Background: Due to the lingering effects of general anesthesia and the administration of medications for pain after surgery, children in the recovery room are at risk for breathing problems. While there are less data specific to children, in general 25% of patients in the recovery room experience complications from anesthesia. The most common complications involve the patient's airway and their ability to breath adequately. Currently, checks of oxygenation with a pulse oximetry monitor and of respiration through nursing assessments are used to detect breathing problems. However, these are believed not to be adequate for reliably recognizing significant respiratory depression until other dangerous events develop such as the cessation of breathing, severe drops in oxygen levels, or cardiac arrest. Capnography is a monitoring device that measures the amount of carbon dioxide being exhaled and assesses the adequacy of respiration. A small plastic cannula sits at the base of the nose and on the lip to continuously monitor the patient's breathing. Most children tolerate this device well and staff consider it easy to use. While capnography is routinely used in the operating room to monitor breathing, it is not used during post-operative care when patients are still at risk of breathing problems.

Objectives: To determine if capnography can detect problems with breathing faster and more often than traditional monitors. To determine if the addition of capnography to routine monitoring will decrease the numbers of additional adverse events that occur in children undergoing post-operative care by allowing nurses to intervene in care faster and more frequently.

Methods: In the first phase of this study, the investigators will apply the capnography monitor to children in the recover room and determine how often they experience breathing difficulties measured by this device. In the second phase, the investigators will educate staff on the use of capnography and what values are considered abnormal. Children will again have the capnography cannula placed on them as they enter the recovery room. They will then be divided into two groups - in one group the nurses in the recovery room will have access to the capnography monitor for their patients, whereas in the other group the nurses will not be able to see the readout from the monitor. The investigators will determine if children have fewer breathing problems and less additional adverse events when nurses use capnography in addition to the routine monitors already in place in the recovery room as compared to when nurses use standard monitoring alone.

Potential Impact: If capnography can detect breathing problems prior to being identified by current monitoring devices, staff may be able to intervene more quickly and before more serious events occur in the children receiving post-operative care. This can reduce adverse events, improve patient safety, and avert harm in children. The adoption of this device for routine monitoring of post-operative care has the potential to save lives.

Study Site: Children will be enrolled in the post-anesthesia care unit (PACU) at Yale-New Haven Children's Hospital, where capnography is not currently utilized. The pediatric PACU is a 12-bed monitored unit and a 4-bed perioperative area where patients receive one-on-one nursing care. Generally, between 20-30 patients are cared for in the PACU each day with a typical length of stay of approximately 2 hours.

Specific Aim 1b: To determine whether the addition of capnography will lead to a decrease in the frequency of oxygen desaturations as measured by pulse oximetry and respiratory depression in children recovering from general anesthesia.

Study Design and Setting: The investigators will conduct a randomized controlled trial in the Yale-New Haven Children's PACU.

The identical population, inclusion and exclusion criteria and recruitment procedures will apply. A random number list will again be generated to guide recruitment. Investigators will be blinded to group designation prior to consent and enrollment.

Study Intervention and Data Collection: Prior to the onset of this study, PACU nursing staff will receive focused training on the application and interpretation of capnography. Proper positioning of the nasal-oral cannula will be demonstrated. Capnography waveforms will be reviewed including normal respiratory patterns and end-tidal CO2 (ETCO2) levels that would indicate hypoventilation, apnea or obstruction. Staff will be instructed that they can provide interventions when ETCO2 values greater than 50mmHg or less than 30mmHg are obtained without a concomitant increase in respiratory rate. Simple noninvasive interventions may include reposition of the head, instructing the patient to take several deep breaths, physical stimulation, or placing a shoulder roll under the patient. Invasive interventions may include administration of reversal medications, bag-valve mask assisted ventilation, the insertion of an oral or nasal airway and endotracheal intubation. Patients who develop oxygen saturations less than 93% may receive an increased amount of supplemental oxygen at the discretion of the treating staff.

Patients will undergo routine care and monitoring while in the PACU. As part of this study, a duo-port nasal-oral cannula for ETCO2 monitoring will be placed in the nostrils and over the lip of the patient prior to transfer from the operating room to the PACU (Filterline Smart CapnoLine Plus) and attached to the portable capnography monitor (Capnostream 20 Bedside Capnography Monitor by Oridion). Subjects will be randomized into 2 groups: those in whom the responsible nurse and physician are blinded to the capnography monitor (controls), and those in whom the capnography monitor is viewed by treating staff during the recovery period (cases). The investigators control group represents the current standard of care for PACU monitoring. Stratified block randomization will take place. Patients will be stratified by type of surgery: otolaryngology (ENT) or non-ENT. The randomization sequence will be placed in sequentially numbered opaque envelopes for each stratum. Although each patient and his/her family will consent to the study prior to surgery, they will not be officially enrolled until the patient demonstrates that he/she can tolerate the nasal-oral cannula upon arrival to the PACU. Thus, when an enrolled subject arrives to the PACU and is successfully tolerating the nasal-oral cannula the study envelope containing their group designation will be opened by the research assistant (RA) or principal investigator (PI) for this study and data collection will commence. All staff will be blinded to group designation until this time.

For the case subjects, alarm settings will be programmed on the monitor to alert the provider to ETCO2 levels below 30mmHg and above 50mmHg. The monitor will be positioned where it can be easily viewed by both the treating staff and the RA/PI for this study. For control subjects, the portable capnography monitor will be positioned by the PI or RA and out of the line of sight of other treating staff. The alarm function will be turned off so as not to alert staff to abnormal readings on the monitor. Since the investigators are not mandating that staff intervene for abnormal ETCO2 values, the RA or PI will be the only participant to view the capnography monitor in the control group and will not inform the treating staff of any abnormal values under 2 minutes in length. The Yale Human Investigations committee was consulted regarding the ethics of withholding information from the control arm of this study and approves this decision. Instances where the PI or RA had to notify staff of events will be recorded on the data form.

Methodology: For the purposes of this randomized controlled trial, heart rate, respiratory rate, and oxygen saturation will be routinely monitored and recorded on a data sheet every 30 seconds while the patient is in the PACU by the PI or RA who will not otherwise be involved in patient care. ETCO2 values are recorded by the capnography monitor every 30 seconds and waveforms are continuously captured. These data will be downloaded from the capnography monitor at the time of discharge for each patient. Adverse respiratory events as well as nurse and physician interventions related to airway and ventilatory management will be documented by the PI/RA. These were described previously. Additional pain or sedative medications may be administered by nursing staff and will be recorded during this study. Patient disposition and total length of stay in the PACU will be recorded.

The investigators primary outcome will be frequency of The investigators previously defined respiratory events that occur with and without the use of capnography.

Sample size calculations: While the investigators will refine their sample size based on event rates found in the first cohort of this proposal, the investigators plan to enroll at least 200 children in this pilot study in order to detect a 30% reduction in adverse event rates.

Statistical Analysis: Variable distributions will first be examined with descriptive statistics, including box-plots and histograms. Any missing data due to artifact or equipment malfunction will be explored for each group. The investigators will describe the two groups of patients, those with blinded capnography readings vs. those without, in terms of their age, gender, ethnicity, type of condition for which sedation was required, and physician performing procedure. The investigators will use Student's t-test for continuous variables and chi-square/Fisher's exact test for categorical variables. Statistically significant (p < 0.05) or clinically meaningful differences (10%) between two randomized groups will be added to the multivariate analysis of the primary outcome. Groups will then be analyzed by strata - patients receiving ENT procedures and those receiving other non-ENT procedures.

The primary outcomes, airway or respiratory adverse events, will be summarized by group in terms of the number and percent of patients in each group who experience these outcomes. Generalized estimating equation (GEE) approach will be used to model the probability of an outcome (two separate models for each outcome of interest) with the randomization group (blinded vs. not to capnography readings) as the main independent predictor. The investigators will adjust the model with other important predictors, especially the ones that were different between the two groups. Odds ratios (OR) and surrounding confidence intervals will be reported for each predictor. The investigators can use the adjusted OR for the group variable to calculate relative risk (RR).37 Significance will be established with a two-tailed alpha=0.05. The investigators will use this data to obtain funding for a larger, randomized trial which will also include cost-effectiveness analyses.

Inclusion:

• Pediatric patients (1-20yo) undergoing a scheduled, elective procedure at Yale-New Haven Hospital.
• An English-speaking parent or guardian must be present in the pre-operative care area in order to provide informed consent.

Exclusion:

• Currently or expected to receive post-operative assisted ventilation via an endotracheal tube (ex. major cardiac surgery) or tracheostomy
• Being directly admitted to the Pediatric Intensive Care Unit or pediatric ward instead of the PACU
• Undergoing urgent or emergent surgical procedures (i.e. non-elective) or surgery that will preclude the use of a nasal-oral cannula (ex. facial reconstruction)
• Any disease or state that may lead to abnormal ETCO2 values or baseline abnormalities in pulse oximetry. These would include evidence of active asthma exacerbation such as wheeze, diabetic ketoacidosis, moderate to severe dehydration, cardiac abnormalities, severe lung disease, and major trauma as the reason for surgery.
• Inability to tolerate the nasal-oral cannula, if it is removed during the study period by the patient or family, or if subjects cry continuously during the study, which inhibits accurate readings on the monitor.

Data were collected from the time of arrival to the PACU to the time of discharge. This was on average 1 hour.