DOS 2017

DOS Kongressen 2017 ·

A restrictive blood transfusion limit does not affect

mortality in hip fracture patients – a regional cohort

study based on national databases

Bjarke Viberg, Per Hviid Gundtoft, Jesper Schønnemann, Lasse Pedersen, Lis Røhl

Andersen, Kjell Titlestad, Jens Lauritsen, Søren Overgaard

Department of Orthopaedic Surgery and Traumatology, Department of Clinical

Research, Institute of Regional Health Research, Kolding Hospital - part of Lil-

lebaelt Hospital, Odense University Hospital, University of Southern Denmark

Background:

In 2014, the Danish National Clinical Guidelines (NCG) for trans-

fusion with blood components using restrictive transfusions limits was intro-

duced. Very few trials have evaluated the influence of restrictive transfusion

limits on hip fracture patients.

Purpose / Aim of Study:

To estimate the association of introducing the NCG

for transfusion with blood components with mortality and transfusion frequen-

cy for hip fracture patients above 65 years.

Materials and Methods:

From 01102015 to 30092016, all hospitals treat-

ing hip fracture patients in the Region of Southern Denmark adhered to the new

NCG and the patients were included as the restrictive transfusion group (RG).

This group was compared to a liberal transfusion group (LG) from 01102012

to 30092013. Data from the Danish Interdisciplinary Registry for Hip Fractures

was collected including age, sex, Charlson Comorbidity Index (CCI), type of

fracture and surgery. Data was merged with data from the Danish Transfusion

Database and the Regional Laboratory Database. Statistic for comparing groups

and Cox proportional hazards model was performed.

Findings / Results:

2,908 patients were included with no major baseline dif-

ference in the two groups concerning age, sex, CCI, type of fracture, or type

of surgery. The 30-day mortality was 12% in LG (n=1,494) and 9% in RG

(n=1,414) yielding a relative risk of 0.74 (0.59; 0.94) for RG compared to LG

(adjusted for age, sex, CCI, and type of surgery). In LG 42% received blood

transfusions compared to 30% in RG (p<0.001). The mean (CI) hemoglobin

at the first transfusion was 5.65 mmol/l (5.57;5.72) in LG and 5.16 mmol/l

(5.08;5.24) in RG (p<0.001).

Conclusions:

The restrictive NCG lowered the percentage of patients trans-

fused with blood components without increasing the overall mortality rate.

Thus, it seems safe to implement this guideline for hip fracture patients.

No conflicts of interest reported

19.