Anemia is defined by the World Health Organization (WHO) as a level of hemoglobin of less than 12 g/dL (7.4 mmol/L) for women and less than 13 g/dL (8.1 mmol/L) for men, but cutoff values may differ between countries and in the literature making comparisons difficult. There seems to be little controversy that perioperative anemia and low hemoglobin may lead to increased transfusion rates, morbidity, and mortality. In a systematic review on the prevalence of anemia in patients operated with THA and TKA, preoperative anemia was found in 24 ± 9 % and postoperative anemia in 51 ± 10 % of patients (Spahn 2010). Perioperative anemia was associated with a blood transfusion rate of 45 ± 25 %, postoperative infections, poorer physical functioning and recovery, and increased length of hospital stay (LOS) and mortality (Spahn 2010). Also, in a recent multicenter study on fast track with a median LOS of 2 days, it was found that 12.8 % of patients had preoperative anemia which was associated with a 4.7-fold increased risk of receiving blood transfusion, a 1.4 times increased risk of readmission within 90 days, and a 2.5-fold increased risk of LOS >5 days (Jans et al. 2014). Earlier fast-track studies also found blood transfusion to be associated with longer LOS (Husted et al. 2008). However, a nationwide study on blood transfusion following THA found variations between departments from 7 to 71 % following this standard procedure illuminating the need for guidelines based on research and identification of an evidence-based transfusion trigger (Jans et al. 2011). Also, this study confirmed transfused patients to stay longer and to have a 5.5-fold higher mortality within 90 days following index operation (Jans et al. 2011). But even though major bleeding or severe postoperative anemia is strongly associated with blood transfusion, direct causality between transfusion and adverse outcomes remains unclear. Breaking down the data in the nationwide Danish study (Jans et al. 2011) showed the transfusion-related mortality to include cases of major perioperative bleeding or severe postoperative anemia with delayed blood transfusion in addition to possible complications to blood transfusion per se (Jans et al. 2012). The risks of blood transfusion are well established and include the risk of transmitted disease, however rare, and the associated costs of blood transfusion (Carless et al. 2012b). Also, studies have found an increased risk of prolonged wound healing and infection associated with blood transfusion (Innerhofer et al. 1999; Weber et al. 2005).

Historically, a hemoglobin cutoff of 10.0 g/dL has been used as an indication for transfusion or a drop in hematocrit of 30 %, the combination often referred to as the 10/30 rule. Nowadays, a more individualized approach is often used without a single threshold for red cell transfusion, but instead recommending a range of hemoglobin values between 6.0 and 10.0 g/dL combined with the presence of comorbidities (especially ischemic heart disease) and symptoms of anemia. To give an example, the Danish National Health Board states in its guidelines that hemoglobin values of less than 7.4 g/dL in all patients and of less than 9.9 g/dL in patients with ischemic heart disease should lead to considerations on giving blood transfusion (Sundhedsstyrelsen 2007).

Although the best transfusion trigger in elective patients remains unknown, restrictive transfusion protocols are recommended – and put forward as guidelines – with cutoffs for most patients of 7.0–8.0 g/dL (Carson et al. 2012a; Spahn and Vamvakas 2013; Carson and Hebert 2014). In accordance with this, the hemoglobin values used as transfusion triggers in most studies varies between 7.0 and 9.0 g/dL (Carson et al. 2012b).

Trying to reduce the transfusion trigger to reduce blood transfusion and the associated risks and costs is not new. It was tried successfully in coronary bypass surgery in 1999: The transfusion trigger was reduced from 9 to 8 g/dL without negative effect on fatigue, morbidity, and mortality (Bracey et al. 1999). Generally, older studies on outcome after THA and TKA have used more liberal transfusion protocols – if any – as opposed to some newer studies where more restrictive transfusion protocols have been applied with the increasing awareness also of the deleterious effects of blood transfusion. Thus, several studies have examined the outcome of a more restrictive transfusion protocol on various aspects.

A study investigating the potential association between postoperative anemia and quality of life (QoL) and fatigue at 4 and 14 days postoperatively found no association between postoperative hgb values (10.5 ± 1.1 and 11.4 ± 1.2 g/dL, respectively) and QoL or fatigue (So-Osman et al. 2011).

Another study in three hospitals compared a liberal and a more restrictive transfusion protocol in elective orthopedic surgery, the latter based on age and medical comorbidity. Results were, as expected, significantly less transfusions in the restrictive group but also included fewer infections and less respiratory complications without affecting LOS, cardiovascular complications, mortality rate, or QoL scores (So-Osman et al. 2013). The hgb trigger resulting in transfusion was set at 6.4, 8.1, and 9.7 g/dL for age less than 50 years, 50–70 years, and more than 70 years, respectively.

So, it seems that moderate postoperative anemia is well tolerated causing no impact on fatigue, QoL, cardiovascular complications, LOS, mortality, and with a potential for less infections. The question is, of course, how low postoperative hemoglobin values are tolerated without affecting early functional recovery? A study differentiated between patients and hemoglobin values (≤8.0, 8.1–9.0, 9.1–10.0, and >10.0 g/dL) finding no difference between groups in the decrease of the distance walked preoperatively versus postoperatively or in the outcome of a 6MWT (6 min walk test). Also, no differences were found with perception of effort, maximal dominant hand strength, and SF-36 QoL scores (Vuille-Lessard et al. 2012).