**1.3 Validation**

6 Infrared Spectroscopy – Life and Biomedical Sciences

nirswkshop1999.htm) Once NIRS monitors became commercially available a few animal and many clinical trials were conducted. The clinical investigations were for the most part small, brief observational prospective studies. Also NIRS was introduced into daily practice by others at that time, years before normative data and validation studies had

There is great potential to use the NIRS technology in the neonatal intensive care unit (NICU) since it is a portable, continuous, non-invasive bedside monitoring technique. Following the development of small and skin friendly sensors and FDA approval of some NIRS monitors for use in neonates, both research and clinical use of NIRS in the NICU increased exponentially. The number of research projects over the last 5-10 years is large. However, the trials, while dealing with questions important to understanding physiology and clinical care in the NICU, are small and almost exclusively conducted at single centers. Often no more than 10-20 patients are being followed. Very large NIRS related studies enrolled 40-90 patients. Many of the observations reported are of brief sampling periods,

This chapter is a limited overview for non-clinicians such as engineers and science students, or clinicians who want to learn about a medical application of NIRS. The recent introduction of the NIRS technology into neonatal medicine is used as an example of how a new device came into use into use in the clinical setting over the last decade. Main areas of clinical use and supporting studies will be mentioned. Limitations of NIRS technology and controversies as well as future directions will be addressed. With the abundance of available literature this chapter cannot claim to be a reference. This is an exciting and rapidly advancing field with new studies published even as this article was sent to press. This chapter will demonstrate how a new technology is adopted into medical care, in this case

Pub Med and Google have been queried regarding NIRS in NICUs, abdominal/splanchnic,

Near-infrared spectroscopy has been used as a tool to determine the redox state of lightabsorbing molecules. This technology is based on the Beer-Lambert Law, which states that light transmission through a solution with a dissolved solute decreases exponentially as the concentration of the solute increases. In mammalian tissue, only three compounds change their spectra when oxygenated: cytochrome *aa3*, myoglobin, and hemoglobin. Because the absorption spectra of oxyhemoglobin and deoxyhemoglobin differ, their relative concentrations within tissue change with oxygenation, and the relative concentrations of the types of hemoglobin can be determined. Because NIRS measurements are taken without regard to systole or diastole, and because only 20% of blood volume is intra-arterial, spectroscopic measurements are primarily indicative of the venous oxyhemoglobin concentration. In the near infrared region (700 –1,000 nm), light transmits through skin, bone, and muscle without attenuation. (Cohn et al., 2003) There are several FDA approved

cerebral and renal measurements, utility, and of NIRS use as prognosticator.

The principle of how NIRS works in humans was excellently summarized by Cohn:

been obtained.

the NICU.

**1.1 Materials** 

sometimes being no more than spot samples.

**1.2 Technology and measurements** 

NIRS was implemented by many enthusiastic clinicians without a vast body of previous research evidence. This phenomenon may be representative of an era of limited funding for larger studies linked with the promise of a non-invasive "safe" monitoring technology.

Before human application the initial research applying NIRS to measure rSO2 technology in the medical field occurred in the laboratory: One of the first examples of validation used a phantom brain model in which O2, N2, and CO2 content of a blood perfusate could be altered during measurements. The results correlated with findings in animal models. (Kurth et al., 1995) Later NIRS was further validated for the neonatologist in a newborn piglet model. The carotid, renal and mesenteric arteries were occluded and reperfused. These interventions led to rapid, simultaneous changes in rSO2 of the affected end-organs. (Wider, 2009) Furthermore, there have been validations in patients during intensive care, extracorporeal membrane oxygenation (ECMO) and cardiac surgery by comparing central blood samples with NIRS values. (Abdul-Khaliq et al., 2002; Benni et al., 2005; Nagdyman et al., 2004; Rais-Bahrami K et al, 2006; Weiss, 2005) Menke found reproducibility to be good as well. (Menke et al., 2003). The accuracy of data is impacted by light scattering, hemoglobin concentration and chromophores such as melanin and bilirubin. In the presence of a thicker overlying tissue layer, such as severe subcutaneous edema or excess subcutaneous fat, it may be impossible for the NIR light beam to reach the target organ. In the newborn modest changes in weight have a small effect on abdominal measurements while changes in hemoglobin over the first weeks of life can change measurements by 30-50%. (Ferrari et al.,

Use of Near-Infrared Spectroscopy in the Management of Patients in

**2.2 Variability** 

amongst the more common:

study of 90 patients. (Soul et al., 2007)

Neonatal Intensive Care Units – An Example of Implementation of a New Technology 9

When reviewing this literature regarding the contradicting study results, possible explanations present themselves: Patient populations are not identical. Protocols vary from study to study. Different sampling times may play an important role in influencing results, especially when spot samples versus long-term continuous data were collected. If studies were not blinded, care giving and subsequently observations might have been influenced. The use of different monitors and probes and probe placement may further lead to different results. Studies were small and data inconclusive. There was some agreement regarding abnormally low values being linked to poor outcome. (Dullenkopf et al., 2003; Sorensen et

Variability is the change in percent of rSO2 away from a calculated baseline. It can be followed over time to know how much time the rSO2 was above or below baseline. The baseline differs from patient to patient. Variability is an area of interest and needs further investigation: Cerebral daily variability is small. Large changes (>20%) off the baseline would raise concern for acute clinical change. (McNeill et al., 2010, 2011) Change in variability may be an indicator of infection (Yanowitz et al., 2006). The change in baseline over the first weeks of life, which is observed in preterm infants, may represent ongoing

In the research setting cerebral blood flow and blood volume measurements, oxy- and deoxy hemoglobin and fractional extraction of oxygen (FTOE) as well as blood gas samples

Adequate O2 delivery to the brain tissue is most critical. Assessment of O2 delivery and consumption help understand clinical scenarios and their underlying pathophysiology: At the bed side this evaluation can occur by following changes in cerebral rSO2, changes in BP, oxygenation and peripheral blood gases. The below clinical scenarios for monitoring are

Cerebral autoregulation is a homeostatic phenomenon controlled by the main capacitance vessels in the cerebral circulation. Through dilatation and constriction of these vessels cerebral blood flow and cerebral rSO2 or TOI are maintained at a steady level over a range of changing mean arterial blood pressures (MABP). This range is narrower in neonates, particularly in preterm infants. Cerebral pressure-passivity or loss of autoregulation is associated with low gestational age, low birth weight and systemic hypotension in a large

If rSO2 or TOI changes correlate with the wave form of MABP autoregulation is lost. Swings in peripheral perfusion will be mirrored in cerebral blood flow and regional saturation readings. This phenomenon, when profound, carries an increased risk for intra-ventricular hemorrhage (IVH) and peri-ventricular leucomalacia (PVL) in preterm infants and generally a poor prognosis for neurodevelopment outcome. The more swings or changes in mean arterial pressure (MAP) and NIRS coincide and mirror each other, the more the waves are in concordance. Several studies link concordance with a more unfavorable prognosis and a higher likelihood of death. (Caicedo et al., 2011; DeSmet et al., 2010; Greisen & Borch, 2001;

developmental maturation independent of feeding status. (McNeill et al., 2010, 2011)

**2.3 Peripheral blood pressure and oxygenation, impact on autoregulation** 

from central catheters added to detailed understanding of physiology.

al., 2008; van Bel et al., 2008; Wolf & Greisen, 2009, also see cerebral hypoxia)

2004; Madsen et al., 2000; McNeill et al., 2010, 2011; Wassenaar et al., 2005) NIRS measurements may differ between probes. (Sorensen et al., 2008)

#### **1.4 Safety and feasibility**

Commercially available sensors for neonates have become well tolerated due to smaller size and being lined with a skin friendly adhesive. To provide further skin protection in extremely premature patients probes can be attached to a light-permeable skin barrier without interference with measurements. (McNeill et al., 2010, 2011)

### **1.5 Monitoring**

Organs which can be monitored in neonates are brain, kidney, gut, liver and muscle. This chapter will comment on the most commonly used sites– the brain, kidney and gut.
