| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Division of Neonatology, Department of Pediatrics, University of Miami Miller School of Medicine, Miami, Florida.
Address correspondence and reprints requests to Nelson Claure, Division of Neonatology, Department of Pediatrics, University of Miami School of Medicine, P.O. Box 016960 R-131, Miami, FL 33101. Address e-mail to NClaure{at}miami.edu.
Abstract
Premature infants are at an increased risk of ophthalmic, neurologic, and respiratory sequelae related to inadequate maintenance of oxygenation and exposure to increased levels of inspired oxygen. Management of inspired oxygen is complicated in this population by an increased variability in oxygenation. Automated regulation of the fraction of inspired oxygen is a technology that has a potential of improving such outcomes as well as impacting personnel workload. This is a review of current experimental evidence on the effectiveness of automated regulation of inspired oxygen and its effects on oxygenation variability and personnel workload during the care of premature infants.
Supplemental oxygen is given to otherwise hypoxemic premature infants to maintain adequate oxygenation and to avoid the deleterious effects of hypoxia. However, delivery of supplemental oxygen is not a risk-free therapy. Because of their premature birth, these infants are often exposed to oxygen at a time when their lungs, antioxidant system, and retina are too immature. Also, these infants often require supplemental oxygen for several days or weeks after birth. Therefore, they are at an increased risk for oxidative stress, lung injury, and retinopathy of prematurity (1–6). Management of the supplemental oxygen aims at minimizing exposure to increased levels of inspired oxygen while maintaining adequate oxygenation.
OXYGENATION VARIABILITY IN PREMATURE INFANTS
Under routine clinical conditions, management of the fraction of inspired oxygen (Fio2) is often complicated by fluctuations in oxygenation. These fluctuations vary in severity and frequency. Fluctuations associated with worsening or improvement of a respiratory condition occur gradually, and are relatively simple to correct while other fluctuations can occur quite frequently and are often characterized by rapid and severe changes in oxygenation (7–11).
Compared to other patient populations in intensive care, premature infants are at increased risk of developing complications related to fluctuations in oxygenation, which are believed to influence the development of retinopathy of prematurity (12–17). Intermittent hypoxemia can also have a negative impact on airways, lung vasculature, and other organs (18–21) while exposure to high concentrations of inspired oxygen can lead to development of permanent respiratory sequelae (1–4), which is associated with impaired neurologic outcome (22,23).
These fluctuations in oxygenation are detected by arterial oxygen saturation monitors (Spo2) and trigger a personnel response. The response to the occurrence of hypoxemia usually consists of a transient increase in Fio2 until normoxemia is restored. In the event of hyperoxemia, Fio2 is gradually reduced. Under standard clinical conditions, personnel-to-patient ratio and increasing amounts of caregiver workload often do not permit full attention to these tasks. As a result, the response time when Spo2 is outside an intended range of oxygenation is affected. Delayed responses can prolong exposure to unnecessarily high concentrations of supplemental oxygen or periods of hypoxemia.
Recent, multicenter data has shown that premature infants who require supplemental oxygen spend only about half of the time within individual center’s intended range of oxygenation. Of the remaining time, about 20% is spent with Spo2 readings below and about 30% above the intended range (24,25). Spo2 readings in hyperoxemia in infants receiving supplemental oxygen could only be attained by providing unnecessarily high concentrations of inspired oxygen.
RATIONALE FOR AUTOMATED REGULATION INSPIRED OXYGEN
Automated regulation of the inspired oxygen in premature infants is intended to address the above-mentioned issues by reducing periods of hyper- and hypoxemia, and limiting the exposure to high concentrations of inspired oxygen. The rationale that supports the development of automatic Fio2 control is based on the timely Fio2 adjustment to increase the alveolar oxygen concentration, followed by prompt weaning as soon as the additional oxygen is no longer needed.
These tasks could also be achieved with manual adjustments by clinical personnel. However, they can become time-consuming and increase workload.
EXPERIENCE WITH AUTOMATED Fio2 REGULATION: OXYGENATION STABILITY AND WORKLOAD
Various studies have shown the feasibility of automated Fio2 regulation for the care of premature infants. Moreover, automated Fio2 regulation shows improvement in oxygenation stability around intended ranges of oxygenation compared to routine care and even when compared to nursing care dedicated to oxygen regulation (26–32).
When assessing the comparisons between automated and manual Fio2 regulation in regard to oxygenation stability and frequency of manual interventions, it is important to consider the basal rate of variability of the studied population and personnel responsiveness. Comparisons of automated Fio2 regulation to manual interventions are not likely to benefit infants who present with minimal or no variability in oxygenation. On the other hand, comparing automated Fio2 regulation to routine care that is relatively unresponsive to fluctuations in oxygenation is likely to show greater benefits in comparison to more dedicated or attentive care. Also, depending on predetermined objectives for the automated system, not all such systems may be designed to address the different forms of oxygenation variability.
Clinically, the most important goal of automated Fio2 regulation is to stabilize oxygenation. Automated Fio2 regulation prolonged the time within an intended range of oxygenation compared to routine care consisting of adjustments at two hourly intervals (26,27,29) or manual adjustments every 20–30 min, even when compared to a more dedicated care with adjustments every 2–3 min (28). Moreover, automated Fio2 prolonged the time within an intended range of oxygenation even when compared to fully dedicated care (29,32). These data illustrate the efficacy of this type of automation.
Infants with very frequent and/or acute fluctuations in oxygenation increase the demand for personnel time and also are a greater challenge for automated Fio2 regulation. Not all systems described in the literature have been developed to address these fluctuations. In some of these, the response of the automated system consists of a user alert (26) or, in others require manual intervention because of inadequate system response (27,30). An automated Fio2 controller designed to assist during acute episodes of hypoxemia increased the time within an intended range in comparison to the time spent by a fully dedicated nurse at bedside in a group of preterm infants who presented with an average of 15 episodes of hypoxemia per hour (32).
Under these challenging conditions, quantification of the effort required for tight control showed that the nurse adjusted Fio2 a mean of 29 times an hour. This frequency does not fully indicate workload since continuous dedicated attention is needed at the onset, throughout, and at the end of the episode of hypoxemia. Reducing the workload resulting from these repetitive tasks may lead to a more effective use of personnel.
ASSESSING OXYGENATION
Perhaps, the most important component in the routine or automated process of Fio2 regulation is continuous assessment of oxygenation. This can be done using indwelling Pao2 electrodes inserted through an umbilical artery catheter, transcutaneous PO2 electrodes (tcPO2) or arterial oxygen hemoglobin saturation measurements by pulse oximetry (Spo2). Indwelling measurements are limited by availability of an invasive line. tcPo2 is noninvasive, but depends on electrode temperature for accuracy. The continuity of tcPo2 measurements is therefore affected by changes in application site to avoid thermal injury. Spo2 provides a continuous noninvasive assessment of oxygenation. Spo2 setup is relatively simple and is widely used in newborn intensive care units.
PULSE OXIMETRY TO DETECT HYPEROXEMIA AND HYPOXEMIA
Interpretation of Spo2 to detect and respond to hypo- or hyperoxemia is done in the context of the relationship between Spo2 and Pao2, as depicted by the sigmoid-shaped O2 dissociation curve. In the range of hyperoxemia, relatively small changes in Spo2 are associated with large changes in Pao2. Data (33) show that a threshold Spo2 around 94% and 96% would correctly classify as hyperoxemic most Pao2 readings above 80 or 90 mm Hg. Most false-positive readings would be in the high–normal range (>60 mm Hg) and none in the low (<40 mm Hg) or low–normal Pao2 range (40–60 mm Hg). Therefore, a gradual reduction in Fio2 in response to high Spo2 will correct most hyperoxemic events and is unlikely to lead to hypoxemia.
In the range of hypoxemia, most readings below a Spo2 threshold around 85% and 88% are associated with Pao2 readings in hypoxemia (<40 mm Hg), a small fraction with readings in the low–normal range (40–60 mm Hg) and none in the high Pao2 range (>80 mm Hg). Therefore, a Fio2 increase in response to hypoxemia detected by Spo2 will be appropriate in most instances and unlikely to result in hyperoxemia.
ARTIFACT OR TRUE HYPOXEMIA?
The reliability of Spo2 in premature infants is especially sensitive to motion because of the relatively low pulse pressure. Motion can produce venous blood pulsation and/or disrupt the optical pathway from the transmitter to receiver side of the probe. Despite documented improvements in Spo2 technology, some hypoxemic episodes detected by pulse oximetry are considered artifactual. Visual inspection of the infant may reveal significant movement of extremities, absence of conclusive signs of cyanosis, and a transient mismatch between oximeter pulse rate and heart rate monitors. However, these observations cannot exclude true hypoxemia. To the contrary, increased infant activity and changes in heart rate, lung volume, and ventilation have been associated with hypoxemia episodes (9,10). Moreover, hypoxemia is more prevalent during arousal and indeterminate sleep states compared to periods of active or quiet sleep (11).
The response to an episode of hypoxemia and simultaneous motion, either by an automated system or by a caregiver, involves the risk of unnecessary oxygen exposure when the episode is artifactual. While the value of avoiding exposure to unnecessary supplemental oxygen is unquestionable, failure to assist a true hypoxemic episode because suspicion of artifact may allow alveolar and tissue hypoxia to increase.
Accuracy of Spo2 is influenced by intrinsic patient conditions, such as low perfusion or inadequate setup of the oximeter (34). These conditions, however, are particular not only to automated regulation of Fio2 but also to routine care. It is ultimately the clinician who decides if the use of pulse oximetry is appropriate and sufficiently reliable to monitor an infant.
When used in infants with frequent episodes of hypoxemia, automated Fio2 regulation during every episode resulted in negligible overshoot with <3% of hypoxemia episodes followed by readings in hyperoxemia (32). If most of these episodes of hypoxemia had been artifactual, the additional oxygen would have had a higher rate of overshoot.
SIGNAL DROP OUT: MISSING Spo2 INFORMATION
Poor signal quality can lead to periods of missing Spo2 information, as determined by the built-in oximeter validation algorithms. This leaves clinical personnel or an automatic system without feedback information. As mentioned above, hypoxemia can be associated with body movement; therefore, there is the possibility that a period of missing Spo2 information caused movement is accompanied by undetected hypoxemia. On the other hand, if the missing Spo2 period is caused by technical factors, it could result in a period of unnecessary oxygen exposure. This, however, may resolve as soon as the clinician responds to the appropriate alarms and warnings given by the pulse oximeter and/or the automatic system. Data obtained during clinical use of automated Fio2 control indicated that most periods of missing Spo2 data were followed by periods of hypoxemia (32).
AUTOMATED SYSTEM RESPONSE
The method of automated Fio2 regulation should modulate its response to severity, duration, rate of change and direction of the fluctuation in oxygenation. Optimization of the timing between Spo2 events and Fio2 adjustments is important when the objective is to assist rapidly ensuing and frequent hypoxemia episodes, while requirements are less for assisting infants who present with mild or gradual changes in oxygenation. A long delay may result in prolongation of hypoxemia or exposure to higher concentration of inspired oxygen when it is no longer needed.
Spo2 data are averaged over a running window to reduce the effect of short variations. A long averaging window can slow detection of changes in oxygenation, and can increase the inertia of automated control leading to overshoot. A shorter averaging window enables the automated system to determine its own response time based on the variables mentioned above.
The mode of delivery should be optimized to produce changes in the inspired gas with minimal delays in setting adjustments at the controller. Optimized gas mixing can be accomplished by increased circulating flow rates in the ventilator and minimizing reservoir spaces.
RISKS AND BENEFITS
Failure or suboptimal function of the measurement, delivery or control components of automated Fio2 control may involve risks of exposure to hypoxemia or unnecessarily high concentrations of inspired oxygen. The use of this technology may result in only a relative increase of these risks, which may be, in part, present during routine care because with the exception of the manual regulation component, the measurement and delivery components are standard.
Monitoring and assistance during routine care may be sufficient for infants with mild or gradual fluctuations in oxygenation or those whose disease condition requires high Fio2 continuously. On the other hand, infants who present with frequent and acute changes in oxygenation may obtain a greater benefit from automated assistance. Dedicated assistance to these infants imposes additional workload on personnel.
Automated Fio2 regulation is aimed mainly at conducting repetitive and time consuming tasks. It is meant to be used in a manner in which the time dedicated by the caregiver to patient care is not affected by these tasks. However, automated Fio2 regulation may in some instances result in an unwanted reduction in the time dedicated by the caregiver to patient monitoring. Since hypoxemia episodes in preterm infants have a wide etiology, automated Fio2 regulation could mask the effects of a triggering event such as hypoventilation. Proper and timely monitoring of cardiorespiratory variables should be assured to avert this type of situation. On the other hand, one could argue that an automated Fio2 increase should at least reduce the detrimental effects until corrective actions are taken.
Assistance for every episode of hypoxemia with an increase in Fio2 could result in additional exposure to oxygen. Timely weaning when hypoxemia resolves should limit the exposure to oxygen. As mentioned above, overshoot (hyperoxemia) was observed in <3% of hypoxemia episodes corrected with an automated Fio2 increase. Nonetheless, compared to no intervention in the form of supplemental oxygen, the lungs would be exposed to higher concentrations of oxygen and therefore an increased risk of lung injury. A predetermined absence of intervention during hypoxemia episodes assumes that the effects of these episodes on the central nervous system and other organs are irrelevant to outcome.
SUMMARY
Supplemental oxygen administration, although a necessary and often life-saving therapy, has side effects that are of relevance in the premature infant population. This population has a high risk for respiratory, neurologic, and ophthalmic compromise. Therefore, there is a need for gentler, yet effective, oxygenation support.
The emergent availability of improved technology for oxygenation monitoring and computing has facilitated the development of automated modes of regulation of supplemental oxygen for the preterm infant. Although this technology has been only used experimentally, preliminary results suggest future impact on patient care and practice. This, however, remains to be tested. Future research will determine the role of this strategy in improving pulmonary, ophthalmic, and neurological outcome in premature infants, as well as in determining the true impact on caregiver workload.
Footnotes
Accepted for publication April 19, 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
