Education in the placement of ultrasound-guided peripheral venous catheters: a systematic review

In collaboration with a research librarian, a systematic literature search of three databases (CINAHL, PubMed, EMBASE) was conducted using relevant search terms related to USG-PVC placement. Full search strategies can be found in Additional file 1. The first search was conducted on December 10, 2017, and a final search was conducted on February 13, 2021, to ensure the results being up to date. Two authors (RJ, PIP) independently screened all articles for eligibility based on titles and abstracts. Relevant studies were then reviewed independently as full-text for final inclusion. A third investigator (CBL) resolved any inclusion disagreements. Finally, a hand search through the reference lists of included articles was conducted and expert recommendations were screened for inclusion as well.

Original research articles were included if both of the following were present:

Articles were excluded if one or more of the following were present

No restrictions were made for either patient population, or the type of health personnel performing the USG-PVC procedure.

The following data were extracted from each article: Study type, characteristics of the study population, characteristics of the patient population, educational program, technique of USG-PVC placement, and results.

A modified risk of bias evaluation based on the criteria of the Cochrane Collaboration risk of bias assessment tool was performed on all included studies [9]. Non-randomized studies were assessed using the Newcastle-Ottawa Scale bias tool [10].

Due to the wide aim of the research question and a significant heterogeneity in study designs, a meta-analysis could not be performed and a descriptive synthesis and approach was therefore applied.

No needs-assessment studies were found, but a few studies evaluated participants’ impression of skills and the self-defined benefits from participation after completing the education [11‐15]. Participants found the education program helpful and felt able to place PVC in patients where it would otherwise not have been possible before education. The participants estimated a need for USG-PVC to be approximately six to seven patient cases every week. They felt an overall improvement of personal skills and considered it possible to acquire the necessary skills for ultrasound guided PVC placement through the utilized educational program [12, 14, 15].

In most of the included studies the curriculum was poorly described but generally included; ultrasound physics, knobology, probe selection, ultrasonic vascular and nerve recognition, preparation for USG-PVC placement, and complications (Additional file 3).

In six studies the USG-PVC education was a part of a general ultrasound curriculum [16‐21]. These educational programs included other ultrasound examinations like echocardiography and focused lung ultrasound among other techniques. The lengths of these general educational programs were between 15 to 20 h or a full rotation at an emergency ultrasound department. Five studies included scheduled follow-up with either more than 100 scans (n = 3) or at least 2 weeks of emergency medicine rotation to ensure competency (n = 5) [16, 18‐21].

In studies only training the USG-PVC, the training comprised of a combination of didactic and hands-on sessions, including education on live models for normal vascular anatomy and cannulation on phantoms (Additional file 3). Fourteen educational programs included video material of the procedure [12‐14, 22‐32] and five included live demonstrations of the procedure [22, 23, 29, 33, 34]. Inter-study duration variations in the training sessions were seen from the shortest of 5 min to the longest of 9 h [35, 36]. In general, the duration of the training was between two to 4 h (Additional file 3).

Three studies were based on the mastery learning approach and two other studies used a similar approach [15, 28‐31]. It took the participants less than 1 h of training before meeting the requirement for passing in two studies [29, 31]. A fixed curriculum time limit was used in two of the five studies, and in these two studies participants failing the first assessment attempt only required additional 15–60 min of training to pass in a second attempt [28, 30]. One study showed that extra training after meeting mastery criteria did not improve the participants’ performance [31].

E-learning, test-enhanced learning, and the composition of phantoms for training were evaluated [37‐39]. None of these were found to have a positive or negative effect on learning competencies. Teaching participants to use colour Doppler for what is called “twinkle artifact” showed an improvement compared with only using cross-section view [40]. Two studies found a relatively steep learning curve after initial training, and that after four to nine real world attempts a nurse’s probability of success was over 70% on average [41, 42]. Likewise, results were seen in paediatric patients with a success rate of 67% after ten supervised attempts, and an increase after ten additional unsupervised attempts to 83% [27]. After ten successful attempts Ault et al. showed that the participants’ learning curve flattened [43].

The participants in the included studies varied from doctors/physicians (n = 12), nurses (n = 24), emergency department technicians (ED-technicians) (n = 3), nurse students (n = 2), medical students (n = 7) and in one study it was unclear. Additional 13 studies included more than one of the groups mentioned above. In-depth information about education participants can be found in Additional file 3.

USG-PVC methods and education were evaluated through different parameters like success rate, time, attempts, and participant rating mainly through Likert-scales. Sixty-one studies assessed the learners using mainly, two objects; either phantoms (n = 17) or living subjects (n = 38). The remaining studies did not use any evaluation objects and only evaluated through participant feedback (n = 6). The living subjects could be patients (n = 36) and some of the studies specified the type of patient. Adult patients were used in 20 studies, whereas children were used in five studies. Fifteen studies included only patients specified as DIVA patients. In other studies, the living subjects were healthy persons (n = 2). The phantoms used in studies were commercially available (n = 10) [13, 23, 34, 35, 38, 40, 44‐47], homemade (n = 5) [26, 37, 39, 48, 49], or a mixture of both types of phantoms [50]. One study did not specify the type of phantom they used [51]. In-depth information about test subject and phantom distribution can be found in Additional file 3.

Assessing the competence of participants through assessment tools was done in seven studies [22, 28, 30, 31, 47, 52, 53]. In all seven studies the used assessment tool was a checklist. Good et al. evaluate if it is meaningful to use motion analysis as a tool for assessing competency for USG-PVC placement [47]. They did this by comparing a group of nurses’ motion analysis before and after USG-PIV education and comparing their improvement to the result of experts’ motion analyses. Good et al. found that 17 out of 21 nurses obtained expert proficiency in at least one of six motion-analysis metrics after education.

Two competence assessment tools were identified: one as a checklist of competence and one as a rating scale for competencies [54, 55]. Validity evidence was only explored for the rating scale [56].

Different types of USG-PVC techniques were evaluated in studies carried out on patients, phantoms, or both [12, 13, 18, 23, 25, 34, 35, 44‐48, 57‐59]. When evaluating the effect of long-axis view or short-axis view, most studies found no difference. Both the long-axis and the short-axis method were found to be better than the oblique method [35]. The differences between using the dominant or non-dominant hand for probe handling were evaluated in one study, that found the dominant to be superior [23]. Results comparing one versus two-person techniques were inconclusive [13, 58] and the implementation of guidance-markers on the screen only showed an effect when used by nurses [48].

Success rate for cannulation, number of attempts or time were the most frequently used variables for evaluating the clinical effects of education and implementation of USG-PVC placement. Likert-scales for pain and participant satisfaction were used as well. The most relevant outcomes are listed in the data-extraction sheet (Additional file 3).

Traditional PVC placement without ultrasound guidance was primarily used as control. In some studies the traditional technique was specified as being anatomically guided. Nine out of ten studies found a significant difference in time, attempts, or cannulation success rate favouring ultrasound guidance compared to conventional methods on patients [16, 19, 32, 52, 60‐64], whereas the remaining one did not find any significant difference [65]. The effect in these studies was only seen on DIVA patients. A decrease in the need for PICC or CVC was seen in several studies after the implementation of education in ultrasound guidance for PVC placement [21, 29, 62, 66, 67].

Lastly, two studies compared different health-professional groups and found no significant difference in success rate between nurses, physicians, or ED-technicians [25, 68].

Only outcomes assessed relevant to the aim of this study are included in this results section. For a full overview of outcomes see the extended data-extraction sheet in Additional file 3.

The Cochrane risk-of-bias tool was used to evaluate all studies included and 33 studies were also evaluated through the Newcastle-Ottawa scale, but the heterogeneity of the included studies made the risk of bias hard to address and compare. The risk-of-bias assessments are shown in Tables 1 and 2.

The Newcastle-Ottawa scale risk assessment is based on three different areas; selection, comparability and outcome. In each a certain amount of stars can be given. These stars represent the quality in that focus area. The maximum amount of stars a study can get is 9 spread over three categories as 4/2/3 (selection/comparability/outcome). For more information, visit the Newcastle-ottawa scales website.

The aim of medical educational programs is to provide its participants with sufficient knowledge or competences in specific areas or procedures defined through a curriculum. Ideally, the educational program should lead to a clinical improvement and ensure a minimum of competence for all participants.

Kern’s six-step model for curriculum development can be used to ensure that an education program fulfils the aim just mentioned above [76, 77]. The six steps are; problem identification and general needs-assessment, targeted needs-assessment, setting goals and measurable objectives, educational strategies, implementation, and evaluation and feedback. In this review, we use Kern’s model to evaluate and discuss the knowledge in USG-PVC education existing at the moment. The first two steps, general and targeted needs-assessment, help to define a demand for a given education and explore the current knowledge limitations within the field [78]. To clarify the extent of the curriculum broad goals and objectives are determined (step three). Assessment tools preferably based on solid evidence of validity should be used to ensure these goals. The educational program is then implemented in a clinical setting, where it is tested for its effect and impact, and lastly evaluating the whole process to chance or refine some of the five previous steps. Throughout each step, the previous steps should be evaluated and changes considered.

When assessing medical education, it is recommended to carry out a general needs-assessment on a national or international level [76, 78]. A national general needs-assessment in anaesthesiology identified the technical procedures that leading key-opinion leaders thought should be trained using simulation [79]. USG-PVC placement was mentioned as one of 30 important procedures.

Edwards et al. and Adhikari et al. investigated nurses’ perceived need for an educational program through questionnaires after completion of USG-PVC training [12, 14]. Nurses expected approximately three DIVA-patients per shift and thought that ultrasound guidance might be an aid. More than half the nurses felt that the biggest barrier for them to use USG-PVC was lack of experience and almost everyone agreed that focused training was adequate to learn this. The two studies addressed a local perception of need for education but in reverse order. A needs-assessment could be put in place to clarify the need for education, preferably with the possibility for generalisation.

General goals and specific measurable objectives are important because they help define the curriculums and direct content and participant focus [76]. Benchmarks can through assessment tools as checklists or global rating scales ensure participants’ competence during and after education programs.

A lack of a defined curriculum and a rare use of any tools for ensuring competencies are seen in the included articles. Furthermore, seven of the included articles use a general US education [16‐21, 71], which makes it harder to address the effect of the USG-PVC training and open up for the possibility for transfer of skills. Ahern et al. investigated general ultrasound educations in America and found a lack of assessment tools and lack of specific curricula [80]. However, there seems to be a general consensus in the included articles regarding which topics are important e.g. ultrasound and machine understanding, knobology, probe handling, ultrasound picture recognition and ultrasound cannulation technique.

The Delphi-method can help define the extent of a curriculum together with clearly defined goals and sub-goals [81]. Messick’s framework can then be used to ensure the assessment tool’s capability to measure what it is supposed to measure [82]. The framework describes five sources of validity in experimental data consist of five groups; content, response process, internal structure, relationship to other variables, and consequences. Each source of validity can demand varying attention depending on the curriculum assessed [83].

Two assessment tools were found; Jung’s 16 items checklist and Primdahl’s rating scale, both developed through the Delphi-method. Only Primdahl et al. explored the validity evidence of the assessment tool using Messick’s framework [54‐56]. The items in these assessment tools could be the inspiration for the items in a curriculum. In the seven studies using an assessment tool all of them were checklists and the validity evidence was not explored for any of these checklists [22, 28, 30, 31, 47, 52, 53].

A key element of curriculum development is addressing different educational strategies or adding educational components to improve the learning process. The mastery learning approach builds on the concept that not time but acquired competencies define the educational process [84]. Usually, by introducing new steps of procedures or topics when a previous step has been mastered. This could be introduction, machine settings and then ultrasound scanning a forearm followed by training cannulation technique on phantoms. The approach can differ and some suggest an effect by letting participants self-evaluate when they are ready to move on based on fixed goals [85, 86]. Others prefer letting an external objective evaluation and feedback by a simulator or instructor decide when to move on [87]. Feedback as a mechanism is a central part of the mastery-learning concept with its possibilities to improve performance by correction. On the other hand, too much feedback might also have its drawbacks as explained by the guidance hypotheses [84, 88, 89].

Mastery learning has proven efficacy but surprisingly only three studies used this approach [28, 30, 31]. A reason could be that it is easier to plan a traditional course using a fixed amount of time than a course that ensures that all trainees acquire the pre-defined proficiency level because trainees learn at different paces. The three studies prove that it is possible to plan an USG-PVC education by mastery learning. Compared to the studies that did not use mastery learning, mastery learning did not take up more time, with an average around 1 h, with approximate 30 min extra education time, if the first assessment was not passed [28, 30, 31]. Additionally, Kule et al. showed that so called overtraining did not increase success rate [31].

Four studies address the learning curve of nurses after an education program [41‐43, 52]. Three studies showed a steep learning curve but a difference in attempts to obtain a success rate of 70–80% between four and ten USG-PVC performed in the clinic [41, 52, 73]. This finding aligns with the superiority of the mastery learning approach; no fixed numbers can ensure competency of all trainees.

Evaluation of different educational strategies and educational initiatives is equally important as investigating the effect of education. Chenkin et al. found that a one-hour web-based learning program was equally efficient compared to a one-hour traditional classroom lecture [38]. Only three of the studies used an e-learning module as a part of their educational program [22, 36, 47]. This indicates that e-learning is not implemented to its capability. The usage of color Doppler function to identify “twinkle artifacts” had an effect on cannulation time and a positive but non-significant impact on the success rate [40]. New studies in the future can help to clarify the full effect of color Doppler in USG-PVC education.

Neither test-enhanced learning nor using different types of phantoms seem to have a significant effect on the learning outcome and therefor none of these can be recommended [37, 39, 50].

The program’s clinical impact is the highest level of evaluation in Kirkpatrick’s method for evaluating the effects of educational programs [90]. The measurable impact of an educational program on an institutional level defines the impact of the program and evaluates the potential benefits and cost-efficiency.

The majority of the included studies found that ultrasound-guidance improved PVC-placement in either attempts, time, or success rate compared to traditional method when used on DIVA patients [19, 32, 52, 60‐63] as it is also mentioned in Van-Loon meta-analysis [91]. Furthermore, a decrease in the need for CVC or PICC was seen in several studies [21, 29, 62, 66, 67]. Carter et al. and Oliveira et al. found that the effectiveness did not depend on profession and education should therefore be considered to a wide range of health care professions [25, 68].

Even though most studies found an effect on success rate, time, or amount of attempts, the results varied a lot between studies and few did not find an effect. The variation of success rate was between 85 and 97.5% in the top group [17, 19, 22, 24, 42, 53, 59, 64, 68], down to only 63% in adult patients [58] and no difference between traditional and ultrasound-guided in two studies [65, 69]. These big differences seem strange and even though some are explained by difference in equipment, patients and practitioners, some of the differences might also be explained by different educational strategies and the lack of evidence-based curricula to ensure competencies in ultrasound-guided PVC-placement. This shows the importance of more studies and easy access to information about proper educational strategies for USG-PVC.

This systematic review was conducted on the premises of the available published articles and their quantity and quality. Only including full text published articles could have resulted in the exclusion of possibly relevant results. Ideally, a meta-analysis of presented data should have been included as a part of this review, but was deemed not clinically meaningful due to the large heterogeneity of the reported methods and results.

In summary there seems to be a clinical effect of using USG-PIV in specific patient groups as DIVA, and maybe also children. An educational program could very well be structured around the mastery-learning program, with an e-learning pre-course introducing the participants to the principals of ultrasound, anatomy and USG-PVC technique. This e-learning could be followed by a hands-on session structured around mastery learning and ending with an assessment test, at the moment preferably the one from Primdahl et al. since it is the only one where validity of evidence has been explored.