We Can Do Better
Manual ventilation is a basic skill that every emergency medical services (EMS) responder is expected to perform proficiently. Improper manual ventilation may result in significant morbidity; however, there is no feedback mechanism or method of control for the volume, pressure, or frequency during manual ventilation. In this study, we aimed to quantify the ventilation parameters of manually delivered breaths using a simulated lung.
One-hundred ninety-nine volunteer EMS responders from the EMS World Expo 2019 and EMS Today 2020 participated in this study. Volunteers manually ventilated a simulated lung using a manual resuscitator bag for 18 breaths. Descriptive statistics were computed for peak pressures (Ppeak) and tidal volumes (VT)), and a multivariable linear regression was conducted to determine whether there was an independent correlation between Ppeak or VT and seven different variables (sex, using one versus two hands, hand length, hand width, self-confidence rating, frequency of manual resuscitator use per month, and years of experience).
Both Ppeak and VT delivered by EMS responders had a high level of variability:
- 9% of providers delivered at least one breath exceeding the recommended safety thresholds of 22 cm H2O Ppeak or 560 mL VT.
- 90% of first responders delivered at least one breath that was inadequate or excessive (VT <420 or >560 mL or Ppeak <10 or >22 cm H2O).
Tidal volumes were significantly higher in:
- Males (p < 0.001, effect size 34.54 mL),
- Those using two hand manual ventilation (p < 0.001, effect size 114.96 mL),
- Shorter hand length (p = 0.013, effect size 5.87 mL per sequential decrease in size),
- Higher confidence (p < 0.001, effect size 23.30 mL per point on a scale of 1-5), and
- More years of experience (p < 0.001, effect size 1.19 mL per increasing years of experience).
Peak pressures were significantly higher in:
- Those using two hand manual ventilation (p < 0.001, effect size 8.92 cm H20),
- Wider hand width (p = 0.004, effect size 1.05 cm H20 per sequential increase in hand width),
- Higher confidence (p < 0.001, effect size 1.28 cm H20 per point increase on a scale of 1-5),
- Less frequent use of the manual resuscitator per month (p < 0.001, effect size .27 cm H20),
- More experience (p < 0.001, effect size .21 cm H20 with increasing years of experience).
Our study demonstrated large variability of VT and Ppeak within and, to a lesser degree, between providers. Our results showed no likely clinical significant role of gender, hand size, frequency of use, or years of experience in determining Ppeak and VT.
Out of the seven variables that might have affected tidal volume or peak pressures, only the use of two hands versus one hand had a potentially clinically significant effect.
Our study identifies a clear need for evolving the manual resuscitator in order to ensure every provider can consistently deliver breaths at appropriate volumes and safe pressures.
The bag-valve mask (BVM), also known as the manual resuscitator, is a critically important piece of medical equipment used to ventilate patients. It is used both in and out of the hospital setting during transportation and emergency resuscitation. Healthcare providers of different disciplines and levels of training routinely use the BVM. Ventilating with a BVM is considered a basic skill that the most inexperienced provider is expected to perform proficiently.
Low tidal volumes (VT ) of 6-8 mL/kg has been reported as having a lower mortality rate compared to the prior convention of 10-12 mL/kg in Acute Respiratory Distress Syndrome (ARDS) patients.
It has been reported that low tidal volumes (VT ) of 6-8 mL/kg have a lower mortality rate compared to the prior convention of 10-12 mL/kg in acute respiratory distress syndrome (ARDS) patients.1 The low VT ventilation strategy was then adopted as the standard of care when low VT was shown to decrease morbidity and mortality in patients without ARDS.2
There have been recommendations of ventilating under 6-8 mL/kg, likely in an effort to avoid high ventilation volumes.3 However, it is important to note that hypoventilation is also harmful to the patient.
Just as high volumes can cause traumatic lung injury, high pressures can also cause complications. Bouvet et al. found pressures delivered via manual resuscitator that were greater than 15 cm H2O resulted in gastric insufflation, which is a known risk factor for aspiration of stomach contents and poor clinical outcomes.4 Additionally, peak pressures greater than 40-45 cm H2O have been associated with barotrauma.5
Peak pressures greater than 40-45 cm H2O have been associated with barotrauma.
The standard manual resuscitator has no feedback mechanism or means to control volume, pressure or frequency of ventilation. This makes it difficult for providers to know if they are ventilating within the recommended parameters.
More importantly, improper manual ventilation may result in significant morbidity to the patient. Increased VT and Ppeak, as well as an increased frequency of breaths delivered, may lead to complications including stomach insufflation with subsequent aspiration; damage to lung parenchyma including pneumothorax; hemodynamic and pulmonary compromise; decreased oxygenation to tissues or increased carbon dioxide retention.4,6 The economic impact of unsafe manual ventilation and subsequent complications is not well studied.
Training among EMTs and paramedics are extremely variable. EMT Basic trainees will typically have about 240 hours of training; EMT A’s have about 800-1,000 hours, while paramedics typically have 1,600 hours of training. (Source: Robert McClintock, deputy director of Fire & EMS Operations, International Association of Fire Fighters, phone call, August 26, 2020).
It is unknown how well EMS responders follow manual ventilation recommendations with regards to Ppeak and VT. In this study, we aimed to quantify the ventilation parameters of each breath delivered by EMS responders in a simulated lung.
Additionally, we analyzed factors that might influence ventilation technique such as confidence, experience, hand size, and hand position (one or two hand bag squeeze). We hypothesized that EMS responders would exhibit ventilation performance outside of the recommendations and that confidence, experience, hand size and hand position would have no effect on manual ventilation performance.
Following the approval from University of Tennessee Healthcare System Institutional Review Board, this study was conducted at the EMS World Expo (New Orleans, Louisiana, October 14-19, 2019) and EMS Today (Tampa, Florida, March 2-6, 2020) conferences. EMS responders gave verbal consent to participate in this study. These study volunteers were asked to manually ventilate a simulated lung (IMT Analytics SmartLung 2000 2L test lung and TSI 5300 Series Gas Flow Multimeter, USA) using an AMBU® SPUR II® (AMBU, Denmark) manual resuscitator while observing chest rise (Figure 1).
Figure 1. Schematic of ventilation simulator.
Participants were instructed to normally ventilate the manikin, which the participants were informed was supposed to represent an average American male patient who was 5-feet 10-inches tall and 200 pounds, for 18 breaths over 90 seconds following a preset variable frequency indicated by a flashing light.
The participants were not told that the ideal VT for this scenario was 420-560 mL based on the height and weight, and they were not told what the purpose of the study was.7 Participants were asked five survey questions while ventilating.
During ventilation, EMS providers were unable to see their performance results and were also blinded to the results of other providers. The variable frequency of ventilation and questioning while performing the ventilation was meant to simulate field conditions.
In addition to the variable rate flashing light, administering the survey during the ventilation served as a distraction to create an environment closer to real clinical practice. Seven independent variables were analyzed: sex, ventilation technique (one versus two handed ventilation), hand length, hand width, confidence, frequency of BVM use per month experience and years of experience.
VT and Ppeak were gathered every 10 milliseconds of ventilation using the TSI 5300 Series Mass Flow Multimeter. Independent variable statistics were computed and multivariable linear regression was conducted to determine statistically significant demographic and participant characteristics associated with delivered VT and Ppeak.
There were 220 attendees who initially volunteered to participate in the study, but 21 providers were excluded based on the following exclusion criteria: retired providers and/or those who live outside the United States of America.
Of the 199 volunteer EMS providers included in this study, 62/199 (31%) were female and 137/199 (68.8%) were male. Providers originated from 34 states with the top five being Florida (37), New York (20), Louisiana (15), Pennsylvania (12), and Texas (10). The average experience with the manual resuscitator was 16.16 years.
Independent variable analysis is displayed in Table 1. On multiple regression using the ordinary least squares method, significance was defined as a p-value <0.05.
Table 1. Independent Variable Analysis (n=199). p-values were generated from multiple regression using the ordinary least squares method evaluating seven variables and their relationship with volume and pressure.
Data (out of 199)
|p-value||Effect Size (mL) and Interpretation
|p-value||Effect Size (cm H2O) and Interpretation
|Sex||M: 137 (68.8%)
F: 62 (31.2%)
|< 0.001||34.54, Males delivered higher VT||0.51||-1.4, No statistically significant correlation|
|One hand vs two hand ventilation||One: 141(70.9%)
Two: 58 (29.2%)
|114.96, Providers using two hands delivered higher VT
|8.92, Providers using two hands delivered higher Ppeak
|Hand Length||XXS: 5 (0.5%)
XS: 27 (13.57%)
S: 49 (24.6%)
M: 51 (25.6%)
L: 36 (18.1%)
XL: 22 (11.1)
XXL: 8 (4%)
XXXL: 1 (0.5%)
|0.013||-5.87, Providers with shorter hands delivered higher VT||0.551||0.13, No statistically significant correlation|
|Hand Width||S: 37 (18.6%)
M: 61 (30.65%)L: 74 (37.2%)
XL: 27 (13.6%)
|0.248||4.69, No statistically significant correlation||0.004||1.0, Providers with wider hands delivered higher Ppeak|
|Self- Confidence Rating (1-5, 5 being most confident)||1: 0 (0%)
2: 5 (2.5%)
3: 29 (14.6%)
4: 71 (35.7%)
5: 94 (47.2%)
|< 0.001||23.30, Providers with higher confidence with their ventilation technique/performance were statistically significant in delivering higher VT||< 0.001||1.28, Providers with higher confidence with their ventilation technique/performance were statistically significant in delivering higher Ppeak|
|Frequency of BVM use per Month||0-5: 137 (68.84%)
5-10: 32 (16.08%)
10+: 30 (15.08)
|0.35, No statistically significant correlation||< 0.001
|-0.27, Providers who bagged less frequently were statistically significant in deliver higher Ppeak|
|Years of Experience||0-5:137 (68.84%)
5-10: 32 (16.1%)
10+: 30 (15.1%)
|<0.001||1.19, Providers with more experience were statistically significant in delivering higher VT||< 0.001
|0.21, Providers with more experience were statistically significant in delivering higher Ppeak|
Higher tidal volumes had statistically significant associations with male sex (p < 0.001, effect size 34.54 mL); two-hand manual ventilation (p < 0.001, effect size 114.96 mL); decreasing hand length (p = 0.013, effect size 5.87 mL per sequential decrease in size), confidence (p < 0.001, effect size 23.30 mL per 1 point on a scale of 1 to 5); and greater experience (p < 0.001, effect size 1.19 mL with each five-year increment of experience).
Higher peak pressures had statistically significant associations with two-hand manual ventilation (p < 0.001, effect size 8.92 cm H20), increasing hand width (p = 0.004, effect size 1.05 cm H20 per sequential increase in hand width), increasing confidence (p < 0.001, effect size 1.28 cm H20 per 1 point on a scale of 1 to 5), less frequent use of the manual resuscitator per month (p < 0.001, effect size .27 cm H20), more years of experience (p < 0.001, effect size .21 cm H20 with each five-year increment of experience).
Figure 2 displays a scatterplot of the performance of breaths delivered by EMS providers compared to that of two commonly used ventilators, a Puritan Bennett 980 (Medtronic, Ireland) and a ReVel® Portable Critical Care Transport Ventilator (Carefusion, USA), used with the same simulated lung.
Table 2 shows the percentages of breaths delivered and providers who were above or below various thresholds for VT and Ppeak. Eight-two point nine percent of our providers delivered a breath over 22 cm H2O or a breath that exceeded the recommended 560 mL VT. 98.0% of first responders studied delivered at least one breath with a volume or pressure that was inadequate or excessive (VT <420 or >560 mL or Ppeak <10 or >22 cm H2O).
Table 2. EMS Provider Performance Summary (n=199).
To our knowledge, this is the largest study specifically designed to evaluate manual ventilation performance among EMS providers. Our study demonstrates considerable inconsistency of manual ventilation within and between providers.
The scatter plot in Figure 2 displays a clear lack of consistency with VT and Ppeak. Over half of the total breaths delivered amongst all providers had a VT that was greater than the recommended maximum of 560 mL. Seventy-one point nine percent of providers delivered at least one breath over 560 mL, which could increase the likelihood of lung injury in real patients.
While the emphasis is normally on dangers of overventilation, under-ventilating during a resuscitation can also cause harm to the patient. Forty-two percent of providers under-ventilated at least one breath below the recommended minimum 420 mL VT.
Seventy-six point nine percent of all of providers involved in the study had at least one breath that crossed the gastric insufflation threshold (22 cm H2O). Fifteen percent of our providers delivered at least one breath at a peak pressure over 50 cm H2O. Peak airway pressure over 50 cm H2O is associated with increased risk of alveolar rupture during mechanical ventilation.8 Nine point one precent of our providers delivered one breath at a pressure over 60 cm H2O.
Complications at approximately 60 cm H2O and beyond are severe and include pneumothorax and pneumomediastinum.9 Pneumothorax has been reported to be associated with a significant increase in the ICU length of stay, hospital stay and mortality, with mortality ranging from 46% to 77%.8 Eighty-two point nine percent of our providers delivered a breath over the gastric insufflation threshold of 22 cm H2O or a breath that exceeded the recommended 560 mL VT. The most concerning statistic is that 98.0% of first responders studied delivered at least one breath with a volume or pressure that was inadequate or excessive.
Despite many of the independent variables having a statistically significant effect on pressure and volume, their effect size was not large enough to make an impact clinically. There was only one variable that was both statistically significant and potentially clinically relevant.
Providers using two hands to ventilate the bag ventilated 114.96 mL more, and at a pressure 8.92 cm H20 higher, than those using one hand. There have been recommendations of using two hands when manually resuscitating a patient.3
As noted previously, elevated peak pressures can cause serious morbidity and our study shows using two hands had the greatest clinical implication of delivering elevated peak pressures in manual ventilation. In comparison to our study, previous literature reported that professional experience, hand size and grip strength had no influence.10,11 Khoury et al. suggest since professional experience does not impact ventilation performance, self-improvement is difficult to achieve with current devices.10 A likely reason that we found statistically significant relationships is the large size of our sample.
Other considerations include potential discrepancy of ventilation between brands and bag size. Khoury et al. reported that the use of different BVM models (Laerdal Bag II and Ambu Spur II) has no impact on manual ventilation performance (𝑝 = 0.79).10 Augustine et al. similarly reported no obvious relationship between various BVM models used and the average VT delivered.12
In a study performed by Dafilou et al., they found that although a pediatric bag performed better compared to an adult bag in an adult simulated patient, the mean VT delivered still exceeded upper limits for lung-protective ventilation.13
Our study participants/providers were limited to a convenience sample of American EMS providers attending the EMS World Expo 2019 and EMS Today 2020 conferences. Therefore, the results might not be generalizable to the global population of EMS providers.
Additionally, our results were obtained from a simulated lung rather than in a clinical setting. Performance variability may exist between these two settings, and ventilation may be even more unsafe in a real clinical setting.
Lastly, most volunteers used the manual resuscitator in their places of work less than five times per month. Results may be different among providers who use manual resuscitators more often.
Whether a patient has ARDS or normal compliant lungs, there is a general consensus for what is considered “safe” when setting VT and Ppeak for a patient on a mechanical ventilator. We believe that ventilating with a manual resuscitator should be held to those same safety parameters, given its use in emergent resuscitations and during transportation.
This study demonstrated the variability and inconsistencies of VT and Ppeak delivered when comparing providers to each other and to themselves. Our results imply that gender, hand size, frequency of use or years of experience has no clinical bearing on how safely a provider ventilates a patient.
Our study also showed that only using one versus two hands to deliver ventilations has some potential clinical relevance regarding volume and peak pressure in our study. Alarmingly, 98.0% of the providers studied delivered at least one breath at an inappropriate VT or Ppeak.
An obvious solution for manual BVM ventilation safety is to create real-time feedback that standardizes both VT and Ppeak delivered, irrespective of who is handling the manual resuscitator.
Our study identifies a need for further innovation to the BVM that ensures any provider can consistently deliver breaths at appropriate VT and safe Ppeak.
- Brower RG, et al. Ventilation with Lower Tidal Volumes as Compared with Traditional Tidal Volumes for Acute Lung Injury and the Acute Respiratory Distress syndrome.” New England Journal of Medicine. 2000;342(18): 1301-1308.
- Neto AS, Cardoso SO, Manetta JA, et al. Association Between Use of Lung-Protective Ventilation with Lower Tidal Volumes and Clinical Outcomes Among Patients Without Acute Respiratory Distress Syndrome: A Meta-Analysis. JAMA. 2012;308(16):1651-1659.
- Rock, M. The Dos and Don’ts of Bag-Valve Mask Ventilation. Journal of Emergency Medical Services. 2014;39(8). https://www.jems.com/2014/08/08/dos-and-don-ts-bag-valve-mask-ventilatio/.
- Weiler N, Latorre F, Eberle B, Goedecke R, Heinrichs W. Respiratory Mechanics, Gastric Insufflation Pressure, and Air Leakage of the Laryngeal Mask Airway. Anesthesia & Analgesia. 1997;84(5):1025–8.
- Esteban A, Anzueto A, Frutos F, et al. Characteristics and outcomes in adult patients receiving mechanical ventilation: a 28-day international study. JAMA. 2002;287(3):345-55.
- 6. Bucher JT, Vashisht R, Ladd M, et al. Bag Mask Ventilation (Bag Valve Mask, BVM). Treasure Island (FL): StatPearls Publishing; 2020 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK441924/.
- Peterson CM, Thomas DM, Blackburn GL, Heymsfield SB. Universal Equation for Estimating Ideal Body Weight and Body Weight at any BMI. Am J Clin Nutr. 2016;103(5):1197-203.
- Hsu CW, Sun SF. Iatrogenic Pneumothorax Related to Mechanical Ventilation. World J Crit Care Med. 2014;3(1):8-14.
- Woodring JH. Pulmonary Interstitial Emphysema in the Adult Respiratory Distress Syndrome. Crit Care Med. 1985;13(10):786-91.
- Khoury A, Sall FS, De Luca A, Pugin A, Pili-Floury S, Pazart L, et al. Evaluation of Bag-Valve-Mask Ventilation in Manikin Studies: What Are the Current Limitations? Biomed Research International. 2016;2016:4521767.
- 11. Otten D, Liao MM, Wolken R, Douglas IS, Mishra R, Kao A, et al. Comparison of Bag-Valve-Mask Hand-Sealing Techniques in a Simulated M Ann Emerg Med. 2014;63:6-12.e3. https://doi.org/10.1016/j.annemergmed.2013.07.014.
- 12 Augustine JA, Seidel DR, McCabe JB. Ventilation Performance Using a Self-Inflating Anesthesia Bag: Effect of Operator Characteristics. Am J Emerg Med. 1987;5(4):267-70.
- Dafilou B, Schwester D, Ruhl N, Marques-baptista A. It’s In The Bag: Tidal Volumes in Adult and Pediatric Bag Valve Masks. West J Emerg Med. 2020;21(3):722-726.