Correlation of Vital Signs and Depth of Sedation by Modified Observer’s Assessment of Alertness and Sedation (MOAA/S) Scale in Bronchoscopy

Abstract

Background: 

With complex, lengthy bronchoscopies, there is a need for safe, effective sedation. Most bronchoscopists strive for moderate sedation, though often difficult without compromising vital signs. The Modified Observer’s Assessment of Alertness and Sedation (MOAA/S) scale is a validated 6-point scale assessing responsiveness of patients coinciding with the American Society of Anesthesiologists (ASA) continuum of sedation. It is commonly used in studying bronchoscopic sedation, but depth of sedation by MOAA/S and correlation with vital signs and adverse events has not been determined.

Methods: 

This study was a post hoc analysis of a prospective, double-blind, randomized trial evaluating the safety and efficacy of remimazolam. MOAA/S and corresponding vital signs were used to assess the effect of the level of sedation on vital signs and adverse events.

Results: 

A total of 23,341 MOAA/S scores from 431 patients were recorded. Older and higher ASA class patients spent more time in deeper sedation (MOAA/S 0 to 1) (6% vs. 2%, P=0.01). Oxygen saturation was equal in deep sedation (MOAA/S 0 to 1) (97±3%) compared with moderate sedation (96±3%) (P=0.11). Mean systolic and diastolic blood pressures were significantly lower when comparing MOAA/S 0 to 1 to MOAA/S 5 (systolic blood pressure: 126±19 vs. 147±24 mm Hg, P<0.01; diastolic blood pressure: 68±14 vs. 84±15 mm Hg, P<0.01). There was a nonsignificant trend towards lower heart rate at deep versus moderate sedation (84±15 vs. 94±18 beats/min, P=0.07). Respiratory rate was also comparable with moderate and deep sedation (17±5 vs. 18±6 beats/min, P=0.94).

Conclusion: 

There was no clinically meaningful correlation between vital signs and depth of sedation assessed by MOAA/S. Older and higher ASA class patients spend more time in deeper sedation. However, when in deep sedation, there was no difference in vital signs other than a slightly increased incidence of clinically insignificant hypotension.

Bronchoscopy is frequently utilized for diagnostic and therapeutic pulmonary interventions. As procedures become more complex and lengthy, bronchoscopists need to provide safe and effective sedation which maintains favorable conditions for a successful procedure. Because sedation is a continuum, it can be difficult to determine the exact depth of sedation during procedures.^1,2 As outlined by the American Society of Anesthesiologists (ASA), sedation includes 4 stages: minimal, moderate, deep, and general anesthesia.³ These stages are defined based on the responsiveness of the patient to verbal or tactile stimulation, which is often difficult to truly ascertain. For example, a key difference between moderate and deep sedation is “purposeful response to verbal or tactile stimulation” as opposed to purposeful response after “repeated or painful stimulation.”³ This distinction is paramount given that deep sedation involves spontaneous ventilation that “may be inadequate” to maintain cardiopulmonary physiology and interventions for airway protection may be required.³ The goal for most bronchoscopists is to achieve and maintain moderate sedation, which is often difficult because patients drift between different depths of sedation throughout the procedure. Therefore, the challenge is to achieve an adequate amount of sedation during the procedure without compromising the patient’s airway, hemodynamics, and oxygenation.¹

The Modified Observer’s Assessment of Alertness and Sedation (MOAA/S) scale is a 6-point scale that assesses the responsiveness of patients and coincides with the ASA continuum of sedation. The MOAA/S scale is a validated efficacy scale commonly used in studying sedation in bronchoscopy.^4,5 A patient with a score of 5 responds readily to name spoken in a normal tone, 4 lethargic response to name spoken in a normal tone, 3 responds only after the name is called loudly and/or repeatedly, 2 responds only after mild prodding or shaking, 1 responds only after painful trapezius squeeze, and 0 does not respond to painful trapezius squeeze (0 to 1 deep sedation, 2 to 4 moderate sedation, and 5 minimal sedation or alert, Table 1).^4–6

TABLE 1 - MOAA/S Scale

MOAA/S Scale	Scale	ASA Classification
Responds readily to name spoken in normal tone	5 (alert)	Minimal
Lethargic response to name spoken in normal tone	4	Moderate
Responds only after name is called loudly and/or repeatedly	3	Moderate
Responds only after mild prodding or shaking	2	Moderate
Responds only after painful trapezius squeeze	1	Deep
Does not respond to painful trapezius squeeze	0	Deep/general anesthesia

ASA indicates American Society of Anesthesiologists; MOAA/S, Modified Observer’s Assessment of Alertness and Sedation.

Vital signs are also widely accepted as clinically meaningful markers when evaluating the level of sedation, particularly as harbingers of adverse events (AEs).⁷ Therefore, bronchoscopists can use vital signs, along with patient responsiveness, as indicators for the depth of sedation. However, it is unclear how the 2 correlate, specifically how the MOAA/S scores actually correlate with vital signs. The primary goal of this study was to assess the effect of deeper sedation (ie, lower MOAA/S scores) on vital signs and AEs during sedation for bronchoscopy. A secondary goal was to determine the correlation of MOAA/S and vital signs in at-risk groups, which include age below 65 versus 65 years and older and ASA class I to II versus III.

METHODS

Vital signs and MOAA/S levels of sedation were obtained from 431 patients in a secondary analysis of a prospective, multicenter, randomized, double-blind study comparing the efficacy of remimazolam to placebo and midazolam for bronchoscopy.⁸ The study protocol has been previously described.⁸

For sedation, there were a total of 3 arms combining fentanyl with either remimazolam, placebo, or an open-label midazolam arm. The latter was dosed strictly according to the US prescribers’ information. If sedation was insufficient to begin the bronchoscopy procedure (MOAA/S>3), patients were declared a “treatment failure” and were given rescue sedative medication. Midazolam (dosed per physician discretion) was the only rescue sedative medication permitted in conjunction with the use of fentanyl solely for analgesia. Medications were administered by dedicated and trained staff under a pulmonologist’s supervision without anesthesia personnel. All sites obtained institutional review board approval. Male and female patients and age 18 years and above with an ASA score 1 to 3 were included though those with ASA class IV were excluded. The dosing protocol is listed in detail in a prior publication.⁸ Enrollment and allocation of patients is shown in Figure 1, which is listed in the primary paper.⁸

FIGURE 1: CONSORT flow diagram of enrollment as listed in the primary paper.⁸ IMP indicates investigational medicinal product.

On the day of their procedure, patients had baseline vital signs recorded as well as the serial recording of baseline MOAA/S scores. The administration of supplemental oxygen 4 L/min began shortly before the procedure and continued until the patient was fully alert. The MOAA/S scores were determined by values documented in the case report forms. The procedure was initiated once the patient had a MOAA/S score of 3 using a double-blinded sedation protocol. The MOAA/S scores were recorded at 1, 1.5, 2, 2.5, and 3 minutes after medication injection.⁸ After 3 minutes, the MOAA/S scores were measured in 1-minute intervals until the patient was fully alert, which was defined as the first of 3 consecutive MOAA/S scores of 5. Vital signs were recorded within 5 hours, 30 minutes, 15 minutes, and 1 minute predose, and then at least every 5 minutes postdose.⁸ Each MOAA/S score was matched with a corresponding vital sign measured within the respective full minute. In case of multiple vital sign measurements within the same full minute, the first recorded observation was used. The same approach for multiple MOAA/S scores was applied. As the amount of missing data was relatively small, no imputation was performed and only vital sign measures with missing covariates were omitted.

AEs, including hypoxia, hypotension, hypertension, bradypnea, and bradycardia, were recorded and listed with their corresponding MOAA/S scores. Bradycardia was defined as <40 beats/min or a decrease in heart rate of >20% from baseline that lasts continuously for at least 30 seconds. Hypertension consisted of an increase in systolic blood pressure (SBP) to ≥180 mm Hg and diastolic blood pressure (DBP) to ≥100 mm Hg, or an increase of blood pressure of 20% over baseline. Hypotension was a fall in SBP to ≤80 mm Hg or a fall in DBP to ≤40 mm Hg, or as a fall in blood pressure of 20% below baseline. Bradypnea was defined as <8 breaths/min, and hypoxia was listed as oxygen saturation (SpO₂) <90% for ≥1 minute.

Each of the aforementioned AEs was matched with the MOAA/S observed at the start of the respective AE. On the basis of the level of the lowest MOAA/S score, the subject was counted within one or another category (Table 2). In case 2 or more of the same AE occurred within the same subject, the subject was counted only once, applying the worst-case scenario (the subject was counted for the lowest MOAA/S score category). Incidence was calculated by dividing the number of subjects with AEs with the number of subjects with the same lowest MOAA/S score. Results of this analysis will be presented only descriptively.

TABLE 2 - Incidence of AEs by Lowest MOAA/S Score

	n (%)
AE	MOAA/S 0 (N=21)	MOAA/S 1 (N=23)	MOAA/S 2 (N=95)	MOAA/S 3 (N=377)	MOAA/S 4 (N=355)	MOAA/S 5 (N=236)
Bradycardia	2 (9.52)	1 (4.35)	6 (6.32)	10 (2.65)	0	0
Hypertension	6 (28.57)	8 (34.78)	45 (47.37)	65 (17.24)	2 (0.56)	0
Hypotension	17 (80.95)	9 (39.13)	44 (46.32)	78 (20.69)	3 (0.85)	1 (0.42)
Hypoxia	4 (19.05)	4 (17.39)	21 (22.11)	63 (16.71)	0	0
Respiratory rate decreased	1 (4.76)	1 (4.35)	8 (8.42)	2 (0.53)	0	0

AE indicates adverse event; MOAA/S, Modified Observer’s Assessment of Alertness and Sedation; N, total number of subjects experiencing a MOAA/S.

Mean, SD, median, minimum and maximum of each of the vitals were analyzed by descriptive statistics overall and grouped by MOAA/S score. The difference between vital signs during MOAA/S 0 to 1 episodes versus MOAA/S 5 was tested using the Kruskal-Wallis rank-sum test at a significance level of P-value <0.05. Association between the incidence of deep sedation and ASA class or age was tested by the “N−1” χ² test as recommended by Campbell and Richardson.^9,10

For analysis of the relationship between MOAA/S score and vital sign measurement, a complex statistical model was built up, using SAS 9.2, where the MOAA/S score was considered as categorical. The correlation analysis was carried out from the start of the procedure to end of the procedure (bronchoscope in to bronchoscope out). The effect of the MOAA/S was controlled by age, sex, ASA, treatment arm, time from first investigational medicinal product dose, and procedure duration.

The univariate linear mixed model with normal distribution was applied using the mixed procedure of SAS 9.2. The model goodness-of-fit was checked by q-q plot and histogram of residuals and scatter plot of residuals versus predict mean. The residuals were studentized to improve their interpretation. The marginal studentized residual was chosen to reveal the relation of vital signs and MOAA/S in the overall study population, instead of the relation within a particular subject (reference: https://support.sas.com/resources/papers/proceedings/proceedings/sugi29/189-29.pdf)

As the residuals of vital sign measures were not normally distributed, log transformation of data was applied. The log-transformed vitals showed an independent and homoscedastic normal distribution for the residuals of all variables except SpO₂. Thus, SpO₂ data was modified to range from 0 to 1 and then modeled using a mixed model with beta distribution and logit link function. For convenience, the transformed SpO₂ values will be continuously called SpO₂.

The covariance structure of vital-sign-measures’-sequence was chosen based on Akaike Information Criterion, to ensure modeling the repeated component-time by subject level. A mixed model, including all controlled variables, was used. The structures tested for log(SYSBP), log(DIASBP), log(HR), and log(Resp) were: first-order Autoregressive AR(1), first-order autoregressive moving-average ARMA(1,1), and Toeplitz with Two Bands TOEP(2). ARMA(1,1) had the best fit for all vital sign models and was further used for the final results. In addition, the intercept of each subject was considered as a random effect with variance components covariance structure.

Exceptionally, the model for SpO₂ as measured by pulse oximetry (SpO₂) did not converge to TOEP(2) and ARMA(1,1) covariance structures. It reached convergence to AR(1).

The backward variable selection was performed to fit the final reduced model using α=5%. The backward selection was, however, stopped at the time MOAA/S should have been removed, to maintain a consistent presentation of the results.

RESULTS

A total of 431 patients participated in this study. The mean age was 62 years and nearly half (209) of patients were 65 years and older. Approximately 63% of patients were ASA class I to II, and 37% were ASA class III. All patients were assigned MOAA/S scores at specific time intervals throughout their procedure. Most patients (83%) spent their time moderately sedated (MOAA/S 2 to 4) compared with 13% awake/minimal sedation (MOAA/S 5) or 4% deeply sedated (MOAA/S 0 to 1).

Comparison of At-risk Groups

Patients were further categorized and analyzed based on subclasses of age (below 65 and 65 y and older) and ASA class (I to II vs. III) with their corresponding MOAA/S scores. Thirty-seven percent of patients were ASA class III and spent more time in deep sedation (6%) than patients in ASA classes I to II (2%) (P=0.01). There was no difference in the amount of time in deep sedation (3% vs. 4%) in those below 65 versus 65 years and above (P=0.33). However, patients above 65 years and ASA III (20% of the total cohort), spent 4 times the amount of time in deep sedation compared with younger and ASA I to II patients (P<0.01).

AEs During Sedation

AEs that occurred during the procedure were paired with the MOAA/S scores at the time the event occurred (Table 2). Each event listed was matched with the closest MOAA/S score at the time of the start of the event. Patients spent most of their time in moderate sedation (91%), and most patients experienced AEs while they were under moderate sedation. Of the 431 subjects, 16 had bradycardia at MOAA/S 2 to 3, and only 3 of them experienced the same AE at MOAA/S 0 to 1. Similar distributions were seen in cases of hypertension, hypotension, and hypoxia (110 vs. 14; 122 vs. 26 and 84 vs. 8, respectively).

Comparing Vital Signs and Depth of Sedation

The mean procedure time was 12.3 minutes (SD: 11.5 min). A total of 5560 MOAA/S scores were recorded. Overall, ∼8% of the time was spent in deep sedation compared with 4% of the entire time period, which included significant postprocedural time. SpO₂ was equal in deep sedation (MOAA/S 0 to 1) (97±3%) compared with minimal sedation (MOAA/S 5) (96±3%) (P=0.11). SBP and DBP had a greater range of change when comparing MOAA/S 0 to 1 to MOAA/S 5 (SBP: 126±19 vs. 147±24 mm Hg, P<0.01; DBP: 68±14 vs. 84±15 mm Hg, P<0.01). There was a trend towards lower heart rate at deep versus moderate sedation (84±15 vs. 94±18 beats/min, respectively) (P=0.07) though it did not reach statistical significance. The respiratory rate was also comparable with minimal and deep sedation using this time frame (17±5 vs. 18±6 beats/min, respectively) (P=0.94).

Patients age below 65 years spent 6% of the time in deep sedation compared with 9% in the patient 65 years and older (P=0.18). ASA class III spent 12% of the time in deep sedation compared with 5% in ASA class I to II (P=0.04). Patients 65 years and older and ASA class III spent 12% in deep sedation compared with 4% in age below 65 years and ASA class I to II (P=0.02).

Correlation of MOAA/S and Vital Signs

Using the mixed model, the correlation between the vital sings and the MOAA/S scores was tested for the time interval of the start of the procedure to the end of the procedure. During the procedure, the impact of MOAA/S score on vital signs was significant only in the case of heart rate (P=0.046) and SBP (P=0.001). None of the other vitals were significantly impacted by changes in the MOAA/S (Table 3). Even in these 2 cases where vital signs were significantly impacted, a statistically significant difference (in comparison to vital signs measured under MOAA/S 5) was detected only in case of heart rate, and only when subjects had MOAA/S 0 (P=0.022). Estimates, SEs, 95% confidence intervals, and corresponding P-values are provided (Table 3).

TABLE 3 - Correlation Between VS and MOAA/S for Start of Procedure to End of Procedure

VS	MOAA/S	Estimate	SE	95% CI	P†	P LMM‡
Systolic blood pressure*	0	−0.080	0.04325	−0.165, 0.005	0.0657	0.001
	1	−0.035	0.04308	−0.120, 0.049	0.4146
	2	−0.025	0.03897	−0.101, 0.052	0.5295
	3	0.009	0.0381	−0.065, 0.084	0.8049
	4	0.002	0.03807	−0.072, 0.077	0.9477
	5	Reference	—	—	—
Diastolic blood pressure*	0	−0.095	0.05606	−0.205, 0.015	0.089	0.153
	1	−0.048	0.05635	−0.159, 0.063	0.395
	2	−0.038	0.05058	−0.137, 0.062	0.4556
	3	−0.027	0.04919	−0.124, 0.070	0.5837
	4	−0.016	0.04932	−0.113, 0.081	0.7448
	5	Reference	—	—	—
Heart rate*	0	−0.090	0.03926	−0.167, −0.013	0.0222	0.046
	1	−0.040	0.03889	−0.116, 0.037	0.3066
	2	−0.056	0.0355	−0.125, 0.014	0.1166
	3	−0.052	0.03472	−0.120, 0.017	0.1384
	4	−0.038	0.03471	−0.106, 0.030	0.2705
	5	Reference	—	—	—
Respiratory rate*	0	−0.028	0.0835	−0.192, 0.136	0.7395	0.285
	1	−0.042	0.08337	−0.206, 0.122	0.6145
	2	−0.049	0.07616	−0.199, 0.100	0.5193
	3	−0.017	0.07455	−0.163, 0.130	0.8232
	4	−0.052	0.07452	−0.198, 0.095	0.4878
	5	Reference	—	—	—
SpO2	0	0.140	0.2913	−0.434, 0.713	0.6325	0.156
	1	0.017	0.2898	−0.555, 0.588	0.9546
	2	0.155	0.2624	−0.362, 0.672	0.5549
	3	0.314	0.2557	−0.190, 0.817	0.2214
	4	0.309	0.256	−0.195, 0.814	0.2283
	5	Reference	—	—	—

Log transformation of the data was needed to test the correlation between MOAA/S and vitals, as the vitals did not show normal distribution.

^*Log-transformed vitals.

^†Estimates and P-values of the reduced model (using vital signs measured under MOAA/S=5 as reference).

^‡P-value resulted from linear mixed model (type 3 tests of fixed effects) reduced model with backward selection.

CI indicates confidence interval; LMM, linear mixed model; MOAA/S, Modified Observer’s Assessment of Alertness and Sedation; SpO₂, oxygen saturation; VS, vital signs.

DISCUSSION

This is the first study we are aware of that evaluates the depth of sedation using a combination of benzodiazepines (midazolam and remimazolam) and/or an opioid analgesic (fentanyl) by MOAA/S score and correlates with vital signs and AEs during bronchoscopy. Overall, this study has 4 important findings. First, there were no clinically significant changes in vital signs with increasing depths of sedation. Second, there was also no correlation between clinically significant AEs and depth of sedation. Third, while older (age above 65 y) and sicker (ASA class III) patients spent more time in deep sedation, there was no corresponding increase in the overall incidence of AEs. Fourth, no matter what the depth of sedation is, bronchoscopy with moderate sedation remains an extremely safe procedure.

Bronchoscopists are constantly challenged when targeting an ideal level of moderate sedation. Too little sedation can cause discomfort and potentially a longer, more painful procedure for the patient, which could also affect the clinicians’ ability to make a diagnosis. When patients are oversedated, their cardiopulmonary status can be compromised, and further complications can result.¹ Therefore, it is imperative to ensure patients receive adequate sedation while also assuring patient safety. The MOAA/S scale has been utilized as a reliable measure in numerous previous studies to assess the depth of sedation.^4–6 In this study, we tested whether the MOAA/S scale correlates with vital signs and whether it is a reliable predictor for adverse effects.

Vital signs during moderate sedation can alert bronchoscopists of patient discomfort and potential AEs. However, this study did not show a correlation with vital signs and depth of sedation, except for blood pressure, which, while significantly lower in deep sedation, was not clinically significant. Current guidelines, however, suggest hemodynamic and SpO₂ monitoring during sedation can serve as harbingers of AEs related to sedation.¹¹ With these results, bronchoscopists should be mindful that stable vital signs should not imply a minimal level of sedation, and future investigation should include adjunctive maneuvers during bronchoscopy to reliably measure the depth of sedation.

This study also showed there is no correlation with deeper sedation and vital signs when comparing at-risk groups. When evaluating subclasses of patients based on age and ASA class, older and higher ASA class patients did spend more time in deep sedation. ASA class III had a greater influence on deeper sedation than older age. Both older age and higher ASA class had increased time in deep sedation compared with younger and lower ASA class patients. There are several considerations given this association. Patients with more comorbidities, such as chronic kidney disease or liver dysfunction, may have difficulty with clearance of the benzodiazepines and opioids that are given. Older patients are known to have a greater sensitivity to sedatives and many sedatives reflect this in their dosing recommendations.¹² Ultimately, bronchoscopists should be aware that at-risk patients are more susceptible to deep sedation and dose adjustment should be considered in this population.

At-risk groups spent more time in deep sedation when given similar doses of sedation compared with younger and healthier patients. However, when examining the entire cohort of patients, there was not an increase in AEs except for hypotension. This finding contrasted from previous studies in the gastroenterology literature, which showed older and higher ASA class patients had a greater risk of sedation-related AEs.¹³ Our findings illustrate that although this at-risk group spends more time in deep sedation, they do not necessarily have increased AEs other than a modest increase in hypotension.

While MOAA/S is designed for determining the depth of sedation, the scale has limitations. The scale did not reliably differentiate deep sedation from general anesthesia. Once a patient has reached a MOAA/S score of 0, it is difficult to determine whether the patient is still in deep sedation sufficient to blunt reactions to a rather small stimulus, such as a biopsy, or has already transitioned into full general anesthesia, which would allow for more intense surgical interventions. Some have suggested that an extended MOAA/S scale be established to help discriminate between these 2 levels.⁶

This study, in particular, has several limitations: (1) MOAA/S scores and vital signs were not recorded strictly at the same time intervals and thus some scores were correlated with a specific vital sign based on proximity of time (though they were close). (2) ASA class IV patients were not included in this study, and future studies should evaluate this population which is considered particularly vulnerable to AEs. (3) The majority of vital signs were taken in moderate sedation with <10% recorded during deep sedation. (4) Even though advanced diagnostic procedures were performed with procedure times up to 68 minutes, shorter procedures were more represented in this study. Finally, the results here can only be generalized to patients receiving opioids and benzodiazepines, and the results may not apply when using a different class of sedatives (ie, propofol) where vitals may not be as stable at deeper levels of sedation.

In conclusion, there was no clinically meaningful correlation between vital signs and depth of sedation as assessed by MOAA/S scores. Thus, vital signs should not be used as the exclusive indicator for depth of sedation. Although older and sicker patients spent more time in deep sedation, there was no corresponding increase in the overall incidence of AEs other than hypotension, which did not adversely affect clinical outcomes. Therefore, bronchoscopy remains an extremely safe procedure at any level of sedation.