imported into an MS ACCESS database and linked by means of a unique laboratory number. Queries were then run to extract the relevant information and processed in Microsoft Excel for graphical presentation. We assessed the utility of this data extraction system to audit transfusion practice in a 600-bed adult tertiary hospital over an 18-month period.

RESULTS

A total of 52 MB of data were extracted from the two laboratory information systems for the 18-month period and together with 2.0 MB theatre ICD10 data enabled case-specific transfusion information to be generated. The audit evaluated 15,992 blood group and antibody screens, 25,344 cross-matched red cell units and 15,455 transfused red cell units. Data evaluated included cross-matched to transfusion ratios and pre- and post-transfusion haemoglobin levels for a range of clinical diagnoses. Data showed significant differences between clinical units and by ICD10 code.

CONCLUSION

This method to electronically extract large amounts of data and linkage with clinical databases has provided a powerful and sustainable tool for monitoring transfusion practice. It has been successfully used to identify areas requiring education, training and clinical guidance and allows for comparison with national haemoglobin-based transfusion guidelines.