Is it time to automate the heart lung machine?

Abstract

For six decades cardiopulmonary bypass has been manually managed by a perfusionist or an extracorporeal specialist trained in the nuances of circulatory devices, human physiology, physics, and engineering. This process allows for blood to be routinely diverted out of the body, altered, and returned to the circulation with the intention of normalizing the patient’s physiology during the period of extracorporeal support for either surgical or medical intervention for an underlying condition(s). It has worked well over time, benefiting patients of all sizes and shapes. Indeed, cardiopulmonary bypass is considered one of the most significant medical developments of the 20th century that permitted cardiac surgery to develop.1 Extracorporeal membrane oxygenation (ECMO) and ventricular assist platforms were similarly extraordinary offshoots of the technology. However, one important aspect is the control and management of the extracorporeal circuit is done manually. Consider the Gibbon-IBM device first reported in 1951.2 It is the historic reference point of all heart-lung machines and, after extensive laboratory work, made cardiopulmonary bypass a clinical reality.3,4 One of the most interesting things about that early device was that most of its controls were automated.5 A “wide variety of devices, sensors and automatic controls” were incorporated into early machines from Gibbon’s group and others. High pressure cut-offs for the systemic or recirculating pumps, high and low reservoir level detectors, automatic vacuum regulators, in line blood filters or bubble traps and passive gravity filling and self-regulating pumps.”6 These facilitated operation of the system at a time when extracorporeal sciences were at their infancy, and users were trying to figure how to manage basic parameters such as anticoagulation and pressure. Blood gas machines in the early years were concepts. Measurements and control of anticoagulation and its reversal, hemodynamics on extracorporeal support, and gas exchange values were often learned through trial and error as clinicians shared their experiences. As the field evolved, the Gibbon heart-lung machine footprint was shared, and other systems came into existence including the Mayo-Gibbon apparatus.7 Later, Richards Sarns in the United States built pump systems for the masses that made it possible to conduct cardiac surgery as it moved from academic medical centers to community hospitals. Heart-lung machines were custom built for the operator. However, one fact remained, regardless of the system, which was it required manual control and operation by a perfusionist. Manual operation led to the possibility of human error and the potential for accidents, so so-called safety devices were developed to reduce the risks of errors. Also, with the development of perfusion management protocols and compliance ranges, consensus standards were developed and promulgated,8,9 and the door for automation may have cracked open. In other industries, semiconductor and software technology today is nothing like it was even ten years ago. This has allowed for use of novel technology and even something called artificial intelligence. With such technological potential in the 21st century, it has been a difficult concept to partially relinquish control of cardiopulmonary bypass to an automated system. In some related circulatory technologies, there are now examples that are more computer-controlled. Modern hemodialysis machines, intra-aortic balloon pumps, and ventricular support devices are fully automated that, other than setup and initiation, allow their use with minimal human intervention. Destination therapy for patients with implanted ventricular assist devices would be impossible without automation. Early ECMO systems would have been impossible to run without a “bladder box” system.10 This simple switching device interrupted power to the roller pump if the pump flow exceeded venous return, thus reducing the risk of entraining air into the circuit. With advent of centrifugal pump technology and better understanding of pump preload, kinetic pumps of many different designs have replaced roller pumps and can be used in short-term and extended support, but still require manual manipulation to control blood flow. An ECMO system might be an ideal platform to be automated for control and regulation according to some preset parameter (e.g. venous oxygen saturation). More contemporary attempts at automating devices have been few, yet Pedersen et al.11 modified a commercial centrifugal pump to ramp up or down to servoregulate the sub-atmospheric pressures generated by the pump. This allowed for more precise control of the pump inlet pressure and ultimately may have prevented excessive hemolysis and possible renal side effects. Is it time to automate the heart lung machine? 1039367 PRF0010.1177/02676591211039367PerfusionEditorial research-article2021