Clinical practice for outpatients that are chronically red cell dependent: A survey in the Netherlands

Abstract Background and Objectives Limited data are available to guide physicians on how to determine the red blood cell (RBC) transfusion regimen in chronically transfusion‐dependent patients. The lack of clarity on thresholds and targets to be used for transfusion could easily result in either under or over transfusion in these patients. The aim of our survey is to investigate (1) transfusion thresholds; (2) number of RBC units given per transfusion episode; (3) interval between transfusions and (4) patient factors, like decreased cardiac function modulating the former. Materials and Methods We sent a web‐based 44‐question survey to members of the Dutch Haematology Association. Results Fifty physicians responded between June and October 2020 (response rate 30%), well‐distributed between community and academic hospitals. A wide variation in transfusion strategies was reported: Most patients have transfused 1–2 RBC units (range: 0–3 units) every 2–4 weeks (range: 1–12 weeks) with a median threshold of 8.0 g/dl ranging from 6.4 to 9.6 g/dl. Patient‐specific clinical factors that are most frequently reported to influence the transfusion strategy are angina pectoris, cardiac failure and dyspnoea, softer parameters that are of influence are the quality of life and self‐sustainability. Conclusion The results of this survey indicate a broad variation in RBC transfusion strategies in Dutch patients with chronic transfusion dependency. While the current variation in transfusion strategies may be unavoidable in an individualized approach, randomized trials and better defined usable parameters to evaluate the effect of transfusion strategies are required to reach a consensus on how to determine the transfusion strategy.


INTRODUCTION
Patients who receive red blood cell (RBC) transfusions on a regular basis typically suffer from diseases like myelodysplastic syndrome (MDS), myeloproliferative diseases (MPN), thalassaemia, sickle cell disease or aplastic anaemia. Although these patients receive 20%-30% of all RBC transfusions in Europe [1,2], limited data are available to guide physicians on how to determine a chronic RBC regimen [3][4][5][6][7]. As studies on restrictive RBC transfusion strategies show no disadvantages in other patient categories [8][9][10][11][12], physicians tend to use a restrictive transfusion policy based on the perceived Quality of Life (QoL) of the individual patient [13][14][15]. The effect of this individualized transfusion policy is hard to quantify and depends on the haemoglobin (Hb) threshold, the timing of transfusion and the number of RBC units administrated [6,7,16]. Factors like anaemia related symptoms, the (estimated) cardiopulmonary compensation capacity and the presence of iron overload might further influence the transfusion policy [17,18]. The lack of clarity on thresholds and targets to be used for transfusion, will easily result in either under or over transfusion in these patients. Both moreover, will have negative effects, with on one hand anaemia-related QoL [19][20][21][22] and on the other ironoverload and other transfusion-specific side effects [23][24][25].
We performed a survey amongst haematologists in the Netherlands to investigate their current standard of care for patients dependent on chronic transfusions, regarding the applied (1) transfusion thresholds; (2) number of RBC units given per transfusion episode; (3) interval between transfusions and (4) patient factors, like decreased cardiac function modulating the former.

Study design
In this cross-sectional study, a structured, 44-question online survey was conducted amongst members of the Dutch Society of Haematology

Study cohort
We sent out a survey to 169 members of the NVvH in the Netherlands of which 50 physicians responded between June and October 2020 (response rate: 30%

General response
In the first part of the survey, we asked physicians to indicate, which of a set of 19 given factors affects their transfusion strategy in general. In Figure 1

Patient-specific response
The reported Hb thresholds for RBC transfusion-dependent patients are depicted in Figure 2. The most-reported threshold ( Table 1 and depicted in Figure 3. The mean threshold/ (interval Â transfusion amount) does not show a particular trend, suggesting that the variation in thresholds is independent of these two factors and more dependent on patient-specific parameters. Although no single strategy prevails most, a common frame can be identified: 63% of the reported patients are transfused 1-2 units every 2-4 weeks.

Parameters associated with changes in transfusion strategies
We   Although most physicians report that they try to adjust transfusion strategies to avoid long-term negative outcomes due to iron overload, they also report not necessarily taking iron parameters into the equation (Figure 2 Because we included incomplete surveys, we performed a nonresponse error analysis to evaluate whether bias might have been introduced by non-responders. A relatively high amount of respondents failed to report the patient's most recent ferritin (5/58; 9%). We did not find a difference between those who did not report the ferritin level and those who did. Other questions were not left unanswered more frequently than others. Regarding the underlying disease, patients suffering from sickle cell disease and thalassaemia had a lower mean threshold (7.2 and 7.5 g/dl, resp.) and higher interval (median 6 weeks) compared to the other included patients (Hb-threshold >8.2 g/dl, interval = 3 weeks).

DISCUSSION
This may be explained by a lifelong need of these patients to cope with lower Hb-levels but also by the known higher alloimmunization levels of these patients and thus more focus on restriction. In addition, transfusion goals differ between different indications: whereas an MDS patient receives transfusions to improve the oxygentransportation capacity, the aim in patients with sickle cell disease is also to decrease the proportion of sickle Hb relative to Hb A to prevent or reverse vaso-occlusive disease.
Patients receiving chemotherapy are transfused at a lower mean threshold (7.6 g/dl) as compared to those with immunomodulatory drugs or no active therapy (8.5 and 8.3 g/dl, resp.). The suppression of haematopoiesis by chemotherapy, although more severe, is also more likely to be reversible. Physicians may thus accept a lower Hb level as a rise is to be expected when the chemotherapy's side effects wear off.
There may be various arguments to transfuse at a higher more liberal threshold: a physician might aim to increase the patient's QoL; or to counter a patient's poor tolerance of low Hb-levels. Vice versa, restrictiveness with thresholds below 8 g/dL seems only enforced if there are no clear risk factors such as a reduced estimated cardiopulmonary compensation capacity ( Figure 5).
Clearly, the exact effect of a transfusion on the patient is hard to quantify. As few physicians report to measure post-transfusion Hblevels, this also appears to be a parameter of little value. Therefore, tages on the accumulated parameters like morbidity and mortality [8][9][10][11][12]. Others may choose a more liberal strategy because a higher Hb-level is believed to be beneficial for the transfusion-dependent patient [27]. While variability in transfusion strategies between patients may not be a problem at all, the lack of guidelines should not lead to large interphysician differences, as this will cause both under and over transfusion. Nevertheless, the observed range of thresholds does not find a base in Dutch or international guidelines [13,28].

SUPPORTING INFORMATION
Additional supporting information may be found in the online version of the article at the publisher's website.