Objective: To assess the effect of introducing consensus guidelines on avoidable blood transfusions in Mwanza region, Tanzania.
Methods: Avoidable blood transfusions were determined among 842 blood transfusion recipients in eight hospitals in 1991. In a workshop with senior health workers from the region, consensus guidelines for the prescription of blood transfusions were developed and introduced in the hospitals and after 7 months intervention data were collected on 1042 blood transfusion recipients. The 1991 and 1992 data were compared to estimate the change in the proportion of avoidable blood transfusions.
Results: In blood transfusion recipients aged < 5 years there was a significant reduction in the proportion of avoidable blood transfusions from 257 (52%) out of 498 to 197 (33%) out of 595 (P < 0.001), especially at the peripheral hospitals. For children the proportion decreased from 25 to 17% (P < 0.05) and for operated patients the percentage remained at 24%. In pregnant women there was a significant increase in the proportion of avoidable blood transfusions from 10 to 27% and in adults from 37 (25%) out of 146 to 121 (50%) out of 242. The improvement in peripheral hospitals was offset by a similar deterioration in the referral hospital, thus no overall reduction was achieved in the proportion of avoidable blood transfusions.
Conclusion: The development and introduction of consensus guidelines was not sufficient to change prescribing practice. The proportion of avoidable blood transfusions decreased only in hospitals where compliance was maintained through regular clinic meetings and strict supervision by senior medical staff.
PIP: In 1991, senior health staff in the Mwanza region of Tanzania achieved consensus guidelines for the prescription of blood transfusions and introduced these guidelines through training workshops to staff of all hospitals providing blood transfusions. Seven months after the workshops, researchers collected data on 1042 blood transfusion recipients to evaluate the effect of the introduction of the consensus guidelines on avoidable blood transfusions. A significant reduction in the proportion of avoidable blood transfusions occurred among the 595 5-year-old children (52% vs. 33%; p .001). The reduction was even greater in the peripheral hospitals (59% vs. 32%; p .001). The proportion of avoidable blood transfusions fell considerably among 5-14 year old children (25% vs. 17%; p .05). Avoidable blood transfusions did not decline among operated patients (24%). They increased significantly among adults (25% vs. 50%) and among pregnant women (10% vs. 27%). The referral hospitals had more avoidable blood transfusions at evaluation than at baseline (45% vs. 26%; p .001), while the peripheral hospitals performed significantly better in 1992 than in 1991 (28% vs. 46%). Most of the improvement in peripheral hospitals was limited to 1 hospital (61% vs. 23%; p .001). Among the remaining peripheral hospitals, the improvement was slight (35% vs. 30%; p .05). Hospitals with regular clinic meetings to discuss the consensus guidelines and senior medical staff who strictly supervised and provided feedback to interns either improved significantly or maintained their already good blood transfusion practices. The hospital that maintained an already low proportion of avoidable blood transfusions required physicians to include the indication for blood transfusion in patient records. These findings show that there was no overall reduction in avoidable blood transfusions, and that the development and introduction of consensus guidelines alone did not change prescribing practices.