A quality improvement strategy to reduce unintended extubation in the very low birth weight infant: A case report

Abstract

BACKGROUND:

Unintended extubations remain a common complication across neonatal intensive care units, with very low birthweight infants being the most vulnerable of them all. Ongoing efforts across different institutions exist with the goal of reducing the rate of unintended extubations to keep a median rate of <2 events per 100 ventilator days as defined by the Vermont Oxford Network. Our objective was to reduce unintended extubations in the very low birthweight infant in a large delivery hospital to ≤2/100 ventilator days.

METHODS:

A collaborative group was formed between two academic health institutions targeting training and implementation of the Children’s National unintended extubation system, focusing on endotracheal tube securement methods and surveillance protocols.

RESULTS:

The unintended extubation rate decreased from 3.23 to 0.64 per 100 ventilator days. Changes were implemented from 2018–2020 with a sustained reduction in the unintended extubation rate of 1.54 per 100 ventilator days. Most events occurred between 12 : 00 pm –4 : 00 pm and the commonest cause was spontaneous (25%) followed by dislodgment during repositioning (19%).

CONCLUSION:

Very low birth weight infants present a challenge to endotracheal tube maintenance due to their developmental and anatomical changes during their neonatal intensive care unit stay. Successful reduction of unintended extubations in the very low birthweight infant can be achieved by adaptation of successful protocols for older infants.

Keywords

Collaborative unplanned extubations very low birthweight infant

1 Introduction

Unintended extubation (UE) are defined as any unintentional dislodgement of an endotracheal tube (ETT) from the airway [1]. Despite different efforts to reduce the incidence of UE events, they remain a common complication across Neonatal Intensive Care Units (NICU). When UE happen, they may lead to the development of major morbidities such as airway trauma, intraventricular hemorrhage and cardiorespiratory collapse events. UE events have also been linked to excessive use of resources and increased costs in the NICU [2 –7].

Very low birth weight (VLBW) infants ≤1500 g, are the most vulnerable of all infants in the NICU for UE when being mechanically ventilated as they present a bigger challenge with adequate ETT positioning and securement due to their smaller faces, presence of abundant amounts of secretions leading to moist and wet tape requiring frequent change, and significant developmental and anatomical changes that occur during their stay in the NICU [3 , 9]. Furthermore, the limited evidence supporting safe use of sedatives in mechanically ventilated premature infants and the wide clinical practice variability, their use in the VLBW infant in efforts to help prevent UE events [9, 10]. Additionally, the fragile and small airway in the VLBW infant is more susceptible to trauma with urgent or prolonged intubations, further increasing their vulnerability to morbidities such as Bronchopulmonary Dysplasia (BPD), present in 16% of our VLBW infants [7 , 11–13].

Ongoing efforts across different institutions exist with the goal of reducing the rate of UE to keep a median rate of <2 events per 100 ventilator days as defined by the Vermont Oxford Network (VON) [14].

In efforts to achieve a sustained reduction in UE from an average of 4 events/100 ventilator days after multiple previous failed interventions, the Children’s National Medical Center (CNMC) was sought out for collaboration shortly after reporting a successful significant sustained reduction to <1 events/100 ventilators days post development of original interventions over a 10-year period span later adopted by larger collaboratives [15].

As a Children’s Hospital with a moderate sized delivery service, we describe our collaborative efforts with CNMC in Washington, DC in successfully adapting their potentially better practices for the VLBW infant in our NICU with an aim of reducing our UE rate to <2 events/100 ventilator days and by 50% within a year of implementing new potentially better practices.

2 Methods

2.1 Context

The Children’s Hospital of Richmond at Virginia Commonwealth University (CHoR at VCU) NICU is a 40-private room, Level IV NICU with ≥450 admissions a year with approximately 67% (300/450) inborn, and 27% (80/300) are inborn VLBW infants. The CNMC NICU is an urban, academic, 66-bed, level IV non-delivery NICU that admits >900 infants per year with 100–150 VLBW infants per year admitted at less than 7 days of age (16%) [15]. At the time of the collaborative, care for the intubated VLBW infant remained constant with no other changes introduced. Our typical RN to intubated VLBW infant is a 1–2 ratio, with hourly airway checks and hands-on care every 3–6 hours. Our use of humidity in the VLBW infants remained constant, starting with 60–80%, using the higher humidity level in the smaller VLBW infants initially, with a gradual reduction over the first week of life. This project was exempted from VCU’s Institutional Review Board or CNMC’s Institutional Review Board.

2.2 Planning of intervention

Baseline data was collected from the years 2015–2017 using chart review and bedside UE event forms that were kept over these years. During this time, monthly ventilator days were not tracked, however, baseline data revealed an average of 4 UE events per month over three years. The different interventions implemented during this time were the following: Two-person care while obtaining imaging at bedside, and/or while repositioning of infant. Use of an ETT tube holder that eventually became discontinued by manufacturer, and lastly the use of 3Mtrademark Duraporetrademark tape (purple tape) which proved to be unsuccessful in adequately securing the ETT tube during humidification (Table 1). Retrospective data was reviewed to determine contributing factors leading to UEs (Fig. 1) anda multidisciplinary team reviewed the data during the baseline period and formulated a key driver diagram to reduce the UE rate (Fig. 2).

Table 1
Interventions were first implemented in November of 2015 starting with adopting a 2-person care protocol

Interventions Date

First 2-person care protocol implemented November 2015

ET-Holder trials May 2016

Cavilon skin prep protocol implemented September 2016

3Mtrademark Duraporetrademark tape (purple tape) trial November 2017

Collaborative Initiative begins September 2018

Training at Children’s National Medical Center September 2018

Training on ETT^* securement method and simulation training at CHoR at VCU^* October 2018

ETT taping protocol implemented November 2018

Reporting of intubated patients and UE events during safety huddles begins November 2018

Two-person handling protocol re-implemented December 2018

Refresher training January 2019

High risk identifying visual tools implemented July 2019

Interventions	Date
First 2-person care protocol implemented	November 2015
ET-Holder trials	May 2016
Cavilon skin prep protocol implemented	September 2016
3Mtrademark Duraporetrademark tape (purple tape) trial	November 2017
Collaborative Initiative begins	September 2018
Training at Children’s National Medical Center	September 2018
Training on ETT^* securement method and simulation training at CHoR at VCU^*	October 2018
ETT taping protocol implemented	November 2018
Reporting of intubated patients and UE events during safety huddles begins	November 2018
Two-person handling protocol re-implemented	December 2018
Refresher training	January 2019
High risk identifying visual tools implemented	July 2019

The collaborative began in September of 2018. ^*ETT: Endotracheal Tube, CHoR at VCU: Children’s Hospital of Richmond at Virginia Commonwealth University.

Fig. 1

Pareto Chart-Contributing factors leading to unintended extubations in the years 2018–2020.

Fig. 2

Driver diagram for reduction of unintended extubations in the NICU.

2.3 Intervention

A quality improvement (QI) multidisciplinary team for UEs was formed at CHoR at VCU’s NICU consisting of neonatologists, nurses (RN), and respiratory therapists (RT) to target reduction and prevention of UE primarily in the intubated VLBW infant. The team decided to approach the problem by adopting previously successful ways of preventing UE by other institutions and identified CNMC as an excellent collaborator due to their successful reduction in and sustained low UE rates over several years. Training was obtained on CNMC’s UE prevention model, with the focus on the VLBW infant through standardization of successful non-existent ETT securing methods at CHoR at VCU and included: standardized ETT taping and maintenance as well as continued monitoring and surveillance protocols. After both health systems agreed to collaborate, both teams met in person and initial training occurred at the CNMC’s NICU utilizing pre-recorded visuals describing their ETT taping method followed by a practical hands-on experience with the use of low fidelity mannequins, and 3Mtrademark cloth tape. Targeted unit wide training was conducted by UE champions representatives of each discipline in two phases: Phase one followed the first training obtained at CNCM and consisted of the usage of pre-recorded material and low fidelity mannequins for hands-on experience with an initial focus on ETT taping method. This phase was completed within a month of first training and allowed the team to train >100 RNs, 22 dedicated pediatric/NICU RTs, and 23 providers (9 neonatologists, 9 neonatal nurse practitioners, and 5 neonatology fellows). Phase two consisted of retraining RT within two months post implementation using low fidelity mannequins. This additional 2nd training refresher was necessary as securement of the ETT is primarily done by RT, sometimes with the aid of an RN, and was aimed to reinforce the knowledge and skills they had obtained from first training. The rest of the staff and provider team members were retrained on a need-to basis. Other important interventions forming part of the CNMC UE prevention model such as 2 people handling of the intubated VLBW infant, 2 person ETT check, and implementation of a high-risk identification tool aimed at identifying individual risk factors and implemented by RN in collaboration with RT at least once per shift, were addressed in the second phase.

The high-risk identification tool consisted of an animal paw printed in three different colors (Green to mark low risk, Yellow to mark medium risk, and Red to mark high risk), that was laminated and displayed at the patient’s door where it could be visible for all those who entered their room.

In efforts to hold accountability and improve the UE tracking method, the team incorporated the number of intubated patients and number of UE within a 24-hour period to be reported to providers during a twice a day multidisciplinary unit safety huddle prior to the start of day and evening patient rounds. The change was first implemented on 11/2018, within a month from initial RT training.

2.4 Measures

The rate of UE defined as the number of UE per 100 ventilator days was the primary measure excluding tracheostomies. Extubation data was derived from event notes entered into the electronic medical record crosschecked with post event forms completed by the bedside RN to confirm accuracy of documentation. Ventilator days were derived from the electronic medical record in the months that they became available, and once both sets of data were obtained, the UE rate was manually calculated. Appropriate ETT securement, implementation of CNMC UE prevention model and use of the high-risk tool were randomly audited by one of the UE champions at least once a month during phase two to measure compliance and recognize opportunities for change.

2.5 Analysis

Statistical analysis of demographic data was performed using chi-square and t-tests where appropriate. Process control charts were developed using QI Macros 2019 (version 2019.06). A U-chart was created to analyze the process as there were different number of ventilator days per month with unequal opportunity of events. Rules for shift were used.

3 Results

During the initiative, there were 1222 admissions with 972/1222 (80%) being inborn. Of the 972, 25% were (244/972) VLBW infants born between the years 2018 and 2020. During their NICU stay 61% (151/244) VLBW infants were intubated at least once and were included in the analysis. During this period, 110/150 (73%) never experienced an UE, while 40/150 (27%) of the infants had at least one UE event with 9/40 (22%) infants having more than one event. Those with an UE were overall of lower gestational age at birth (25±2 vs 26±3, p = 0.04) and lower birth weight (740±207 g vs 935±318, p = 0.001). They were on mechanical ventilation for longer periods of time (84 days±85 vs 9 days±12, p = 0.0002), and 14/22 (63%) required immediate reintubation (Table 2). At the time of the first UE the mean postmenstrual age (PMA) was 31 weeks with a mean weight of 1314 g (Table 2).

Table 2
Clinical characteristics of infants with and without Unintended Extubations

Clinical Characteristics No UEs UE p-value

(N = 110) (N = 41)

Gestational Age at birth (weeks), mean (SD) 26±3 25±2 0.04^

Birth weight at birth (g), mean (SD) 935±318 740±207 0.001^

Gender Male, n (%) 54(49) 24 (59) 0.33

Race

White n (%) 42(38) 10 (24) 0.09

Black n (%) 49 (45) 24 (59)

Apgar 5-min, mean (SD) 7±2 7±2 0.35

Ventilator days, mean (SD) 9±12 84±85 0.0002^**

Outborn, n (%) 22(20) 9 (22) 0.81

Postmenstrual age (weeks) at time of first UE^*, 28 (26;33)

median (IQR)

Weight (g) at time of first UE, median (IQR) 710 (628;841)

Clinical Characteristics	No UEs	UE	p-value
Gestational Age at birth (weeks), mean (SD)	26±3	25±2	0.04^**
Birth weight at birth (g), mean (SD)	935±318	740±207	0.001^**
Gender Male, n (%)	54(49)	24 (59)	0.33
Race
White n (%)	42(38)	10 (24)	0.09
Black n (%)	49 (45)	24 (59)
Apgar 5-min, mean (SD)	7±2	7±2	0.35
Ventilator days, mean (SD)	9±12	84±85	0.0002^**
Outborn, n (%)	22(20)	9 (22)	0.81
Postmenstrual age (weeks) at time of first UE^*,	28 (26;33)
median (IQR)
Weight (g) at time of first UE, median (IQR)		710 (628;841)

^*UE-Unintended extubation, ^**p-value < 0.05. Gestational age, birthweight and ventilator days were reported to be significantly different in those with and without UE. Data is described as either total (n) and percentages (%), means and standard deviations (SD) or median and interquartile range (IQR).

Over the period of Nov 2018–Nov 2020, the intervention resulted in a 69% reduction of the UE rate from 3.23 to 1.01 per 100 ventilator days (Fig. 3) and an average of 16±21 days in between events (Not Shown). Most events during the study period were noted to occur from 12 : 00 pm to 4 : 00 pm (p = 0.12) (Not Shown). A total of 20 random audits were performed during phase 2 of the initiative, which identified adequate compliance with proper securement of the ETT, but inconsistency with completion of the high-risk tool in 50% of the time (Not Shown).

Fig. 3

U-chart showing number of unintended extubation events during 2018–2020.

4 Discussion

Presented is a collaborative QI project amongst two distinct academic institutions that led to a successful sustained reduction in the VLBW infant UE rate to <2 events per 100 ventilator days surpassing the set goal of 50% reduction.

Studies have targeted reduction of UE events in the NICU with the use of bundles of care with some demonstrating success in their use, and some demonstrating no change in the outcome of reducing UE events, but none to our knowledge have specifically addressed the challenges faced in reducing UE events in institutions with a large inborn population of the VLBW infant or addressed their unique challenges [15 –19].

Klugman et al. through a multicenter quality improvement initiative showed the benefits of implementing a quality improvement bundle in the reduction of UE at several Children’s Hospitals through a Solutions for Patient Safety collaborative, but their study did not report on patient-specific data and demographics, not allowing for a better understanding of specific associations in different patient populations, and each participating NICU adapted their own UE bundle [1]. Galiote et al. showed a significant reduction in UE rates to below the suggested benchmark of <1 event per 100 ventilator days via standardization of interventions in a mixed population of infants, but this was in a level IV NICU without a delivery service [15].

Although most of our events occurred at a later PMA and increased weight in comparison to birth, most infants in contrast to those described by Galiote et al. (median PMA of 35 weeks and mean weight of 2108 g) and most recently by Pavlek et al. (median PMA of 34 weeks and weight of 1775 g), were of smaller gestational age (median PMA of 28 weeks) and remained below 1500 g at time of first event (median weight at time of event of 710 g) [6, 15].

Addressing this issue becomes of outmost importance for the VLBW infant despite the PMAat which the event might occur, due to the already increased risk of significant morbidities seen with repeated UE events such as airway trauma, prolonged ventilator days, BPD, intraventricular hemorrhage, and prolonged hospitalization as described by Pavlek et al. most recently and others [6 , 20].

Creating a successful collaboration amongst two distinct institutions and standardizing the ETT fixation method are the major strengths of this quality improvement initiative and drivers to achieving the set goal. Through collaboration, the time needed to create and initiate interventions was shortened by adopting already successful ones. Success was maintained through standardization of the ETT fixation method decreasing challenges faced with retaping and readjustment of the ETT by RT. This led to a significant change in the unit’s culture and awareness regarding the importance of maintaining and adequately securing an ETT in an already high-risk population.

The benefits of standardizing ETT fixation targeting a broader neonatal population have been addressed by others such as Galiote et al, and Crezee et al., with Merkel et al. describing the importance of culture change to attain success, but none to the group’s knowledge have specifically addressed the VLBW infant [15 , 21].

As noted in the data, inappropriate fixation of the ETT was a cause of UE, a challenge described by others such as Veldman et al., Ridore et al., and Powell et al. [7 , 22]. This finding led efforts and focus in achieving the set goal by accomplishing appropriate ETT fixation and standardizing management in the VLBW infant who already poses a unique challenge compared to larger infants due to their anatomical and physiological limitations, such as having smaller faces and increased oral secretions [9].

The taping method was preferred over other securing devices for two main reasons: 1) CHoR at VCU’s NICU lacked one standard way of taping the ETT and has had previous failed attempts at usage of other securement devices which led to trialing a more simplistic and readily available method. 2) There has been inconsistent success with the use of other securement devices other than the use of tape, such as that described by Loganathan et al. who demonstrated no benefit in the use of commercial ETT securement devices over taping in prevention of UE event in contrast to Pavlek et al’s experience with the use of a commercial ETT holder [6, 23].

Sustainability as described by Mortimer et al. is of importance when one does QI [24]. The following changes were critical to maintain sustainability and another strength noted in this QI collaborative: 1) The simultaneous use of simulation with a low fidelity mannequin. This served a pivotal role in obtaining and maintaining the necessary skills and techniques in a short period of time to properly care for the intubated VLBW infant. 2) Retraining was necessary to reinforce the skills and knowledge of the RT, who are the primary individuals securing and maintaining the ETT. This was done as well in efforts to sustain a downtrend in the number of UE events. 3) Incorporating the routine report of the total number of intubated VLBW infants and total number of UE events that occurred in the previous 24 hours into the twice a day safety huddle was essential in creating situational awareness and change in culture. 4) The addition of a high-risk tool adapted to identify individual risk factors in the VLBW infant for UE, used at a minimum of twice a day by RN and RT focused on improving multidisciplinary communication and increase awareness to other disciplines participating in care of the infant. This was reinforced using the colored visuals at the patient’s room entrance. 5) As noted in the audit results, compliance during phase 2 was at 50%. This led to re-education of staff on the utility of the tool leading to an increase in report of the score during provider rounds.

Despite being able to significantly reduce the UE rate to below the set goal, there were limitations. Achieving a benchmark rate of <1 UE/100 ventilator days as set in the literature has been the most important limitation [17]. Our limitation was a consequence of the following: 1) Method of communication: communication with the rest of the staff to refresh their knowledge and demonstrate the result of their efforts has been via electronic mail (e-mail), but due to the high volume of e-mails staff receive during the day, it has created a challenge in effectively mass deliver the information needed to sustain numbers and maintain momentum. 2) RN workload: the importance of the effect that RN workload has on reducing UE events has been described by Barber J. A. and Veldman et al. [3, 7]. As noted in the data, although not statistically significant most events occurred during the hours of 12 : 00 pm to 4 : 00 pm followed by 8 : 00 am to 12 : 00 pm. It is during this time that most of the staff are either on a break or changing shifts, reducing the number of RN available to care for these infants. This creates an opportunity to further investigate the effect of time of day on the risk of UE. 3) Post event feedback: There is a need for a standardized approach in addressing the event immediately after it occurs. 4) Lastly, lack of staff satisfaction measurement and sedation use. Addressing these limitations are believed to lead to better sustainability and continued reduction in the UE rate.

5 Conclusion

Focused collaboration amongst two distinct institutions with the goal of reducing UE, a significant deleterious event in the VLBW infant, is feasible, effective, and highly encouraged. The benefits and spread of knowledge learned from one another allow for a significant reduction in the amount of time needed to implement successful change. It also allows for exploration of different avenues for a focused common problem. Future efforts should continue to focus on prevention of UE in the VLBW infant by addressing appropriate ETT fixation and RN workload as well as improving effective communication amongst different teams to sustain a positive change.

Footnotes

Acknowledgments

The authors want to thank the staff at both academic institutions for the continuous efforts in promoting best practice for our patients, and to our patients and their families for trusting us in providing best practice.

Funding

No funding was secured for this study.

Financial disclosures

All authors have no financial relationships relevant to this article to disclose.

Ethics approval

This project was conducted according to local institutional standards for QI initiatives and therefore IRB approval was not needed.

Conflict of interest

All authors have no conflicts of interest to disclose.

Availability of data

All study data are available upon request by contacting the corresponding author.

References

Klugman

, Melton

, Maynord

, Dawson

, Madhavan

, Montgomery

et al. Assessment of an unplanned extubation bundle to reduce unplanned extubations in critically ill neonates, infants, and children. JAMA Pediatr. 2020;174(6):e200268.

Meyers

, Pinheiro

, Nelson

. Unplanned extubation in NICU patients: Are we speaking the same language? J Perinatol. 2015;35(9):676–7.

Barber

. Unplanned extubation in the NICU. J Obstet Gynecol Neonatal Nurs. 2013;42(2):233–8.

Hatch

, 3rd Scott

, Slaughter

, Xu

, Smith

, Stark

et al. Outcomes, resource use, and financial costs of unplanned extubations in preterm infants. Pediatrics. 2020;145(6):e20192819.

Sharek

, Horbar

, Mason

, Bisarya

, Thurm

, Suresh

, et al. Adverse events in the neonatal intensive care unit: Development, testing, and findings of an NICU-focused trigger tool to identify harm in North American NICUs. Pediatrics. 2006;118(4):1332–40.

Pavlek

, Dillard

, Ryshen

, Hone

, Shepherd

, Moallem

Short-term complications and long-term morbidities associated with repeated unplanned extubations. J Perinatol [Internet]. 2021;41(3):562–70.

Veldman

, Trautschold

, Weiss

, Fischer

, Bauer

Characteristics and outcome of unplanned extubation in ventilated preterm and term newborns on a neonatal intensive care unit. Paediatr Anaesth.. 2006;16(9):968–73.

Silva

, Reis

, Aguiar

, Fonseca

Unplanned extubation in the neonatal ICU: A systematic review, critical appraisal, and evidence-based recommendations. Respir Care. 2013;58(7):1237–45.

Hatch

, Grubb

, Markham

, Scott

, Walsh

, Slaughter

, et al. Effect of anatomical and developmental factors on the risk of unplanned extubation in critically ill newborns. Am J Perinatol. 2017;34(12):1234–40.

10.

Zimmerman

, Smith

, Benjamin

, Laughon

, Clark

, Traube

, et al. Sedation, analgesia, and paralysis during mechanical ventilation of premature infants. J Pediatr. 2017;180, 99–104.e1

11.

Chawla

, Natarajan

, Shankaran

, Carper

, Brion

, Keszler

, et al. Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network. Markers of successful extubation in extremely preterm infants, and morbidity after failed extubation. J Pediatr. 2017;189, 113–19.e2.

12.

Berger

, Mehta

, Bucholz

, Dziura

, Bhandari

Impact of early extubation and reintubation on the incidence of bronchopulmonary dysplasia in neonates. Am J Perinatol. 2014;31(12):1063–72.

13.

Adams

, Bassler

, Bucher

, Roth-Kleiner

, Berger

, Braun

, et al. Swiss Neonatal Network and the Vermont Oxford Network. Variability of very low birth weight infant outcome and practice in swiss and us neonatal units. Pediatrics. 2018;141(5):e20173436.

14.

Aydon

, Zimmer

, Sharp

Reporting the incidence of unplanned extubation in the neonatal intensive care unit. J Paediatr Child Health. 2018;54(7):784–7.

15.

Galiote

, RidoréM , Carman

, Zell

, Brant

, Gayle

, et al. Reduction in unintended extubations in a level IV neonatal intensive care unit. Pediatrics. 2019;143(5):e20180897.

16.

Powell

, Gilbert

, Volsko

. Reducing unplanned extubations in the NICU using lean methodology. Respir Care. 2016;61(12):1567–72.

17.

Merkel

, Beers

, Lewis

, Stauffer

, Mujsce

, Kresch

Reducing unplanned extubations in the NICU. Pediatrics.e. 2014;133(5):1367–72.

18.

Morii

. Prevention strategies for unplanned extubation in NICU –A literature review. J Neonatal Nurs. 2016;22(3):91–102.

19.

Ferraz

, Barros

, Miyoshi

, Davidson

, Guinsburg

. Bundle to reduce unplanned extubation in a neonatal intensive care unit. J Matern Fetal Neonatal Med. 2020;33(18):3077–85.

20.

Greenough

, Prendergast

. Difficult extubation in low birthweight infants. Arch Dis Child Fetal Neonatal Ed.. 2008;93(3):F242–5.

21.

CrezeéKL , DiGeronimo

, Rigby

, Carter

, Patel

Reducing unplanned extubations in the NICU following implementation of a standardized approach. Respir Care. 2017;62(8):1030–5.

22.

Ridore

, Carman

, Zell

, Brown

A-M

, Blake

, Gayle

, et al. Reducing unintended extubation rates in the children’s national neonatal intensive care unit (NICU): A quality improvement project. Pediatrics. 2018;143(5):e20180897.

23.

Loganathan

, Nair

, Vine

, Kostecky

, Kowal

, Soraisham

. Quality Improvement Study on New Endotracheal Tube Securing Device (Neobar) in Neonates. Indian J Pediatr. 2017;84(1):20–4.

24.

Mortimer

, Isherwood

, Wilkinson

, Vaux

. Sustainability in quality improvement: Redefining value. Future Healthc J. 2018;5(2):88–93.