Use of Oxygen Cannulas in Extremely Low Birthweight Infants is

Original Article . . . . .

. . . . . . . . .

Use of Oxygen Cannulas in Extremely Low Birthweight Infants is Associated with Mucosal Trauma and Bleeding, and Possibly with Coagulase-negative Staphylococcal Sepsis Arthur E. Kopelman, MD Donald Holbert, PhD

OBJECTIVE: We studied the association between the use of oxygen cannulas (OCs) and (1) nasal bleeding and (2) coagulase-negative staphylococcal sepsis (CNSS).

STUDY DESIGN: Review of care sheets, with w2 or sign-test group comparisons. RESULTS: Infants treated with OCs were suctioned more frequently (2.6 vs 1.3 times per day, po0.001), and had more bloody nasal secretions (34.6% vs 4.6%, po0.05) that increased with increasing OC days. By 10 days, 90% of infants had experienced bloody secretions. CNSS occurred less often in infants treated with oxyhoods than those on OC or CPAP (1 of 13, 8%, vs 10 of 44, 23%), but the difference was not significant. Eight of the 10 CNSS episodes clustered within 3 and 7 days of starting CPAP or cannula treatments. CONCLUSION: OC use in extremely low birthweight infants is associated with nasal mucosal injury and bleeding. Studies are needed to see if use of OCs is a risk factor for CNSS. Journal of Perinatology (2003) 23, 94–97. doi:10.1038/sj.jp.7210865

INTRODUCTION After experience with extremely low birthweight (ELBW) infants treated with nasal oxygen cannulas (OCs), who developed increased nasal secretions, nasal bleeding, and occasionally nasal obstruction, we performed a retrospective chart review to determine

Department of Pediatrics-Neonatology (A.E.K.), Brody School of Medicine at East Carolina University, Greenville, NC, USA; and Department of Biostatistics (D.H.), East Carolina University, School of Allied Health Sciences, Greenville, NC, USA. Address correspondence and reprint requests to Arthur E. Kopelman, MD, Brody School of Medicine at East Carolina University, Pediatrics-Neonatology, 600 Moye Boulevard, Greenville, NC 27858-4354, USA.

how frequently the use of OCs in ELBW infants was associated with injury to the nasal mucosal. After that study showed that use of OCs was often associated with nasal mucosa bleeding, we hypothesized that use of OCs would be a risk factor for the development of coagulase-negative staphylococcal sepsis (CNSS). The nares of premature infants are heavily colonized with coagulase-negative staphylococcus (CNS) after the first week,1 so we hypothesized that erosion of the nasal mucosa (manifested as bleeding) would provide a route for blood-stream invasion by this organism. A second retrospective study was then performed to test that hypothesis. We present here the results of both studies: the first study showing an association between the OC use in ELBW infants and nasal mucosal injury with bleeding, and the second looking for an association between use of OCs and CNSS.

METHODS Study 1: Nasal Mucosal Injury and OCs The purpose of this study was to look for a possible association between the use of OCs in ELBW infants and trauma to the infant’s nares recognized by increased suctioning of nasal secretions and blood in the secretions. As neonatologists in our NICU differ in their preference of OCs, nasal CPAP, or oxygen hoods following extubation of ELBW infants, and as the neonatologists rotate on service monthly, we reviewed data for the first two ELBW infants extubated each month during 1997, a total of 24 infants. For each of the 28 days following the infants’ first extubation, we abstracted data from the nursing flow sheets regarding the number of times per day that nasal suctioning was performed and the number of times blood was noted in the nasal secretions as well as the mode of treatment in use: oxygen cannula, nasal CPAP, or oxyhood. We then compared the number of times per day that nasal suctioning was performed and the number of times per day that bloody secretions were noted for infants being treated with an OC, nasal CPAP, or oxyhood. We used the sign test to test the hypotheses that the rates of nasal suctioning and the appearance of blood in the secretions were greater when infants were treated with OCs or nasal CPAP than when they were under oxyhoods. To determine whether the rates of nasal suctioning and of bloodtinged secretions increased over time, we plotted their rates of occurrence against consecutive days of treatment with OCs, nasal Journal of Perinatology 2003; 23:94–97 r 2003 Nature Publishing Group All rights reserved. 0743-8346/03 $25

94

www.nature.com/jp

Cannulas, Bleeding and Staphylococcal Sepsis in ELBW Infants

CPAP, or neither. This retrospective study was approved by our institutional review board. Clinical Procedures Routine nursery practices were followed. Oxygen was administered through the nasal cannulas as a humidified, unheated blend of oxygen and air at 1 l/min through an infant nasal cannula (Salter Labs). The nasal prong outer diameter was 3 mm tapering down to 2 mm at its tip and the prong length was 8 mm. Nasal CPAP was delivered using a size 0 Infant Nasal CPAP Cannula (Hudson RCI). The CPAP prongs had an outer diameter of 3 mm and a length of 12 mm. Oxygen delivered by CPAP was blended with room air, heated, and humidified, as was oxygen delivered by an oxygen hood. Our NICU did not have a policy for routine nasal suctioning, which was performed with a suction catheter only when the nurse felt it necessary to clear secretions. The nurses routinely recorded nasal suctioning and noted the character of the nasal secretions including the presence of blood on the nursing flow sheets. Study 2: Association Between Use of OCs and the Development of CNSS In this retrospective study, we determined the rate of CNSS following the first extubation for all ELBW infants in our NICU during 1999. The infants were assigned to one of three groups based upon their treatment at extubation, as determined by the neonatologist, with OC, nasal CPAP, or oxyhood. A research nurse who was not aware of the study hypothesis abstracted data from the charts including the occurrence of sepsis. Sepsis was defined as the occurrence of clinical findings consistent with that diagnosis, a positive blood culture, and treatment with antibiotics for at least 7 days. Groups were compared for their rates of CNSS by w2 analysis. As a temporal relation between starting nasal prongs and developing CNSS would support a causal relation, we also recorded the time intervals between extubation, the time when infants were begun on OCs, nasal CPAP, or oxyhoods, and the occurrence of CNSS. This retrospective study was approved by our institutional review board, and clinical practice was unaffected by the study.

RESULTS Study 1: Nasal Mucosal Injury and OCs The 24 infants studied had a mean birthweight of 791 g (range 510 to 998 g) and a mean gestational age of 26.6 weeks (range 24 to 29 weeks). The rates of nasal suctioning, and appearance of blood in the nasal secretions while infants were on OCs, nasal CPAP, or oxyhoods are shown in Table 1. Few days were spent on CPAP, so that despite the apparently higher rates of suctioning and of bloody secretions while on nasal CPAP, statistical analysis was limited to a comparison of days on an OC with days under an oxyhood. Both the frequency of nasal suctioning and the percent of blood-tinged nasal secretions were significantly greater for infants treated with OCs than for those under oxyhoods (2.6 vs 1.3 times per day, Journal of Perinatology 2003; 23:94–97

Kopelman and Holbert

Table 1 Comparison of Frequency of Nares Suctioning and Bloody Secretions during Different Treatments

Days on modality No. of suctions/day Blood-tinged (% of all secretions)

Cannula

CPAP

Oxyhood

438 2.6* 34.6w

17 3.5 22.9

108 1.3* 4.6w

Significant differences occurred between OC and oxyhood: *po0.001, wpo0.05.

Figure 1. Percent of secretions with visible blood plotted against number of consecutive days the infants remained on an OC. Bloodtinged secretions increased by 11% per week, po0.001. Three infants were placed on an OC twice so the 24 infants had 27 courses on OC.

po0.001, and 34.6 vs 4.6%, po0.05, respectively). Blood-tinged secretions only occurred in three infants who were under oxyhoods, and two of them had been treated with OCs within 24 hours before the bloody secretions were observed. The frequency of nasal suctioning did not change with time on a nasal cannula. However, the percent of nasal secretions containing visible blood continued to increase linearly by 11% per week as long as the infants remained on an OC, po0.001. (Figure 1). Almost 90% of infants who were on a nasal cannula for 10 consecutive days had experienced blood-tinged secretions on at least one occasion (Figure 2). In contrast there was no increase in blood-tinged secretions over time for infants treated with oxyhoods. Study 2: Association Between Use of OCs and the Development of CNSS We reviewed the charts of all 57 ELBW infants who were extubated within 28 days of birth during 1999 for episodes of CNSS following extubation. A total of 11 infants were placed directly on nasal cannulas, 33 on nasal CPAP, and 13 under oxyhoods. Table 2 shows that at extubation the groups were comparable for most recognized risk factors for nosocomial infection. Infants treated with OCs were heavier than those treated with CPAP or oxyhoods, 95



Figure 2. Cumulative percent of infants who ever had blood-tinged nasal secretions plotted against number of days on an OC. Table 2 Comparison of ELBW Infants Extubated to Each Modality

Birth weight (g) Gestation age (weeks) Central line (%) Extubation age (median days) CNSS (%)

Cannula (11)

CPAP (33)

Oxyhood (13)

919784* 27.671.9 9 (82) 3 3 (27)

7957104 26.571.6 17 (52) 20* 7 (21)

7937140 26.771.7 8 (62) 6 1 (8)

*Signifies significant difference from other groups, po0.01.

and infants treated with CPAP were extubated at a later postnatal age than infants in the other groups. Analysis was done on the basis of the intent to treat with each of the modalities. The rates of CNSS following extubation were higher for infants on OCs (three of 11, 27%) or nasal CPAP (seven of 33, 21%), than for infants treated with oxyhoods (one of 13, 8%), but the differences did not reach statistical significance. The 10 episodes of CNSS in infants treated with nasal prongs (either OCs or nasal CPAP) occurred between 6 and 24 days of postnatal life. When the occurrence of CNSS was plotted as days after starting on nasal cannulas or CPAP, eight of the 10 CNSS episodes were clustered within a 5-day period, between 3 and 7 days after being started on an OC or CPAP (Figure 3). DISCUSSION We became interested in the association between the use of OCs and nasal mucosal trauma in ELBW infants after observing that some infants treated with OCs had increased nasal secretions or nasal bleeding, and on occasion even developed upper airway obstruction. In the first study, we found that ELBW infants treated with OCs required more frequent nasal suctioning to clear secretions, and were more likely to have visible blood in their nasal secretions than infants treated with oxyhoods. The longer the infants were treated with OCs the more frequently they had bloodtinged secretions, and 90% of those treated with OCs for 10 consecutive days had bloody secretions at least once. 96

It is important to emphasize that the association between the use of nasal OCs and increased nasal secretions and bleeding that we found does not mean that nares injury is an inevitable consequence of OC use. Nasal injury and bleeding was associated with our technique for OC use as described in the Methods section and might have been avoided with a different technique for OC use. We suspect but cannot prove that the flow of unheated humidified gas at 1 l/min through the OCs may have resulted in drying and injury to the infants’ nasal mucosa. It is also possible that too frequent or overly aggressive suctioning of the nares in an attempt to clear nasal secretions contributed to, or even caused, the mucosal injury. This retrospective study did not evaluate different techniques for providing cannula oxygen to ELBW infants or address the question of whether injury to the nasal mucosa can be avoided. From personal experience, working in other NICUs, we are aware that nasal injury including bleeding in infants treated with OCs is not unique to our NICU. This is the first report of an association between the use of OCs and nasal mucosal injury, but use of nasal CPAP has been associated with injury to the nares of premature infants.2–4 We also observed an increase in nasal suctioning and blood-tinged secretions in infants treated with nasal CPAP, but too few infants were treated with CPAP to include them in the statistical analysis. Since the nares of ELBW infants are smaller and their tissues are more delicate, they are likely more vulnerable to nasal mucosal trauma and erosion from use of OCs than larger preterm or term infants. The small diameter of their nares also makes ELBW infants especially vulnerable to nares obstruction from mucosal edema, secretions, or blood. Nasal resistance can be assumed to be higher in ELBW infants because of the smaller cross-sectional area of their nares as, with laminar gas flow, resistance varies inversely with the radius of the airway raised to the fourth power. Even partial nares obstruction would greatly increase the work of breathing in these infants as nasal resistance accounts for up to 50% of total lung resistance in newborn infants.5 Injury to the nasal mucosa is also more likely to cause airway obstruction in premature infants because they are obligate nose breathers.5,6 Many of the infants treated with OCs developed blood-tinged nasal secretions, and since the ruptured blood vessels course through the submucosa, this indicates that they had experienced mucosal erosions. The second study tested the hypothesis that use of OCs in ELBW infants would be associated with an increased risk of CNSS. This possibility of such an association occurred to us because the nares of premature infants are heavily colonized with CNS after the first week1 and mucosal erosion might provide a portal of entry into the circulation. This possible route of infection was not considered in a previous study that found a four-fold difference in CNSS rates between NICUs, which was not explained by known clinical risk factors.7 While the rates of CNSS were higher for infants treated with OCs (27%) or nasal CPAP (21%) than for infants nursed in oxygen hoods (8%), the differences did not reach statistical significance, Journal of Perinatology 2003; 23:94–97



thoracoabdominal asynchrony, oxygen requirement, and probably increases work of breathing.8 Our study showing that the use of high-flow OCs is associated with increased nasal secretions and bleeding which would likely increase the work of breathing and may prove to be a risk factor for the development of CNSS provides an additional reason to question the routine use of OCs in ELBW infants. OCs largely replaced oxyhoods in NICUs because they facilitated nursing care. As OCs were not considered invasive, they were widely used without evaluation of potential risks and benefits. We suggest that such an evaluation of this long-standing practice is appropriate.

Figure 3. CNSS episodes following extubation plotted against postnatal days (above) and postextubation days (below).

possibly because of the small number of infants studied. The infants in the three groups were comparable for most risk factors for CNSS such as gestational age and presence of a central line at the time of extubation. The infants treated with oxyhoods weighed significantly less than those treated with OCs, and as nosocomial sepsis correlates strongly with lower birth weight one would have anticipated that they would have more episodes of CNSS. Instead they had fewer CNSS episodes, consistent with our hypothesis. The observation that eight of their 10 episodes of CNSS were clustered between 3 and 7 days after nasal prongs (cannulas or CPAP) were started is also consistent with our hypothesis that nasal mucosal injury from OC use is a risk factor for the development of CNSS. We believe these preliminary findings are sufficiently suggestive to justify a prospective trial to test the hypothesis. While high-flow cannula oxygen has been used in preterm infants for many years, and has been suggested as a means to provide CPAP, the practice has recently been questioned as it results in an increase in respiratory rate, minute ventilation,

Journal of Perinatology 2003; 23:94–97

References 1. D’Angio CTR, McGowan KL, Baumgart S, St Geme J, Harris MC. Surface colonization with coagulase-negative staphylococci in premature neonates. J Pediatr 1989;114( 6 ):1029–34. 2. Ahmuda CA, Goldsmith JP. Continuous distending pressure. In: Goldsmith JP, Karotkin EH, editors. Assisted Ventilation of the Neonate, 3rd ed. Philadelphia, PA: Saunders; 1996. 3. Sreenan C, Lemke RP, Hudson-Mason A. High-flow nasal cannula in the management of apnea of prematurity: A comparison with conventional positive airway pressure. Pediatrics 2001;107:1081–3. 4. Vermont Oxford Network Second Annual Quality Congress for Neonatology, Washington, DC, December 9, 2001, poster # 23. 5. Carlo WA, Martin RJ, Bruce EN, Strohl KP, Fanaroff AA. Aloe nasi activation (nasal flaring) decreases nasal resistance in preterm infants. Pediatrics 1983;72( 3 ):338–43. 6. Polgar G, Kong GP. The nasal resistance of newborn infants. J Pediatr 1965;67( 4 ):557–67. 7. Brodie SB, Sands KE, Gray JE, et al. Occurrence of nosocomial bloodstream infections in six neonatal intensive care units. Pediatr Infect Dis J 2000;19( 1 ):56–65. 8. Courtney SE, Pyon KH, Saslow JG. Lung recruitment and breathing pattern during variable versus continuous flow nasal continuous positive airway pressure in premature infants: An evaluation of three devices. Pediatrics 2001;107:304–8.

97