Near-infrared spectroscopy in perioperative medicine

Near-infrared spectroscopy (NIRS), originally described in 1977, entered clinical practice as a non-invasive method to assess regional tissue oxygenation.1 This monitoring technique was initially used in high risk surgeries to measure cerebral oxygenation in real time. After encouraging results that linked intraoperative NIRS guided, goal directed treatment to improved neurological outcomes in cardiac surgery, NIRS use quickly expanded.234 This increased use has sparked the development of multiple competing clinically approved NIRS monitoring systems, highlighting the substantial market interest in this technology.5

Despite the widespread use of NIRS, negative findings from subsequent trials have raised concerns about whether NIRS monitoring truly improves outcomes.678 These concerns have prompted a debate regarding the surgical populations who might benefit from NIRS and the best practices for its use.9 In a linked research paper (doi:10.1136/bmj-2024-082104), Han and colleagues present noteworthy findings from the Bottomline-CS trial, which examined whether perioperative care guided by cerebral and peripheral tissue oximetry enhanced clinical outcomes in patients undergoing off-pump coronary artery bypass graft surgery (CABG).10 This trial is remarkable in multiple ways. It is by far the largest randomized controlled trial to study this question in surgeries associated with substantial hemodynamic fluctuations.1112 NIRS was also paired with continuous hemodynamic monitoring to obtain estimates of cardiac output, stroke volume, and systemic vascular resistance, which were used to guide treatment. While a pragmatic design was adopted to approximate the real world use of NIRS where anesthesiologists make individualized care decisions, a goal directed diagnostic and interventional framework was provided to keep tissue oxygenation within +/−10% of presurgical values. Finally, the trial randomized 1960 patients with sufficient power to detect a clinically relevant difference in the incidence of a meaningful composite outcome capturing one or more serious complications at 30 days. Considering the size and design of the Bottomline-CS trial, the authors’ major conclusion is consequential. The findings indicate that the routine use of NIRS during off-pump CABG and similar types of surgery does not reduce the incidence of common postoperative complications, a conclusion well supported by the data.

Every trial has limitations. As a single center study, results of the Bottomline-CS trial might not generalize to other practice settings. The use of a composite outcome, a commonly used technique for comprehensively capturing postoperative complications with sufficient statistical power, can highlight more frequent but less severe complications. In this context, it is notable that the 95% confidence intervals of the reported risk ratios were quite narrow for several neurological, cardiac, pulmonary, renal, and infectious complications captured by the composite outcome. The study groups did not differ significantly in any specific complication, and narrow confidence intervals suggest that negative findings are more likely to be true than falsely negative. However, for complications associated with wider confidence intervals, including some infectious complications, subsequent studies are required before making such conclusions.

Notably, the Bottomline-CS trial does not address whether the routine use of NIRS in clinical conditions with no or faint pulsatile flow reduces postoperative complications. Although off-pump CABG surgery is an excellent clinical model for studying NIRS in a scenario with hemodynamic instabilities, the majority of CABG and other cardiac surgeries involve the use of cardiopulmonary bypass.13 During cardiopulmonary bypass, circulatory flow is non-pulsatile, rendering standard monitors inadequate to assess oxygenation and perfusion in real time. In this scenario, NIRS might be particularly useful for the early detection of catastrophic equipment failures or bypass cannula malposition, allowing for quick intervention and prevention of irreversible end organ damage.1415 NIRS guided, goal directed treatment might also prove advantageous during prolonged periods of non-pulsatile flow and improve clinical outcomes.11 Furthermore, NIRS could enhance the detection and management of differential hypoxia syndrome or extremity ischemia associated with non-pulsatile flow. Patients undergoing long term extracorporeal membrane oxygenation are a relevant example.16 Examining the potential benefits of NIRS in these clinical settings is an essential next step.

Finally, current NIRS monitoring techniques are limited. Although the most common use of NIRS is for monitoring cerebral oxygenation, most devices only monitor frontal brain regions. They do not provide information about regional oxygenation in other brain areas, which can differ, especially in patients with cerebrovascular disease. This limitation has led to criticism that current NIRS monitoring techniques have a considerable event rate of false negatives where patients with normal NIRS-derived oxygenation values have, in fact, experienced cerebral ischemia.17 Multichannel NIRS arrays that offer better representation of the entire brain are a common tool in neuropsychology and have made an appearance in neurocritical care.18 It seems appropriate to assess the usefulness of these devices in the perioperative setting.

NIRS remains a powerful tool to assess regional oxygenation, but its current clinical implementation might not yield benefit in most perioperative settings. However, cardiac surgery and monitoring of patients on mechanical circulatory support are areas where NIRS could be valuable. A nuanced approach to its clinical use is needed, along with further trials to evaluate its use in specific patient populations and to investigate the clinical usefulness of next-generation systems.