Pre-operative anemia and peri-operative transfusion are associated with poor oncologic outcomes in cancers of the esophagus: potential impact of patient blood management on cancer outcomes

BACKGROUND

Both Red Blood Cell (RBC) transfusion and anemia are thought to negatively impact cancer survival. These effects have been reported with mixed findings in cancer of the esophagus. The potential impact of the application of restrictive transfusion strategies on this patient population has not been defined.

MATERIALS AND METHODS

We conducted a retrospective study of esophagectomies and studied cases based on whether they were anemic or were transfused peri-operatively. Clinical characteristics and known clinicopathologic prognosticators were compared between these groups. Survival was compared by Cox proportional hazard modeling. Post-operative transfusions were assessed for compliance with restrictive transfusion thresholds.

RESULTS

Three-hundred ninety-nine esophagectomy cases were reviewed and after exclusions 348 cases were analyzed. The median length of follow-up was 33 months (range 1-152 months). Sixty-four percent of patients were anemic pre-operatively and 22% were transfused. Transfusion and anemia were closely related to each other. Microcytic anemia was uncommon but was evaluated and treated in only 50% of cases. Most anemic patients had normocytic RBC parameters. Transfusion but not anemia was associated with a protracted/prolonged post-operative stay. Transfusion and anemia were both associated with reduced survival however only anemia was associated with decreased survival in multi-variable modeling. Sixty-eight percent of patients were transfused post-operatively and 11% were compliant with the restrictive threshold of 7 g/dL.

CONCLUSIONS

Pre-operative anemia and transfusion are closely associated, however only anemia was found to compromise survival in our esophageal cancer cohort, supporting the need for more aggressive evaluation and treatment of anemia. Adherence to restrictive transfusion guidelines offers an opportunity to reduce transfusion rates which may also improve short-term outcomes.

Clinicians' Perceptions of an Artificial Intelligence-Based Blood Utilization Calculator: Qualitative Exploratory Study

BACKGROUND

According to the US Food and Drug Administration Center for Biologics Evaluation and Research, health care systems have been experiencing blood transfusion overuse. To minimize the overuse of blood product transfusions, a proprietary artificial intelligence (AI)-based blood utilization calculator (BUC) was developed and integrated into a US hospital's electronic health record. Despite the promising performance of the BUC, this technology remains underused in the clinical setting.

OBJECTIVE

This study aims to explore how clinicians perceived this AI-based decision support system and, consequently, understand the factors hindering BUC use.

METHODS

We interviewed 10 clinicians (BUC users) until the data saturation point was reached. The interviews were conducted over a web-based platform and were recorded. The audiovisual recordings were then anonymously transcribed verbatim. We used an inductive-deductive thematic analysis to analyze the transcripts, which involved applying predetermined themes to the data (deductive) and consecutively identifying new themes as they emerged in the data (inductive).

RESULTS

We identified the following two themes: (1) workload and usability and (2) clinical decision-making. Clinicians acknowledged the ease of use and usefulness of the BUC for the general inpatient population. The clinicians also found the BUC to be useful in making decisions related to blood transfusion. However, some clinicians found the technology to be confusing due to inconsistent automation across different blood work processes.

CONCLUSIONS

This study highlights that analytical efficacy alone does not ensure technology use or acceptance. The overall system's design, user perception, and users' knowledge of the technology are equally important and necessary (limitations, functionality, purpose, and scope). Therefore, the effective integration of AI-based decision support systems, such as the BUC, mandates multidisciplinary engagement, ensuring the adequate initial and recurrent training of AI users while maintaining high analytical efficacy and validity. As a final takeaway, the design of AI systems that are made to perform specific tasks must be self-explanatory, so that the users can easily understand how and when to use the technology. Using any technology on a population for whom it was not initially designed will hinder user perception and the technology's use.

Factors influencing clinicians' willingness to use an AI-based clinical decision support system

Background

Given the opportunities created by artificial intelligence (AI) based decision support systems in healthcare, the vital question is whether clinicians are willing to use this technology as an integral part of clinical workflow.

Purpose

This study leverages validated questions to formulate an online survey and consequently explore cognitive human factors influencing clinicians' intention to use an AI-based Blood Utilization Calculator (BUC), an AI system embedded in the electronic health record that delivers data-driven personalized recommendations for the number of packed red blood cells to transfuse for a given patient.

Method

A purposeful sampling strategy was used to exclusively include BUC users who are clinicians in a university hospital in Wisconsin. We recruited 119 BUC users who completed the entire survey. We leveraged structural equation modeling to capture the direct and indirect effects of “AI Perception” and “Expectancy” on clinicians' Intention to use the technology when mediated by “Perceived Risk”.

Results

The findings indicate a significant negative relationship concerning the direct impact of AI's perception on BUC Risk (ß = −0.23, p < 0.001). Similarly, Expectancy had a significant negative effect on Risk (ß = −0.49, p < 0.001). We also noted a significant negative impact of Risk on the Intent to use BUC (ß = −0.34, p < 0.001). Regarding the indirect effect of Expectancy on the Intent to Use BUC, the findings show a significant positive impact mediated by Risk (ß = 0.17, p = 0.004). The study noted a significant positive and indirect effect of AI Perception on the Intent to Use BUC when mediated by risk (ß = 0.08, p = 0.027). Overall, this study demonstrated the influences of expectancy, perceived risk, and perception of AI on clinicians' intent to use BUC (an AI system). AI developers need to emphasize the benefits of AI technology, ensure ease of use (effort expectancy), clarify the system's potential (performance expectancy), and minimize the risk perceptions by improving the overall design.

Conclusion

Identifying the factors that determine clinicians' intent to use AI-based decision support systems can help improve technology adoption and use in the healthcare domain. Enhanced and safe adoption of AI can uplift the overall care process and help standardize clinical decisions and procedures. An improved AI adoption in healthcare will help clinicians share their everyday clinical workload and make critical decisions.

Effect of risk, expectancy, and trust on clinicians' intent to use an artificial intelligence system -- Blood Utilization Calculator

A gap exists between the capabilities of artificial intelligence (AI) technologies in healthcare and the extent to which clinicians are willing to adopt these systems. Our study addressed this gap by leveraging 'expectancy-value theory' and 'modified extended unified theory of acceptance and use of technology' to understand why clinicians may be willing or unwilling to adopt AI systems. The study looked at the 'expectancy,' 'trust,' and 'perceptions' of clinicians related to their intention of using an AI-based decision support system known as the Blood Utilization Calculator (BUC). The study used purposive sampling to recruit BUC users and administered a validated online survey from a large hospital system in the Midwest in 2021. The findings captured the significant effect of 'perceived risk' (negatively) and 'expectancy' (positively) on clinicians' 'trust' in BUC. 'Trust' was also found to mediate the relationship of 'perceived risk' and 'expectancy' with the 'intent to use BUC.' The study's findings established pathways for future research and have implications on factors influencing BUC use.

Electronic clinical decision support to facilitate a change in clinical practice: Small details can make or break success

BACKGROUND

The use of electronic clinical decision support (CDS) is becoming common to change historically common clinical practices considered outdated by current guidelines. Preimplementation design of CDS tools is key to their success in changing clinical behaviors. Unfortunately, there are no established protocols for CDS tool development, and CDS failure can result from even small design flaws. This paper describes an example of a design oversight and how correction resulted in CDS success.

STUDY DESIGN AND METHODS

We performed a retrospective review of compliance with a CDS tool to encourage the use of prothrombin complex concentrate over plasma transfusion for the emergent reversal of warfarin. We identified a potential design flaw, made the necessary modifications, and repeated the compliance review.

RESULTS

After CDS, plasma orders declined by 150 units/mo; however, 48% of orders placed for non-warfarin coagulopathy were still for warfarin reversal. Hospital-wide, this noncompliance was 36% and was 80% in the emergency department. By simply relocating the qualifier "NOT on warfarin" from the end to the beginning of the order, noncompliance for warfarin reversal was reduced to 5% (P < .0001 by chi-square).

CONCLUSIONS

The successful use of electronic clinical decision support in the electronic medical record can depend on optimal design. Missing even small design elements such as the positioning of key terms within the tool can result in an ineffective CDS. Important design strategies to avoid poor performance are discussed as they relate to the CDS tool we describe.

The blood utilization calculator, a target-based electronic decision support algorithm, increases the use of single-unit transfusions in a large academic medical center

BACKGROUND

Electronic decision support has been used to reduce use of red blood cell (RBC) transfusion. With the goal of reducing transfusions, we modified our RBC orders to default to 1 unit. Next, we created a target-based algorithm, the blood utilization calculator or BUC, to calculate a dose in units, based on initial hemoglobin or hematocrit and weight.

STUDY DESIGN AND METHODS

RBC orders defaulted to 1 unit in March 2016 and the BUC was implemented in July 2016. This gave three periods to compare old orders (before intervention), new orders (1-unit default), and the BUC period. A hospital dashboard that tracks blood product orders was queried to determine changes in single-unit transfusions between periods. Changes in transfusions were compared by analysis of variance. Acceptance of the BUC dosage recommendation was studied in both medical-based and surgical-based specialties.

RESULTS

The number of single-unit transfusions showed significant increases after each of the two interventions studied from 247 ± 19 before interventions to 358 ± 19 and then to 445 ± 141-unit transfusions/month (p < 0.0001). The ratio of 1-unit to 2-unit transfusions increased from 0.72 to 1.67 (p < 0.0001) and we observed a 19% overall reduction in units transfused. The BUC recommendation was accepted in 49% of orders.

CONCLUSIONS

One-unit default orders and implementation of the BUC resulted in a significant increase in the use of single-unit transfusions. Improvement in the rate of acceptance of the BUC recommendation should further increase the use of single-unit transfusions.

Outcomes of red blood cell transfusions prescribed in organ donors by the Digital Intern, an electronic decision support algorithm

BACKGROUND

The Digital Intern (DI) is an electronic decision support tool for the management of organ donors. One algorithm determines the dose, in units of red blood cells to be transfused, based on hematocrit (Hct) thresholds and targets. The effectiveness of the transfusion dose calculated by the DI in terms of achieving the selected Hct target and the duration of the targeted dose is not known.

STUDY DESIGN AND METHODS

This was a retrospective study to describe the outcomes of transfusions prescribed by the DI. Pre- and posttransfusion Hct levels were compared to define response and all posttransfusion Hct values were plotted to evaluate the duration of the prescribed dose.

RESULTS

A total of 120 organ donors were studied and 22 donors had 28 transfusions (six were transfused twice). The transfused donors were a mix of trauma and medical admissions and brain death and cardiac death donors. The transfusion target of 24% Hct was attained in 96% of transfusions. The mean number of units transfused was 1.4 and the mean time from transfusion to procurement was 19.8 hours. There was a decline in Hct over time after transfusion in all but one case with a mean decline of 1.9% Hct over 13 hours. Six donors were transfused twice, likely due to a longer donor time period (41.7 hr vs. 27 hr).

CONCLUSIONS

The DI provided transfusion dosing that achieved the desired threshold in the majority of organ donors transfused. Ongoing work focuses on application of this technology to transfusions in general patient populations.