Decreasing Pre-Procedural Fasting Times in Hospitalized Children

Abstract

Objective:

Prolonged pre-procedural fasting in children is associated with decreased patient and family satisfaction and increased patient hemodynamic instability. Practice guidelines recommend clear liquid fasting times of 2 hours. We aimed to decrease pre-procedural clear liquid fasting time from 10 hours 13 minutes to 5 hours for pediatric hospital medicine (PHM) patients.

Methods:

All children admitted to the PHM service at a quaternary care children’s hospital with an NPO (nil per os) order associated with a procedure requiring general anesthesia or sedation from November 2, 2017 to September 19, 2021 were included. The primary outcome measure was the average time from clear liquid fasting end time to anesthesia start time. The process measure was the percent of NPO orders including a documented clear liquid fasting end time. Balancing measures were aspiration events and case delays/cancellations. Statistical process control charts were used to analyze outcomes.

Results:

Shortly after implementation of a SmartPhrase in the NPO order, there was special cause variation resulting in a centerline shift from a mean of 10 hours 13 minutes to 6 hours 37 minutes and an increase in the process measure from a baseline of 2% to 52%. Following implementation of a hospital-wide change to the NPO order format, another centerline shift to 6 hours 7 minutes occurred which has been sustained for 6 months. No aspiration events and 4 NPO violations occurred during the intervention period.

Conclusion:

Quality improvement methodology and higher reliability interventions safely decreased the average pre-procedural fasting time in hospitalized children.

INTRODUCTION

Pre-procedural fasting is a routine strategy to reduce the risk of pulmonary aspiration of gastric contents in children and adults undergoing anesthesia. The American Society of Anesthesiologists (ASA) 2017 guidelines outline pre-procedural fasting times before anesthesia which include 2 hours for clear liquids.¹ Despite these guidelines, many hospitals use an “NPO at midnight” strategy for procedures the following day.² For children, studies demonstrate this practice results in fasting times between 6 and 15 hours, far exceeding the ASA recommended fasting time for clear liquids.^3–6

Prevention of prolonged fasting for hospitalized children without increasing complications, delays, or cancellations can improve the pre-procedural experience for children and families while maintaining hospital efficiency. Prolonged fasting in children increases thirst and irritability and results in detrimental physiologic and metabolic effects including reductions in systolic blood pressure during anesthesia induction and increased catabolic state.^6–9 Both child and parent perceptions of hunger and thirst are lower for children who are able to drink clear liquids 2 hours before anesthesia.¹⁰ Allowing clear fluids closer to the procedure time results in children being less thirsty, hungry and anxious and improves hemodynamic conditions by preserving intravascular volume.^{4,6–9,11–13} Our hospital’s Family Advisory Council identified reducing pre-procedural fasting for hospitalized children as a priority, which served as a “call-to-action” for our team.

To safely improve the overall pre-procedural experience for hospitalized children who require previously unscheduled procedures, our aim was to decrease clear liquid fasting time prior to anesthesia or sedation from an average of 10 hours 13 minutes to 5 hours for patients admitted to the PHM service over one year.

METHODS

Setting and Context

This study was conducted at a 307 bed, university affiliated, quaternary care children’s hospital with over 15,000 annual hospitalizations. Approximately 3500 children per year are admitted to one of four PHM teams. The teams are composed of 2–4 first year residents, 1 second or third year resident, 1–2 medical students, and a PHM faculty physician. Residents rotate monthly. Family-centered, bedside rounds are conducted each morning. An encrypted hospital-provided pager or an encrypted Smartphone application (MH-Cure, Mobile Heartbeat^™, Waltham, MA) are used for team communication.

Our hospital has a specialized anesthesiology team comprised of 13 nurse practitioners (NPs), who are a part of the Pediatric Anesthesiology division, that conducts pre-procedural evaluations on all hospitalized children requiring sedation or general anesthesia. However, responsibility for scheduling a procedure differs based upon procedure location (operating room versus radiology). Procedures for PHM patients are considered to be “add-on procedures” given the acute nature of the hospitalizations.

Interventions

A multidisciplinary team consisting of nurses, residents, a PHM fellow, PHM faculty physicians, anesthesiology faculty and NPs, and hospital family advisory council members were engaged to address the stated aim. Using the Model for Improvement as the team’s framework, a key driver diagram (Figure 1) was created and modified throughout the study using knowledge gained through multiple plan-do-study-act (PDSA) cycles.¹⁴

Figure 1.

Key driver diagram.

The ASA fasting guidelines, known as the “6,4,2” rule, align with our hospital NPO policy and guided key driver diagram development.¹ The “6,4,2” rule refers to the number of hours before anesthesia that certain food consistencies should be stopped: 6 hours for solid food (including infant formula), 4 hours for breastmilk, and 2 hours for clear liquids.

Shortly after development of the project’s aim in September 2019, a meeting with pediatric anesthesiology providers revealed that the earliest start time for a non-emergent inpatient sedated procedure in our hospital is 7 am. This served as a standardized starting point for interventions targeting the first key driver of knowledge of procedure start time among multiple teams. Interventions included communication with our anesthesiology colleagues who complete the pre-procedure evaluations because they have the most updated procedure time information. For hospitalized patients, procedure time is fluid given that all PHM procedures are “add-ons.”

For patients whose procedure was not scheduled until after 12 pm, clear liquid fasting end time was expected to be extended from 5 am to 10 am. In order to facilitate this, several PDSA cycles were completed. We encouraged residents to contact the anesthesiology team before rounds to request updated timing information. Reminders to communicate daily with the anesthesiology NPs were sent to residents via e-mail at the beginning of inpatient rotations and included the NP’s pager number. In addition, the anesthesiology NP’s pager number was placed on physician computer workstations. Our hospital system recently began using an encrypted Smartphone application (MH-CURE) to securely text instead of using a pager and phone. However, this technology is still inconsistently used making pagers the most reliable communication method. Therefore, residents would page the anesthesia team which, in turn, requires a response via phone. This was often challenging for both the anesthesiology NPs who were seeing multiple inpatients and pediatric residents who were rounding and completing other tasks. Finally, we tried individual audits with feedback about communication challenges with residents. Efforts to improve communication within workflows is ongoing.

The second key driver was focused on establishing a shared mental model about our hospital’s NPO guidelines (“6,4,2” rule), and the third key driver was proper communication among all care team members and family. Interventions included periodic education, reminder emails, and signage for bedside nurses, residents, and faculty. Champions of each group helped remind key stakeholders of the project goal.

In December 2019, prototype NPO patient door signs were shared with nursing champions and nursing managers. Nursing champions provided sign design feedback and adjustments were made accordingly. In January and February 2020, updated NPO door signs were trialed with 1–3 nurses on two different units. After receiving feedback, nursing management formally approved new signs in March 2020. These signs were laminated to allow for easy modification of times to stop solid foods, breastmilk, and clear liquids as new procedure time information evolved (Appendix Figure 1). While this intervention was not directly associated with a shift in our primary outcome measure, it likely helped reinforce to hospital staff that “NPO at midnight” was no longer the PHM services’ standard of care.

The fourth key driver was flexibility with NPO status based on potential changes or uncertainty of procedure timing, and the fifth key driver was an electronic health record (EHR) system that was supportive of NPO guidelines. Interventions included daily communication with anesthesia as previously described in addition to changes within the EHR.

Despite updated NPO door signs and education of residents, attendings, and nurses, the “6,4,2” rule was inconsistently applied using the existing EHR NPO order. Therefore, we created an EHR SmartPhrase. SmartPhrases for commonly used language can be made by EHR users and shared with others. The SmartPhrase was originally named “.ARCNPO” (utilizing the first author’s initials) to specify stop times more clearly for solid food, breastmilk, and clear liquids using 7 am as the default earliest anesthesia start time. This SmartPhrase was shared with residents and its use was encouraged in the comment section of the NPO order (Figure 2a). However, the name of this SmartPhrase was difficult for providers to remember, so it was later changed to the more generic “.NPO.”

Figure 2a.

Original NPO order requiring providers to use the Epic Smartphrase titled “.NPO” to populate the “Diet Cmmts” section with the default NPO times using 7 am as the earliest possible anesthesia start time.

This SmartPhrase was a “work-around” as we sought to increase buy-in from stakeholders to test and implement a higher reliability intervention—modification of the hospital-wide NPO order. Given our success in decreasing pre-procedural fasting time using the SmartPhrase, we received approval from hospital administration to change the NPO order format by adding an “NPO for Future Anesthesia/Sedation” order option. When selected, this option defaults the effective solid food NPO starting time to 1 am, breastmilk to 3 am, and clear fluids to 5 am to comply with the “6,4,2” rule using 7 am as the earliest possible start time (Figure 2b). The modified NPO order also provides examples of allowable clear liquids which are available on the inpatient floors. The times are all editable and teams are encouraged to adjust them to allow patients to eat and drink for as long as possible.

Figure 2b.

Newly developed hospital-wide NPO order with an option for “NPO for Future Anesthesia/Sedation.” Choosing this option defaults the NPO start time to 1 am and specifies the end times that solid food, breastmilk and clear liquids can be provided to the patient using 7 am as the earliest possible start time. These times are editable if the exact procedure start time is known..

Study of the Interventions

Data were collected and analyzed weekly from the EHR (Epic, August 2020) using a Tableau (Seattle, WA) dashboard that extracts data from the Epic Clarity data warehouse. NPO start time is recorded in Epic as the time the order should start or, more specifically, the time that clear liquids should end. Anesthesia start time is documented with a date and time in an “anesthesia post procedure evaluation” note which is generated for any encounter in which a patient is given sedation medications. The NPO duration was calculated as the time difference between clear liquid end time and anesthesia start time in hours and minutes.

All children admitted to the PHM service with an NPO order associated with a procedure or radiologic study requiring anesthesia (i.e., had an NPO order that started before the documented anesthesia start time) from November 2, 2017 to September 19, 2021 were included. Children with fasting times <2 hours were excluded as these likely represented emergent cases which needed to proceed despite suboptimal fasting. Children with NPO times >24 hours were also excluded as these likely represented children NPO for medical reasons (e.g., bowel rest) who subsequently required a procedure.

Measures

Primary Outcome Measure

The primary outcome measure was the average time from documented clear liquid fasting end time to anesthesia start time.

Process Measure

The process measure was the percent of NPO orders that included a documented clear liquid fasting end time using either text in the comment box of the existing NPO order or the newly created NPO order option.

Balancing Measure

Balancing measures were aspiration events and case delays/cancellations due to pre-procedural fasting violations. Aspiration events are charted by anesthesiology team members within the patient chart in the Anesthesia Quality Institute Quality Capture Application. Case delays/cancellations are documented in patients’ charts by a nurse using standardized codes (i.e., Delay to OR - Patient not NPO or Patient: NPO Violation). Both were charted prior to initiation of this project and are used for internal quality improvement (QI) purposes.

Analysis

An X-bar and S-chart (for the primary outcome measure, S-chart not shown) and a p-chart (for the process measure) were constructed and updated weekly to track outcomes using QI Charts (Process Improvement Products, Austin, TX). For weeks with only one instance of an anesthesia event associated with an NPO order, that event was added to the prior week’s data since X-bar and S-charts require >1 value. Control charts were annotated with interventions to visually inform the team’s improvement theories. The rules of interpretation for Shewhart charts were used to identify special cause variation.^14,15

Our institutional review board determined the project was not human subjects’ research.

RESULTS

A total of 579 NPO/anesthesia events were included in the baseline period from November 2017 to September 16, 2019 and 586 NPO/anesthesia events in the intervention period from September 17, 2019 to September 19, 2021. During the intervention period, an average of six children per week required anesthesia or sedation. For both the baseline period and the intervention period the average age of the patient was 6 years and 26% of children in the intervention period were <1 years old.

Primary Outcome Measure

Shortly after the development and sharing of the SmartPhrase, special cause variation occurred resulting in a mean NPO duration reduction from a baseline of 10 hours 13 minutes to 6 hours 37 minutes beginning the week of January 13, 2020. System-wide updates to the NPO order within the EHR to include default clear liquid fasting times were associated with a second occurrence of special cause variation reducing the mean to 6 hours 7 minutes beginning the week of March 1, 2021. Overall, improvement has been sustained for 20 months from the initial centerline shift in January 2020 with the most recent, small improvement sustained for the last 6 months (Figure 3). Although the newly created NPO order went live for the entire children’s hospital in November 2020, it took some time to educate providers about the order update. The process measure improved in January 2021 when orders with a documented clear liquid ending time increased from 64% to 84% (Figure 4). A shift in our primary outcome measure followed in March 2021 when the average NPO duration decreased from 6 hours 37 minutes to the current average of 6 hours 7 minutes (Figure 3). Special cause variation occurred the week of July 5, 2021 with a single point outside of the control limits corresponding to the new academic year.

Figure 3.

Primary outcome measure of weekly average NPO duration prior to sedation for patients on the hospital medicine service (X-bar chart). LCL, lower control limit; UCL, upper control limit.

Figure 4.

Process measure of weekly percent NPO orders with specific end time for clear liquids in order comment box (p-chart). LCL, lower control limit; UCL, upper control limit.

Process Measure

The baseline percent of NPO orders with a documented clear liquid fasting end time was 2%. SmartPhrase use was associated with a centerline increase to a mean of 64% the week of January 13, 2020. The system-wide change to the NPO order the week of November 16, 2020 was associated with improvement to 84% the week of January 18, 2021. This has been sustained for 8 months at 84% of NPO orders with documented clear liquid fasting end (Figure 4).

Balancing Measures

There have been no aspiration events throughout the intervention period. Four NPO violations occurred during the intervention period leading to minor start time delays but no case cancellations. None of the NPO violations were related to the order itself (NPO order was in the EHR correctly but was inadvertently not followed).

DISCUSSION

We used improvement methodology to decrease pre-procedural fasting time in hospitalized children from 10 hours 13 minutes to 6 hours 7 minutes without any adverse events or case cancellations. Overall improvement was sustained for 20 months and the more recent, smaller improvement has been sustained for 6 months. Although the stated goal of 5 hours (reducing baseline fasting time by 50%) has not yet been reached, we made significant improvements towards this aim. Implementing a change to the hospital-wide NPO order format created a reliable method to standardize desired NPO times and allowed patients to have clear liquids far longer than the commonly ordered “NPO at midnight.”

Through structured PDSA cycles, we tested our hypotheses using a stepwise approach. Lower reliability interventions included education and signage to promote a shared NPO timing mental model. Translating the “6,4,2” rule into the EHR order system using a SmartPhrase with standardized language was associated with an improvement in both our process measure and primary outcome. This evidence allowed us to gain buy-in to implement a higher reliability intervention which provides the same NPO verbiage with a single order option within the EHR NPO order that serves as both a prompt and task simplification.

Few studies have sought to test interventions to decrease pre-procedural fasting times for children hospitalized for a non-surgical primary condition. One pediatric QI study conducted in the United Kingdom at a same-day surgery center increased the percent of children with clear liquid fasting times <4 hours from 19% to 72%.³ In the United States, one QI study focused on children presenting from home for elective procedures which successfully decreased the mean liquid fasting time from 9 hours to 6 hours which was attributed to the adoption of a 1-hour minimum clear liquid fasting policy.¹⁶ To our knowledge, our study is the first which sought to decrease pre-procedural fasting for hospitalized children for whom scheduling is a more complex process than for patients undergoing outpatient procedures.

While allowing clear fluids overnight may not be as beneficial for older patients who sleep through the night, allowing infants (26% of our intervention group) and young children to consume breastmilk and/or clear fluids for as long as possible is clinically meaningful. Specifically, studies have shown improved hemodynamic stability during sedation for young children with limited glycogen stores.¹⁷ Shorter fasting times also improve patient, family, and healthcare staff satisfaction. While not formally evaluated in our study, anecdotally patients and families were happier with shorter fasting times which correlates with previous studies of decreased perceived hunger when clear liquids are provided closer to sedation.¹⁰

Complexities of large medical centers can, at times, complicate providers’ efforts to provide efficient, patient-centered care. This is particularly true for hospitalized children who may require sedation during admission. Multiple teams must balance the uncertainty of procedure schedules with the patients’ ability to eat and drink as long possible. Our team created a shared mental model around earliest procedure start time (7 am) which allowed higher-reliability EHR interventions to more easily follow the “6,4,2” rule. These institutional level changes produced an environment which allowed children to drink clear fluids for longer and maintained safe sedation and scheduling flexibility resulting in only 4 minor case delays (not related to the NPO order) and no aspiration events. However, as a quaternary care level 1 trauma center, the entire system has to accommodate everything from scheduled outpatient procedures to acute needs of PHM patients to the potential trauma patient requiring emergent imaging and/or surgery. Such complexities in accommodating emergencies may impede achieving a fasting goal of 5 hours for our patients on average, but is still possible for individual patients.

This study has several limitations. First, because it was performed at a single university affiliated, quaternary care children’s hospital with a strong QI infrastructure and culture, it may not be generalizable to all institutions. Additionally, EHR changes may not be feasible in all settings. However, even without the use of an EHR, creating standardized procedures, language, and order sets can create higher-reliability environments which make it easier for people to perform desired actions. Finally, we do not know the exact time that children took their last bite of food or sip of a liquid as we were limited to using data that could be pulled directly from the EHR which may not always reflect patient-specific behaviors.

CONCLUSIONS

Utilizing a high-reliability change in the EHR ordering system, our QI initiative resulted in a 4 hour decrease in pre-procedural fasting time for hospitalized children. Future work involves improving communication between teams to allow clear liquid fasting time extensions for procedures scheduled later in day. Our hope is that a paradigm shift away from a non-evidence based medical dogma (“NPO at midnight”) bolstered by QI methodology will improve the pre-procedural experience for all hospitalized children at our hospital.

Supplementary Material

Appendix Figure 1

Updated NPO patient door signs.

Abbreviations:

PDSA

Plan-Do-Study-Act

PHM

Pediatric Hospital Medicine

NPO

Nil per os

QI

Quality Improvement

ASA

American Society of Anesthesiologists

EHR

Electronic health record

NP

Nurse Practitioner

US

United States