Programmatic Approach to Outpatient Cardiac Catheterization in Children With Stable Pulmonary Arterial Hypertension

ABSTRACT

When performing cardiac catheterization in pediatric outpatients with pulmonary arterial hypertension (PH), our approach is to allow spontaneous ventilation, minimize procedural length, and evaluate for same‐day discharge whenever safe and feasible. We describe our experience with this approach and identify clinical characteristics that influenced procedural safety. Outpatients < 21 years who underwent catheterization for PH from 2009 to 2018 were included in the retrospective cohort. Demographic, clinical, and procedural data were collected. Data were modeled using a mixed effects logistic regression for correlated data, and a patient random effect was included to account for multiple procedures in the same patient. Of 409 catheterizations screened, 250 procedures in 118 outpatients were included. Of the 250 procedures, 185 (74.0%) were discharged on the same day. There were no major adverse events within 48 h of discharge in any of the 185 nor in the 12 (197 total, 78.8%) admitted for medication titration or an unrelated procedure (i.e., could have otherwise been discharged). Median procedural duration was 51.0 (33.0, 76.8) minutes. Endotracheal intubation, younger age, longer procedural duration, and worse functional status were associated with higher odds of admission. In a prospective secondary cohort of 39 procedures in 34 patients, 32 (82%) were discharged same‐day without complication, including over 90% of children over 3 years of age who were managed without endotracheal intubation. By prioritizing spontaneous ventilation and procedural efficiency, outpatient pediatric PH patients who undergo catheterization, emergence, and a 4‐h observation with no complications may be considered for same‐day discharge or observation in a low‐acuity bed.

Cardiac catheterization is the gold standard for the diagnosis, classification, risk stratification, and longitudinal follow‐up of pediatric patients with pulmonary arterial hypertension (PH) [1, 2, 3]. Although PH is a relatively uncommon diagnosis in children, it is a common indication for catheterization at regional congenital heart centers [4], and is associated with major adverse events [1, 4, 5, 6, 7, 8, 9, 10]. Reports of individual catheterization laboratory experiences [11, 12], administrative databases [9, 13], and multicenter clinical registries [4, 8, 10] have identified patient and procedural factors associated with major adverse events, such as younger age, the use of general anesthesia or endotracheal intubation, increasing PH severity, and preprocedural dialysis or vasopressor requirements, among others. Factors reported to decrease risk have included prior initiation of PH medications and a higher volume of PH cardiac catheterizations at the reporting heart center [9, 10, 13].

Many outpatient children with PH undergo elective cardiac catheterization to confirm the diagnosis and for interval hemodynamic assessments. Though outpatients have a different risk profile than children requiring inpatient care before the procedure, prior reports of catheterization outcomes in children with PH have either excluded outpatients [9, 13] or studied both outpatients and inpatients in the same cohort [4, 8, 10, 11, 12]. It is our impression that many pediatric programs will proactively intubate PH patients undergoing catheterization, which in turn requires a deeper anesthetic and can be the catalyst for instability. Our institutional approach for over a decade in PH catheterizations is to utilize a native airway or laryngeal mask airway with spontaneous ventilation, minimize procedural length, and target same‐day discharge when feasible. The aims of the present study were to describe outcomes of this outpatient pediatric PH catheterization experience, elucidate risk factors for major adverse procedural events, and to identify characteristics of children who underwent safe same‐day discharge.

1. Methods

Pediatric PH data were extracted from an institutional database within the Vera Moulton Wall Center for Pulmonary Vascular Disease at Stanford University. This study was approved by our institution's review board with a waiver of consent (Protocol #49231). The database was queried for children < 21 years who underwent cardiac catheterization between January 2009 and July 2018 and were outpatients before the procedure. Variables recorded included demographics, medical comorbidities, and World Health Organization (WHO) PH diagnosis group. Pre‐catheterization specific variables included preprocedural medications and right ventricular size and systolic function from the most recent echocardiogram within 6 months of catheterization. All echocardiograms were performed at our center using pediatric guidelines to determine right ventricular size and systolic function. Catheterization data included intracardiac pressures, cardiac index (by Fick and/or thermodilution methods), indexed pulmonary and systemic vascular resistances (PVRi, SVRi), and descriptions of interventional procedures.

For each catheterization, the primary author reviewed the hemodynamic values for internal consistency, and any catheterizations with inconsistent data not able to be reconciled were excluded from analysis. In instances where the cardiac index was reported by both thermodilution and Fick methods, the thermodilution values were selected and also utilized for PVRi and SVRi calculations. All catheterizations with a reported PVRi under 3 Wood Units × m² (WU) or a mean PA pressure (PAm) < 25 mmHg were excluded from analysis. Since children with and without PH are routinely admitted for observation following interventional procedures, all procedures involving transcatheter intervention were excluded.

Anesthetic data were collected from patients' medical record and reviewed by an attending pediatric cardiac anesthesiologist (ZK, CR). Variables included American Society of Anesthesiology (ASA) physical status, induction technique and agents, airway management, and maintenance of anesthesia. Airway management included nasal cannula, anesthesia mask, laryngeal mask airway, or endotracheal intubation. Additional details recorded from the anesthetic record included opioid usage and pharmacologic interventions for hemodynamic derangements. Since there is a preference at our center to use only ketamine and propofol when feasible in children with PH, the anesthesia regimen was grouped as to whether patients did or did not also receive volatile anesthetic (other than what was used for induction).

Outpatient status, preprocedural PH medication regimen, WHO PH diagnostic group [14], and procedural duration (from induction to sheath removal) were confirmed by chart review for all cases. The catheterization and anesthetic records, as well as the clinical documentation notes, were reviewed for peri‐ and postprocedural complications. Patient postprocedural disposition (discharge following post‐anesthesia recovery unit stay vs. hospital admission) was recorded. For patients who were discharged home following the procedure, all clinical, telephone, and documentation encounters were reviewed to screen for any complications at home within 48 h of discharge.

1.1. Outcome Definition

The primary outcome was admission to the hospital following catheterization. The decision to admit for observation could occur at any time between the end of the case and the routine 4‐h recovery period, at the discretion of the attending cardiac anesthesiologist and/or cardiologist. The secondary outcome was major adverse events occurring either during the procedure or recovery period. For admitted patients, in addition to hospital notes, the inpatient medical record was reviewed for medication delivery, oxygen administration, or other change in care that could be attributable to PH and/or the catheterization. Adverse event severity was assigned using previously published guidelines [15, 16] (Supporting Information S1: Table S1).

1.2. Statistical Analyses

Patient demographics were summarized based on the first catheterization for each patient, whereas catheterization‐related variables were summarized at the procedural level. Categorical variables were presented as number and percentage of cohort, and continuous variables as mean and standard deviation or median (Q1, Q3) if the data were not normally distributed. To estimate the probability for admission post‐catheterization, the data were modeled using a mixed‐effects logistic regression for correlated data. Fixed effects included sex, race, weight, age, most recent WHO PH diagnosis group, ASA status, induction type, airway management technique, right ventricular size, right ventricular function, use of volatile anesthetic, PH therapy (categorized as none, single, double, or triple therapy), PAm, PVRi, and procedural duration. Variables with large scales (i.e., procedural duration) were standardized by subtracting the mean and dividing by the standard deviation. A random effect for the patient was included to account for multiple catheterizations within the same patient.

Missing data were imputed using multivariate imputation by chained equations [17]. The number of imputations was based on the amount of missing values in the data. Estimates calculated from each of 10 imputation sets (5 iterations for each imputation) were ultimately combined using Rubin's rule [18]. Sensitivity analysis was performed using only the first outpatient catheterization for each patient.

Given the limited number of major adverse events, only univariate comparisons were used to show the relationship of clinical and hemodynamic variables and the outcome of a major adverse events at the catheterization level. All tests were two‐sided and p‐values less than 0.05 were considered to be statistically significant. Retrospective analyses were performed using the R statistical package, version 3.5.2 [19] and SPSS version 29.0.

1.3. Prospective Secondary Cohort

Following analysis and interpretation of the retrospective data, the decision was made to follow a prospective secondary cohort, specifically to confirm the safety of patients who are discharged home. From July 1, 2020 through April 30, 2023, data were collected for all outpatient children (age < 21 years) with a diagnosis of PH undergoing cardiac catheterization. All complications were prospectively recorded. Discharged patients were screened for post‐catheterization complications either by phone follow‐up or at the post‐catheterization clinic visit. For admitted patients, the location (ICU vs. floor) and rationale for admission were recorded. Continuous variables were compared using Student's t‐test or Mann–Whitney test, as appropriate. Categorical variables were compared using the Chi‐square test.

2. Results

In the PH database, between January 1, 2009 and July 31, 2018, 158 patients age 0–21 years underwent 409 outpatient catheterizations. Based on the exclusion criteria, 159 catheterizations and 40 patients were excluded, leaving 118 patients (50% female, 56% identified as Caucasian) and 250 cases included for analysis (Figure 1). Of the 86 catheterizations excluded for a PVRi < 3 WU, 33 (38%) were in 14 patients who previously had a PVRi > 3 WU and had been treated with PH medications.

Figure 1.

Diagram of inclusion and exclusion criteria. Caths, catheterizations; ICD, internal cardiac defibrillator; PVRi, indexed pulmonary vascular resistance.

Of the 118 patients included in the retrospective analysis, 46% had 1 catheterization, 20% had 2, 20% had 3, and 14% had 4 or more. Patient variables are shown in Table 1. Among the 250 procedures, 184 (73.6%) were in children over the age of 3 years, 74 (29.6%) were ASA status 4, 119 (47.6%) were on double or triple medication PH therapy, 85 (34%) on subcutaneous treprostinil and 20 (8%) had moderate to severe right ventricular dysfunction. Only 45 (19%) procedures were managed with endotracheal intubation, 41 of which were elective intubations as the intended ventilation strategy, and four of which were transitioned from a laryngeal mask airway following an adverse event. The remaining 187 (81%) procedures were managed with either nasal cannula (n = 87) or laryngeal mask airway (n = 100), without escalation to endotracheal intubation. The median procedural duration was 51 (33, 77) minutes. There was a median PVRi of 6.9 (4.6, 12.7) WU, PAm was 39 (28, 55) mmHg, and Rp:Rs was 0.53 (0.33, 0.93). Procedural data are shown in Table 2.

Table 1.

Patient demographics (retrospective cohort).

Patient‐level variables (n = 118)
Male gender	59 (50%)
Age at first catheterization (years)	6.5 (1.7, 11.6)
Age > 3 years at first catheterization	69 (59%)
Weight at first catheterization (kg)	15.5 (9.7, 34.4)
WHO PH diagnosis group
I	88 (75%)
II	9 (8%)
III	19 (16%)
IV	0 (0%)
V	2 (2%)
Number of catheterizations per patient
1	54 (46%)
2	24 (20%)
3	24 (20%)
4	10 (9%)
5 or more	6 (5%)

Note: Data presented as n (%) or median (Q1, Q3).

Abbreviations: PH, pulmonary hypertension; WHO, World Health Organization.

Table 2.

Procedural demographics, clinical status, anesthesia management, and hemodynamics (retrospective cohort).

Catheterization‐level variables (n = 250)
Male gender	125 (50%)
Age (years)	6.8 (2.9, 12.4)
Age greater than 3 years	184 (74%)
Weight (kg)	20.1 (11.6, 43.2)
WHO PH diagnosis group
I	193 (77%)
II	14 (6%)
III	38 (15%)
IV	0 (0%)
V	5 (2%)
ASA status (n = 232)
1	4 (2%)
2	2 (1%)
3	137 (59%)
4	74 (32%)
5 or 6	0 (0%)
Pulmonary hypertension medications
None	49 (20%)
Single therapy	82 (33%)
Double therapy	75 (30%)
Triple therapy	44 (18%)
RV size (n = 237)
Normal or mildly enlarged	188 (79%)
Moderately or severely enlarged	49 (21%)
RV function (n = 243)
Normal or mildly depressed	223 (92%)
Moderately or severely depressed	20 (8%)
Anesthesia management
Induction type (n = 228)
Intravenous	78 (34%)
Inhaled	150 (66%)
Neuromuscular blockade	37 (15%)
Airway type (n = 232)
Nasal cannula, mask, or LMA	187 (81%)
Endotracheal intubation	45 (19%)
Propofol and ketamine only (n = 223)	104 (47%)
Catheterization data
Procedural duration (minutes)	51.0 (33.0, 76.8)
Cardiac Index (L/min/m2)	3.9 (3.2, 4.6)
Right atrial pressure (mmHg)	7.0 (5.0, 9.5)
Systolic PA pressure (mmHg)	57.0 (42.8, 82.0)
Mean PA pressure (mmHg)	39.0 (28.0, 55.0)
Diastolic PA pressure (mmHg)	25.0 (17.0, 39.3)
PVRi (Wood Units × m2)	6.9 (4.6, 12.7)
Rp:Rs	0.53 (0.33, 0.93)
PA wedge pressure (mmHg)	10.0 (8.0, 12.0)

Note: Data presented as n (%) or median (Q1, Q3). The patient number is 250 for each variable unless otherwise indicated.

Abbreviations: ASA, American Society of Anesthesiologists; LMA, laryngeal mask airway; PA, pulmonary artery; PH, pulmonary hypertension; PVRi, indexed pulmonary vascular resistance; Qp:Qs, ratio of pulmonary to systemic blood flow; RV, right ventricular; WHO, World Health Organization.

About one‐quarter of cases (65, 26.0%) were admitted after the procedure, of which 39 (60.0%) were for observation without adverse event, 9 (13.8%) for a major adverse event, 5 (7.7%) for a minor adverse event, 7 (10.8%) for a medication initiation or adjustment, and 5 (7.7%) for an unrelated procedure or diagnostic workup. No patient required ECMO and there were no deaths. In total, there was a 3.6% (9/250) major adverse event rate, and details of all adverse events are shown in Table 3. Clinical, telephone, and documentation encounters (including follow‐up progress notes) for all 185 discharged patients revealed no adverse events within 48 h of discharge. There were also no major adverse events in the 7 patients admitted for medication initiation or adjustment, or the 5 admitted for an unrelated procedure or diagnostic workup. Two of the 7 patients admitted for medication initiation had side effects possibly related to treprostinil (1 with headache, 1 with nausea).

Table 3.

Description of all adverse events (retrospective cohort).

Timing	Adverse event (1–5)	Age (years)	WHO group	Description	Intervention
Induction	3‐Moderate	13.3	1	Upper airway obstruction	Changed airway from LMA to ETT
Induction	4‐Severe	12.5	2	Pulmonary hypertensive crisis	Sedation, intubation, vasopressin
Case	2‐Minor	4.4	1	Transient desaturation and bronchospasm	Albuterol
Case	2‐Minor	0.3	2	Hemodynamically insignificant atrial tachycardia	None
Case	3‐Moderate	3.1	3	Hypoxemia and difficulty ventilating	Changed airway from LMA to ETT
Case	3‐Moderate	0.4	1	Pulmonary edema with iNO testing	Remained intubated, admitted to ICU
Case	4‐Severe	4.7	1	Aspiration event followed by bradycardic arrest	CPR with ROSC; Changed airway from LMA to ETT
Emergence	3‐Moderate	3.7	1	Aspiration	Antibiotics
Emergence	4‐Severe	6.8	1	Hypotension with near cardiac arrest	Sedation, intubation, vasopressin
Emergence	4‐Severe	1.5	1	Bradycardic arrest	CPR with ROSC; Changed airway from LMA to ETT
Recovery	1‐None	12.2	1	Rash, admitted for possible allergic reaction	None
Recovery	2‐Minor	1.7	1	Fever	Laboratory evaluation
Recovery	2‐Minor	0.6	3	Transient desaturation	Blow‐by oxygen
Recovery	3‐Moderate	3.9	1	Hypotension	Normal saline bolus

Abbreviations: CPR, cardiopulmonary resuscitation; ETT, endotracheal tube; iNO, inhaled nitric oxide; LMA, laryngeal mask airway; ROSC, return of spontaneous circulation; WHO, World Health Organization class.

Review of the anesthesia and hospitalization records for the 39 post‐catheterization admissions for observation revealed that 8 were admitted for a history of obstructive sleep apnea or airway abnormalities, 11 for extracardiac disease (i.e., liver or kidney dysfunction), 6 for an interval significant worsening of PH found at catheterization, 4 due to left sided heart disease, 4 for age under 3 months, and the remaining 7 had no clear documentation of the rationale for admission. The hospital admission records for all 39 patients who were admitted for observation revealed no inpatient major adverse events.

The results for the primary analysis are shown in Figure 2. Children less than 3 years independently had higher odds of admission compared to kids who were older than 3 [OR with 95% CI: 5.0 (1.9, 12.5), p = 0.001]. Increasing procedural duration was associated with higher patients' odds of admission [OR with 95% CI: 1.52 (1.01, 2.28), p = 0.037]. Additionally, patients with ASA status 4 or 5 [OR with 95% CI: 2.33 (1.10, 4.92), p = 0.027], and those managed with endotracheal intubation [OR with 95% CI: 4.50 (1.83, 11.02), p = 0.001] had higher odds of admission. WHO PH diagnosis group, PVRi, and PAm were not associated with higher odds of admission. Of the 65 admissions, 14 (22%) had a PVRi between 10 and 20 WU, and 6 (9%) had a PVRi greater than 20 WU. Of the 188 same‐day discharges, 44 (23%) had a PVRi between 10 and 20 WU and 15 (8%) had a PVRi greater than 20 WU. In the sensitivity analyses and using only data from the first catheterization for each patient, only endotracheal intubation [OR with 95% CI: 4.57 (1.26, 16.59), p = 0.02] and PAm [OR with 95% CI: 3.25 (1.17, 9.02), p = 0.02] significantly increased the odds of hospital admission. In the univariable analysis of adverse events, major adverse events were associated with higher ASA status and moderate to severely depressed right ventricular function (Table 4).

Figure 2.

Probability of admission following cardiac catheterization by variable, using a mixed effects logistic regression for correlated data. ASA, American Society of Anesthesiology; PA, pulmonary artery; RV, right ventricular; WHO, World Health Organization. Bold values denote statistical significance at the p < 0.05 level.

Table 4.

Adverse event analysis (retrospective cohort).

	No major adverse event (n = 241)	Major adverse event (n = 9)	p value
Age (years)	7.3 (2.4, 13.0)	5.74 (3.0, 12.7)	0.23
Age greater than 3 years	177 (73%)	7 (78%)	0.77
Weight (kg)	21.5 (11.3, 46.5)	18.0 (10.7, 37.8)	0.22
WHO PH diagnosis group			0.87
I	186 (77%)	7 (78%)
II	13 (5%)	1 (11%)
III	37 (15%)	1 (11%)
IV	0 (0%)	0 (0%)
V	5 (2%)	0 (0%)
ASA status			0.028
1	4/223 (2%)	0 (0%)
2	17/223 (8%)	0 (0%)
3	135/223 (61%)	2 (22%)
4	67/223 (3%)	7 (78%)
Pulmonary hypertension therapy			0.84
None	46 (19%)	3 (33%)
Single therapy	78 (32%)	4 (44%)
Double therapy	75 (31%)	0 (0%)
Triple therapy	42 (17%)	2 (22%)
RV moderately or severely dilated	46/229 (20%)	3/8 (38%)	0.33
RV moderately or severely depressed	16/234 (7%)	4 (44%)	< 0.001
Endotracheal intubation	41/223 (18%)	4 (44%)	0.06
Propofol and ketamine only	98/223 (44%)	6 (67%)	0.11
Hemodynamic data
Cardiac Index (L/min/m2)	3.8 (3.1, 4.5)	3.3 (2.9, 3.8)	0.30
Right atrial pressure (mmHg)	7.0 (5.0, 9.0)	9.0 (7.3, 12.5)	0.11
Systolic PA pressure (mmHg)	56.0 (44.0, 83.0)	82.0 (58.0, 139.5)	0.22
Mean PA pressure (mmHg)	40.0 (29.0, 58.8)	58.5 (38.0, 91.8)	0.10
PA wedge pressure (mmHg)	10.0 (8.0, 12.0)	18.5 (9.5, 20.0)	0.11
PVRi (Wood Units × m2)	7.1 (5.0, 13.7)	10.9 (6.1, 22.8)	0.30

Note: Data presented as n (%) or median (Q1, Q3). Bold values denote statistical significance at the p < 0.05 level.

Abbreviations: ASA, American Society of Anesthesiology; PA, pulmonary artery; PH, pulmonary hypertension; PVRi, indexed pulmonary vascular resistance; RV, right ventricle; WHO, World Health Organization.

2.1. Prospective Secondary Cohort

From July 1, 2020 to April 30, 2023, there were 39 catheterizations in 34 outpatients with PH not included in the primary cohort, with data summarized in Table 5. When compared to the primary cohort, patients in the prospective secondary cohort were on average older [10.6 (8.0, 16.2) vs. 6.5 (1.7, 11.6) years, p < 0.01], but had similar PVRi [8.3 (7.3, 11.2) WU vs. 6.9 (4.6, 12.7) WU, p = 0.16], and similar incidence of endotracheal intubation [6/39 (15%) vs. 45/250 (19%), p = 0.82]. Arterial access was obtained in 24 (62%) cases, vasoreactivity testing was performed in 28 (72%), and average procedure length was 62 (39, 85) minutes. There were no major adverse events, and there was 1 (3%) minor adverse event (hypotension managed with a vasopressin infusion, resolved by the end of the procedure).

Table 5.

Patient characteristics and outcomes (prospective secondary cohort).

Catheterization‐level variables (n = 39)
Male gender	17 (44%)
Age (years)	10.6 (8.0, 16.2)
Age greater than 3 years	37 (95%)
Weight (kg)	32.6 (20.8, 44.4)
WHO PH diagnosis group
I	24 (71%)
II	5 (15%)
III	5 (15%)
IV	0 (0%)
V	0 (0%)
ASA status (n = 232)
1	0 (0%)
2	0 (0%)
3	32 (82%)
4	7 (18%)
5 or 6	0 (0%)
Pulmonary hypertension medications
None	10 (26%)
Single therapy	10 (26%)
Double therapy	10 (26%)
Triple therapy	9 (23%)
RV size (n = 237)
Normal or mildly enlarged	24 (62%)
Moderately or severely enlarged	15 (38%)
RV function (n = 243)
Normal or mildly depressed	32 (82%)
Moderately or severely depressed	7 (18%)
Anesthesia management
Induction type (n = 228)
Intravenous	20 (51%)
Inhaled	19 (49%)
Airway type (n = 232)
Nasal cannula or LMA	33 (85%)
Endotracheal intubation	6 (15%)
Propofol and ketamine only	21 (54%)
Catheterization data
Procedural duration (minutes)	61.5 ± 31.1
Cardiac Index (L/min/m2)	3.4 (3.0, 4.2)
Right atrial pressure (mmHg)	7.0 (5.0, 8.0)
Systolic PA pressure (mmHg)	59.0 (46.0, 72.5)
Mean PA pressure (mmHg)	40.0 (32.0, 49.5)
Diastolic PA pressure (mmHg)	25.0 (19.5, 35.0)
PVRi (Wood Units × m2)	8.3 (7.3, 11.5)
PA wedge pressure (mmHg)	8.0 (7.0, 10.0)
Arterial access	24 (62%)
Vasoreactivity (nitric oxide) testing	28 (72%)
Minor adverse events	1 (3%)
Major adverse events	0 (0%)
Discharged same‐day	32 (82%)

Note: Data presented as n (%) or median (Q1, Q3). The patient number is 250 for each variable unless otherwise indicated.

Abbreviations: ASA, American Society of Anesthesiologists; LMA, laryngeal mask airway; PA, pulmonary artery; PH, pulmonary hypertension; PVRi, indexed pulmonary vascular resistance; Qp:Qs, ratio of pulmonary to systemic blood flow; RV, right ventricular; WHO, World Health Organization.

There were 32 (82%) same‐day discharges, with documented follow‐up telephone or in‐person encounter with the PH clinical team, with no post‐discharge complications. There were 7 (18%) admissions: 5 for inpatient initiation or modification of PH medications for disease progression, 1 for a minor complication, and 1 for initiation of anti‐arrhythmic medications. Of 24 (62%) procedures with arterial access, 20 (83%) were discharged same day. Of 28 (72%) procedures with acute vasoreactivity testing, 24 (86%) were discharged same day. In this smaller prospective evaluation, no measured variables had a statistically significant association with same‐day discharge (Table 6).

Table 6.

Factors associated with same day discharge (prospective secondary cohort).

	Same‐day discharge (n = 32)	Admission (n = 7)	p value
Age (years)	11.5 ± 5.6	10.7 ± 4.6	0.68
Age > 3	1 (3%)	1 (14%)	0.23
Weight (kg)	30.1 (19.8, 56.1)	36.5 (22,4, 43.5)	0.90
WHO Class I	23 (72%)	5 (71%)	0.98
ASA 4 or 5	6 (19%)	1 (14%)	0.78
IV induction	18 (56%)	1 (14%)	0.09
Endotracheal intubation	4 (13%)	2 (29%)	0.29
Volatile anesthetic	7 (22%)	2 (29%)	0.65
RV moderately or severely enlarged	11 (34%)	4 (57%)	0.26
RV function moderately or severely depressed	4 (13%)	3 (43%)	0.09
Mean PA pressure	37.0 (31.3, 47.8)	48.0 (43.0, 78.0)	0.08
PVRi	8.2 (6.9, 10.5)	11.4 (9.1, 13.6)	0.09
Procedure length	58.7 ± 27.1	74.6 ± 45.5	0.22
Vasoreactivity testing	24 (75%)	4 (57%)	0.34
Arterial access	20 (63%)	4 (57%)	0.79

3. Discussion

Over a 10‐year period, about 75% of children with PH undergoing outpatient catheterization procedures under general anesthesia were discharged same‐day, with no readmissions and no important adverse events at home. In a prospective secondary cohort over a nearly 3‐year period, 82% of patients were discharged same‐day without complications or readmissions. This is the result of a multidisciplinary effort by the catheterization, cardiac anesthesiology, and PH teams to optimize pre‐procedural planning, procedural efficiency, and anesthetic optimization and minimization of airway intervention. The primary goal is to limit the procedural risk to the patient, and a same‐day discharge is an unambiguous marker of patient stability during and after the procedure.

Patients younger than 3 years of age, those managed with endotracheal intubation for the procedure, and/or those with longer procedural duration (and therefore longer anesthetic duration) were more likely to be admitted post‐procedure over the 10‐year historical dataset. The study cohort had significant pulmonary vascular disease, as evidenced by a median PVRi of 6.9 WU, with a quarter of patients having a PVRi over 12 WU and the majority (~80%) on outpatient pulmonary vasodilator therapies. Although higher mean PA pressure was associated with admission in the sensitivity analysis, we conclude that the primary drivers for admission were related to airway management and other anesthesia considerations. This may be of particular significance for WHO Group III patients, where a baseline oxygen requirement and/or airway abnormalities will factor into postprocedural disposition.

Though the primary goal is minimizing both procedural and post‐procedural risk to the patient, there are also obvious financial, logistical, and emotional benefits to minimizing hospital length of stay. Cumming and colleagues demonstrated reduced patient and parental anxiety, less exposure to nosocomial infection, and less missed work and school for same‐day discharges when compared to those admitted [20, 21]. Waldman and colleagues discharged about 50% of their over 600 pediatric patients following catheterization, and demonstrated an average savings of 80% in overall charge and 29% in total cost when compared to admission, though these data are not contemporary [21]. Colombo and colleagues reported safe same‐day discharge for two‐thirds of their high‐risk single ventricle infant patients, and demonstrated a considerably lower cost when compared to universal overnight admission [22]. In addition, a lack of elective intensive care units and general hospital beds may lead to the cancellation of procedures for other patients in the hospital system, thereby affecting patient satisfaction and delaying therapy. Given the relatively large volume of PH catheterizations performed at regional heart centers, a program to minimize intensive care unit bed utilization and hospital length of stay may have a positive impact on efficiency, cost, and patient satisfaction.

3.1. Adverse Events

Despite focusing exclusively on stable outpatients, we do report a significant 3.6% major adverse event rate in the retrospective cohort, generally occurring during induction or emergence from anesthesia. It is not our intent to downplay the acute risk of general anesthesia or catheterization in this population, but rather to report that in both our retrospective and prospective experience, no new adverse events occurred following the 4‐h recovery period from anesthesia (Table 3). All patients who experience a major adverse event during the procedure or 4‐h recovery period, including hypotension or pulmonary hypertensive crisis, are admitted for observation. Similar to other reports [23], we found that significant RV dysfunction and higher ASA status are both associated with major adverse events. However, this association is limited by low numbers of adverse events, and merits more study. There were no major adverse events in the lower volume prospective cohort, but these patients were older when compared to the retrospective group and airway management differed.

The safety concerns of catheterization in children with PH are widely reported in single institution studies, administrative database reports, and multicenter registries, but may not be generalizable to the stable outpatient. For example, O'Byrne and colleagues studied over 8000 catheterization procedures from the IMPACT database and reported a 1.4% rate of catastrophic events and a 5.2% rate of death before discharge [10], but over 10% of these patients were receiving inotropic support and about 15% were mechanically ventilated before the procedure. Reports from the PHIS database exclude outpatients [9], and reports from disease‐specific multicenter registries have included critically ill inpatients without a sub‐analysis of outpatients [4]. It is known that the risk of both anesthesia and catheterization is increased in newly diagnosed and treatment naïve severe PH patients [23], all of whom would likely be inpatients.

3.2. Anesthetic Considerations

The majority of catheterization risk for patients with PH surrounds the anesthetic management. All drugs used for induction and maintenance of anesthesia can perturb pulmonary and systemic pressures [24], but the drugs with the least effects on PVRi in children when used judiciously are fentanyl [25], ketamine [24, 26], and propofol [27]. Care must be taken to avoid systemic hypotension, particularly in patients with marked elevation of pulmonary artery pressures and low cardiac output [1], and when hypotension occurs it should be treated promptly.

While catheterization can be performed on an awake or lightly sedated adult patient, most children require either moderate to deep sedation or general anesthesia with or without an established airway. Coughing, laryngospasm, hypoventilation, and hypercapnia can occur during induction or emergence from anesthesia and can precipitate a PH crisis. Therefore, published guidelines recommend that anesthesiologists with experience in pediatric PH be involved in the sedation and airway management of children with PH during cardiac catheterization [1, 2, 28].

3.3. Institutional Anesthesia Approach

When managing children with PH in the catheterization laboratory, our approach is to maintain sedation and anesthesia with spontaneous respiration while avoiding endotracheal intubation. A preoperative anxiolytic is typically administered, and younger patients undergo inhalational induction with sevoflurane before placing an intravenous line. Unless there is known airway disease, spontaneous ventilation is supported with a nasal cannula, anesthetic mask, or laryngeal mask airway. Maintenance of anesthesia is accomplished mostly with ketamine [26] and propofol infusions, with a 0.5–2 mg/kg ketamine bolus at the beginning of the procedure. Additionally, less than one minimum alveolar concentration of sevoflurane may be added. Boluses of ketamine, propofol or opioid are administered at the discretion of the anesthesiologist, and if hypotension occurs, the current strategy is to bolus 0.03 units/kg of vasopressin [29]. In pediatric patients with PH, though vasopressin, phenylephrine, and epinephrine all similarly increase the SVRi, only vasopressin results in a decrease in the PVRi to SVRi ratio [29].

Hemodynamic changes during emergence from anesthesia and/or with rebound from vasodilator testing can precipitate a PH crisis [28]. All patients are therefore monitored in the catheterization laboratory after emergence to ensure normal hemodynamics and oxygenation before transfer to the post‐anesthesia care unit for 4 h of additional observation. If an adverse event were to occur either during the procedure or in the immediate postoperative period, the child is admitted for overnight monitoring.

3.4. Length of Procedure

Since anesthesia duration is an important driver of risk in children with PH, the focus is to acquire data expeditiously and efficiently. In many instances, outpatient children with PH have undergone a prior MRI and/or catheterization to evaluate for pulmonary thromboembolic disease, peripheral pulmonary artery stenosis, pulmonary vein stenosis, pulmonary veno‐occlusive disease, and parenchymal lung disease, as per published guidelines [1, 7]. Therefore, many stable outpatients are referred for only a right heart catheterization with vasoreactivity testing, with or without arterial access for invasive blood pressure monitoring. The median procedural duration for the primary cohort was 50 min (interquartile range of 33–76 min). Patients who were admitted tended to have longer procedural times, and therefore a longer anesthetic, which may have contributed in some part to the decision to admit for observation.

3.5. Limitations

This study had several limitations, including the single‐center study design. The decision to admit was at the discretion of the anesthesiologist and varies between providers. An admission was not synonymous with an adverse outcome, and in many cases was a result of our preconceived ideas of what makes a patient higher risk. Therefore, variables associated with admission were a description of the behavior of providers at our institution and not necessarily a specific reflection of patient or procedural variables that are directly associated with safe discharge. Though the chart review for adverse events following same‐day discharge was thorough, it is possible that in the retrospective cohort minor adverse events at home could have been missed. Major adverse events were not likely to have been missed, since all discharged patients continued outpatient follow‐up at our institution with documented follow‐up notes. These limitations are alleviated in part by the follow‐up prospective secondary cohort, which included screening for adverse events. Due to the small number of major adverse events, mixed modeling and/or multivariable analysis were not performed.

3.6. Recommendations

Outpatient pediatric PH patients undergoing cardiac catheterization should be managed with spontaneous ventilation and avoidance of endotracheal intubation when feasible. In our experience, stable outpatients with PH who undergo uncomplicated diagnostic catheterization, emergence, and a 4‐h observation can be evaluated for post‐catheterization monitoring in a low‐acuity bed or possible same‐day discharge.

Author Contributions

Gregory T. Adamson: study concept and design, data collection, initial manuscript draft. Minnie N. Dasgupta: study concept and design, data collection, manuscript revision. Zachary Kleiman: study concept and design, data collection, manuscript revision. Lynn F. Peng: study concept and design, manuscript revision. Michelle Ogawa: study concept and design, manuscript revision. Vidhya Balasubramanian: statistical analysis, manuscript revision. Chandra Ramamoorthy: study concept and design, manuscript revision. Jeffrey A. Feinstein: study concept and design, manuscript revision.

Ethics Statement

Institutional Review Board approved (Protocol #49231) with waiver of consent.

Consent

Waiver of consent.

Conflicts of Interest

The authors declare no conflicts of Interest.

Supporting information

Supplementary Table.