Intubation Setting, Aspiration, and Ventilator-Associated Conditions

Abstract

Background

Patients experience endotracheal intubation in various settings with wide-ranging risks for postintubation complications such as aspiration and ventilator-associated conditions.

Objectives

To evaluate associations between intubation setting, presence of aspiration biomarkers, and clinical outcomes.

Methods

This study is a subanalysis of data from the NO-ASPIRATE single-blinded randomized clinical trial. Data were prospectively collected for 513 adult patients intubated within 24 hours of enrollment. Patients with documented aspiration events at intubation were excluded. In the NO-ASPIRATE trial, intervention patients received enhanced oropharyngeal suctioning every 4 hours and control patients received sham suctioning. Tracheal specimens for α-amylase and pepsin tests were collected upon enrollment. Primary outcomes were ventilator hours, lengths of stay, and rates of ventilator-associated conditions.

Results

Of the baseline tracheal specimens, 76.4% were positive for α-amylase and 33.1% were positive for pepsin. Proportions of positive tracheal α-amylase and pepsin tests did not differ significantly between intubation locations (study hospital, transfer from other hospital, or field intubation). No differences were found for ventilator hours or lengths of stay. Patients intubated at another hospital and transferred had significantly higher ventilator-associated condition rates than did those intubated at the study hospital (P = .02). Ventilator-associated condition rates did not differ significantly between patients intubated in the field and patients in other groups.

Conclusions

Higher ventilator-associated condition rates associated with interhospital transfer may be related to movement from bed, vehicle loading and unloading, and transport vehicle vibrations. Airway assessment and care may also be suboptimal in the transport environment.

Oral endotracheal intubation is performed in a variety of settings with different airway access complexities. Operating room intubations are usually more controlled, whereas emergency department and prehospital (field) intubations are more urgent and less controlled. Environmental conditions, lighting, and the working space further complicate oral endotracheal intubation and airway management in the prehospital and interfacility transfer environments.^1–3

Oral endotracheal intubation is not without risks, with aspiration being one of the most common risks.^4–7 Clinical complications associated with aspiration include ventilator-associated conditions (VACs) and ventilator-associated pneumonia, which are leading causes of increased morbidity and mortality in critically ill patients.^4,8,9 Detection of aspiration events is often difficult. One emerging technique is analyzing tracheal secretions for α-amylase (an oral digestive enzyme) and pepsin A (a gastric digestive enzyme).^10–13 The presence of these enzymes may reflect aspiration events in which microbes, acids, and other oral and gastric contents enter the lungs.

Studies evaluating associations between tracheal α-amylase or pepsin levels and clinical outcomes following prehospital intubation are limited,⁴ and no published studies have examined these biomarkers in patients experiencing interfacility transport. Differences in urgency and intubation circumstances of the prehospital setting and the environmental conditions associated with interfacility transport may put patients intubated in these settings at greater risk for aspiration and aspiration-associated complications. The purpose of this study was to evaluate the association between the presence of aspiration biomarkers and clinical outcomes in patients intubated in different settings.

Methods

This study is a subanalysis using data collected for the NO-ASPIRATE (Oral Suction Intervention to Reduce Aspiration and Ventilator Events) single-blinded randomized controlled trial (NCT02284178). A truncated version of the full methods, containing only elements relevant for this study, is presented here. Please refer to the published protocol for further details on the study methods.¹⁴

Setting and Patients

The primary study facility is an 800-bed community hospital in central Florida. It serves the region as a level I trauma center. It also provides medical education residencies and fellowships.

A power analysis computed for the full randomized controlled trial produced a final sample size of 400 patients necessary to detect differences in mean values of α-amylase with an effect size (Cohen d) of 0.25, α of .05, and power of 0.80. After receiving institutional review board approval from all appropriate entities, we obtained consent to participate from each patient or their legally authorized representative.

A convenience sample of 513 eligible patients was randomized with a block randomization technique to either the intervention (NO-ASPIRATE) group or the control group. For this subanalysis, patients were eligible to participate if they were at least 18 years of age, had received oral endotracheal tube (ETT) intubation with mechanical ventilation, and were intubated for 24 hours or less. Exclusion criteria were documented aspiration at the time of intubation, intubation to treat known aspiration, treatment with rescue mechanical ventilation therapies (high-frequency oscillator ventilation or extra-corporeal membrane oxygenation), reintubation for any reason, contraindications to receiving the intervention (eg, oral injuries), history of lung or head/neck cancers that might produce α-amylase in the lungs, history of disease that affects saliva production (eg, Sjögren syndrome), and being a prisoner. Although 410 patients completed the NO-ASPIRATE trial, all 513 enrolled patients were eligible for this subanalysis because it used only initial tracheal enzyme concentrations obtained at study enrollment.

Data and Specimen Collection

Members of the research team prospectively abstracted data from electronic medical records. Demographic data included age, sex, race, and ethnicity. Intubation-specific data included location, ETT size, reason, and urgency. The primary diagnosis was classified as being medical, surgical, neurological, or trauma. Acute Physiology and Chronic Health Evaluation II scores were recorded for all patients. The injury severity score was recorded for all trauma patients.

Trained research assistants obtained tracheal specimens for α-amylase testing at study enrollment, which was within 24 hours of intubation. Midway through enrollment, measurements of pepsin were added to obtain data related to aspiration of gastric contents (n = 293). Following standardized oral care, the research assistants suctioned each patient’s ETT per procedure to retrieve tracheal secretions into a specimen trap. All tracheal specimens were stored at −20 °C until assays were performed. Although tracheal specimens for α-amylase and pepsin evaluation were obtained every 12 hours during study enrollment, results from the initial specimen were used for this study. The initial specimen was considered the best indicator of events at the intubation location. Study data were collected and managed with REDCap electronic data capture tools.¹⁵

All specimens were analyzed following standard protocols by laboratory personnel who were blinded to study group. Tracheal α-amylase values were based on an assay measuring the absorbance at 405 nm (ELx808IU absorbance reader, BioTek). Linear increase of the absorbance at 405 nm is largely dependent on the presence of active α-amylase in the clinical sample. α-Amylase values greater than 396 U/L were considered positive. The pepsin enzymatic assay was modified according to the method developed by Krishnan et al.¹⁶ The assay was performed in a 96-well plate, and the plate was read in a spectrofluorometer at excitation (485 nm) and emission (530 nm) (Synergy H1 hybrid reader, BioTek). The final pepsin concentration of the sample (ng/mL) was determined on the basis of the net fluorescent intensity of the known concentrations of the standards. Detected pepsin was considered a positive specimen.

Outcomes

The primary outcomes were tracheal α-amylase and pepsin levels obtained from the initial tracheal suctioning event. Specimens were coded as being positive or negative. Secondary outcomes were ventilator hours, length of intensive care unit stay, length of hospital stay, and VACs. To determine VACs, we used Centers for Disease Control and Prevention criteria: at least 2 calendar days of stable or decreasing daily minimum positive end-expiratory pressure or fraction of inspired oxygen followed by a rise in positive end-expiratory pressure of at least 3 cm H₂O or a rise in fraction of inspired oxygen of at least 20 percentage points, sustained for at least 2 days.^17,18

Statistical Analysis

Demographic, intubation, and clinical data were summarized as proportions for categorical variables and as means and SDs for continuous variables. Categorical variable comparisons between groups were made with the χ² test with a Bonferroni adjustment for multiple comparisons. One-way analysis of variance with Bonferroni post hoc analysis or the Kruskal-Wallis test was used to compare continuous-level variables. A P value of .017 or less was required for significance with Bonferroni adjustments.

Results

A summary and a comparison of demographic data and intubation variables are shown in Table 1. The study population was mostly male and white. However, African American and Hispanic patients also were represented. Medical conditions were the most common primary diagnosis, but neurological and trauma conditions also were prevalent. Most patients were urgently intubated for airway protection.

Table 1.

Demographics, intubation location, and clinical summary

Variablea	Total Sample (n = 513)	Study hospital (n = 426)	Transfer (n = 46)	Field (n = 41)	P
Age, mean (SD), y	58.77 (18.83)	59.14 (18.97)	60.22 (14.61)	53.22 (20.99)	.22
Sex					.004
Female	217 (42.3)	194 (45.5)	13 (28)	10 (24)b
Male	296 (57.7)	232 (54.5)	33 (72)	31 (76)b
Race					.36
African American	104 (20.3)	90 (21.1)	10 (22)	4 (10)
White	382 (74.5)	312 (73.2)	34 (74)	36 (88)
Other	27 (5.3)	24 (5.6)	2 (4)	1 (2)
Ethnicity					.10
Hispanic	95 (18.5)	86 (20.2)	5 (11)	4 (10)
Non-Hispanic	418 (81.5)	340 (79.8)	41 (89)	37 (90)
Endotracheal tube size, mm					<.001
7.0	40 (7.8)	25 (5.9)	4 (9)	11 (27)b
7.5	221 (43.1)	195 (45.8)	14 (30)	12 (29)
8.0	251 (48.9)	205 (48.1)	28 (61)	18 (44)
8.5	1 (0.2)	1 (0.2)	0 (0)	0 (0)
Urgency					<.001
Routine	53 (10.3)	52 (12.2)	1 (2)	0 (0)
Urgent	413 (80.5)	359 (84.3)	40 (87)	14 (34)b
Emergency	47 (9.2)	15 (3.5)	5 (11)	27 (66)b
Reason					<.001
Airway protection	262 (51.1)	205 (48.1)	32 (70)b	25 (61)
Respiratory distress	156 (30.4)	146 (34.3)	8 (17)	2 (5)b
Surgery	58 (11.3)	57 (13.4)	1 (2)	0 (0)b
Cardiopulmonary arrest	37 (7.2)	18 (4.2)	5 (11)	14 (34)b
Primary diagnosis					<.001
Medical	194 (37.8)	173 (40.6)	9 (20)b	12 (29)
Surgical	28 (5.5)	24 (5.6)	4 (9)	0 (0)
Neurological	154 (30.0)	118 (27.7)	26 (57)b	10 (24)
Trauma	137 (26.7)	111 (26.1)	7 (15)	19 (46)b
H2 blocker or PPI use during study period	504 (98.2)	416 (97.7)	45 (98)	41 (100)	.61
APACHE II score, mean (SD)	22.85 (7.58)	22.98 (7.67)	21.40 (5.73)	23.10 (8.41)	.40
Injury Severity Score, mean (SD)	22.99 (10.90)	22.89 (10.68)	20.00 (5.42)	24.55 (13.43)	.63
Body mass index,c mean (SD)	29.15 (9.01)	29.05 (9.07)	31.29 (10.75)	27.47 (4.82)	.35

Abbreviations: APACHE, Acute Physiology and Chronic Health Evaluation; PPI, proton pump inhibitor.

^aValues are No. (%) unless otherwise noted.

^bSignificantly different from study hospital (P≤.017 for pairwise comparison).

^cCalculated as weight in kilograms divided by height in meters squared.

Comparisons between intubation locations showed no differences for age, race, ethnicity, Acute Physiology and Chronic Health Evaluation II scores, or injury severity scores across sites. A higher proportion of men were intubated in the field than in the study hospital. Differences in admitting diagnoses were noted across sites. Medical diagnoses were highest among patients intubated at the study hospital, neurological conditions were highest among patients who were transferred, and trauma was highest among patients intubated in the field.

Field intubation was associated with ETT size of 7.0 mm, intubation under emergency conditions, and intubation because of cardiopulmonary arrest (Table 1). Patients intubated at the study hospital were significantly more likely than others to be intubated for respiratory distress. Patients undergoing transfer were significantly more likely than others to be intubated for airway protection.

Aspiration Biomarkers (α-Amylase and Pepsin)

Most patients (76.4%) had positive tests for α-amylase, and about one-third (33.1%) had positive tests for pepsin on the baseline tracheal specimen. Comparisons between intubation locations showed no differences for proportions of positive tracheal α-amylase and pepsin test results or for mean baseline tracheal α-amylase and pepsin levels (Table 2).

Table 2.

Tracheal α-amylase and pepsin levels by intubation location^a

Outcome	Study hospital (n = 426)	Transfer (n = 46)	Field (n = 41)	P
Positive tracheal α-amylase	325 (76.3)	38 (83)	29 (71)	.42
Positive tracheal pepsinb	80 (32.7)	12 (41)	5 (26)	.52
Tracheal α-amylase value, mean (SD), U/L	11 486 (32 647)	9965 (30 044)	15 709 (34 212)	.68
Tracheal pepsin value, mean (SD), ng/mL	4.90 (22.14)	2.23 (5.19)	6.48 (23.54)	.76

^aValues are No. (%) unless otherwise noted.

^bPepsin values determined for a total of 293 patients (245 at study hospital, 29 in transfer group, and 19 in the field).

Clinical Outcomes

We found no significant differences among intubation locations for ventilator hours, intensive care unit length of stay, or hospital length of stay. The VAC rate for patients who were transferred was more than double the rate for patients intubated at the study hospital (Table 3).

Table 3.

Ventilator hours, lengths of stay, and rates of ventilator-associated conditions by intubation location

Outcome	Study hospital (n = 426)	Transfer (n = 46)	Field (n = 41)	P
Ventilator hours, mean (SD)	125.78 (100.29)	148.22 (88.73)	146.45 (88.24)	.18
Days in intensive care unit, mean (SD)	9.22 (7.55)	9.39 (6.52)	9.03 (5.84)	.97
Days in hospital, mean (SD)	20.16 (19.90)	18.17 (17.43)	16.12 (13.84)	.38
Ventilator-associated conditions,a No. (%) of patients	49 (11.5)	12 (26)b	6 (15)	.02

^aAt least 2 calendar days of stable or decreasing daily minimum positive end-expiratory pressure (PEEP) or fraction of inspired oxygen (FIO₂) followed by increase in PEEP ≥ 3 cm H₂O or increase in FIO₂ ≥ 20 percentage points sustained for ≥ 2 days.

^bSignificantly different from study hospital (P ≤ .017 for pairwise comparisons).

Discussion

We hypothesized that field intubations would be associated with significantly higher levels of α-amylase and pepsin and higher VAC rates than intubations performed in hospital settings. We also hypothesized that patients who required interhospital transport would have higher levels of α-amylase and pepsin than those intubated at the study hospital. Literature comparing these environments with one another is scant,¹⁹ so these hypotheses were driven by a combination of clinical experience in all environments and studies of intubation complexities and conditions within each environment.^1,2,4–7 Factors that may have contributed to our results include the characteristics of the sample, timing of specimen collection relative to the intubation event, and inequality of group sizes.

Unlike many other published studies, a documented aspiration event at the time of intubation was an exclusion criterion for our larger parent study. This exclusion criterion complicates comparison across studies because we intended all patients in our study to be as aspiration free as possible at baseline. Vomiting, aspiration, and multiple intubation attempts are well-documented complications associated with field intubations.¹⁹ Our exclusion criterion eliminated most patients with field intubation complications from this study and created a more homogeneous baseline patient profile with regard to α-amylase and pepsin exposure. The baseline characteristics of our sample were probably different from those of most studies involving field intubations. Despite these exclusions, our analysis found that more than three-fourths of all patients had positive tracheal α-amylase values and one-third had positive tracheal pepsin values in the baseline specimen. The presence of aspiration during intubation may be much higher than initially suspected according to clinical assessment at time of the procedure.

Detection of α-amylase and pepsin may be influenced by the amount of enzyme present and the timing of specimen collection relative to the aspiration event.²⁰ In animal studies, pepsin has been detected in tracheal aspirates up to 4 hours after exposure,^21,22 with minimal loss of detection at 6 hours.²² Reported rates of detection loss vary, but aspiration volume is not consistent between studies.²¹ An inclusion criterion for the larger parent study was enrollment within 24 hours of intubation, so the timing of baseline specimen collection relative to the intubation event could have varied by as much as 23 hours. Patients who require interhospital transfer have a known delay between the intubation event and baseline specimen collection. Varying timing may influence baseline tracheal aspirate values. Published cutoffs for α-amylase levels classified as aspiration events vary across studies and by specimen location. Our study used a lower α-amylase cutoff level than some of those reported in the literature. Dewavrin et al¹³ stated that a cutoff level of 1688 U/L had moderate performance for identifying microaspiration (area under the curve, 0.72; 95% CI, 0.61–0.83). Similar values have been used by other authors.^23,24 When we applied a cutoff of 1688 U/L to our sample, the proportion of positive α-amylase cases decreased to approximately 47% for all groups (P = .98).

Unequal group sizes can influence power, variance, and type II error rates.^25,26 The group of patients intubated at our study hospital (n = 426) was much larger than the groups of patients transferred from other hospitals (n = 46) and intubated in the field (n = 41). A major contributing factor to type II errors is inadequate sample size, and the possibility of some type II errors must be considered.²⁶ Although the larger parent trial was adequately powered, this subanalysis may be underpowered. We addressed nonnormal distributions and unequal variance by using nonparametric tests.²⁷

Significant differences between groups for some demographic variables must be investigated further with regard to clinical outcomes, especially VACs. We found no associations between VACs and ETT size, Acute Physiology and Chronic Health Evaluation II score, intubation reason, race, or primary diagnosis. Compared with patients without VACs, those with VACs were significantly younger (51.6 vs 59.9 years, P = .001), more likely to be male (16.2% vs 8.8%, P = .01), and more likely to be intubated under emergency conditions rather than urgent or routine conditions (25.5% vs 11.6% and 13.2%, P = .03). Field intubations were also associated with being younger, male, and being intubated under emergency conditions. However, the incidence of VACs for patients intubated in the field was no different from that for patients intubated in other locations. Patients intubated in the field may receive aggressive early clinical management (including airway clearance and care) in the emergency department, which may contribute to the comparable clinical outcomes despite the association with other VAC risk factors.

Several factors may have contributed to the higher VAC rates associated with transfer from another hospital. The highest proportion of patients with neurological diagnoses was in the transfer group, which may be a reason for the higher VAC rate. However, a neurological diagnosis was not associated with a higher VAC rate in the study population (P = .27). Additionally, patients intubated at another hospital experience a substantial amount of movement that may have compromised the ETT cuff seal. Transfers from the hospital bed to the transport stretcher, from the stretcher to the transport vehicle, and from the transport stretcher to another hospital bed are all opportunities for compromise of the ETT cuff seal.²⁸ Vibrations associated with the transport vehicle and transport conditions (eg, road surface irregularities or turbulence) may also provide opportunities for micro aspiration. Airway assessment may be limited by transport conditions (eg, ambient noise) that can result in less frequent suctioning. Airway care equipment in the transport environment may also be more limited than in a hospital. Although patients intubated in the field would experience some of these conditions, the duration of transport is probably shorter than that of an interhospital transfer (transport time was not available for analysis). Patients intubated at the study hospital would experience none of these conditions.

Our collective VAC rate was 13.1% among all enrolled patients and was lowest for patients intubated at the study hospital. Our VAC rate was lower than the rate reported in some previous studies. For instance, another randomized controlled trial assessing the impact of subglottic secretion suctioning reported VAC rates of 22.0% in the intervention group and 22.9% in the control group (P = .84).²⁹ Several studies found VAC rates ranging from 4.0% to 14%, comparable to or lower than ours.^21–25 Our VAC rate can likely be attributed to the standard oral care provided to all patients every 4 hours as part of the NO-ASPIRATE protocol.¹⁴

Although in our study VAC rates were lowest among patients intubated at the study hospital, complications associated with intubation in emergency settings within the hospital (eg, emergency department) should be noted. A retrospective cohort study found that of patients with initiation of mechanical ventilation in the emergency department, nearly 1 in 4 were readmitted to the hospital within 30 days of discharge.³⁰ In our study, patients intubated in the study hospital had a nonsignificantly longer mean hospital stay than did patients intubated in other hospitals or in the field. We did not explore the reasons for different lengths of stay, which may include transfers back to other hospitals, transfers to skilled nursing and rehabilitation facilities, and death. Longer stays contribute to higher health care costs and an increased risk of hospital-acquired conditions. Further examination of the impact of intubation location on discharge outcomes is warranted.

Using biomarkers such as α-amylase or pepsin to establish occurrence of aspiration may still be premature because much is still unknown. Future research can help establish clearly defined cutoffs for diagnosis, explore dose effects related to VACs and ventilator-associated pneumonia, identify high-risk patient populations, and recommend specific clinical actions once biomarker levels are obtained.

Limitations

It was not possible to know complete medical histories and therapies for all patients before enrollment in the study. Conditions such as reflux disease and therapies such as bilevel positive airway pressure and protein pump inhibitor use may increase aspiration risk and confound enzyme analyses.^12,31 In the parent trial, no significant differences in demographics or other key study patient characteristics were noted between treatment groups.³² With the exception of sex, no differences in demographics were found in this study. However, we found significant differences between location groups for ETT size, urgency of intubation, reason for intubation, and primary diagnosis.

Another potential limitation is the nature of assays used to determine the presence of enzymes such as α-amylase and pepsin. Pepsin has several isoenzymes but is broadly classified as either pepsin A or C. Pepsin A is a protease stored in its inactive form (pepsinogen) in gastric mucosal chief cells. Pepsinogen is converted to pepsin in the decreasing pH levels of the stomach,³³ where it exerts proteolytic actions on cells, especially those outside the gastric environment.^34,35 Pepsin’s actions are pH dependent in the pH 1.0 to 5.5 range.³⁴ Pepsin C is found in nongastric organs such as the lungs and pancreas.³⁶ Studies of diseases such as reflux disease and chronic respiratory conditions have found that small amounts of pepsin routinely occur in the lungs but are neutralized and removed.^33,37 Unfortunately, assays may not be able to distinguish between the A and C forms of pepsin, possibly complicating interpretation of results.³⁶

This was a single-center study conducted at a tertiary care, level I trauma center with an active residency program. Therefore, generalizability of these results to children, nonteaching hospitals, and hospitals without a trauma designation may be limited.

Conclusions

Early identification of aspiration in intubated patients may be important for reducing clinical complications such as VACs and ventilator-associated pneumonia. Our findings indicate that aspiration as indicated by tracheal α-amylase and pepsin levels may be much more frequent than originally thought. Future research opportunities include determining aspiration diagnostic cutoff points for α-amylase and pepsin levels, establishing specific assays for the most informative enzymes, and determining if a dose-response effect for aspiration is present. Once diagnosis of aspiration is clearly identified, clinical research efforts can focus on targeted interventions to reduce respiratory complications. We found that patients intubated at other hospitals and transported to the study hospital had higher VAC rates than patients intubated at the study hospital. Future research can evaluate measures to reduce aspiration during interhospital transports.