This study received Institutional Review Board approval (CERAR IRB00010254-2020-131 - June 16, 2020). All eligible participants submitted a digital form of informed consent.

Voluntary clinicians were recruited in 3 ways to participate to this web-based questionnaire study. First, 2000 flyers with a QR code were distributed during the 3 days of the national congress of the French Society of Anaesthesia and Intensive Care (SFAR) in September 2021. Usually, this congress includes about 5000 to 6000 participants, including physicians, residents, and anesthetic nurses. Second, in the 2 months following the congress, we sent an email via the SFAR network to all members with a brief explanation and the same link. In parallel, the SFAR and the Association of Young Anesthesiologists (AJAR) shared the link on their social networks (Facebook, Instagram, and Twitter [subsequently rebranded as X]). All clinicians recruited were invited to share the link to the questionnaire with other anesthesiologists.

Once on the website, care providers selected whether or not to participate in this study. The first web page explained the aim of the study and reminded the reader of inclusion and exclusion criteria. Participants included only French-speaking anesthesiologists, including senior on-going physicians, fellows, and residents. Assistants, specialized nurse anesthesiologists, and medical students were not allowed to participate in the study.

If the participant was eligible, a digital form of informed consent was submitted. There was a brief explanation of the study, which specified that only anesthesiologists could participate and that the results would be anonymized. They agreed that their data would be anonymously used. We also provided the reference number of the Institutional Review Board approval so the participants could check it if they wanted. Then, access to the questionnaire was opened via a public web portal using an internet-based polling software called Typeform [22,23], which allowed participants to anonymously complete the questionnaire from different devices or navigators. Only 1 access per participant was available based on the participant’s IP address. Once consent was obtained, the participants approved their email to finally be randomized. Some questions about demographics were asked at this step.

The survey was created and validated by a committee of PBM experts. The final survey included 7 short clinical case vignettes containing up to 3 multiple-choice questions each (totaling 12 questions). These cases specifically involved potential preoperative and postoperative anemia situations. Some of them were built according to existing guidelines, while others led to an open discussion due to a lack of adapted recommendations. Moreover, no questions about intraoperative management of patients with anemia or control of bleeding were proposed to focus on the perioperative course (Table 1). Following receipt of informed consent, participants were randomly and equally divided into 2 groups: one with decision support and one without (the control group). The decision support group benefited from cognitive assistance during the survey; this was not available to the control group. The cognitive assistance consisted of the current recommendations about anemia management. This assistance was only visible if an error was made in the clinician’s responses. The clinician then had the choice to select a different answer or not.

The following demographic characteristics of the population were collected: age, experience, specialties, and site of practice (public, private, or mixed). The time to complete the survey was recorded. Moreover, the quality of the responses was analyzed (false or correct). Cognitive assistance consisted of the current recommendations about anemia management. In fact, if a participant in the decision support group gave a wrong answer, a window appeared showing the current recommendations and their references (Table 1). The participants then had the choice to follow the recommendations and change their answer or to continue with their original answer. In the control group, participants were not able to go back and change answers to previous questions during the study. At the end of the survey, scores were made available to the participants. Individual participant data will not be shared.

The number of participants was calculated to allow for interpretation of the results using parametric tools. The main hypothesis was superiority in the rate of correct responses among physicians in the decision support group, with a planned difference of 25%. The estimated rate of correct response in the control group was 66%. With an α score of 5% and a power of 90%, at least 110 physicians were required. Moreover, to ensure a national representation of experience and geography, we determined the required number of completed forms to be 500 out of 9000 active physicians.

The data are expressed as the median (IQR) and mean (SD) according to their nonnormal or normal distribution. The quality of the questionnaire was determined based on the percentage of correct responses, and the comparison was mainly performed using chi-square and Wilcoxon-Mann-Whitney tests, as appropriate. The analysis was performed using SPSS 12.0 (IBM).

This survey demonstrated that better performance in the management of preoperative anemia can be achieved if the anesthesiologists, regardless of experience, have at their disposal a cognitive aid (P<.001). Usually, guidelines require a certain time before being implemented, and decision support may accelerate this shift and help physicians make a decision, especially in rare situations or situations with a low level of prior knowledge.

The World Health Organization defines PBM as “a patient-focused, evidence-based, and systematic approach to optimize the management of patients and transfusion of blood products for quality and effective patient care. It is designed to improve patient outcomes through the safe and rational use of blood and blood products and by minimizing unnecessary exposure to blood products” [29]. Indeed, the incidence of preoperative anemia (about 30% regardless of the type of surgery) is critically important and increases with age. Additionally, preoperative anemia, which represents an independent risk factor of morbidity and mortality, may delay functional recuperation with a major risk of loss of autonomy in older patients [8,30]. The interest in the preoperative field is the potential reversibility of this factor if recognized with a different line of therapy, as included in different guidelines.

Despite their clinical and physiological fundamental importance, the strategies for optimizing the management of perioperative anemia are not always well-known and applied by anesthesiologists in daily practice. It is essential to understand the underlying reasons to improve practices and quality of care. This was the primary aim of the present study, and we observed weak knowledge of the recommendations among active anesthesiologists, with a mean rate of 6.2 (SD 2.1) out of 12 correct responses.

The second conclusion is the potential value of an immediately available cognitive aid, which significantly improved the rate of correct responses in our study. Several reasons may explain the moderate adherence to guidelines about perioperative anemia management. First, the importance of PBM is a relatively new topic that requires visibility and is not well-known by clinicians. In fact, the occurrence of complications in patients with anemia are mid-term or long-term complications, and the delay between general anesthesia and occurrence of complications may disconnect the physician from the patient’s daily analysis and prescriptions. Moreover, at the time of our study, some situations had been noted in the last recommendation, especially in the Frankfurt Consensus Conference published in 2018, and there was probably a need for updating to include additional situations, such as chronic renal failure and functional surgery [17,25]. Finally, strategies of PBM take time to be fully efficient, sometimes requiring a few weeks, and if the experts suggest postponing an elective surgery to correct iron deficiency or anemia, this is still difficult in many departments due to the typical course of patient care, the need to respect surgical planning, and the patient’s needs and wishes. To postpone an elective surgery remains difficult if the parties involved (eg, surgeon and anesthesiologist) do not share the same view of the importance of anemia management, and this is still the case in many teams. Furthermore, these strategies represent additional costs in the patients’ pathway, even if they reduce morbidity and mortality on balance [31].

The other aspect concerns the implementation of a cognitive aid. On one hand, there is communication work to be carried out to formulate clear and applicable recommendations in daily practice, which is a work in progress [32]. On the other hand, the addition of a decision support tool significantly improved patient management by reminding practitioners of the current recommendations. A simple line of work to help improve the quality of care could then be an implementation of these kinds of aids within the software used by anesthesiologists for their consultations, for example. Interest in such a solution was recently explored regarding the decision to perform transfusion. One single center randomized controlled trial compared young physicians who received computerized decision support about transfusion to those who did not [33] and showed an increase in the appropriate transfusion rate (RBCs, platelets, fresh frozen plasma) in the computerized decision support group (40.4%) compared with the control group (32.5%; risk ratio 1.24, 95% CI 1.13-1.37). Additionally, 3 cohort studies assessed RBC usage before and after the intervention [34-36] and showed a significant reduction in overall or inappropriate RBC usage (RBC transfusions per 100 inpatient days, P<.001) after computerized decision support was implemented. A statistically significant reduction in overall or inappropriate RBC usage over time (P=.01) was also reported. Furthermore, reduced 30-day re-admission (5.2%; risk ratio 0.62, 95% CI 0.56-0.69) and mortality (2.2%; risk ratio 0.60, 95% CI 0.51-0.71) were found in one single center trial. This cognitive reminder could be applied not only to anesthesiologists but to surgeons or any specialists involved in the medical management of the patients to share the crucial role of optimization. In view of its demonstrable benefits, there is an increasing awareness of the need to integrate the pillars of PBM within routine surgical care. In the United States, PBM has been successfully introduced in some centers [37], while in western Australia, it has become the standard of care [38]. PBM initiatives in Europe, however, have been variable and inconsistent, reflecting the difficulties that can be met with its implementation [39]. Multiple barriers, including lack of knowledge, interdisciplinary commitment, resources, and concerns, limit the translation of PBM guidelines into clinical practice [33-35,40]. Education and training should be considered to increase awareness of the clinical implications of anemia and the need for alternatives to transfusion. Typically, the lowest rate of correct responses in the control group concerned the management of anemia in patients with chronic kidney failure experiencing some trouble with iron supplements, erythropoietin, and situations with hemoglobin levels greater than 11 g/dL and high ferritin. Other cases with a low rate of correct responses were orthopedic cases and the management of anemia for pregnant women. Certain situations are well-described in the guidelines, but some patients fall in grey areas, warranting the use of cognitive aids to make the most pertinent choice. Once any deficiencies in the underlying knowledge have been addressed, attitude and behavior should change [41]. All recommendations and standard operating procedures must be easily accessible and aimed at supporting clinical judgment as the cornerstone of patient care.

The main population who completed our survey was composed of 763 active anesthesiologists; this is a large sample size with a high rate of complete answers and represents all our care areas and experience levels. This was permitted by using different channels of diffusion. We created the cases with the help of experts in PBM and made them representative of real-life, everyday clinical practice, which is an important strength.

However, there are several limitations of our study. The questionnaire study was destined only for French participants. Furthermore, there was a bias in participant recruitment, and we can assume a worse rate of correct responses among nonresponding physicians or incomplete questionnaires. Third, we did not track the rate of response changes in the decision support group to objectively measure the effect of receiving help. Additionally, we could not verify that all the participants truly were anesthesiologists; if they declared that they were we approved their participation without any further control. Finally, recommendations evolve quickly, which may have misled some participants.

The recommendations about current PBM are not well-known by practicing physicians. We found that the implementation of decision support is very useful to face frequent and diverse situations of anemia and to improve the quality of patient management. Collaborative and continuous efforts to translate PBM guidelines into clinical practice, if done in an engaged, multidisciplinary, organized, and structured way, could make PBM the norm.