Paediatrics and International Child Health
ISSN: 2046-9047 (Print) 2046-9055 (Online) Journal homepage: http://www.tandfonline.com/loi/ypch20
Impact of an Emergency Triage Assessment and Treatment (ETAT)-based triage process in the paediatric emergency department of a Guatemalan public hospital Heather L. Crouse, Francisco Torres, Henry Vaides, Michael T. Walsh, Elise M. Ishigami, Andrea T. Cruz, Susan B. Torrey & Miguel A. Soto To cite this article: Heather L. Crouse, Francisco Torres, Henry Vaides, Michael T. Walsh, Elise M. Ishigami, Andrea T. Cruz, Susan B. Torrey & Miguel A. Soto (2016) Impact of an Emergency Triage Assessment and Treatment (ETAT)-based triage process in the paediatric emergency department of a Guatemalan public hospital, Paediatrics and International Child Health, 36:3, 219-224, DOI: 10.1179/2046905515Y.0000000026 To link to this article: http://dx.doi.org/10.1179/2046905515Y.0000000026
Published online: 06 Sep 2016.
Submit your article to this journal
Article views: 109
View related articles
View Crossmark data
Citing articles: 1 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ypch20 Download by: [Southern Cross University]
Date: 10 October 2017, At: 00:43
Original Research Paper
Impact of an Emergency Triage Assessment and Treatment (ETAT)-based triage process in the paediatric emergency department of a Guatemalan public hospital Heather L. Crouse1, Francisco Torres2, Henry Vaides2, Michael T. Walsh3, Elise M. Ishigami3, Andrea T. Cruz1, Susan B. Torrey1,4, Miguel A. Soto2 Department of Pediatrics, Section of Emergency Medicine, Baylor College of Medicine, Houston, Texas, USA, Department of Pediatrics, Hospital Nacional Pedro Bethancourt, La Antigua, Guatemala, 3Global Health Initiative, Texas Children’s Hospital, Houston, Texas, USA, 4Department of Emergency Medicine, Division of Pediatric Emergency Medicine, New York University School of Medicine, New York City, USA
Downloaded by [Southern Cross University] at 00:43 10 October 2017
1
2
Background: Triage process implementation has been shown to be effective at improving patient outcomes. This study sought to develop, implement and assess the impact of an Emergency Triage Assessment and Treatment (ETAT)based emergency triage process in the paediatric emergency department (PED) of a public hospital in Guatemala. Methods: The study was a quality improvement comparison with a before/after design. Uptake was measured by percentage of patients with an assigned triage category. Outcomes were hospital admission rate, inpatient length of stay (LOS), and mortality as determined by two distinct medical record reviews for 1 year pre- and post-intervention: a random sample (RS) of all PED patients and records for all critically-ill (CI) children [serious diagnoses or admission to the paediatric intensive care unit (PICU)]. Demographics, diagnoses and disposition were recorded. Results: The RS totalled 1027 (51.4% male); median ages pre- and post-intervention were 2.0 and 2.4 years, respectively. There were 196 patients in the CI sample, of whom 56.6% were male and one-third were neonates; median ages of the CI group pre- and post-intervention were 3.1 and 5.6 months, respectively. One year after implementation, 97.5% of medical records had been assigned triage categories. Triage categories (RS/CI) were: emergency (2.9%/54.6%), priority (47.6%/44.4%) and non-urgent (49.4%/1.0%). The CI group was more frequently diagnosed with shock (25%/1%), seizures (9%/0.5%) and malnutrition (6%/0.5%). Admission rates for the RS (8% vs 4%, P=0.01) declined after implementation. For the CI sample, admission rate to the PICU (47% vs 24%, P=0.002) decreased and LOS (7.3 vs 5.7 days, P=0.09) and mortality rates (12% vs 6%, P=0.15) showed trends toward decreasing post-implementation. Conclusions: Paediatric-specific triage algorithms can be implemented and sustained in resource-limited settings. Significant decreases in admission rates (both overall and for the PICU) and trends towards decreased LOS and mortality rates of critically ill children suggest that ETAT-based triage systems have the potential to greatly improve patient care in Latin America. Keywords: Paediatric emergency medicine, Paediatric triage, International emergency medicine, Emergency Triage Assessment and Treatment (ETAT), Clasificación, Evaluación y Tratamiento de Emergencias Pediátricas (CETEP) Abbreviations (in order of reference: ETAT; Emergency Triage Assessment and Treatment; HCW, shealthcare workers; HNPB, Hospital Nacional Pedro Bethancourt; BCM/TCH, Baylor College of Medicine/Texas Children’s Hospital; MoH, Guatemalan Ministry of Health; PAHO, Pan-American Health Organization; CETEP, Clasificación Evaluación y Tratamiento de Emergencias Pediátricas; PED, paediatric emergency department; PICU, paediatric intensive care unit; QI, quality improvement; LOS, inpatient length of stay; RS, random sample; CI, critically ill sample
There are no known conflicts of interest associated with this publication and there has been no significant financial support for this work which could have influenced its outcome. Sources of support: (internal grant) Texas Children’s Hospital Faculty Educational Grants 2010 and 2011 Correspondence to: Heather Crouse, 6621 Fannin Street, Suite A2210, Houston, Texas, USA 77030. Fax: 832 825 5424; email: [email protected]
© 2016 Informa UK Limited, trading as Taylor & Francis Group DOI 10.1179/2046905515Y.0000000026
Paediatrics and International Child Health 2016 VOL. 36
NO.
3
219
Crouse et al. Implementation of paediatric triage
Introduction
Downloaded by [Southern Cross University] at 00:43 10 October 2017
Early recognition and stabilization of acutely ill infants and children improves outcomes in all settings, regardless of resources. When resources are constrained, interventions which include the introduction of triage training and process, the use of clinical practice guidelines and the supervision and monitoring of patients lead to reduced mortality rates in critically ill children.1−5 In a cluster randomized trial in district hospitals in rural Kenya, similar interventions reduced mortality and were cost-effective. 6,7 Although triage is an effective process that requires few resources to implement, most healthcare facilities in developing countries have no formal triage system.8−11 The acuity of patients presenting to paediatric emergency departments varies widely. Although many children require only basic care, those with emergency medical conditions often experience delays in initial treatment. Consequently their conditions deteriorate, resulting in admissions that may have been avoidable. None of the 18 randomly selected facilities described in an assessment of the quality of care for hospitalized children in Bangladesh had a functioning triage system.12 These reports suggest that locally driven efforts to implement triage systems could be effective and feasible.13,14 Reports describing triage systems in Latin America are limited. Interventions (including prioritizing care for children in emergency departments) implemented at 14 Nicaraguan hospitals improved outcomes without increasing hospital costs.15 Previous studies of triage interventions in resource-limited settings have used the Emergency Triage Assessment and Treatment (ETAT) programme developed by the World Health Organization. This strategy was developed to promote improved assessment, triage and initial management of critically ill children in resource-limited settings by teaching healthcare workers (HCWs) and those with limited healthcare training how to rapidly assign a triage category of ‘emergency’, ‘priority’ or ‘non-urgent’ to patients based on easily identified risk factors and clinical signs.16−18 Limited experience suggests that implementation of ETAT based triage systems improves outcomes in critically ill children in Africa;4,19,20 however, it is not available in Spanish. As part of a collaborative effort between the Hospital Nacional Pedro Bethancourt (HNPB, a public hospital in Antigua, Guatemala), Baylor College of Medicine/ Texas Children’s Hospital (BCM/ TCH), the Guatemalan Ministry of Health (MoH) and the Pan-American Health Organization (PAHO), a self-sustaining Spanish ETAT [Clasificación Evaluación y Tratamiento de Emergencias Pediátricas (CETEP)] training programme was developed and implemented at HNPB.21 The aim was to implement a collaboratively developed CETEP-based triage process and to evaluate its impact on process and patient outcomes in the paediatric emergency department (PED) at HNPB.
220
Paediatrics and International Child Health 2016 VOL. 36
NO.
3
Methods
This quality improvement study compared a retrospective pre-intervention sample with a similar prospective sample which was identified and analysed after the triage process was implemented in the PED. The two primary objectives of the study were to assess uptake of the triage process and to assess its impact on patient outcomes. The choice of measures to assess clinical uptake and patient outcomes was limited to data that were routinely collected in the PED at HNPB. The number of PED patients in the post-intervention sample to whom a triage category has been assigned was used as a process measure to indicate uptake. Hospital admission rate was chosen as the primary patient outcome measure. Investigators in sub-Saharan Africa have used death within 2 days of hospital admission as a surrogate for the effectiveness of initial emergency care.1,20 However, child mortality rates in Central America generally and in Guatemala specifically are lower than those in sub-Saharan Africa.22 As a result, the sample size required or order to use mortality rate as an outcome measure would have been unrealistic for this project.
Setting and study population
The study was conducted at Hospital Nacional Pedro Bethancourt (HNPB), a secondary level public hospital in La Antigua, Guatemala, located approximately 45 kilometres from Guatemala City. It is the referral hospital for the Department of Sacatepe′quez and surrounding areas serving a population of approximately 329,547. The PED evaluates approximately 20,000 children aged 0–12 years annually. The PED is open 24 hours/day and is staffed primarily by medical students, paediatric residents, one supervising general paediatrician and one or two nurses per shift. HNPB inpatient paediatric services include a 30-bed inpatient paediatric ward, a 30-bed neonatology ward and small neonatal and paediatric intensive care units. Before the intervention, the average rate of admission to the paediatric ward from the PED was 10%. All sub-speciality needs are referred to Roosevelt Hospital in Guatemala City. The hospital has an established paediatric residency training programme and has developed several innovative programmes including development of the first certified human milk bank in Latin America. Two groups of patients who presented to the PED for care between 1 November 2009 and 31 October 2011 were selected for the study: a random sample (every 40th patient) of all patients (defined by HNPB as age 0–12 years and identified here as RS) and all who were critically ill [defined as those with an emergency diagnosis per CETEP or who were admitted to the Paediatric Intensive Care Unit (PICU) as recorded in the PED logbook and identified here as CI]. It was decided to evaluate all critically ill children (CI) separately because the number of
Downloaded by [Southern Cross University] at 00:43 10 October 2017
Crouse et al. Implementation of paediatric triage
critically ill children presenting to the PED is typically small and therefore there might not have been enough emergency patients in the random sample to adequately assess the impact of the triage process on this important population.
after implementation, paper supplies ran low. As a result, triage forms were unavailable and HCWs used the new triage system less often. In response, the team developed a low-cost triage stamp containing the same information that was in all the PED records.
Intervention
Measurements
A practical application of CETEP training is the implementation of a triage process that defines specific triage categories (red for emergency, yellow for priority and green for non-urgent). Training of paediatric HCWs at HNPB in CETEP began in June 2010. By November 2010, 80% (n=590) of paediatric HCWs were trained. The accuracy of HCW triage assessed immediately after CETEP training was found to be 95.1% (CI 91.9–97.3).21 Preparatory to development of the triage system, in November 2010, a pilot study was initiated by HNPB HCWs in the PED to monitor patient flow and acuity, staffing and room availability for triage. With this information, a multi-disciplinary team developed and implemented a formal three-level, CETEP-based triage process in the PED as part of a larger, institutional quality improvement (QI) project. The team consisted of HNPB PED nurses, paediatric residents, general paediatricians, the head of paediatrics and representatives from TCH/BCM. First, he team defined ideal patient flow for the PED on the basis of CETEP principles: all patients presenting to the PED were initially evaluated by a designated ‘triager’, usually a senior medical student or resident. Patients classified as ‘emergency’ were put directly in the ‘shock’ room for immediate stabilization. Patients classified as ‘priority’ were put in a room, if available, or in the waiting room, but were moved ahead in the queue so that management could begin as quickly as possible. Non-urgent patients presenting to the PED between 7 and 11 a.m. were sent directly to the paediatric outpatient department for evaluation. At other times, they were put in the PED waiting room for consultation in order of presentation. Next, the team developed a paper triage form to be attached to all PED patient records; it included patient name, date of birth, time of triage, triage category and reason for that level. It is important to recognize that this new protocol was managed entirely by existing staff with no formal task shifting between team members and no new employees required. In addition, there was no change to the physical space in the PED. The only small notable change was a redefinition of role as one of the senior medical students or residents was tasked with triage duties for their shift rather than evaluating patients in order of the queue with the rest of the team. Implementation included CETEP training for all HCWs in the PED and education of staff about the new triage process. Posters of the triage algorithm were placed on the walls of the PED. Members of the team met periodically to evaluate uptake and effectiveness of the interventions in order to address any problems identified. Two months
Indicators to measure the impact of the triage process were defined by key local and regional stakeholders (HNPB, BCM/TCH, MoH and PAHO). Basic demographics including age, gender, triage category, diagnosis and disposition were collected for both study groups. All diagnoses including ‘malnutrition’ and ‘shock’ were made using standard national guidelines for Guatemala. Triage categories for the pre-intervention group were assigned retrospectively using an algorithm based on the presenting and vital signs and results of physical examination in the records. This same triage algorithm was used to review post-intervention records. Clinical uptake, defined as the number of PED patients with an assigned triage category, was used as the process measure. The primary outcome measure was hospital admission rate. Secondary outcome measures included inpatient length of stay (LOS) and mortality.
Data collection
Two paediatric residents in HNPB used a standardized data collection form that was created by BCM/TCH to extract all data from both PED and inpatient paper medical records. They also assigned all triage categories to the pre-intervention group. BCM/TCH worked with the HNPB residents on site to pilot the data forms, create the electronic database and complete a sample of 30 medical record reviews side-by-side to assure quality of extraction. These two paediatric residents reviewed all records and cross-checked their results with each other before data entry; if there was variation between categories assigned, they discussed the case together until reaching consensus.
Statistical analyses
The sample size was calculated on the basis of the primary outcome measure, the hospital admission rate. It was hypothesized that early recognition and stabilization of critically ill children would decrease overall admission rates. A decrease from 10% to 5% in the frequency of admissions was deemed to be clinically important. Therefore, a minimum of 434 patients was needed in each group (pre- and post-intervention) to detect a difference. Based on the anticipated volume of paediatric patients over 2 years (approximately 40,000 patients), it was determined that review of every 40th patient visit would meet this requirement. Any incomplete medical records or obvious inaccurately transcribed data were excluded from statistical analysis. Unless otherwise stated, t-tests were used to compare means between groups. All data were analysed using SPSS
Paediatrics and International Child Health 2016 VOL. 36
NO.
3
221
Crouse et al. Implementation of paediatric triage
(IBM SPSS Statistics 21.0, August 2012). For all comparisons, statistical signifi- cance was P
ISSN: 2046-9047 (Print) 2046-9055 (Online) Journal homepage: http://www.tandfonline.com/loi/ypch20
Impact of an Emergency Triage Assessment and Treatment (ETAT)-based triage process in the paediatric emergency department of a Guatemalan public hospital Heather L. Crouse, Francisco Torres, Henry Vaides, Michael T. Walsh, Elise M. Ishigami, Andrea T. Cruz, Susan B. Torrey & Miguel A. Soto To cite this article: Heather L. Crouse, Francisco Torres, Henry Vaides, Michael T. Walsh, Elise M. Ishigami, Andrea T. Cruz, Susan B. Torrey & Miguel A. Soto (2016) Impact of an Emergency Triage Assessment and Treatment (ETAT)-based triage process in the paediatric emergency department of a Guatemalan public hospital, Paediatrics and International Child Health, 36:3, 219-224, DOI: 10.1179/2046905515Y.0000000026 To link to this article: http://dx.doi.org/10.1179/2046905515Y.0000000026
Published online: 06 Sep 2016.
Submit your article to this journal
Article views: 109
View related articles
View Crossmark data
Citing articles: 1 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ypch20 Download by: [Southern Cross University]
Date: 10 October 2017, At: 00:43
Original Research Paper
Impact of an Emergency Triage Assessment and Treatment (ETAT)-based triage process in the paediatric emergency department of a Guatemalan public hospital Heather L. Crouse1, Francisco Torres2, Henry Vaides2, Michael T. Walsh3, Elise M. Ishigami3, Andrea T. Cruz1, Susan B. Torrey1,4, Miguel A. Soto2 Department of Pediatrics, Section of Emergency Medicine, Baylor College of Medicine, Houston, Texas, USA, Department of Pediatrics, Hospital Nacional Pedro Bethancourt, La Antigua, Guatemala, 3Global Health Initiative, Texas Children’s Hospital, Houston, Texas, USA, 4Department of Emergency Medicine, Division of Pediatric Emergency Medicine, New York University School of Medicine, New York City, USA
Downloaded by [Southern Cross University] at 00:43 10 October 2017
1
2
Background: Triage process implementation has been shown to be effective at improving patient outcomes. This study sought to develop, implement and assess the impact of an Emergency Triage Assessment and Treatment (ETAT)based emergency triage process in the paediatric emergency department (PED) of a public hospital in Guatemala. Methods: The study was a quality improvement comparison with a before/after design. Uptake was measured by percentage of patients with an assigned triage category. Outcomes were hospital admission rate, inpatient length of stay (LOS), and mortality as determined by two distinct medical record reviews for 1 year pre- and post-intervention: a random sample (RS) of all PED patients and records for all critically-ill (CI) children [serious diagnoses or admission to the paediatric intensive care unit (PICU)]. Demographics, diagnoses and disposition were recorded. Results: The RS totalled 1027 (51.4% male); median ages pre- and post-intervention were 2.0 and 2.4 years, respectively. There were 196 patients in the CI sample, of whom 56.6% were male and one-third were neonates; median ages of the CI group pre- and post-intervention were 3.1 and 5.6 months, respectively. One year after implementation, 97.5% of medical records had been assigned triage categories. Triage categories (RS/CI) were: emergency (2.9%/54.6%), priority (47.6%/44.4%) and non-urgent (49.4%/1.0%). The CI group was more frequently diagnosed with shock (25%/1%), seizures (9%/0.5%) and malnutrition (6%/0.5%). Admission rates for the RS (8% vs 4%, P=0.01) declined after implementation. For the CI sample, admission rate to the PICU (47% vs 24%, P=0.002) decreased and LOS (7.3 vs 5.7 days, P=0.09) and mortality rates (12% vs 6%, P=0.15) showed trends toward decreasing post-implementation. Conclusions: Paediatric-specific triage algorithms can be implemented and sustained in resource-limited settings. Significant decreases in admission rates (both overall and for the PICU) and trends towards decreased LOS and mortality rates of critically ill children suggest that ETAT-based triage systems have the potential to greatly improve patient care in Latin America. Keywords: Paediatric emergency medicine, Paediatric triage, International emergency medicine, Emergency Triage Assessment and Treatment (ETAT), Clasificación, Evaluación y Tratamiento de Emergencias Pediátricas (CETEP) Abbreviations (in order of reference: ETAT; Emergency Triage Assessment and Treatment; HCW, shealthcare workers; HNPB, Hospital Nacional Pedro Bethancourt; BCM/TCH, Baylor College of Medicine/Texas Children’s Hospital; MoH, Guatemalan Ministry of Health; PAHO, Pan-American Health Organization; CETEP, Clasificación Evaluación y Tratamiento de Emergencias Pediátricas; PED, paediatric emergency department; PICU, paediatric intensive care unit; QI, quality improvement; LOS, inpatient length of stay; RS, random sample; CI, critically ill sample
There are no known conflicts of interest associated with this publication and there has been no significant financial support for this work which could have influenced its outcome. Sources of support: (internal grant) Texas Children’s Hospital Faculty Educational Grants 2010 and 2011 Correspondence to: Heather Crouse, 6621 Fannin Street, Suite A2210, Houston, Texas, USA 77030. Fax: 832 825 5424; email: [email protected]
© 2016 Informa UK Limited, trading as Taylor & Francis Group DOI 10.1179/2046905515Y.0000000026
Paediatrics and International Child Health 2016 VOL. 36
NO.
3
219
Crouse et al. Implementation of paediatric triage
Introduction
Downloaded by [Southern Cross University] at 00:43 10 October 2017
Early recognition and stabilization of acutely ill infants and children improves outcomes in all settings, regardless of resources. When resources are constrained, interventions which include the introduction of triage training and process, the use of clinical practice guidelines and the supervision and monitoring of patients lead to reduced mortality rates in critically ill children.1−5 In a cluster randomized trial in district hospitals in rural Kenya, similar interventions reduced mortality and were cost-effective. 6,7 Although triage is an effective process that requires few resources to implement, most healthcare facilities in developing countries have no formal triage system.8−11 The acuity of patients presenting to paediatric emergency departments varies widely. Although many children require only basic care, those with emergency medical conditions often experience delays in initial treatment. Consequently their conditions deteriorate, resulting in admissions that may have been avoidable. None of the 18 randomly selected facilities described in an assessment of the quality of care for hospitalized children in Bangladesh had a functioning triage system.12 These reports suggest that locally driven efforts to implement triage systems could be effective and feasible.13,14 Reports describing triage systems in Latin America are limited. Interventions (including prioritizing care for children in emergency departments) implemented at 14 Nicaraguan hospitals improved outcomes without increasing hospital costs.15 Previous studies of triage interventions in resource-limited settings have used the Emergency Triage Assessment and Treatment (ETAT) programme developed by the World Health Organization. This strategy was developed to promote improved assessment, triage and initial management of critically ill children in resource-limited settings by teaching healthcare workers (HCWs) and those with limited healthcare training how to rapidly assign a triage category of ‘emergency’, ‘priority’ or ‘non-urgent’ to patients based on easily identified risk factors and clinical signs.16−18 Limited experience suggests that implementation of ETAT based triage systems improves outcomes in critically ill children in Africa;4,19,20 however, it is not available in Spanish. As part of a collaborative effort between the Hospital Nacional Pedro Bethancourt (HNPB, a public hospital in Antigua, Guatemala), Baylor College of Medicine/ Texas Children’s Hospital (BCM/ TCH), the Guatemalan Ministry of Health (MoH) and the Pan-American Health Organization (PAHO), a self-sustaining Spanish ETAT [Clasificación Evaluación y Tratamiento de Emergencias Pediátricas (CETEP)] training programme was developed and implemented at HNPB.21 The aim was to implement a collaboratively developed CETEP-based triage process and to evaluate its impact on process and patient outcomes in the paediatric emergency department (PED) at HNPB.
220
Paediatrics and International Child Health 2016 VOL. 36
NO.
3
Methods
This quality improvement study compared a retrospective pre-intervention sample with a similar prospective sample which was identified and analysed after the triage process was implemented in the PED. The two primary objectives of the study were to assess uptake of the triage process and to assess its impact on patient outcomes. The choice of measures to assess clinical uptake and patient outcomes was limited to data that were routinely collected in the PED at HNPB. The number of PED patients in the post-intervention sample to whom a triage category has been assigned was used as a process measure to indicate uptake. Hospital admission rate was chosen as the primary patient outcome measure. Investigators in sub-Saharan Africa have used death within 2 days of hospital admission as a surrogate for the effectiveness of initial emergency care.1,20 However, child mortality rates in Central America generally and in Guatemala specifically are lower than those in sub-Saharan Africa.22 As a result, the sample size required or order to use mortality rate as an outcome measure would have been unrealistic for this project.
Setting and study population
The study was conducted at Hospital Nacional Pedro Bethancourt (HNPB), a secondary level public hospital in La Antigua, Guatemala, located approximately 45 kilometres from Guatemala City. It is the referral hospital for the Department of Sacatepe′quez and surrounding areas serving a population of approximately 329,547. The PED evaluates approximately 20,000 children aged 0–12 years annually. The PED is open 24 hours/day and is staffed primarily by medical students, paediatric residents, one supervising general paediatrician and one or two nurses per shift. HNPB inpatient paediatric services include a 30-bed inpatient paediatric ward, a 30-bed neonatology ward and small neonatal and paediatric intensive care units. Before the intervention, the average rate of admission to the paediatric ward from the PED was 10%. All sub-speciality needs are referred to Roosevelt Hospital in Guatemala City. The hospital has an established paediatric residency training programme and has developed several innovative programmes including development of the first certified human milk bank in Latin America. Two groups of patients who presented to the PED for care between 1 November 2009 and 31 October 2011 were selected for the study: a random sample (every 40th patient) of all patients (defined by HNPB as age 0–12 years and identified here as RS) and all who were critically ill [defined as those with an emergency diagnosis per CETEP or who were admitted to the Paediatric Intensive Care Unit (PICU) as recorded in the PED logbook and identified here as CI]. It was decided to evaluate all critically ill children (CI) separately because the number of
Downloaded by [Southern Cross University] at 00:43 10 October 2017
Crouse et al. Implementation of paediatric triage
critically ill children presenting to the PED is typically small and therefore there might not have been enough emergency patients in the random sample to adequately assess the impact of the triage process on this important population.
after implementation, paper supplies ran low. As a result, triage forms were unavailable and HCWs used the new triage system less often. In response, the team developed a low-cost triage stamp containing the same information that was in all the PED records.
Intervention
Measurements
A practical application of CETEP training is the implementation of a triage process that defines specific triage categories (red for emergency, yellow for priority and green for non-urgent). Training of paediatric HCWs at HNPB in CETEP began in June 2010. By November 2010, 80% (n=590) of paediatric HCWs were trained. The accuracy of HCW triage assessed immediately after CETEP training was found to be 95.1% (CI 91.9–97.3).21 Preparatory to development of the triage system, in November 2010, a pilot study was initiated by HNPB HCWs in the PED to monitor patient flow and acuity, staffing and room availability for triage. With this information, a multi-disciplinary team developed and implemented a formal three-level, CETEP-based triage process in the PED as part of a larger, institutional quality improvement (QI) project. The team consisted of HNPB PED nurses, paediatric residents, general paediatricians, the head of paediatrics and representatives from TCH/BCM. First, he team defined ideal patient flow for the PED on the basis of CETEP principles: all patients presenting to the PED were initially evaluated by a designated ‘triager’, usually a senior medical student or resident. Patients classified as ‘emergency’ were put directly in the ‘shock’ room for immediate stabilization. Patients classified as ‘priority’ were put in a room, if available, or in the waiting room, but were moved ahead in the queue so that management could begin as quickly as possible. Non-urgent patients presenting to the PED between 7 and 11 a.m. were sent directly to the paediatric outpatient department for evaluation. At other times, they were put in the PED waiting room for consultation in order of presentation. Next, the team developed a paper triage form to be attached to all PED patient records; it included patient name, date of birth, time of triage, triage category and reason for that level. It is important to recognize that this new protocol was managed entirely by existing staff with no formal task shifting between team members and no new employees required. In addition, there was no change to the physical space in the PED. The only small notable change was a redefinition of role as one of the senior medical students or residents was tasked with triage duties for their shift rather than evaluating patients in order of the queue with the rest of the team. Implementation included CETEP training for all HCWs in the PED and education of staff about the new triage process. Posters of the triage algorithm were placed on the walls of the PED. Members of the team met periodically to evaluate uptake and effectiveness of the interventions in order to address any problems identified. Two months
Indicators to measure the impact of the triage process were defined by key local and regional stakeholders (HNPB, BCM/TCH, MoH and PAHO). Basic demographics including age, gender, triage category, diagnosis and disposition were collected for both study groups. All diagnoses including ‘malnutrition’ and ‘shock’ were made using standard national guidelines for Guatemala. Triage categories for the pre-intervention group were assigned retrospectively using an algorithm based on the presenting and vital signs and results of physical examination in the records. This same triage algorithm was used to review post-intervention records. Clinical uptake, defined as the number of PED patients with an assigned triage category, was used as the process measure. The primary outcome measure was hospital admission rate. Secondary outcome measures included inpatient length of stay (LOS) and mortality.
Data collection
Two paediatric residents in HNPB used a standardized data collection form that was created by BCM/TCH to extract all data from both PED and inpatient paper medical records. They also assigned all triage categories to the pre-intervention group. BCM/TCH worked with the HNPB residents on site to pilot the data forms, create the electronic database and complete a sample of 30 medical record reviews side-by-side to assure quality of extraction. These two paediatric residents reviewed all records and cross-checked their results with each other before data entry; if there was variation between categories assigned, they discussed the case together until reaching consensus.
Statistical analyses
The sample size was calculated on the basis of the primary outcome measure, the hospital admission rate. It was hypothesized that early recognition and stabilization of critically ill children would decrease overall admission rates. A decrease from 10% to 5% in the frequency of admissions was deemed to be clinically important. Therefore, a minimum of 434 patients was needed in each group (pre- and post-intervention) to detect a difference. Based on the anticipated volume of paediatric patients over 2 years (approximately 40,000 patients), it was determined that review of every 40th patient visit would meet this requirement. Any incomplete medical records or obvious inaccurately transcribed data were excluded from statistical analysis. Unless otherwise stated, t-tests were used to compare means between groups. All data were analysed using SPSS
Paediatrics and International Child Health 2016 VOL. 36
NO.
3
221
Crouse et al. Implementation of paediatric triage
(IBM SPSS Statistics 21.0, August 2012). For all comparisons, statistical signifi- cance was P
