Abstract
BACKGROUND: In mechanically ventilated subjects, intra-tracheal secretions can be aspirated with either open suction systems (OSS) or closed suction systems (CSS). In contrast to CSS, conventional OSS require temporarily disconnecting the patient from the ventilator, which briefly diminishes PEEP and oxygen supply. On the other hand, CSS are more expensive and less effective at aspirating secretions. Thus, it was hypothesized that the 2 procedures differentially affect pulmonary and cardiovascular parameters after suction.
METHODS: Subjects in the ICU (N = 66) were quasi-randomized for initial treatment with OSS or CSS in a crossover design. To compare the potential for these suction systems to compromise cardiorespiratory stability, changes in cardiopulmonary physiology were assessed from before to just after use of each suction system (three 10-s aspirations).
RESULTS: For most pulmonary and cardiovascular parameters (ie, peak inspiratory pressure, airway resistance, pressure plateau, heart rate, and arterial pressures), the effects of aspiration inversely correlated with baseline values for that parameter, with a similar regression slope between suction systems. However, when controlling for baseline values, OSS caused significantly greater increases in airway resistance and peak inspiratory pressure (P < .001 and < .01 vs CSS, respectively).
CONCLUSIONS: Elevated airway resistance prior to endotracheal suction may justify use of a CSS and contraindicate a conventional OSS in mechanically ventilated subjects. Adoption of this approach into clinical guidelines may prevent suction-induced pulmonary injury in subjects, especially for those with underlying diseases involving increased airway resistance or increased alveolar pressure. (ClinicalTrials.gov registration: NCT03256214.)
- endotracheal aspiration
- airway resistance
- pulmonary compliance
- mechanical ventilation
- pulmonary pressure
- lung
Introduction
Mechanical ventilation is used to rescue patients from respiratory failure, maintain oxygen and carbon dioxide, reduce intracranial pressure, and prevent or reduce atelectasis and hypoxemia. However, mechanical ventilation often requires intubation or tracheostomy (or both) as well as sedation,1,2 which promote the accumulation of secretions within airways. Consequently, endotracheal suction systems are essential for routinely aspirating airway secretions to prevent bronchial obstruction3,4 and its attendant complications, including alveolar hypoventilation, hypertension, increasing the stretch of the alveolar wall, emphysematous bubbles, atelectasis, short-term arterial desaturation, and alterations in cardiopulmonary measures such as blood pressure and heart rate.5–9
Endotracheal suction also bears several risks, including hypoxemia, arrhythmia, nosocomial pneumonia, lung hypoinflation, bronchospasm, increased intracranial pressure, and pneumothorax.10–13 Suction, as a short-term effect, can reduce lung compliance3,14 and intrapulmonary pressure, leading to a decrease in oxygen saturation and retention of carbon dioxide.15,16 Moreover, endotracheal aspiration may directly stimulate the trachea or indirectly lead to lung hyperinflation, which can rapidly alter heart rate and blood pressure through autonomic reflexes.16–18 Thus, numerous end points of cardiopulmonary function are used to monitor patient stability during mechanical ventilation and endotracheal aspiration. Arterial blood gas analysis provides measures of and to guide prevention or reversal of acidosis/alkalosis, hypoxemia, central nervous system injury, and gastrointestinal and renal ischemic injury.3,17–20 Indices of lung mechanics, oxygen saturation, and cardiovascular hemodynamics are also routinely evaluated to monitor underlying disease and cardiopulmonary stability while optimizing adjustments in mechanical ventilation to prevent barotrauma, volutrauma, and hypoxemia.2,17,21
In mechanical ventilation, endotracheal aspiration can be performed with open suction systems (OSS) or closed suction systems (CSS). Conventional OSS require disconnection of patients from the ventilator to allow catheter insertion. This diminishes PEEP and oxygen supply and may lead to changes in cardiopulmonary function.3,5,16,22 In contrast, CSS do not require disconnection and hence allow maintenance of PEEP and oxygen supply during suction and improving blood oxygenation.3,5,8,14,22,23 Because CSS use a catheter connected between the endotracheal tube and ventilator circuit in a constantly sterile route of entry, some have speculated that such systems protect the patient from contamination relative to OSS.19,22,24 In contrast with prior negative findings,8,25,26 a recent meta-analysis associated CSS with a lower risk for ventilator-associated pneumonia.27 Others have noted that CSS are easier and faster to use and cause fewer physiologic disturbances.27,28 Indeed, CSS can modestly attenuate aspiration-induced increases in heart rate and arterial pressure and declines in oxygen saturation relative to OSS.8,16,26,28 Further, a systematic review reported that CSS maintain a higher end-expiratory lung volume than OSS,29 but this assessment included only 2 studies and 19 total subjects.5,30 However, most comparisons have neglected fundamental parameters of pulmonary physiology (eg, resistance and peak inspiratory pressure [PIP]), used small sample sizes, and, did not address the effectiveness of suctioning on sputum volume or sputum wet weight. With its reduced hemodynamic effects, CSS might also acutely burden pulmonary physiology less than OSS. However, the relative impacts of CSS and OSS on respiratory physiology remain unclear. CSS have been shown to increase airway resistance (R) in cases of incomplete catheter withdrawal, a complication specific to CSS that often evades detection.31 Therefore, to resolve whether OSS and CSS differentially affect cardiopulmonary physiology differentially affect, this study used a crossover design with 66 subjects to test the hypothesis that CSS presents better pulmonary and cardiovascular parameters after suction than OSS. To compare how these techniques affect pulmonary mechanics and hemodynamics, we assessed cardiopulmonary physiology before and after endotracheal suction in mechanically ventilated subjects in an ICU.
QUICK LOOK
Current Knowledge
Open suction uniquely impairs pulmonary compliance. Closed suction systems may maintain and restore pulmonary function more effectively compared to open suction systems.
What This Paper Contributes to Our Knowledge
During endotracheal aspiration with a closed circuit system post-aspiration decreases in airway pressures were greater than use of the open circuit system. Both techniques had similar impact on hemodynamic variables.
Methods
Subjects
This study was approved by the Research Ethics Committee of Heliopolis Hospital, São Paulo, Brazil (Approval #467). Prior to enrollment, the primary investigator (RD Raimundo) screened the medical records of all mechanically ventilated patients in the ICU (> 18 y old and intubated with a 7.5-mm endotracheal tube) for past or current lung disease, upper gastrointestinal bleeding, hemodynamic instability, or airway bleeding. Subjects were not using vasoactive drugs but were under deep sedation (scores of 5 or 6 on the Ramsay scale) and analgesic effects. Subjects were quasi-randomized, and those with ≤ 90% immediately before aspiration were excluded from analyses ex post facto (Fig. 1).
Subject Involvement
Because sedation had rendered these subjects unresponsive, the primary investigator identified for all eligible subjects a surrogate legally able to represent that subject and provide consent for procedures and research in Brazilian hospitals (eg, a spouse, relative, or guardian ≥ 18 y old) per the approved protocol. Subjects were involved in the design and conduct of this research. During the subject selection stage, priority of the research question, outcome measures, and methods of recruitment were informed by discussions with subjects' surrogates.
Cardiopulmonary Measures and Suction
After confirming absence of respiratory drive, systolic blood pressure, diastolic blood pressure, mean blood pressure, heart rate, breathing frequency, , tidal volume (VT), PIP, plateau pressure (Pplat), PEEP, and inspiratory flow were measured; in addition, PIP, Pplat, PEEP, inspiratory flow, and VT were used to calculate compliance and airway resistance (R). Subjects receiving volume-control continuous mandatory ventilation were screened in the morning (∼7:00 am) while supine after ≥ 1 h free of procedures or postural changes. Immediately thereafter (ie, just before suction), physiologic data were collected. Suction involved either an OSS or a CSS according to a quasi-random crossover sequence and with data recorded 3 min later. Subjects were returned to nursing care until the alternate suction was performed approximately 6 h later (∼1:00 pm), with the same timing for data collection. The procedures required 10 s to insert and remove the aspiration catheter, with each aspiration performed 3 times (see the supplementary materials at http://www.rcjournal.com).
Results
After exclusion, 66 subjects (Fig. 1) were included in the study. Baseline characteristics are summarized in Table 1, and physiology parameters prior to suction are detailed in the supplementary materials (available at http://www.rcjournal.com).
Pulmonary Physiology
The sequence in which suction types were administered did not alter the cardiopulmonary effects of either suction type. In this study, OSS increased R by 1.3 ± 0.5 cm H2O/L/s and PIP by 1.4 ± 0.5 cm H2O, whereas CSS decreased R by 0.9 ± 0.3 cm H2O/L/s and PIP by 1.0 ± 0.5 cm H2O relative to baseline (all P < .05). Moreover, the 2 suction types had significantly opposing mean effects (P < .05); however, because baseline values of R, PIP, and Pplat correlated inversely with suction-induced changes (Fig. 2, Table 2), models that control for baseline were used. This approach yielded results analogous to our initial observations (Fig. 2, Table 2) and revealed that CSS attenuated preexisting elevations in R and PIP more robustly and consistently than OSS (P < .05). Based on the x intercepts in the linear models (Fig. 2), which account for physiologic values prior to suction, CSS decreased R for individuals with baseline > 8.8 cm H2O/L/s (76% of cases), whereas OSS increased R for subjects with baseline < 15.6 cm H2O/L/s (87% of cases). Moreover, CSS diminished PIP in subjects with baseline PIP > 22.4 cm H2O (83% of cases), whereas OSS increased PIP for those with baselines < 31.5 cm H2O (83% of cases). Both suction treatments similarly increased Pplat if a subject's baseline undershot 14.3 cm H2O but decreased this parameter if a subject's baseline exceeded this threshold. Interestingly, CSS did not alter compliance, regardless of baseline value (Fig. 2, Table 2), whereas OSS progressively decreased compliance the more a subject's baseline compliance exceeded 61.6 cm H2O/L (33% of cases); this effect differed significantly from CSS (P = .030) (Fig. 2, Table 2).
Cardiovascular Physiology
With the exception of , aspiration significantly altered all cardiovascular parameters both in paired t tests (data not shown) and in regressions accounting for baseline values and independent of suction type (Fig. 3, Table 3). Suction increased heart rate among subjects with baseline < 115 beats/min (most subjects) and increased systolic blood pressure in those with baseline < 146 mm Hg (most subjects), with more pronounced effects in subjects markedly underneath these threshold values. For instance, the regressions indicated that suction would induce an increase in heart rate of 7 beats/min for a given subject with a baseline heart rate of 70 beats/min. Likewise, suction would cause an estimated increase in systolic blood pressure of 10 mm Hg in a subject with a baseline systolic blood pressure of 120 mm Hg, and an increase of 6 mm Hg in a subject with a baseline systolic blood pressure of 130 mm Hg.
Discussion
The findings of this study indicate that open suction imposes short-term adverse effects on pulmonary physiology that are distinct from those due to closed suction, including increases in pulmonary resistance and peak pressure and decreases in compliance. Baseline pulmonary physiology also influences the effects of suction. After adjusting for baseline physiology, open suction impaired these measures in > 80% of cases, whereas closed suction improved these parameters in > 75% of cases. These effects collectively suggest that open suction may transiently compromise pulmonary function in sedated patients in the ICU. In contrast, in individuals with high resistance or high PIP, only closed suction improved these indices, suggesting that closed systems may uniquely benefit patients with impaired pulmonary function. Although long-term health outcomes were not assessed, the findings of this study bolster prior observations that open suction, but not closed suction, is associated with potentially adverse short-term cardiopulmonary effects.
Another interesting result is that both suction types increased blood pressure and heart rate statistically equally, which may be explained by a decrease in parasympathetic activity after suctioning.32 This should be taken into consideration by clinicians because a decrease in parasympathetic activity in the sinoatrial node decreases the ability of the heart to adapt rapidly to changing stressors, compromises arterial baroreflex sensitivity by decreasing the beat-by-beat heart rate response to changes in arterial pressure, and increases the risk of sympathetically mediated heart rate and rhythm disturbances and sudden death.32
Several studies have compared open and closed suction systems, with the majority examining incidence of ventilator-associated pneumonia, mortality, length of mechanical ventilation, and time of intensive care.10,19,22,30,33–37 Although multiple studies have reported a higher bacterial colonization with CSS, these observations seem to bear little clinical relevance.10,25,35–37 A recent meta-analysis by Kuriyama et al27 associated CSS with a lower risk of ventilator-associated pneumonia; however, similar to other meta-analyses,8,25 the authors reported no differences in mortality or length of ventilation. In our study, OSS and CSS increased heart rate and arterial pressure equally. These effects, which were relatively modest, diminished with increasing baseline values and were thus likely of little clinical concern. Furthermore, these observations agree with at least 1 prior report32 but conflict with other observations that OSS causes greater increases in heart rate3,16,18,37 or arterial pressure16,23,38; however, these earlier studies recruited fewer subjects (n = 10–19 per suction type) and typically did not control for sampling error by adjusting for baseline physiology in their statistical models. Notably, in supplementary analyses, the same statistical approach as that used in these prior studies (ie, 2-tailed paired t tests) revealed no difference between suction types in chronotropic or hemodynamic changes from baseline in our study.
No changes in peripheral arterial oxygen saturation were found with suctioning. However, because this study was limited to subjects lacking prior pulmonary disease and sedated at 5 or 6 on the Ramsay scale, such observations may not apply to patients with respiratory disease or under deeper anesthesia. Moreover, in contrast to data being collected at 1–2 min after suction,30,39,40 we collected these data 3 min after suction. These distinctions may explain how other studies demonstrated that OSS, but not CSS, decreased after aspiration3,30,38,41 and why several studies thus recommend preemptive hyperoxygenation or manual hyperventilation before aspiration.3,30,37,38
OSS and CSS divergently affected R (P < .001) and PIP (P = .003) but had comparable effects on Pplat (P = .78) and compliance (P = .12) when controlling for the significant influence of baseline physiology. In individuals with moderate baseline values, OSS uniquely increased R and PIP, which is directly influenced by R. Thus, the present data indicate that OSS may disproportionately increase R. Furthermore, CSS more consistently returned elevated PIP and R to normal pulmonary function. Together, these findings indicate that CSS poses fewer risks and bears greater salutary effects on pulmonary function. In contrast, Cereda et al30 reported that both suction types negligibly affected airway pressure and PIP after suction, but OSS uniquely increased mean arterial pressure and decreased . Similarly OSS was not observed to affect R in young children.17 However, neither study adjusted their statistical models by baseline pulmonary physiology, which, as noted in our study, significantly influenced the directionality of responses to suctioning. Elevated R can indicate airway obstruction, which may provoke bronchospasm, which is an occasional complication during endotracheal suction in the ICU. The effect of OSS on R exceeded that of CSS by 8%, which in susceptible individuals (eg, patients with asthma and elevated R) may further compromise ventilation and counteract any intended improvement in R via aspiration.17,26,42
Several studies3,18,22,43 have reported specific advantages of CSS over OSS with respect to physiologic impact or potential for infection, but these have not translated to clear differences in subject outcomes. Kaur et al20 noted that, relative to OSS, CSS decreased contamination and improved preservation of lung volumes and oxygenation, especially in severely hypoxemic subjects. Although there is no scientific evidence for the lower efficiency of secretion removal for CSS, users of CSS in ICUs report lower volumes of aspirated secretion and greater propensity for catheter obstruction by airway foreign bodies.20,44 Dong et al26 noted in mechanically ventilated subjects that, relative to OSS, CSS minimized the adverse effects of suction on arrhythmia, arterial pressure, and , and CSS expedited removal from mechanical ventilation. However, the 2 suction types did not differ in ICU length of stay, risks of ventilator-associated pneumonia, microbial colonization, or adverse outcome.
Unlike OSS, CSS allows uninterrupted ventilation and maintains positive pressure, potentially contributing to the improvements in R and PIP in contrast with worsening of these parameters with OSS. In addition to demonstrating that baseline pulmonary physiology can influence the effects of suction, our results suggest that CSS may especially benefit patients with high baseline PIP and R. Conversely, OSS may uniquely impair pulmonary compliance, R, and PIP. We thus speculate that, in individuals with impaired compliance or elevations in PIP or R, CSS may provide greater therapeutic benefits than OSS. These benefits may result particularly from improved R. Conversely, increases in R with OSS may have resulted from bronchospasm, mucus obstruction, or peripheral airway edema secondary to less efficient suction or loss of positive pressure with interruption in ventilation. Although this study did not measure the aspirate collected for comparison, some studies have noted no difference in secretion volume between OSS and CSS,10 while others report that OSS more effectively reduces secretion volume than CSS.35 We can speculate that the increases in R and PIP and the reductions in compliance with OSS in this study may actually indicate movement of secretions into more central airways, potentially indicating that CSS is less effective, as reported by Lindgren et al.45 Because OSS and CSS similarly affected hemodynamic end points that are particularly sensitive to airway irritation,46 the divergent impacts of OSS and CSS on R were likely independent of airway irritant reflexes. Ultimately, the increase in R with OSS may promote pulmonary hyperinflation, increase dead space, impair ventilation, and (with increased intra-alveolar pressures) diminish alveolar perfusion, potentially explaining prior reports that OSS disproportionately decreases .8,16,26,28 Given that patients with asthma, lung disease, or obesity are more susceptible to increases in R, our findings suggest that they may especially benefit from aspiration with CSS.
This study provides novel findings of differential pulmonary effects of OSS and CSS, but they may not be generalizable to all mechanically ventilated patients. First, the OSS used here lacked recent design improvements (eg, Bodai PEEP-SAFE) that reduce the temporary loss in pressure and risk for attendant complications. However, clinics in lower- and middle-income countries (eg, Brazil) commonly reserve these accessories for unstable patients or lack them altogether due to their added costs. Moreover, because the subjects who either declined consent (via surrogates) or did not meet initial criteria were not quantified, the generalizability of this study remains unclear. Finally, a focus on acute responses and the neglect of any long-term outcomes limit the clinical importance of our findings. Future studies involving larger subject populations, longer observation windows, and recent technological improvements may further elucidate the differential impacts of the 2 suction systems.
Conclusions
This study demonstrates that endotracheal suction with a CSS alleviated elevations in airway resistance and PIP more consistently and effectively. In mechanically ventilated patients, high airway resistance or PIP prior to suction may clinically indicate aspiration with a CSS and contraindicate suction with an OSS. Additional studies are required to resolve whether such practices prevent pulmonary injury during mechanical ventilation. In addition, the clear effects of both suction systems on pulmonary physiology immediately after suctioning suggest that, when assessing pulmonary physiology in ventilated patients, clinicians should account not only for the baseline pulmonary function (ie, prior to suctioning) but also the time relative to last suctioning; patients should be allowed some time for recovery after suctioning, even with CSS, before acquiring measures that inform a patient's general pulmonary stability.
Acknowledgments
We thank Dr Joseane Elza Tonussi Mendes Rossette and Dr Andres R Perez-Riera for their valuable contributions to the data collection and analysis.
Footnotes
- Correspondence: Talita Dias da Silva PhD, Harvard School of Public Health, Department of Environmental Health, 677 Huntington Ave, Boston, MA 02115. E-mail: talita.dias{at}unifesp.br
Drs Riamundo and Sato are co-first authors.
The authors have disclosed no conflicts of interest.
Supplementary material related to this paper is available at http://www.rcjournal.com.
- Copyright © 2021 by Daedalus Enterprises