Management of oxygen saturation monitoring in preterm newborns in the NICU: the Italian picture

Background Although many studies emphasize the importance of using oxygen saturation (SpO2) targets in the NICUs, there is a wide variability in used saturation ranges among centers. Primary aim was to draw a representative picture on how the management of oxygen monitoring is performed in the Italian NICUs. Second aim was to identify healthcare-professionals related factors associated with oxygen targeting in the preterm population. Methods Cross-sectional study with data collection via an electronic survey form. A questionnaire containing pre-piloted and open questions on monitoring and management of the SpO2 was administered to neonatologists across the network of the Italian Society of Neonatology. The questions focused on: the infrastructure, specific training, healthcare professionals and patients-related factors. The results of the survey were anonymously collected, summarized and analyzed. Results Out of 378 questionnaires, 93 were correctly filled. Thirty-six different SpO2 ranges were observed. Centers using written standard operating procedures on oxygen management and SpO2 monitoring maintained a correct average range of SpO2 90–95%, avoided hyperoxia and reconsidered saturation targets in relation to comorbidities. 39.8% of responders disabled alarms during neonatal care. One center used biomarkers for complete monitoring of neonatal oxygenation status. Conclusions There is considerable variation in SpO2 targets for preterm infants in the Italian NICUs. Standard operating procedures and specific training for health care personnel are the main factors playing a role for the correct maintenance of the recommended oxygen targets in preterms.


Background
Oxygen is essential for aerobic life, but it can be considered a double-edged sword in the perinatal period having both positive biological benefits and toxic effects. Oxygen toxicity is due to the development of reactive oxygen species (ROS), potent oxidants in biological fluids that may damage tissues, through reaction with lipids, proteins, DNA, amino acids and several other molecules [1]. An imbalance between oxidants and antioxidants is called oxidative stress: a potential cause of cell damage [2]. Newborns, especially if preterm, are particularly susceptible to oxidative stress due to the immature antioxidant capacity and the likely exposure to many processes such as hypoxia, hypoxia-ischemia, hyperoxia and infections, leading to high levels of free radicals' production [3]. Hypoxia and hyperoxia predispose preterm newborns to oxidative stress, through the free radical generation. Antioxidant capacity is lower in the newborn and particularly in the premature infant in comparison to term newborn [2,3]. Therefore, premature infants are especially prone to oxidant injury, in various organs and systems such as lungs, retina and erythrocytes [4][5][6] with short and long term effects. The careful monitoring of oxygen saturation (SpO 2 ) levels during neonatal intensive care unit (NICU) admission is of outmost importance, in order to avoid excessive and undesired exposure to hypoxia/hyperoxia [7,8]. Clinical studies underlined how the level of generally accepted saturation targets for children and adults could not be considered acceptable for preterm and low birth weight infants [9,10]. Recently, multicenter randomized controlled trials support the recommendation to keep SpO 2 between 90 and 95%, in infants with a gestational age less than 28 weeks, up to 36 weeks post-menstrual age [11][12][13]. However, wide variability in SpO 2 ranges has been reported and there is no consensus yet on the specific prerequisites for the management of SpO 2 monitoring in the NICUs [14,15]. There is current evidence on the presence of the condition of "alarm fatigue" which is the desensitization of health care practitioners to the thousands of alarms of a single day in the NICU [16]. Fatigue is related to the many false alarms generated by pulse oximeters; each alarm is supposed to be associated with active nursing intervention with a frequency of around 5-10 min [17,18]. Nurses have to face this dilemma, and previous papers already showed that they tend to disregard alarm policy, due to the high number of false alarms [19]. More attention is needed in selecting reasonable alarms' ranges and levels, while the risk of missing or delaying response to important events must be also considered [20]. Moreover, there is an increasing tendency to disregard the high alarm limit, with the assumption that hypoxemia is more detrimental than hyperoxemia. A recent survey on pulse oximeter saturation target limits for preterm infants in European NICUs pointed out the present climate of uncertainty regarding the optimum range of pulse oximeter SpO 2 for preterm infants, reporting wide institutional variations on SpO 2 targets [21]. The aim of this study is to characterize how SpO 2 is monitored in preterm infants in the Italian NICUs. A further aim is to identify healthcare-professionals related factors associated with different oxygen targeting in preterm newborns.

Materials and methods
This is a national, multi-center, cross-sectional study, implemented by the scientific working group of "Clinical neonatal Biochemistry" of the Italian Society of Neonatology.
Data collection was performed using an electronic survey (eCRF) sent to all the chairs of italian NICUs. Contacts were obtained from the registry of the members of the Italian Society of Neonatology. The questionnaire was also sent to all the neonatologists working in these units.
Data collection was performed using Survey Monkey software (SurveyMonkey Inc. San Mateo, California, USA). The first survey was sent in December 2017. A reminder with a new link was sent in February 2018. Questionnaires were administered and collected anonymously. In order to check reliability and consistency of the answers, redundant questions were included in the survey (i.e. question number 18, q 2 in training related variables, q 9 in staff variables etc.). Exclusion criteria were: an incomplete survey for more than 50%, inconsistent answers to redundant questions more than two times in the whole survey.
The eCRF consisted of two parts (Table 1). Part one contained general information about the hospital, the NICU, the patient-nurse ratio and implemented treatment strategies (biomarkers measurement, standard operating procedures, guidelines). Part two enquired specific information about the management of SpO 2 monitoring currently used in the unit. The following variables were also assessed through the eCRF: infrastructure variables, training, technology variables, staff variables and patient-related variables. For each section multiple choice as well as open questions were included. The respondent neonatologists were instructed, via an introductory email, to provide answers reflecting their unit practice, based on local protocols/standard care, and not personal preferences. A reminder questionnaire was sent once to the nonresponding neonatologists.

Statistical analysis
Data analysis was performed using IBM SPSS Statistics (IBM Corp., Armonk, NY, USA). Normal distribution of data was evaluated by Kolmogorov Smirnov test. The two-proportions z-test was used to compare two observed proportions and the two-sample t-test to test whether the means from the two populations were equal or not. All tests were conducted two-sided in an explorative manner on a 5% significance level. Logistic regression model was performed to identify factors independently associated with the use of oxygen saturation monitoring modalities. Based on these statistical models, odds ratio estimates (OR) were calculated with 95% confidence intervals (CI).

Results
Out of a total of 378 questionnaires sent, 104 replies were received. Of these only 93 were complete and showed consistent answers and therefore assessable for   our study (24.6% of the questionnaires sent). The characteristics of participant centers are described in Table 2. Over the total 93 different NICUs, 36 different SPO 2 ranges were reported, with wide variability between the minimum and maximum target levels. The most frequently used range of SpO 2 was 90-95% (16 centers), 88-95% (8 centers) and 88-94% (seven centers) (Fig. 1). The range of maximum SpO 2 used levels varied from 100% (four centers) to 92% (10 centers) (Fig. 2). The range of minimum SpO 2 levels varied from 75% (two centers) to 93% (one center) (Fig. 3). In 64.4% of the centers the lower limit of SpO 2 is set below 90%; in contrast, in 24.4% of the centers the upper limit of SpO 2 is set above 95%. NICUs using written standard operating procedures or specific local guideline on oxygen management and SpO 2 monitoring, maintained average maximum desired levels of SpO 2 of 94% + 2 versus 96% + 2 of centers that did not have a written local protocol (p = 0,003). However, there were no statistically significant differences with regards to the mean values set for the desired minimum SpO 2 (88% ± 2 versus 88% ± 3, respectively; p = 0,143). Similarly, centers performing specific training of health care personnel on management of oxygen monitoring, would set the desired maximum values of SpO 2 on average at 94 compared to the average values of 95% of the centers not performing any training (p = 0.037). There were no statistically significant differences in the mean values set for the desired minimum SpO 2 (minimum SpO 2 = 88% ± 2 versus 88% ± 3, respectively; p = 0,662). SpO 2 desired range could be changed in 56% of participating NICU in case of associated comorbidity (e.g. anemia, cardiopathy, ROP, BPD, surgery, sepsis), in 13.2% of cases also if the ventilatory support mode was changed, in 12.1% of cases in other conditions, such as in case of both comorbidity and variation of the ventilatory support mode (seven centers, 70%), in case of no oxygen supplementation (three centers, 30%). One center modified the maximum limits, by increasing it, in case of corrected gestational age > 32 weeks. In 18.7% of the centers, SpO 2 limits were never changed. SpO 2 alarms were never turned off in 63.4% of participating centers (n = 59). They were disabled during care maneuvers in 16% (n = 15), during invasive procedures in 11.8% (n = 11), excessive instability of the patient in 4.3% (n = 4), other reasons in 4.3% of the centers (one center in the assistance of terminal patient to death, two centers in case of no supplemental oxygen, one center in case of both invasive procedures and care maneuvers) (Fig. 4). Centers that performed staff training on the monitoring of the SpO 2 turned off the alarms during the assistance maneuvers less frequently compared to those with the opposite attitude (23% versus 76%, respectively; OR (95% CI) =0,368 (0,001-0,269); p = 0.047). Twenty-six centers (five of second level and 21 of third level) had NIRS technology available to study cerebral SpO 2 . Four third level centers had the possibility to dose VEGF. Thirteen centers (three of second level and ten of third level) measured biomarkers of oxidative stress. In a single third level and university center all the three adjunctive technologies were available for the complete study of the oxygenation status of the newborn.

Discussion
To date there is no standardized SpO 2 interval, universally recognized as excellent by the scientific community, to guarantee tissue oxygenation in the NICU setting. Multicenter, randomized and large-scale clinical trials have shown SpO 2 below 90%, during intensive care stay, are associated with a higher mortality, and saturation values higher than 95% are associated with increased morbidity [11,12,22]. The present study allowed to draw a representative picture of how the management of oxygen monitoring is performed in 93 neonatal intensive care units in Italy and which factors independently favored/refrained the use of the recommended targets.
The study revealed a wide variability in the utilized ranges for the surveillance of critically ill patients. Failure to comply with the upper and lower limits inevitably exposes the baby to either hypoxia or hyperoxia. Our study showed that in 64.4% of the centers the lower limit of SpO 2 is set below 90%, highlighting a propensity to tolerate hypoxia; in contrast, in 24.4% of the centers the upper limit of SpO 2 is set above 95% reflecting a permissive attitude towards hyperoxia. In newborns hyperoxia leads to persistent inflammation with impaired innate immune response and increased airway reactivity and susceptibility to respiratory virus infections in adulthood [23]. To limits the dangerous effects of hypoxia or hyperoxia a strict control of oxygen administration is mandatory.
A recent paper investigating the signal type of alarms in the NICU, reported that over 60% of alarms were related to oxygenation monitoring; thus, these represent the major burden of all alarms in newborns population [24]. In this survey, the setting of the maximum value is influenced by the presence or absence of both written operational procedures and staff training, with attention to hyperoxia conditions. NICUs that are particularly sensitive to staff training on the monitoring of SpO 2 , choose more frequently not to disable the alarms during care maneuvers, compared to the centers with opposite attitude. In this way the risk of missing clinically relevant alarms is reduced. So, it is clear how scientific updating and the active involvement of the trained staff in guiding daily clinical decisions play a decisive role in the quality of patient care. In this context, the so called "alarm fatigue" or alarm desensitization is also a key factor. In the NICU environment, the number of alarm signals may reach several hundred per day determining a huge alarm burden with the effect of staff desensitization, disabling of alarm signals and missing upper alarms [25]. The high rates of false or nonactionable alarms may also be involved. In most cases, nurses/neonatologists can adjust the SpO 2 alarm limits and that is reasonable to individualize care to specific patients. However, the presence of local standard protocols may be desirable to avoid individual operator changes to alarm settings. The optimal SpO 2 range for the newborn is also a dynamic value that can change (minimum, maximum or both) in relation to comorbidity, the inspired fraction of oxygen, the gestational age of the newborn and the corrected age. In 18.7% of the centers, saturation limits were never changed. This data challenge current recommendation to increase the saturation target to> 95% in those newborns who, at 32 weeks of correct gestational age, still need oxygen supplementation or to reconsider saturation targets in babies with pulmonary hypertension [26,27]. All these aspects are fundamental not only to guarantee optimal neonatal care but also for later follow-up assessment. The use of specific local protocols and ad hoc personnel training were independently associated with the use of recommended targets limits, indicating the use of these tools as a good local guideline for better neonatal care. A recent European survey, performed only one year before the one presented in this paper, reported a wide variability in ranges similar to our data [21]. Consistently with their findings, the present survey also highlighted a lack of consensus regarding SpO 2 target limits for preterm infants. However, they did not take into account the nurses/infants ratio and the "alarm fatigue" and it was not clear if there was a tendency to tolerate hypoxia or a permissive attitude towards hyperoxia, as suggested by the present findings. Moreover, we could not draw conclusions regarding the changes in the neonatologists' view on SpO 2 monitoring since we did not take into account in our survey when the written internal protocol was eventually introduced.

Limitations
The present study has some limitations. First, the questionnaire gives an instant picture of only 25% of the Italian centers. Unfortunately, we were only allowed to send a single reminder to the invited responders. This means that the present study gives only a partial view on the actual situation of SpO 2 targets in the Italian NICUs. Furthermore, the absence of more detailed background information on the specific SpO 2 monitoring protocols used in each individual institution, and above all the lacking information of clinical outcome limits further conclusions. Also, the absence of a standard local protocol cannot be automatically equated with the absence of knowledge on current guidelines or lower quality of patient care. However, the presence of ad hoc protocols was identified as one of the independent factors associated with the use of recommended SpO 2 targets.

Conclusions
The study reveals that SpO 2 monitoring, although available and performed in all units, still lacks specific local ad hoc protocols which need to be implemented for a correct surveillance of critically ill patients. The use of local protocols, and specific personnel training would possibly allow the wider use of recommended targets limits. These data provide an important overview on the current situation on Italian NICUs on SpO 2 monitoring and management.