We conducted a descriptive, prospective, non-randomized and single-center study of consecutive patients undergoing TEE in our Cardiac Imaging Unit from September 2018 to September 2019. All patients were aged>18 years and signed the informed consents. The study was conducted in accordance with the principles of the Declaration of Helsinki. The study protocol was accepted as a protocol assistance in the Quality Care Program of our hospital, and has not required approval by the Ethics Committee.

Prior to its start, an analysis of the workspace and the necessary material to carry out the program was made. Following international guidelines for procedural sedation and analgesia in adults,8 the training focused on:

• 1.

  Assessment of the airway and comorbidities that could affect sedation.

• 2.

  Competence in recognizing the different levels of sedation.

• 3.

  Knowledge of the most commonly used drugs in sedation and advanced life support training with special focus on advanced airway management.

• 4.

  Competence in recognizing total recovery of consciousness required to leave the monitoring area.

Finally, practical skills were acquired through supervised live cases. Training program lasted for 1 week.

The baseline characteristics of the patients and the presence of a priori pathologies associated with events during sedation, such as underlying lung disease (asthma, chronic obstructive pulmonary disease or sleep-apnea) or severe valvular disease were recorded. Adverse events throughout the sedation procedure until recovery were recorded on a sedation form. These events were classified as respiratory (pulse oximetry oxygen saturation<90%, need for endotracheal intubation [EI]), hemodynamic (hypotension<90mmHg, need for cardiopulmonary resuscitation), arrhythmic (bradycardia<40bpm or sustained new tachyarrhythmia), digestive (non-self-limited digestive bleeding) and neurological events (new-onset focal neurological deficits). The monitoring of the procedure and its complications were recorded in a sedation in TEE register for publication in the patient's medical history.

With the objective of planning sedation, this program designed a decision-making algorithm to perform a pre-anesthetic evaluation. AP sedation was performed if any of the following criteria was present. Otherwise, sedation was performed by NAP:

• -

  Age>70 years-old

• -

  Morbid obesity with BMI>40kg/m2

• -

  Left ventricle ejection fraction<30%

• -

  Chronic liver disease Child C

• -

  Complex studies in American Society of Anesthesiology Physical Status Classification (ASA) III or ASA IV

• -

  Chronic lung disease (sleep-apnea, chronic obstructive pulmonary disease or asthma)

• -

  Chronic kidney disease with estimated glomerular filtration rate<30ml/min

For the sedation procedure the use of propofol was prioritized. The target was to keep patients in a moderate degree of sedation, also called conscious sedation, which enables spontaneous ventilation with preserved cardiovascular function.9 In all cases of NAP sedation an anesthesiologist was previously informed and available to provide support or switch to AP sedation if needed.

After the TEE, patients were transferred to a recovery area. The modified Aldrete10 post-anesthetic recovery score was used to authorize discharge, considering a score>9/10 points in hospitalized patients and>17/18 points in outpatients to finish the procedure. The time to reach the adequate Aldrete score was recorded as recovery time.

Categorical variables are expressed as absolute frequencies and percentages. Quantitative variables are expressed as mean±interquartile range [IQR, 25–75%]. Quantitative variables distribution assessed using the Shapiro–Wilk test showed a significant violation of normality assumption. To compare variables between the 2 groups, Chi-square and Fisher test were performed for categorical variables and Wilcoxon test for quantitative variables.

A logistic regression model was performed to explain the independent association between events, sedation group and ASA grade. Selected multivariate model had moderate discrimination (area under the curve 0.62) and moderate calibration (P value for the Hosmer–Lemsehow test=.12). The results are expressed as odds ratios (OR), with their 95% confidence intervals (95%CI). In all hypothesis tests, the null hypothesis was rejected with a type I error or alpha error < 0.05. Stata 12.1 software was used for all analysis.

There was a crossover of 56 patients between the two initially defined groups, probably due to errors in the functioning of the circuit in its first year. Characteristics of the discordant patients are detailed in Table 2. Safety in the group of 56 discordant patients, and particularly in the subgroup of 28 patients assigned to AP sedation but sedated by NAP was tested. The global prevalence of complications in the 56-patients discordant group was 3%, whereas the prevalence of complications in the discordant subgroup sedated by NAP was 1.87%. We did not detect a significant increase in complications in misclassified patients.

The results of this study show the expansion of TEE in cardiac imaging units, with a new profile of patients with advanced heart disease directed to percutaneous or surgical intervention. This may require the reorganization of the units to respond to new demands, including the establishment of collaboration circuits between medical specialties as well as the incorporation of cardiac imaging specialists trained in the interventionist field. Also, our safety data support the expansion of propofol sedation in TEE procedures performed by NAPs in selected patients.

Our work is a registry of patients included in a pilot sedation program with a selection algorithm that results in two clearly differentiated groups. This may limit the comparability of results between the 2 groups. However, safety data, events and recovery time are similar to previously published work.

Moreover, a standardized definition and classification of adverse events has not been used, thus limiting the comparison with other studies.