Identifying barriers to the use of ultrasound in the perioperative period: a survey of southwestern Ontario anesthesiologists

This survey was targeted to practicing anesthesiologists in southwestern Ontario working at either an academic, community or private healthcare centers. The geographic boundary of southwestern Ontario is bounded by Lake Huron to the north; the St. Clair River to the west; and Lake Erie to the south; Central Ontario and the Golden Horseshoe to the east. The region had a population of 2.5 million in 2016.¹ We targeted only anesthesiologist working in southwestern Ontario because this was the only feasible sample we could obtain a complete contact of the all potential respondents to avoid sampling errors. We did not survey through the professional society/organization membership email list across Canada and Ontario because their membership email lists only represent a fraction of working anesthesiologists in Canada or Ontario (see Discussion). A 20–40% response rate was expected based on the response rate of the previous electronic survey in Ontario [6].

The survey was first constructed by the investigators, and then modified with the recommendations from a survey methodology expert at Western University, Ontario. We further validated the survey by pilot testing on 5 anesthesiologists and 5 anesthesia residents/fellows to ensure appropriate sentence construction and sufficient data collection. The survey was hosted on Qualtrics survey™ (https://​qsharingeu.​eu.​qualtrics.​com). The online administration of the survey in Qualtrics was tested in different browsers and platforms to ensure smooth conduct.

The survey was composed of 48 questions and was estimated to take approximately 30–45 min to complete. The questions were designed to address five areas: i) participant characteristics, ii) institution characteristics, iii) participants’ practice, iv) barriers of using US (e.g. anesthesiologists’ perceived need, logistics and institutional policy), v) training preferences (e.g. preferred learning method).

Contact information for all Site Chiefs of Anesthesia Departments in southwestern Ontario was obtained from our administrator. We manually telephoned all department executives or site chiefs to introduce the study and requested local promotion of the study. We were successfully able to establish a connection to the executives of 196 anesthesiologists working in 28 institutions in southwestern Ontario, who agreed to both distribute the survey on the teams behalf and promote the study locally. We were further able to contact the local executives of additional 70 anesthesiologists in 7 institutes, who agreed to forward the survey to the working anesthesiologists in their Departments but refused to promote the study locally on behalf of the study team. We sent the survey directly to these 70 anesthesiologists. There were no email addresses or contact information available for 36 anesthesiologists. Thus, we were able to deliver the survey invitation to 266 practicing specialists out of a total of 302 anesthesiologists in our region.

The survey was distributed by e-mail in January 2018, with a subsequent reminder e-mail sent every 2 months thereafter for a total of 3 reminders. To optimize the response rate, we provided an incentive (a draw for an iPad Mini) for completing the survey.

The survey data were stored on the Qualitrics™ host server and downloaded for analysis following completion of data collection. The data were analyzed using Stata® [version 14.0] (StataCorp LLP, College Station, Texas). Descriptive statistics, including mean, standard deviation (SD), frequencies, and percentages, were reported as appropriate to assess the qualitative aspects of the data. No inferential statistical analysis was performed.

A vast majority of respondents (89%) reported “always” or “frequent” uses of US for CVC insertion. The use of US for arterial cannulation was quite variable and was less common than for CVC insertion (Table 3).

Despite the well-established benefit of US-guided regional anesthesia techniques, almost half of the respondents did not routinely use US to perform regional blocks. A small minority of the respondents (18%) reported they never use US for regional nerve blocks. The vast majority (85%) of the respondents never or seldom used US for neuraxial anesthesia. No respondent used US routinely for neuraxial blocks. Half of the respondents never or seldom used US for the assessment of cardiac and pulmonary abnormalities.

The majority (68%) of the respondents agreed that US was used on a regular basis at their institution. The majority of the respondents wanted to use ultrasound guidance more frequently as part of their perioperative care and felt that they already have adequate training to use US regularly. None agreed that they do not need to use ultrasound guidance in their practice. However, more than half (59%) of the respondents reported inadequate access to the US machines on a regular basis (Table 4). The majority (58%) of the institutes did not have an existing policy on the use of US guidance; 29% of respondents were not sure whether there was an existing policy in their working institutes. Only a small proportion (21%) of respondents felt they were always encouraged to use US by the clinical leadership at their institutions.

Some respondents have provided written (free text) feedback in the survey expressing the lack of resources was the major hindrance of US applications in their practise. They were frequently discouraged by the lack of availability of US equipment (e.g. borrowing US machine/probe from the intensive care unit). This general lack of resources also resulted in a lack of momentum within the local anesthesia group to incorporate US into their practice.

All respondents reported they have received some forms of US training in the past. This included during their residency or fellowship training, peer-to-peer learning, participating in single/multiple day workshops, reading textbooks or electronic journals (Table 5). The vast majority of the respondents agreed that US guidance is an important skill for all anesthesiologists and they could see the benefit of US training. Most respondents (72%) would like to receive additional ultrasound training; however only half reported they had planned to acquire additional training in the coming 12 months. All respondents agreed that they would be benefit from training via E-learning modules. For those who did not plan to acquire more training in the coming 12 months, most reported they had sufficient training and some reported there was not enough time or resources available.

The current practice pattern of the US utilization in anesthesia in southwestern Ontario was not entirely surprising; US was more frequently used for CVC insertion and regional nerve block than either arterial cannulation or cardiopulmonary assessment. A previous survey [6] of 203 anesthesiologists conducted in Ontario in 2011, found approximately 30–40% of the anesthesiologists used US routinely for CVC insertion and regional anesthesia. Our survey showed an increase in the adoption of US since the last survey [6]. The NICE guidelines state that “US guidance should be used in most clinical circumstances where CVC insertion is necessary and that all those involved in placing CVCs using US guidance should undertake appropriate training to achieve competence.” [4] Similarly, the use of US in peripheral nerve block have been shown to reduce block performance time, increase block success, improve block quality and allow adequate visualization of surrounding structures, needle and catheter (ASRA guideline: Level Ib evidence) [16]. These benefit have resulted in better patient safety by reducing local anesthetic systemic toxicity (ASRA guideline: Level Ia evidence) and reducing hemi-diaphragmatic paresis (ASRA guideline: Level Ib evidence) [16]. Our current practice pattern appears discordant with the current recommendation.

The adoption rates of US for arterial cannulation, neuraxial blockade and cardiopulmonary assessment were low, which might be related to these being relatively new techniques in Anesthesia, have fewer demonstrated benefits as well as limited recommendations from clinical guidelines. Although US-guided arterial cannulation has been shown to improve first-pass success and reduce the number of attempts [17, 18], the current guidelines [19, 20] do not recommend routine real-time US. Similarly, the use of US in neuraxial blockade shortens the procedural time, improves block success and allows better prediction of epidural depth (ASRA guideline: Level 1b evidence). The current guidelines do not strongly advocate the routine use of US. Point-Of-Care assessment of the heart, lungs and volume status with US has become more frequent among Anesthesiologists. Early evidence suggests that POCUS is associated lower mortality [21] and modifies anesthetic technique in hip fracture surgery [22], and frequently alters management in patients with suspected cardiac disease in preoperative clinic [23].

One major finding of this survey was the lack of US resources is a key barrier of using US in the perioperative period. Contrary to the Canadian Anesthesiologist Society (CAS) practice guideline [5] which stated states “For the placement of central venous catheters, dedicated ultrasound capability must be provided.”, this general lack of US resource that impedes the safe conduct of anesthetic procedures should call attention to the relevant administrative bodies for appropriate resources allocation. However, the lack of US resources does not only occur within the subspecialty of Anesthesia in Ontario. A survey [24] of rural Ontario physicians reported that approximately 40% respondents did not have an ultrasound machine available in their emergency department. Ironically, the uses of US guided CVC insertion [25] and regional nerve block [26] were found to be cost-saving compared with landmark technique, corresponding to saving £ 2000 per 1000 CVC insertion [25] and $47.3 per regional nerve block [26]. Given the rapidly diminishing cost of small hand held ultrasound machines economics should not be a factor in the acquisition of ultrasound devices.

Our study was limited by the low response rate but comparable to the response rates in other published electronic surveys in the Canadian anesthesia population [6]. We have attempted to use a more focused survey in a small geographic region with efforts to make personal contact, to improve the response rate; however, this had limited success. As with other surveys, our results suffer from non-responder bias, the non-responders of the survey were likely the one who had less interest in US.

One possible strength of this study was the known denominator of all working anesthesiologists in southwestern Ontario for this survey. Compared to the methodology of the previous survey that was conducted based on the professional society available contact information, the sample might not truly represent the target population. We did not survey through the CAS or Ontario Medical Association membership email list across Canada and Ontario because their membership email lists only represent a fraction of working anesthesiologists in Canada or Ontario. For instance, there were approximately 2600 members of Canadian Anesthesiology Society at 2018 but there was 3744 Anesthesia Specialist registration at the Royal College of Canada Directory in the same year. We did not use the email list on the College of Physicians and Surgeons of Ontario because many community anesthesiologists did not list their personal email contact information in the public domain, resulting in a potential biased and incomplete sample. Arguably, the results of our survey may not necessarily be generalizable to the all other geographic regions or the whole country; however, it may provide a reasonable insight to the regions with similar resources and settings.