To describe the methodology undertaken to provide guidance on the appropriateness, as well as inappropriateness, of vascular access device selection, characteristics, and insertion technique for pediatric patients.
The RAND Corporation–University of California, Los Angeles Appropriateness Method was used. After definition of key terms and scope, a systematic review of the pediatric vascular access literature was undertaken. Clinical scenarios were developed to reflect the common indications for vascular access across pediatric health care. These were sectioned according to (1) device selection, (2) device characteristics, and (3) insertion technique. An interdisciplinary panel of experts (N = 14) consisting of leading experts representing diverse pediatric clinical disciplines including anesthesiology, cardiology and cardiac surgery, critical care and emergency, general surgery, hematology and oncology, hospital medicine, infectious disease, interventional radiology, pharmacology, regional pediatric hospitalist, and vascular access nursing specialties was convened. The scenarios were rated for appropriateness by the panel over 2 rounds (1 [highly inappropriate] to 9 [highly appropriate]). Round 1 ratings were completed anonymously and independently by panel members and classified into 3 levels of appropriateness: appropriate, uncertain, and inappropriate, or disagreement. For round 2, panelists met in-person to discuss the round 1 ratings and independently rerated the indications. All indications were reclassified into 3 levels of appropriateness or disagreement.
The RAND Corporation–University of California, Los Angeles Appropriateness Method provides a rigorous, in-depth and transparent methodology to develop the first appropriateness criteria for the selection of pediatric vascular access devices in a range of patient groups.
No evidence-based criteria have been created to inform the selection and insertion of vascular access devices in pediatrics. The RAND Corporation–University of California, Los Angeles Appropriateness Method is a reliable and validated approach for determining evidence-based appropriateness criteria.
Using the RAND Corporation–University of California, Los Angeles Appropriateness Method, appropriate selection and insertion practices for vascular access devices across a range of clinical indications in pediatrics were determined.
The selection and insertion of an appropriate vascular access device (VAD) in pediatrics is a complex decision. Clinicians must weigh multiple, often conflicting factors—clinical indication, treatment duration, resource availability, complication risk, and patient and/or family preference to name but a few. The most easily available VAD (eg, a peripherally inserted central catheter) may not be the best choice to prevent complications. At the core of this decision-making is a focus on vessel preservation, that is, ensuring vessel health for future treatments both known and unknown.1
Clinicians making such decisions could benefit from a standardized evidence-driven approach. Poor vascular access selection and insertion decisions result in patient harm, including vessel damage, infections, emotional distress, and repeat procedures as well as additional health care costs and treatment delays.2 Clear guidance on the appropriateness, as well as inappropriateness, of VADs across common pediatric clinical indications is necessary to improve outcomes for patients, families, and health care services.
Methods
Following the RAND Corporation–University of California, Los Angeles (RAND-UCLA) Appropriateness Method,3 we systematically determined criteria for the appropriate selection and insertion of VAD in pediatrics across Australia and the United States. We named these criteria the Michigan Appropriateness Guideline for Intravenous Catheters in Pediatrics (miniMAGIC), after the success of the adult version (Michigan Appropriateness Guide for Intravenous Catheters), which provided similar recommendations for hospitalized adult patients.4
Design
The RAND-UCLA Appropriateness Method was used for this study.3 This method has been successfully used across medical specialties to reduce the overuse and underuse of medical and surgical procedures such as medications,5,6 devices,4,7 and diagnostic tests.8–10 The methodology has also been used internationally and in various contextual settings, thus ensuring generalizability for the proposed recommendations. In the RAND-UCLA method, the concept of appropriateness refers to the relative weight of the benefits and harms of a medical or surgical intervention, regardless of the associated costs.3 The method thus balances the best available scientific evidence with the collective judgement of experts to form a statement regarding the appropriateness of procedures.3
The RAND-UCLA Appropriateness Method involves sequential phases, which are detailed in Fig 1. The method was beneficial for developing appropriateness criteria for the selection and insertion of pediatric VAD for many reasons. First, the approach combines systematically identified high-quality evidence with the pragmatic expertise of clinicians to develop a set of recommendations on the appropriateness of indications for insertion and specificity of a VAD. Second, by including a large multidisciplinary international panel, we were able to incorporate the diverse views of a range of clinical specialties involved in pediatric vascular access decision-making. Finally, the method does not mandate consensus; rather, it allows for surfacing and discussion of disagreements between panelists. This was especially valuable for miniMAGIC considering the limited availability of high-quality evidence surrounding pediatric VAD usage and the heterogeneity of the pediatric population requiring VADs.
Scope
miniMAGIC was focused on defining the appropriateness of the selection and insertion of VADs that are commonly used across pediatric health systems (hospitalized and ambulatory care) in Australia and the United States. These regions were chosen because the collaborators who designed and conceived the project (A.J.U., V.C.) wished to generate recommendations that would consider varying contexts of care in these locations. miniMAGIC was focused on inpatient and outpatient settings, including management from maternity hospital or equivalent to discharge, until the patient was 18 years of age.11 The recommendations intentionally did not seek to provide appropriateness recommendations for neonates admitted to the NICUs or special care nurseries immediately after birth because they require a specialist approach outside the expertise of the research team and panel. The devices and specific populations (eg, hospitalized patients, oncology and hematology, critical care, congenital cardiac conditions, long-term vascular access–dependent conditions) included in miniMAGIC were identified as core pediatric VAD consumers across Australia and the United States.12 Finally, miniMAGIC sought to provide guidance on 3 important clinical questions surrounding pediatric vascular access that were relevant to practitioners, regardless of evidence availability. These clinical questions were as follows:
device selection;
device characteristics; and
insertion technique.
Definition of Key Terms
Venous Access
Venous access is an indication for VAD placement when access to the venous system is needed for either infusion of therapeutics and/or laboratory blood draws.4 Access can be classified as being central (a venous access device with the tip residing in the superior or inferior vena cava) or peripheral (a venous access device with the tip positioned outside of the vena cava).13
VAD and Infusion Catheter Types
A range of catheter types are used to gain access to the venous system. Each included device is displayed in Fig 2 and defined in Table 1.14 To ensure relevance to general pediatric health care, the project focused on the appropriateness of catheter types that are commonly available across health institutions rather than those that are currently emerging or only available in superspecialty institutions.
VAD and Infusion Catheter Types
| Device | Definition |
|---|---|
| Intraosseous catheter | Catheter inserted into the spongy, cancellous bone of the epiphysis and the medullary cavity of the diaphysis. |
| Peripheral device | A venous access device with a tip positioned outside of the vena cava. |
| Peripheral intravenous catheter | Typically, a 2–6-cm device that enters and terminates in the peripheral veins (also known as an IV or PIV). |
| Midline catheter | A catheter inserted into a peripheral vein (most commonly the upper arm via the basilic, cephalic, or brachial vein) and threaded to an area of greater blood flow in the proximal portion of the extremity, preferably with the internal tip located level at or near the level of the axilla and distal to the shoulder (in neonates they may also be inserted into a scalp vein and threaded into the jugular vein). |
| Central device | A venous access device with the tip residing in the superior or inferior vena cava. |
| Umbilical catheter | A catheter that is inserted into the vein of the umbilical cord, which is only available for a limited time immediately after birth (tip placement proximal to the ductus venosus indicates a central position). |
| Nontunneled CVAD | Also referred to as “acute,” “short-term” or “percutaneous” device, it is typically 5–15 cm long and placed via direct puncture through the skin and cannulation of the internal jugular, subclavian, or femoral veins, without a portion of the device allowed to remain in a subcutaneous tract. |
| Peripherally inserted central catheter | This is a catheter inserted through veins of the upper extremity or neck in children; for infants, it may be inserted through veins of the scalp or lower extremity. The catheter tip is located in the superior or inferior vena cava, preferably at its junction with the right atrium. |
| Tunneled (and cuffed) CVAD | A CVAD with a segment of the catheter lying in a subcutaneous tunnel with the presence of a cuff into which the subcutaneous tissue grows to offer security for the catheter (indicates that the skin exit site and vein entry site are separated by the subcutaneous tunnel). |
| Totally implanted venous device | Commonly known as a port or mediport, this device is implanted in the subcutaneous tissue of the chest or abdominal wall. It features a reservoir for injection or aspiration, which is inset, and a catheter that communicates from the reservoir to a deep vein of the chest or inguinal region. |
| Device | Definition |
|---|---|
| Intraosseous catheter | Catheter inserted into the spongy, cancellous bone of the epiphysis and the medullary cavity of the diaphysis. |
| Peripheral device | A venous access device with a tip positioned outside of the vena cava. |
| Peripheral intravenous catheter | Typically, a 2–6-cm device that enters and terminates in the peripheral veins (also known as an IV or PIV). |
| Midline catheter | A catheter inserted into a peripheral vein (most commonly the upper arm via the basilic, cephalic, or brachial vein) and threaded to an area of greater blood flow in the proximal portion of the extremity, preferably with the internal tip located level at or near the level of the axilla and distal to the shoulder (in neonates they may also be inserted into a scalp vein and threaded into the jugular vein). |
| Central device | A venous access device with the tip residing in the superior or inferior vena cava. |
| Umbilical catheter | A catheter that is inserted into the vein of the umbilical cord, which is only available for a limited time immediately after birth (tip placement proximal to the ductus venosus indicates a central position). |
| Nontunneled CVAD | Also referred to as “acute,” “short-term” or “percutaneous” device, it is typically 5–15 cm long and placed via direct puncture through the skin and cannulation of the internal jugular, subclavian, or femoral veins, without a portion of the device allowed to remain in a subcutaneous tract. |
| Peripherally inserted central catheter | This is a catheter inserted through veins of the upper extremity or neck in children; for infants, it may be inserted through veins of the scalp or lower extremity. The catheter tip is located in the superior or inferior vena cava, preferably at its junction with the right atrium. |
| Tunneled (and cuffed) CVAD | A CVAD with a segment of the catheter lying in a subcutaneous tunnel with the presence of a cuff into which the subcutaneous tissue grows to offer security for the catheter (indicates that the skin exit site and vein entry site are separated by the subcutaneous tunnel). |
| Totally implanted venous device | Commonly known as a port or mediport, this device is implanted in the subcutaneous tissue of the chest or abdominal wall. It features a reservoir for injection or aspiration, which is inset, and a catheter that communicates from the reservoir to a deep vein of the chest or inguinal region. |
Adapted from Chopra V, Flanders SA, Saint S, et al; Michigan Appropriateness Guide for Intravenous Catheters (MAGIC) Panel. The Michigan Appropriateness Guide for Intravenous Catheters (MAGIC): results from a multispecialty panel using the RAND/UCLA Appropriateness Method. Ann Intern Med. 2015;163(6 suppl):S1–S40. Adapted from Gorski L, Hadaway L, Hagle M, McGoldrick M, Orr M, Doellman D. Infusion therapy standards of practice. J Infus Nurs. 2016;39:S1–S159. Adapted from Wyckoff MM, Sharpe EL. Peripherally Inserted Central Catheters: Guideline for Practice. 3rd ed. Chicago, IL: National Association of Neonatal Nurses; 2015. CVAD, central venous access device; IV, intravenous; PIV, peripheral intravenous.
Catheter Insertion Location
Vessels and locations commonly accessed for VAD and infusion catheter placement are displayed in Fig 3. Catheter insertion location refers to the entrance point in which the vessel is accessed, not where the catheter tip resides.
Populations
The recommendations focused on informing pediatric (from maternity hospital, or equivalent, to discharge at 18 years of age)11 VAD selection and insertion decisions. The population definitions included age and specialty categories, which are displayed in Table 2.
Population Categories
| Definition | |
|---|---|
| Population | |
| Neonates | Birth to first 30 d15 |
| Infants | 31 d to <1 y |
| Children | 1 y to <12 y |
| Adolescents | 12–18 y |
| Setting and diagnosis | |
| General hospitalized patients | Those receiving care in an inpatient setting on a general pediatrics, pediatric subspecialty, or surgical service; excludes critical (ICU and emergency department) and ambulatory patients.4 |
| Specialty populations16 | |
| Congenital cardiac conditions | This includes cardiovascular disorders present in the perinatal or neonatal period, including structural developmental anomaly of the heart and great vessels.15,17 For single ventricle physiology, the conditions are further categorized into repair stage 1–3.15 |
| Critically ill | Those receiving care in an emergency department, ICU, or other critical care setting because of single- or multiorgan dysfunction; can include patients who are unstable (abnormal vital signs, disruption of vital functions, or injuries of a critical nature) or stable.18 |
| Oncology and hematology | Those undergoing treatment of neoplasms and diseases of the blood or blood-forming organs17 ; includes active cancer, which is a patient receiving active chemotherapy or admitted to the hospital for a cancer-related diagnosis.4 |
| Long-term vascular access–dependent condition | Patients with conditions that require access to the vascular system for administration of intermittent or continuous therapy for extensive periods of time (eg, enzyme replacement therapy, chemotherapy, immunotherapy, total parenteral nutrition, plasmapheresis, or hemodialysis).4,19,20 Includes long (2 mo–1 y) and very long (≥1 y) therapy duration. |
| Definition | |
|---|---|
| Population | |
| Neonates | Birth to first 30 d15 |
| Infants | 31 d to <1 y |
| Children | 1 y to <12 y |
| Adolescents | 12–18 y |
| Setting and diagnosis | |
| General hospitalized patients | Those receiving care in an inpatient setting on a general pediatrics, pediatric subspecialty, or surgical service; excludes critical (ICU and emergency department) and ambulatory patients.4 |
| Specialty populations16 | |
| Congenital cardiac conditions | This includes cardiovascular disorders present in the perinatal or neonatal period, including structural developmental anomaly of the heart and great vessels.15,17 For single ventricle physiology, the conditions are further categorized into repair stage 1–3.15 |
| Critically ill | Those receiving care in an emergency department, ICU, or other critical care setting because of single- or multiorgan dysfunction; can include patients who are unstable (abnormal vital signs, disruption of vital functions, or injuries of a critical nature) or stable.18 |
| Oncology and hematology | Those undergoing treatment of neoplasms and diseases of the blood or blood-forming organs17 ; includes active cancer, which is a patient receiving active chemotherapy or admitted to the hospital for a cancer-related diagnosis.4 |
| Long-term vascular access–dependent condition | Patients with conditions that require access to the vascular system for administration of intermittent or continuous therapy for extensive periods of time (eg, enzyme replacement therapy, chemotherapy, immunotherapy, total parenteral nutrition, plasmapheresis, or hemodialysis).4,19,20 Includes long (2 mo–1 y) and very long (≥1 y) therapy duration. |
Clinical Indication for the VAD and/or Infusion Therapy
The clinical indications for needing the device, defined in Table 3, include the infusate characteristics, difficult insertion, and blood sampling.
Clinical Indications
| Clinical Indication | Definition |
|---|---|
| Non-peripherally compatible therapy | An irritant (an agent that induces inflammation, tenderness, warmth, or redness along the vein) or vesicant (an agent that induces blistering or redness) therapy not compatible with administration in peripheral veins for brief or extended periods of time, which includes continuous vesicant chemotherapy, parenteral nutrition (exceeding 10% dextrose and/or 5% protein), therapy with pH <5 or >9, and therapy with osmolarity >600 mOsm/L.1,4 |
| Peripherally compatible therapy | Therapy that is able to be safely administered via a peripheral vessel, without irritant or vesicant characteristics.4 |
| Difficult vascular access | Current or historical difficulties in achieving vascular access,1,4,13 which include: |
| ≤2 visible or palpable veins | |
| History of previous difficulties in venous access per patient or caregiver | |
| History of comorbidities that lend themselves to difficult venous access (eg, morbid obesity, contractures). | |
| Blood sampling | Aspiration of blood from the VAD for the purposes of diagnostic testing. |
| Clinical Indication | Definition |
|---|---|
| Non-peripherally compatible therapy | An irritant (an agent that induces inflammation, tenderness, warmth, or redness along the vein) or vesicant (an agent that induces blistering or redness) therapy not compatible with administration in peripheral veins for brief or extended periods of time, which includes continuous vesicant chemotherapy, parenteral nutrition (exceeding 10% dextrose and/or 5% protein), therapy with pH <5 or >9, and therapy with osmolarity >600 mOsm/L.1,4 |
| Peripherally compatible therapy | Therapy that is able to be safely administered via a peripheral vessel, without irritant or vesicant characteristics.4 |
| Difficult vascular access | Current or historical difficulties in achieving vascular access,1,4,13 which include: |
| ≤2 visible or palpable veins | |
| History of previous difficulties in venous access per patient or caregiver | |
| History of comorbidities that lend themselves to difficult venous access (eg, morbid obesity, contractures). | |
| Blood sampling | Aspiration of blood from the VAD for the purposes of diagnostic testing. |
Vessel Visualization
Vessel visualization includes the devices and techniques used to identify appropriate veins, size, depth, location, and potential complications, which are listed and defined in Table 4.
Vessel Visualization Devices and Techniques
| Device and Technique | Definition |
|---|---|
| Near-infrared light | A device using near-infrared light, which works by either transilluminating the extremity and projecting the vessel image to a screen or by capturing an image of the superficial veins and reflecting it to the skin surface.13 |
| Ultrasound | A device using sound waves at frequencies greater than the limit of human hearing (sound waves directed into human tissue to identify and display physical structures).16 |
| ECG tip guidance | Intracavitary ECG-guided placement provides real-time tip confirmation during the insertion procedure by monitoring changes of the P-wave morphology, tracking the central device tip along the superior vena cava to the right atrium. As the central device tip approaches the sinoatrial node at the cavoatrial junction, the P-wave elevates.21 |
| Fluoroscopy | An imaging technique that uses radiographs to obtain real-time moving images of the interior of an object.13 |
| Surgical cutdown | A procedure during which the vein is exposed surgically and then a catheter is inserted into the vein under direct vision.13 |
| Catheter-to-vein ratio | The indwelling area occupied by an intravascular device positioned within a vessel without a tourniquet (commonly expressed in percentage [eg, 45%]).22 |
| Device and Technique | Definition |
|---|---|
| Near-infrared light | A device using near-infrared light, which works by either transilluminating the extremity and projecting the vessel image to a screen or by capturing an image of the superficial veins and reflecting it to the skin surface.13 |
| Ultrasound | A device using sound waves at frequencies greater than the limit of human hearing (sound waves directed into human tissue to identify and display physical structures).16 |
| ECG tip guidance | Intracavitary ECG-guided placement provides real-time tip confirmation during the insertion procedure by monitoring changes of the P-wave morphology, tracking the central device tip along the superior vena cava to the right atrium. As the central device tip approaches the sinoatrial node at the cavoatrial junction, the P-wave elevates.21 |
| Fluoroscopy | An imaging technique that uses radiographs to obtain real-time moving images of the interior of an object.13 |
| Surgical cutdown | A procedure during which the vein is exposed surgically and then a catheter is inserted into the vein under direct vision.13 |
| Catheter-to-vein ratio | The indwelling area occupied by an intravascular device positioned within a vessel without a tourniquet (commonly expressed in percentage [eg, 45%]).22 |
ECG, electrocardiogram.
Evidence Synthesis
As recommended by the RAND-UCLA method, a systematic review of the literature was undertaken to summarize the scientific evidence available on pediatric VAD selection and insertion. This was undertaken by using high-quality systematic review procedures,3,23 including registration on the International Prospective Register of Systematic Reviews (CRD42019094286). Methods included detailed inclusion or exclusion criteria, trial identification via systematic search of electronic databases with the assistance of a health librarian (Cochrane Central Register of Controlled Trials, Ovid Medline, Ovid Embase, EBSCO, clinical trial databases, and relevant guidelines) using key medical subject heading terms, independent selection of studies, and data extraction by 2 review authors (A.J.U., E.B.). The literature was synthesized and provided to the expert panel before the appropriateness rating. The literature review is available as an article in this supplement.24
Panel Selection and Engagement
To ensure the perspectives and requirements of specialist providers who commonly make decisions regarding pediatric VADs were incorporated into the appropriateness ratings we engaged key experts representing relevant interdisciplinary specializations as voting panel members. As recommended by the RAND-UCLA method, leading Australian and US experts across disciplines who are respected scholars or researchers, represent professional societies, or have substantial clinical experience in the field were invited and provided informed consent to participate.3,4 Nonvoting panel members who participated in discussions included a patient representative to ensure a patient focus was evident throughout deliberations, panel facilitators (including a methodologist), funders, and support staff. Panel representation, including specialty, is listed alphabetically in Table 5.
Panel Representation
| Name | Pediatric Specialty | Location | Hospital and/or University Affiliations |
|---|---|---|---|
| Voting Panelists | |||
| Ranjit Aiyagari | Cardiology | Michigan | CS Mott Children’s Hospital of Michigan |
| University of Michigan | |||
| Darcy Doellman | Vascular access nursing | Ohio | Cincinnati Children’s Hospital |
| E. Vincent Faustino | Critical care | Connecticut | Yale New Haven Children’s Hospital |
| Yale School of Medicine | |||
| Jeffrey Jacobs | Cardiac surgery | Florida | Johns Hopkins All Children's Hospital |
| Julie Jaffray | Hematology and oncology | California | Children’s Hospital of Los Angeles |
| University of Southern California | |||
| Tricia Kleidon | Vascular access nursing | Queensland, Australia | Queensland Children’s Hospital |
| Griffith University | |||
| Prashant Mahajan | Emergency | Michigan | CS Mott Children’s Hospital |
| University of Michigan | |||
| Craig McBride | General surgery | Queensland, Australia | Queensland Children’s Hospital |
| Griffith University; University of Queensland | |||
| Kayce Morton | Regional pediatric hospitalist | Missouri | CoxHealth |
| University of Missouri | |||
| Elizabeth Prentice | Anesthesiology | Victoria, Australia | Royal Children’s Hospital |
| Doug Rivard | Interventional radiology | Missouri | Children’s Mercy Kansas City |
| University of Missouri–Kansas City | |||
| Erin Shaughnessy | Hospital medicine | Arizona | Phoenix Children’s Hospital |
| University of Arizona | |||
| Marc Stranz | Pharmacology | Pennsylvania | Stranz Crossley Inc |
| Joshua Wolf | Infectious disease | Tennessee | St Jude Children’s Research Hospital |
| Nonvoting panelists | |||
| Steve Bernstein | Panel facilitator, methodologist | Michigan | VA Ann Arbor Healthcare System |
| University of Michigan | |||
| Vineet Chopra | Panel facilitator | Michigan | VA Ann Arbor Healthcare System |
| University of Michigan | |||
| Beth Gore | Patient representative | Florida | Association for Vascular Access |
| Ramzy Nasrallah | Funder | New Jersey | Association for Vascular Access Foundation |
| Stephanie Pitts | Vascular access nursing | Florida | St Joseph’s Children’s Hospital |
| B. Braun Medical | |||
| Amanda Ullman | Panel facilitator | Queensland, Australia | Queensland Children’s Hospital |
| Griffith University |
| Name | Pediatric Specialty | Location | Hospital and/or University Affiliations |
|---|---|---|---|
| Voting Panelists | |||
| Ranjit Aiyagari | Cardiology | Michigan | CS Mott Children’s Hospital of Michigan |
| University of Michigan | |||
| Darcy Doellman | Vascular access nursing | Ohio | Cincinnati Children’s Hospital |
| E. Vincent Faustino | Critical care | Connecticut | Yale New Haven Children’s Hospital |
| Yale School of Medicine | |||
| Jeffrey Jacobs | Cardiac surgery | Florida | Johns Hopkins All Children's Hospital |
| Julie Jaffray | Hematology and oncology | California | Children’s Hospital of Los Angeles |
| University of Southern California | |||
| Tricia Kleidon | Vascular access nursing | Queensland, Australia | Queensland Children’s Hospital |
| Griffith University | |||
| Prashant Mahajan | Emergency | Michigan | CS Mott Children’s Hospital |
| University of Michigan | |||
| Craig McBride | General surgery | Queensland, Australia | Queensland Children’s Hospital |
| Griffith University; University of Queensland | |||
| Kayce Morton | Regional pediatric hospitalist | Missouri | CoxHealth |
| University of Missouri | |||
| Elizabeth Prentice | Anesthesiology | Victoria, Australia | Royal Children’s Hospital |
| Doug Rivard | Interventional radiology | Missouri | Children’s Mercy Kansas City |
| University of Missouri–Kansas City | |||
| Erin Shaughnessy | Hospital medicine | Arizona | Phoenix Children’s Hospital |
| University of Arizona | |||
| Marc Stranz | Pharmacology | Pennsylvania | Stranz Crossley Inc |
| Joshua Wolf | Infectious disease | Tennessee | St Jude Children’s Research Hospital |
| Nonvoting panelists | |||
| Steve Bernstein | Panel facilitator, methodologist | Michigan | VA Ann Arbor Healthcare System |
| University of Michigan | |||
| Vineet Chopra | Panel facilitator | Michigan | VA Ann Arbor Healthcare System |
| University of Michigan | |||
| Beth Gore | Patient representative | Florida | Association for Vascular Access |
| Ramzy Nasrallah | Funder | New Jersey | Association for Vascular Access Foundation |
| Stephanie Pitts | Vascular access nursing | Florida | St Joseph’s Children’s Hospital |
| B. Braun Medical | |||
| Amanda Ullman | Panel facilitator | Queensland, Australia | Queensland Children’s Hospital |
| Griffith University |
VA, Veterans Affairs.
Scenario Development
The conceptual framework used to develop the clinical scenarios and appropriateness criteria is displayed in Fig 4. The clinical scenarios and appropriateness criteria for miniMAGIC were informed by Michigan Appropriateness Guide for Intravenous Catheters4 but prioritized by the results of the systematic review and the expert opinion of the panelists within the boundaries of the project scope. The clinical scenarios were sectioned into (1) device selection (across VAD types), (2) device characteristics (including lumen number, size, insertion location), and (3) insertion technique (attempts, image guidance). Device selection was further categorized into chapters for age and specialty populations. A process of validation was completed, including internal review by the investigators and panelists and external review by independent pediatric physicians and vascular access specialist nurses, to ensure the scenarios reflected the predefined aims and scope.25
Conceptual framework for the development of scenarios and indications of appropriateness. The scenarios developed for Michigan Appropriateness Guide for Intravenous Catheters (MAGIC)4 were used as a foundation for the miniMAGIC clinical scenario development, but these were reframed to take into consideration the results of the systematic review and the expert opinion of the panelists. The major clinical questions within the clinical scenarios were device selection, device characteristics, and insertion technique. Device selection was further categorized into age and specialized populations. The clinical scenarios were further modified with the panelists after the round 1 rating.
Conceptual framework for the development of scenarios and indications of appropriateness. The scenarios developed for Michigan Appropriateness Guide for Intravenous Catheters (MAGIC)4 were used as a foundation for the miniMAGIC clinical scenario development, but these were reframed to take into consideration the results of the systematic review and the expert opinion of the panelists. The major clinical questions within the clinical scenarios were device selection, device characteristics, and insertion technique. Device selection was further categorized into age and specialized populations. The clinical scenarios were further modified with the panelists after the round 1 rating.
Round 1 Appropriateness Rating
Per the RAND-UCLA Appropriateness Method, the round 1 appropriateness ratings were made by the 14 panelists voting anonymously and independently. Appropriateness for each clinical scenario was described by considering the benefit/harm ratio on a scale of 1 to 9, in which 1 indicates that harm outweighs benefit (highly inappropriate) and 9 signifies that benefit outweighs harm (highly appropriate).
The voting panelists were provided with the findings of the systematic literature review (including a quality-ranked summary and full-text articles), definition of terms, and clinical scenarios. They were instructed to begin the appropriateness rating process by familiarizing themselves with the literature review and then the structure of the clinical scenario indications document. They were requested to complete the ratings chapters sequentially and to take regular breaks during completion so as to minimize fatigue.
The panelists were instructed to rate the indications using their best clinical judgement and the evidence available in the systematic review. They were specifically asked not to base ratings on their perception of how experts may respond, but rather to consider each scenario as pertaining to an average patient presenting to an average physician to perform the procedure at an average (Australian or US) hospital. Finally, they were instructed not to consider cost implications.
The panelists completed the round 1 ratings on paper, which was electronically returned (scanned) to the investigators; these data inputted into a dedicated Microsoft Excel document, with duplication of entry for at least 20% to minimize transcription errors. The overall ratings for round 1 for each indication were summarized descriptively as a median, frequency, and range. As recommended by the RAND-UCLA method, indications were classified into 3 levels of appropriateness:
appropriate: panel median score of 7 to 9 without disagreement;
uncertain: panel median score of 4 to 6 or with disagreement regardless of median; and
inappropriate: panel median score of 1 to 3 without disagreement.
Disagreement existed if 5 or more panelists rated in each extreme (1–3 and 7–9).3
Round 2 Appropriateness Rating
After round 1, panelists traveled to Ann Arbor, Michigan, for an in-person meeting at which time individualized documents revealing each panelist’s own ratings in comparison with the group median and range were provided. Over 2 days, the panel discussion of all the indications was moderated by the panel facilitators. The sessions were structured to encourage debate and discussion regarding ratings for which disagreement occurred or when scenarios were rated as uncertain. In some situations, the panel recommended rewording, contextual modifications, or removal of certain indications, resulting in a reduction in the final number of clinical scenarios and improvement in clarity of the ratings. After discussion to ensure there was no misunderstanding of what was being rated, each clinical scenario was rerated independently by the voting panelists. The round 2 data were inputted into the revised Microsoft Excel document, and finally, all indications were reclassified into 3 levels of appropriateness or disagreement (as previously described).3
Ethics
The study received ethics approval from Griffith University (2018/207) and was deemed exempt from review by the University of Michigan (HUMM00144945). All panel members provided written, informed consent before panel participation. The funder had no role in the design or results of the study.
Discussion
High-quality evidence is crucial to inform practice. Although randomized controlled trials are the gold standard for determining the safety and effectiveness of medical and surgical procedures, in the absence of this evidence, clinical decisions still need to be made. The RAND-UCLA Appropriateness Method provides a reliable and valid method to systematically assess variation in the use of procedures by defining the appropriateness of these procedures within clinical situations.10 The method also highlights practices in need of further evidence by revealing areas of uncertainty and disagreement.3
This study is the first time the RAND-UCLA Appropriateness Method has been used to define the appropriate selection and insertion of pediatric VADs. The description of how this method has been applied provides clinicians with an understanding of the quality and scope of the minMAGIC recommendations and thereby provides confidence for implementing them.
Although the methods for the RAND-UCLA appropriateness approach are well defined, barriers to performing this project existed. First, choosing the miniMAGIC panelists was not easy. In comparison to other RAND-UCLA projects (in which panel sizes range from 5 to 11), our panel was considerably larger. We consciously made this decision because although VADs are used by almost all pediatric clinical disciplines, there is no formal discipline that is responsible. This reality likely underpins the current wide variation in practice. Many pediatric disciplines insert some or many VAD types, including interventional radiology, ICU, emergency, hospital medicine, surgery, anesthesiology, and nursing. Other clinical disciplines (eg, nephrologist, oncology) request VADs to be inserted, manage their use, and rely on their function. A new specialist group is also emerging in tertiary centers and is known as the vascular access specialist. Thus, to truly determine what constitutes appropriate use in pediatrics, many disciplines and specialties must be involved. Although we could not include all pediatric disciplines, we believe a balanced and informed panel relevant to the project scope was constituted. Second, evidence was limited in certain areas, and panelists freely admitted variable practice patterns. Although this created divergence, it also allowed for a robust discussion to identify what experts were doing where, why, and how. Ultimately, areas of uncertainty or disagreement were felt as important for inclusion because these areas identify gaps for clinical studies and funding agencies to focus. Third, for some scenarios, the panel relied on the expertise of a few individuals (eg, cardiac surgery or cardiology). Although their individual views helped inform the discussion, this relatively singular viewpoint may be seen as a limitation of our approach. Because we adhered to the methodology of the RAND-UCLA Appropriateness Method to ensure there was no artificial disagreement and broadened our areas of ratings, we believe this discussion enriched the understanding of the material. Finally, panelists were only from the United States and Australia. This strengthened the relevance of the recommendations to these health care settings; however, we acknowledge it diminishes the generalizability of these recommendations to other countries where VAD insertion and management resources may differ.26,27
These barriers notwithstanding, the methodology outlined highlights the rigor, depth, and transparency of the methods used to develop the recommendations within miniMAGIC. These methods extended the work of previously impactful projects, including those within vascular access and wider health services research.4,7,9 miniMAGIC has a clearly defined scope and approach, which should provide confidence to clinicians and health services ready to implement the miniMAGIC recommendations.
Conclusions
The RAND-UCLA Appropriateness Method was applied to develop appropriateness criteria for selection and insertion of pediatric VADs across a range of patient groups.
Acknowledgments
We thank the Association for Vascular Access Foundation, Griffith University, and the University of Michigan for providing funding to support this project. We also acknowledge the considerable work by the miniMAGIC panelists in undertaking this project.
Drs Ullman and Chopra conceptualized and designed the study and drafted the initial manuscript; Dr Brown designed the data collection instruments, collected data, and conducted the initial analyses; Prof Cooke, Prof Rickard, Prof Bernstein, and Ms Kleidon designed the study; and all authors reviewed and revised the manuscript, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.
FUNDING: Supported by grants from the Association for Vascular Access Foundation, Griffith University, and the University of Michigan.
- miniMAGIC
Michigan Appropriateness Guide for Intravenous Catheters in Pediatrics
- RAND-UCLA
RAND Corporation–University of California, Los Angeles
- VAD
vascular access device
Comments