Achieving the Recommended Endotracheal Tube Cuff Pressure: A Randomized Control Study Comparing Loss of Resistance Syringe to Pilot Balloon Palpation

Background Both under- and overinflation of endotracheal tube cuffs can result in significant harm to the patient. The optimal technique for establishing and maintaining safe cuff pressures (20–30 cmH2O) is the cuff pressure manometer, but this is not widely available, especially in resource-limited settings where its use is limited by cost of acquisition and maintenance. Therefore, anesthesia providers commonly rely on subjective methods to estimate safe endotracheal cuff pressure. This study set out to determine the efficacy of the loss of resistance syringe method at estimating endotracheal cuff pressures. Methods This was a randomized clinical trial. We enrolled adult patients scheduled to undergo general anesthesia for elective surgery at Mulago Hospital, Uganda. Study participants were randomized to have their endotracheal cuff pressures estimated by either loss of resistance syringe or pilot balloon palpation. The pressures measured were recorded. Results One hundred seventy-eight patients were analyzed. 66.3% (59/89) of patients in the loss of resistance group had cuff pressures in the recommended range compared with 22.5% (20/89) from the pilot balloon palpation method. This was statistically significant. Conclusion The loss of resistance syringe method was superior to pilot balloon palpation at administering pressures in the recommended range. This method provides a viable option to cuff inflation.


Background
High-volume low-pressure cu ed endotracheal tubes (ETT) are the standard of airway protection. However, they have potential complications [1][2][3]. Underin ation increases the risk of air leakage and aspiration of gastric and oral pharyngeal secretions [4,5]. Cu pressures less than 20 cmH 2 O have been shown to predispose to aspiration which is still a major cause of morbidity, mortality, length of stay, and cost of hospital care as revealed by the NAP4 UK study. In this cohort, aspiration had the second highest incidence of primary airway-related serious events [6].
On the other hand, overin ation may cause catastrophic complications. It has been demonstrated that, beyond 50 cmH 2 O, there is total obstruction to blood ow to the tracheal tissues. is has been shown to cause severe tracheal lesions and morbidity [7,8]. However, less serious complications like dysphagia, hoarseness, and sore throat are more prevalent [9][10][11]. ere are a number of strategies that have been developed to decrease the risk of aspiration, but the most important of all is continuous control of cu pressures.
A newer method, the passive release technique, although with limitations, has been shown to estimate cu pressures better [21][22][23][24]. is method has been achieved with a modi ed epidural pulsator syringe [13,18], a 20 ml disposable syringe, and more recently, a loss of resistance (LOR) syringe [21,23,24]. Compared with the cu manometer, it would be cheaper to acquire and maintain a loss of resistance syringe especially in low-resource settings.
Our study set out to investigate the e cacy of the loss of resistance syringe in a surgical population under general anesthesia. ETTcu pressures would be measured with a cu manometer following estimation by either the PBP method or the LOR method. Our secondary objective was to determine the incidence of postextubation airway complaints in patients who had cu pressures adjusted to 20-30 cmH 2 O range or 31-40 cmH 2 O range.

Study Design and Setting.
is single-blinded, parallelgroup, randomized control study was performed at Mulago National Referral Hospital, Uganda. e hospital has a bed capacity of 1500 inpatient beds, 16 operating rooms, and a mean daily output of 90 surgical operations. Anesthesia services are provided by di erent levels of providers including physician anesthetists (anesthesiologists), residents, and nonphysician anesthetists (anesthetic o cers and anesthetic o cer students). In the early years of training, all trainees provide anesthesia under direct supervision. In the later years, however, they can administer anesthesia either independently or under remote supervision.

Study Population.
We included ASA class I to III adult patients scheduled to receive general anesthesia with endotracheal intubation for elective surgical operation. All patients with any of the following conditions were excluded: known or anticipated laryngeal tracheal abnormalities or airway trauma, preexisting airway symptoms, laparoscopic and maxillofacial surgery patients, and those expected to remain intubated beyond the operative room period.
At the time of the intervention, the study investigator retrieved the next available envelope, which indicated the intervention group, from the next available block envelope and handed it to the research assistant. e patient was the only person blinded to the intervention group. e anesthesia providers were either physician anesthetists (anesthesiologists or residents) or nonphysicians (anesthetic o cer or anesthetic o cer student). Students were under the supervision of a senior anesthetic o cer or an anesthesiologist.

Sample Size.
To detect a 15% di erence between PBP and LOR groups, it was calculated that at least 172 patients would be required to be 80% certain that the limits of a 95%, two-sided interval included the di erence.

Intervention and Anesthetic
Technique. Every patient was wheeled into the operating theater and transferred to the operating table. Basic routine monitors were attached as per hospital standards. A wide-bore intravenous cannula (16-or 18-G) was placed for administration of drugs and uids. e patient was then preoxygenated with 100% oxygen and general anesthesia induced with a combination of drugs selected by the anesthesia care provider. ese included an intravenous induction agent, an opioid, and a muscle relaxant. All patients received either suxamethonium (2 mg/kg, max 100 mg to aid laryngoscopy) or cisatracurium (0.15 mg/kg at for prolonged muscle relaxation) and were given optimal time before intubation. Using a laryngoscope, tracheal intubation was performed, ETTposition con rmed, and secured with tape within 2 min. e size of ETT (POLYMED Medicure, India) was selected by the anesthesia care provider. With the patient's head in a neutral position, the anesthesia care provider in ated the ETT cu with air using a 10 ml syringe (BD Discardit II). A research assistant (di erent from the anesthesia care provider) read out the patient's group, and one of the following procedures was followed.
PBP group (active comparator): in this group, the anesthesia care provider was asked to reduce or increase the pressure in the ETT cu by in ating with air or de ating the pilot balloon using a 10 ml syringe (BD Discardit II) while simultaneously palpating the pilot balloon until a point he or she felt was appropriate for the patient. When this point was reached, the 10 ml syringe was then detached from the pilot balloon, and a cu manometer (VBM, Medicintechnik Germany. Accuracy −2 cmH 2 O) was attached. e pressure reading of the VBM was recorded by the research assistant. LOR group (experimental): in this group, the research assistant attached a 7 ml plastic, luer slip loss of resistance syringe (BD Epilor, USA) containing air onto the pilot balloon. e cu was then brie y overin ated through the pilot balloon, and the loss of resistance syringe plunger was allowed to passively draw back until it ceased. is point was observed by the research assistant and witnessed by the anesthesia care provider. e loss of resistance syringe was then detached, the VBM manometer was attached, and the pressure reading was recorded. Cu pressure adjustment: in both arms, very high and very low pressures were adjusted as per the recommendation by the ethics committee. In case of a very low pressure reading (below 20 cmH 2 O), the ETT cu pressure would be adjusted to 24 cmH 2 O using the manometer. On the other hand, high cu pressures beyond 50 cmH 2 O were reduced to 40 cmH 2 O.
Also, at the end of the pressure measurement in both groups, the manometer was detached, breathing circuit was attached to the ETT, and ventilation was started. e patient was maintained on iso urane (1-1.8%) mixed with 100% oxygen owing at 2 L/min. Anesthesia continued without further adjustment of ETT cu pressure until the end of the case. Nitrous oxide and medical air were not used as these agents are unavailable at this hospital. All patients who received nondepolarizing muscle relaxants were reversed with neostigmine 0.03 mg/kg and atropine 0.01 mg/kg at the end of surgery.
2.6. Primary Outcome. Cu pressure reading of the VBM manometer was recorded by the research assistant. e individual anesthesia care providers participated more than once during the study period of seven months.

Secondary Outcome.
e patients were followed up and interviewed only once at 24 hours after intubation for presence of cough, sore throat, dysphagia, and/or dysphonia. Only two of the four research assistants reviewed the patients postoperatively, and these were blinded to the intervention arm. is outcome was compared between patients with cu pressures from 20 to 30 cmH 2 O range and those from 31 to 40 cmH 2 O following the initial correction of cu pressures.

Data Management.
All data were double entered into EpiData version 3.1 software ( e EpiData Association, Odense, Denmark), with range, consistency, and validation checks embedded to aid data cleaning. e data were exported to and analyzed using STATA software version 12 (StataCorp Inc., Texas, USA).
Categorical data are presented in tabular, graphical, and text forms and categorized into PBP and LOR groups. e chi-square test was used for categorical data. Continuous data are presented as the mean with standard deviation and were compared between the groups using the t-test to detect any signi cant statistical di erences. e magnitude of e ect on the primary outcome was computed for 95% CI using the t-test for di erence in group means. An intention-to-treat analysis method was used, and the main outcome of interest was the proportion of cu pressures in the range 20-30 cmH 2 O in each group.
For the secondary outcome, incidence of complaints was calculated for those with cu pressures from 20 to 30 cmH 2 O range and those from 31 to 40 cmH 2 O.

Data Safety Management Board.
e Data Safety Management Board (DSMB) comprised an anesthesiologist, a statistician, and a member of the SOMREC IRB who would be informed of any adverse event. e study would be discontinued if 5% of study subjects in one study group experienced an adverse event associated with the study interventions as determined by the DSMB, or if a p value of <0.001 was obtained on an interim analysis performed halfway through patient accrual. None of these was met at interim analysis.

Ethical Considerations.
e study was approved by the School of Medicine Research and Ethics Committee, Makerere University, and registered with http://www.clinicaltrials.gov (NCT02294422). All patients provided informed, written consent before the start of surgery.  Anesthesiology Research and Practice distribution between arms as shown in the CONSORT diagram in Figure 1. e study comprised more female patients (76.4%). In addition, most patients were below 50 years (76.4%). e study groups were similar in relation to sex, age, and ETT size (Table 1).

ETT Cu Pressures.
Generally, the proportion of ETT cu s in ated to the recommended pressure was less in the PBP group at 22.5% (20/89) compared with the LOR group at 66.3% (59/89) with a statistically signi cant positive mean di erence of 0.47 with p value < 0.01 (0.343-0.602). Also to note, most cu s in the PBP group were in ated to a pressure that exceeded the recommended range in the PBP group, and 51% of the cu pressures attained had to be adjusted compared with only 12% in the LOR group (Table 2). e distribution of cu pressures (unadjusted) achieved by the di erent care providers is shown in Figure 2. e di erence in the number of intubations performed by the di erent level of providers is huge with anesthesia residents and anesthetic o cers performing almost all intubation and initial cu pressure estimations.

Postextubation Airway Symptoms.
e total number of patients who experienced at least one postextubation airway symptom was 113, accounting for 63.5% of all patients. e incidence of postextubation airway complaints after 24 hours was lower in patients with a cu pressure adjusted to the 20-30 cmH 2 O range, 57.1% (56/98), compared with those whose cu pressure was adjusted to the 30-40 cmH 2 O range, 71.3% (57/80). is however was not statistically signi cant (p value 0.053) ( Table 3).

Study Population.
We conducted a single-blinded randomized control study to evaluate the LOR syringe method in accordance with the CONSORT guideline (CONSORT checklist provided as Supplementary Materials available here).
At the study hospital, there are more females undergoing elective surgery under general anesthesia compared with males. In addition, over 90% of anesthesia care at this hospital was provided by anesthetic o cers and anesthesia residents during the study period. Anesthetic o cers provide over 80% of anesthetics in Uganda.  e primary outcome of the study was to determine the proportion of cu pressures in the optimal range from either group. e initial, unadjusted cu pressures from either method were used for this outcome. When considering this primary outcome, the LOR syringe method had a signi cantly higher proportion compared to the PBP method. is adds to the growing evidence to support the use of the LOR syringe for ETT cu pressure estimation. ere are data regarding the use of the LOR syringe method for administering ETT cu pressures [21,23,24], but studies on a perioperative population are scanty. In our study, 66.3% of ETT cu pressures estimated by the LOR syringe method were in the optimal range. Kim and coworkers, who evaluated this method in the emergency department, found an even higher percentage of cu pressures in the "normal range" (22-32 cmH 2 O) in their study. Another study, using nonhuman tracheal models and a wider range (15-30 cmH 2 O) as the optimal, had all cu pressures within the optimal range [21]. It is however di cult to extrapolate these results to the human population since the risk of aspiration of gastric contents is zero while working with models when compared with patients.
ere is a relatively small risk of getting ETT cu pressures less than 30 cmH 2 O with the use of the LOR syringe method [23,24], 12.4% from the current study. Secondly, this method is still provider-dependent as they decide when plunger drawback has ceased. Precaution was taken to avoid premature detachment of the loss of resistance syringe in this study. e PBP method, although commonly employed in operating rooms, has been repetitively shown to administer cu pressures out of the optimal range (20-30 cmH 2 O) [2,3,25]. Findings from this study were in agreement, with 25.3% of cu pressures in the optimal range after estimation by the PBP method. It should however be noted that some of these studies have been carried out in di erent environments (emergency rooms) and on di erent kinds of patients (emergency patients) by providers of varying experience [2].  Another viable argument is to employ a more pragmatic solution to prevent overly high cu pressures by in ating the cu until no air leak is detected by auscultation. is method is cheap and reproducible and is likely to estimate cu pressures around the normal range. Perhaps the LOR syringe method needs to be evaluated against the "no air leak on auscultation" method.
is study shows that the LOR syringe method is better at estimating cu pressures in the optimal range when compared with the PBP method but still falls short in comparison to the cu manometer. In low-and middleincome countries, the cost of acquiring ($ 250-300) and maintaining a cu manometer is still prohibitive.

Postextubation Airway Complaints.
e secondary objective of the study evaluated airway complaints in those who had cu pressure in the optimal range (20-30 cmH 2 O) and those above the range (31-40 cmH 2 O). e di erence in the incidence of sore throat and dysphonia was statistically signi cant, while that for cough and dysphagia was not. e overall trend suggests an increase in the incidence of postextubation airway complaints in patients whose cu pressures were corrected to 31-40 cmH 2 O compared with those corrected to 20-30 cmH 2 O.
is however was not statistically signi cant (p value 0.052).
Considering that this was a secondary outcome, it is possible that the sample size was small, hence leading to underestimation of the incidence of postextubation airway complaints between the groups. It is however possible that these results have a clinical signi cance. e complaints sought in this study included sore throat, dysphagia, dysphonia, and cough. ese were adopted from a review on postoperative airway problems [26] and were de ned as follows: sore throat, continuous throat pain (which could be mild, moderate, or severe), dysphagia, uncoordinated swallowing or inability to swallow or eat, dysphonia, hoarseness or voice changes, and cough (identi ed by a discomforting, dry irritation in the upper airway leading to a cough). All these symptoms were of a new onset following extubation.
ere is consensus that keeping ETT cu pressures low decreases the incidence of postextubation airway complaints [11].
Previous studies have shown that the incidence of postextubation airway symptoms varies from 15% to 94% in various study populations [7,9,11,27] and could be a ected by the method of interview employed, such as the one used in our study (yes/no questions). is study was not powered to evaluate associated factors, but there are suggestions that the levels of anesthesia providers with varying skill set and technique at direct laryngoscopy may be associated with a high incidence of complications. However, this could be a site-speci c outcome.

Conclusion
Although this was a single-blinded, single-centre study, results suggest that the LOR syringe method was superior to PBP at administering pressures in the optimal range. e high incidence of postextubation airway complaints in this study is most likely a site-speci c problem but one that other resource-limited settings might identify with.
We recommend the use of the cu manometer whenever available and the LOR method as a viable option. Alternatively, cheaper, reproducible methods, like the minimum leak test that limit overly high cu pressures should be sought and evaluated.

Disclosure
is work was presented (and later published) at the 28th European Society of Intensive Care Medicine congress, Berlin, Germany, 2015, as an abstract. e poster can be accessed by following the link: https://pdfs.semanticscholar. org/c12e/50b557dd519bbf80bd9fc60fb9fa2474ce27.pdf.

Conflicts of Interest
e authors declare that they have no con icts of interest.

Authors' Contributions
Fred Bulamba, Andrew Kintu, Arthur Kwizera, and Arthur Kwizera were responsible for concept and design, interpretation of the data, and drafting of the manuscript. Charles Kojjo, Agnes Wabule, and Nodreen Ayupo were responsible for patient recruitment and data collection and analysis. All authors have read and approved the manuscript.