Abstract
Background There is a dearth of literature about the safety and practicality of intubation performed by an internal medicine (IM) or any other nonanesthesia, nonemergency physician.
Objectives The objectives of the study were to describe abbreviated airway management training guidelines for IM physicians staffing the emergency department and to compare the success rates between intubations performed by anesthesia and IM physicians, respectively.
Methods In this study, 272 consecutive out-of-operating room intubations performed by anesthesia and IM physicians were evaluated after creating and implementing an abbreviated intubation training protocol.
Results Of 165 intubations attempted by IM physicians and 107 intubations attempted by the anesthesia service, the rates of successful intubation were 93% and 99%, respectively (P = 0.02). There were no other statistically significant differences in outcomes.
Conclusions Procedurally oriented IM fellows could provide a temporary solution to hospitals that currently do not have the resources to provide full-time, in-house anesthesiology or emergency physicians for management of the emergent airway.
Airway management is a critical service administered in acute care settings by a disparate group of health care providers including anesthesiology (AN) physicians, emergency physicians (EPs), and respiratory therapists.1Maintaining a well-qualified staff that has trained for and demonstrated proficiency in airway skills and proactive airway assessment is essential. Despite the importance of this educational process, there is no clear consensus on the most effective training techniques.2
General airway management has historically been the purview of the AN physician, but recently, the routine management of the emergent airway has become the responsibility of the EP.3–5Often, the AN physician is available either in-house or on call to provide assistance when difficult airway management issues arise, and many hospitals continue to provide 24-hour, 7-days-a-week, in-house anesthesia coverage. However, there are still many practice settings, including Veterans Affairs (VA) and rural hospitals, where employing full-time, board-certified EPs is challenging.6–9
There is much literature to support EPs providing airway expertise, but a dearth of literature concerning non–anesthesiology-trained, non–emergency medicine (EM)–trained physicians working in an emergency department (ED).3,4,10This employment scenario is particularly still prevalent with the VA. In this article, we describe abbreviated airway management training guidelines for internal medicine (IM) physicians staffing the ED. We then compare the success rates between intubations performed by the AN and IM physicians, respectively. There are 3 goals of this research. The first goal was to determine if a more rapid training regimen could prepare non–EM-trained physicians to manage an emergent airway. Second, we compared intubation success of non–traditionally trained physicians with a historical criterion standard, the AN specialist. Lastly, we considered if teaching this skill to these physicians could provide a temporary, but practical solution for other EDs to utilize as EM continues to grow.
MATERIALS AND METHODS
Training Protocol
This study was performed in a large-scale, tertiary, VA hospital in Chicago, IL. All clinical staff who perform out-of-operating room (OOR) intubations must demonstrate competency. In 2005, there was a change in policy on emergency airway management requiring live intubation training for practitioners involved in OOR airway management at VA hospitals.11As a result of the mandate, all clinical staff whose normal or traditional duties did not include multiple intubations underwent training and credentialing process where they demonstrated sufficient subject matter expertise and procedural competency before being granted OOR intubation privileges.
The program began with didactic sessions that covered basic airway rescue techniques and difficult airway recognition. The didactic curriculum was designed in concordance with VA policy. At the completion, learners were required to demonstrate the following:
knowledge of the major anatomic structures of the airway
ability to formulate and verbalize an appropriate alternative plan, if initial attempts at intubation were unsuccessful
knowledge of the indications and contraindications for pharmaceutical agents, especially muscle relaxants, for use in airway management
Didactics were followed by practical sessions that guided trainees through airway management techniques on a simulator or had an attending anesthesiologist closely supervise live intubation experiences in the operating room (OR). Trainees were also instructed in the use of alternative airway management techniques, specifically mask ventilation and laryngeal mask airway (LMA) placement. Learners were required to complete at minimum the following:
Five successful (ie, without complications) endotracheal intubations with actual patients, not a mannequin. These patients were previously scheduled for the OR and were deemed low risk by an attending anesthesiologist.
Five successful (ie, without complications) cases of ventilating unconscious patients using a bag and mask. Adequate ventilation was determined by an attending anesthesiologist and was based on percent oxygen saturation and minute ventilation.
Demonstrate one example of the use of alternative methods of intubation that are in practice at our hospital with an actual patient, not a mannequin. The specific method instructed was on the use of LMA. This instruction was performed during OR cases that were already planned to be managed by LMA. Trainees were instructed that in real-world scenarios, if the alternative method of intubation failed, specifically LMA placement, the next step in the pathway was to notify backup, specifically the AN physician. Although the notification of backup does somewhat run counter to the stated aim of the study, in the interest of patient care until the hypothesis is proven, it would be prudent to have backup on standby. At issue is the success rate without intervention from the AN backup so this does not compromise the study results.
Requirements could be met with the same patients at the same time. Competency was determined by the chief of the anesthesiology service or designee.11
Intubations
After the institution of the new mandated training protocol, we collected and evaluated data pertaining to all consecutive OOR intubations from July 2006 to July 2008. The physician primarily responsible for the intubation scenario was required to complete an intubation progress note. This note was a preexisting template already used as a progress note for documentation of intubations and was already in place as a tool when the study was prospectively initiated. All notes were reviewed monthly by a member of the cardiopulmonary resuscitation committee to confirm that all intubation attempts were completely and accurately recorded in the medical record. When an intubation progress note was incomplete or inappropriately completed, the head of the cardiopulmonary resuscitation committee would notify the practitioner and the practitioner’s department head of the insufficiency and then verify that the template was eventually completed appropriately. An intubation attempt was defined as insertion of the laryngoscope into the oropharynx regardless of whether visualization was achieved. Successful intubation was defined as direct visualization of endotracheal tube placement through the vocal cords with distal tube placement in the middle of the trachea by “positioning the upper end of the cuff 2 cm below the cords during direct laryngoscopy” or 4 cm above the tracheal “carinae with aid of a fiberoptic scope.”12,13Additional confirmation of placement was obtained by “clinical and/or radiological assessment,” which may include chest and epigastric auscultation, chest x-ray, cuff maneuvers, fiberoptic scopes, and transillumination.12,13
Two prospective cohorts were compiled and followed. The first cohort consisted of all OOR, non-ED intubations performed by the AN service during the daytime on-call hours. This cohort included intubations performed by AN residents and nurse anesthetists. The second cohort consisted of intubations performed by IM physicians working in the ED during the evening and overnight hours when the in-house AN physician was unavailable. The vast majority of intubations performed by the IM cohort were not in the ED (>95%). Most were done on the hospital floors or intensive care unit. Intubations performed in the controlled setting of the OR were excluded from the data analysis. All remaining intubations performed within the study period were included for study evaluation.
Data Gathering
Data were collected prospectively from the code records and intubation progress note templates in the Computerized Patient Record System (Fig. 1). Data collection was performed by 1 author, and all data were arranged in tabular format in Microsoft Excel. Approval and oversight of data collection were granted by the local institutional review board.
Outcomes
Data were evaluated for successful achievement of emergent airway, number of intubation attempts and rescue airways, survival of code, and 7-day survival. Data were also gathered on the frequency the AN physician was called for assistance with intubations (Fig. 1).
Data Analysis
Data analysis and interpretation were blinded and performed by 2 separate researchers not involved in data collection. All statistical calculations were evaluated and performed by a representative of the Department of Biostatistics and member of the Clinical Translational Research Center through an interprofessional center agreement. Pearson χ2 tests were performed to test the differences between AN and IM physicians in the aspects of success rate of intubation, proportion of surviving the initial code, and 7-day survival rate. Two-sample t test was performed to test the difference of the number of intubation attempts between the 2 groups. PROC Freq and PROC TTest in SAS 9.1 (Cary, NC) were used to perform these analyses.
RESULTS
There were a total of 272 OOR intubations attempted from July 2006 to June 2008 (Table 1). Of those intubations, 165 were performed by the IM physicians working in the ED and 107 were performed by the AN physician.
Internal medicine physicians successfully completed 154 intubations, with a success rate of 93%, and the AN physicians successfully completed 106 intubations, with a success rate of 99%. The IM physicians were unable to successfully perform intubation on 11 patients. The calculated odds ratio was statistically significant at 12.08 (95% confidence interval [CI], 1.4 100; P = 0.02).
Of successful intubations, 92% (141/154) and 95% (101/106) of IM and AN patients, respectively, survived the initial code (Table 2). There was no statistically significant difference in the percentage of patients who initially survived the code, with the calculated odds ratio of 1.5 (95% CI, 0.5–4.7; P = 0.45). Seven-day survival rates after successful intubation by an IM or AN physician were 70% (108/154) and 73% (77/106), respectively (odds ratio, 0.76, 95% CI, 0.41–1.4; P = 0.88). The average number of intubation attempts by the IM physician was 1.33 ± 0.21 compared with the AN physician (average, 1.28 ± 0.3; P [t test] = 0.50).
Three specialties were represented among participating IM fellows: pulmonary/critical care (n = 4), cardiology (n = 3), and gastroenterology (n = 2). The IM fellows had a range of 9 to 23 live intubation experiences (median = 15) and 5 simulated intubations each. The AN residents completed approximately 300 intubations at the start of clinical anaesthesia-2 (postgraduate year-3) year and approximately 600 at the start of clinical anaesthesia-3 (postgraduate year-4) year.
DISCUSSION
In this study, the authors found that IM and AN physicians had successful intubation rates of 93% and 99%, respectively. The absolute difference in success rates was 6% and was found to be statistically significant for this sample size, despite an underpowered design. The IM success rate was achieved after instruction that included as few as 5 simulated intubations in conjunction with each fellow’s own intubation experience. The experience level of the IM fellows and AN residents was discussed above. Of successfully completed intubations, 92% of IM patients and 95% of AN patients survived the initial code. The 3% difference was statistically insignificant. Differences in 7-day survival and average number of intubation attempts were also statistically insignificant.
Previous study of the learning process of first-year AN residents showed that the intubation learning curve reached a 90% success rate after a mean of 57 attempts. After 80 intubations, 18% of residents still needed assistance.14Another prospective study evaluated medical, paramedic, and respiratory therapy students to determine how much direct laryngoscopic tracheal intubation training experience was necessary before a novice could be considered competent. The study calculated an 80% probability of performing a “good” intubation after 35 practice trials and a 90% probability of performing a good intubation after 47 trials.15
In our study, a 90% success rate was achieved with a relatively abbreviated curriculum, in addition to the physicians’ experience. Currently, there are no other published studies that show achievement of similar success rates following an abbreviated training program. There is literature that discusses appropriate methods for teaching and learning airway management, but there is no definitive guidance for whether simulated intubations on models or controlled-setting intubations on live patients best balance the benefits and risks to students and patients.16
The IM fellows were all being trained in procedurally oriented fellowships. This likely contributed to the strong success rate. Outside of the simulated models, the IM fellows had more endotracheal intubation experience and more advanced dexterity from their fellowships in pulmonary critical care, gastroenterology, and cardiology.
In circumstances when in-house AN physicians or EPs are unavailable, trained, procedurally oriented IM fellows could provide a temporary solution to OOR emergencies.
Limitations
There are several limitations to this study. First, AN service was responsible for intubations performed during the daytime, whereas the IM physicians were responsible for nighttime and weekend intubations with anesthesia backup. Also, there was no control for the severity level of patient sickness or presenting airway difficulty level. Emergency medicine staff was present for daytime ED intubations, meaning that the AN physician performed OOR intubations only on the other hospital floors and did not perform intubations within the ED or on the ED patient population. The IM physician was responsible for ED intubations during the evening in addition to the OOR intubations in the rest of the hospital. This contributed a possible timing or location bias to the study.
Second, this study was underpowered, and it was calculated that in order to have 80% power to detect a 10% difference between groups would have required a minimum of 300 intubations per group. As a pilot study, power could not be achieved; however, success rate differences were still found to be statistically significant despite the study being underpowered.
Third, we did not maintain data on the patient population. Because this study was completed at a VA hospital, we acknowledge that the results may not be generalizable to certain outside institutions with more diversity in certain components of its population (ie, women, children). Nevertheless, the training protocol is certainly generalizable to VAs and noteworthy for other institutions that may consider modifications as necessary.
Lastly, there is a potential cofounder in that the IM fellows were all receiving training in arguably the most procedurally oriented IM specialties. This training could have aided the IM fellows developing additional dexterity and intubation skills that could have artificially raised the IM success rate. It is possible that this training protocol built on a preexisting foundation of knowledge. Nevertheless, it is unlikely that they received an amount of previous instruction similar to that provided to AN physicians or even EPs.
CONCLUSIONS
This study evaluated the success and safety of non–EM- and non–anesthesia-trained personnel performing OOR intubations during off-hours and without an in-house AN service in a large, urban, tertiary-care VA hospital. Here, data demonstrated that procedurally oriented IM physicians achieved a greater than 90% success rate, an accepted benchmark for competence set by previous studies of intubation training. Despite inadequate power, we demonstrated that a specific subgroup of IM physicians can likely provide airway management in a large hospital with clinical outcomes similar to AN-trained residents. With our abbreviated training model and previous experience, intubations classified as OOR were safely done by IM fellows in our institution, even though IM residency training does not specifically require a certain level of endotracheal intubation experience to be American Board of Internal Medicine accredited.17The explanation for this achievement is at least partially explained by their previous intubation experience and developed procedural skills as fellows but warrants more investigation in future studies. Further studies of the process of accelerated airway management training could help elucidate the key components of success as well as delineate necessary prerequisites essential for candidate selection. Future research will also have to investigate the feasibility of implementing similar abbreviated airway training programs in other practice environments where there is limited availability of AN- or EM-trained personnel and where the population may be vastly different from a large, urban, tertiary-care VA hospital.