Ⅰ. INTRODUCTION
1. Background
Peripheral intravenous catheter (PIVC) insertion is a clinical skill required by nurses, and it is a major measurement in 80% of inpatients for hydration, electrolyte balance, drug injection, transfusion, and nutrition supply (Jamieson, Whyte, & McCall, 2019). Although infusion therapy through PIVC provides a faster treatment effect than other methods, there is always the risk of complications such as phlebitis or infiltration (Ray-Barruel, Polit, Murfield, & Rickard, 2014). PIVC insertion is an invasive procedure performed by nurses and has potential complications, but its risks tend to be under-recognized within the complexity of the clinical setting (Nickel, 2019).
The failure rate of PIVC insertion was reported to be 35–50% (Keleekai et al., 2016), which was mainly due to the lack of skill of the health care provider as well as the patient's condition (Carr, Higgins, Cooke, Rippey, & Rickard, 2017). Since PIVC insertion technique is an essential clinical skill in the nursing field, the lack of competence not only makes nurses anxious but also an important source of patient dissatisfaction. Furthermore, adverse outcomes in patients with PIVC insertion can cause complications, poor quality of care, long-term hospitalization, and increased health-related costs (Abolfotouh, Salam, White, & Balkhy, 2014;Nickel, 2019). Hence, it is crucial for them to thoroughly observe the whole process and predict, find, and manage any possible problems and safety issues that might occur; nurses should therefore have exquisite techniques and competence regarding the infusion (Gorski, 2017).
There have been many studies on complications associated with infusion therapy (Guanche-Sicilia et al., 2021) as well as insertion skills by health care providers, including clinical nurses (Denton et al., 2016). Most PIVC complications are ascribed to insertion- related factors such as insertion site preparation and insertion technique (Abolfotouh et al., 2014), and older, experienced nurses and IV team nurses were more likely to succeed in PIVC insertion and could reduce associated complications (Jacobson & Winslow, 2005;Marsh et al., 2021). In addition, in a study on nurses working in medical/surgical units or a study on pediatric interns, it was confirmed that knowledge, confidence, and skills were significantly improved in the group that used the simulation-based learning program for PIVC insertion (Gaies et al., 2009;Keleekai et al., 2016). However, since the effectiveness of education decreased after 7 months (Gaies et al., 2009), continuous education and training related to PIVC insertion for nurses may contribute to lowering the PIVC insertion failure rate and complications.
Recently, guidelines for intravenous infusion therapy have been developed and revised. Infusion therapy practice standards, published every five years by the Infusion Nurses Society (INS), are widely cited and used in many procedure manuals to develop and support clinical procedures (Gorski, 2017). The PIVC insertion skills checklist (28-item) was developed and validated based on the hospital PIVC insertion protocol (Schuster, Stahl, Murray, Keleekai, & Glover, 2016), which was redefined as the short PIVC insertion skill assessment tool (Gaies et al., 2009). The PIVC Insertion Skills Checklist was developed for evaluating PIVC insertion and monitoring the procedures in a simulated environment, with long items (28-item), but it does not include patients’ comfort and educational contents (Schuster et al., 2016).
In Korea, the Korean Hospital Nurses Association (KHNA) developed the evidencebased clinical nursing practice guideline- intravenous infusion (Gu et al., 2012) in 2012 consisting of 19 categories and 180 recommendations. When the guidelines spread to hospitals, it was expected that nursing would be standardized and performed based on evidence, and would greatly contribute to improving nursing efficiency. However, nurses at domestic tertiary hospitals and large general hospitals were at the level of 'sometimes use' of the guidelines, and the main hindrance was the lack of time to use the guidelines (Eun et al., 2015). The guidelines of KHNA are extensive as they provide overall principles for intravenous infusion. In addition, although research on the actual condition of small and medium-sized hospitals is insufficient, the use of evidence-based guidelines is considered to be very insignificant (Hong & Eun, 2020). Therefore, there is a need for an evidence-based, concise and standardized tool that enables nurses to actively, easily, and quickly evaluate the process of performing PIVC insertion in most common clinical situations.
2. Purpose
This study aimed to develop a nursing performance measuring tool for PIVC insertion performance (PIVC-IP) by nurses. The specific objectives of this study were ⅰ) to generate pooled items based on the guidelines and prior studies, ⅱ) to perform content validity, ⅲ) to evaluate the construct using item analysis and exploratory factor analysis (EFA), ⅳ) to calculate the correlation coefficient between PIVC-IP and concurrent variable, and ⅴ) to calculate the reliability of PIVC-IP.
Ⅱ. METHODS
1. Study design
This cross-sectional descriptive study was designed to develop and test the psychometric properties of the measuring tool for PIVC-IP.
2. Item development and psychometric evaluation process
The process for developing PIVC-IP measurement tools was composed of six steps (Song, Son, & Oh, 2015): ⅰ) investigating the guidelines and previous studies, ⅱ) item generation, ⅲ) content validity, ⅳ) construct validity, ⅴ) reliability, and ⅵ) confirming the final item.
Investigating the guidelines and previous studies: Various guidelines and literature on peripheral intravenous catheterization and monitoring were searched.
Item generation: INS’s ‘The Infusion Therapy Standards of Practice’ (Gorski et al., 2016), Royal College of Nursing (RCN)’s ‘Standards for Infusion Therapy’ (Denton et al., 2016), KHNA’s ‘Evidence-Based Clinical Nursing Practice Guideline-Intravenous Infusion’ (Gu et al., 2012), and ‘Short Peripheral Intravenous Catheter Insertion Skills Checklist’ (Schuster et al., 2016) was reviewed. To expand the use of the tool, we focused on generating minimum items for the most common situations.
Content validity: Content validity evaluates whether the measurement tool's items adequately represented what the tool intended to measure. The generated preliminary items were reviewed by an expert panel consisting of 17 clinical nurses, a physician specializing in infectious diseases, and a nursing professor, using a 4-point scale. By calculating the proportion of agreement on the relevance of the expert panel, items with an Item-level Content Value Index (I-CVI) of .78 or higher were selected for the scale (Waltz, Strickland, & Lenz, 2010) and a Scale-level Content Validity Index (S-CVI) was confirmed (Polit, Beck, & Owen, 2007).
Construct validity: Construct validity indicates the accuracy of the measurement (Netemeyer, Bearden, & Sharma, 2003). After item analysis for preliminary items with content validity, an EFA was conducted. The item-total correlation was calculated to determine whether each item reflected the concept of the measurement tool and checked whether the value was greater than .30 (Nunnally & Bernstein, 1994). Prior to confirmation by EFA, for the suitability test of factor analysis, the Kaiser–Meyer–Olkin (KMO) index (values between 0.5 and 1.0) and the significance of Bartlett's spherical test were confirmed (Field, 2013). For EFA, a principal component factor analysis with orthogonal rotation was performed. Principal component analysis is useful for reducing the dimensions of interrelated measurement variables to independent principal components (Kang, 2013). After that, rotation was performed to simplify the structure of the factors. Orthogonal rotation is a meaningful method for maintaining factor independence (Nunnally & Bernstein, 1994). The factors were initially extracted based on the Kaiser criteria (eigenvalue 1 or higher), and scree plots were performed to determine the number of factors (Field, 2013).
Concurrent validity: Concurrent validity can be demonstrated when a test correlates with a measure or factor that has previously been evident (LoBiondo-Wood & Haber, 2010;Netemeyer et al., 2003). Pearson’s correlation between PIVC-IP scores and the nursing career period was used. A correlation coefficient of .3 to .6 with the developed tools and criteria is appropriate (Nunnally & Bernstein, 1994).
Reliability: Internal consistency was used to confirm the reliability, which means the degree to which items in the tool measure the same attribute (DeVellis & Thorpe, 2021).
3. Participants and data collection
Participants were recruited by convenience sampling method from nurses working in a tertiary hospital in D City. The inclusion criteria were as follows: ⅰ) registered nurses with more than 1 year of clinical experience, ⅱ) clinical nurses in the Medical-Surgical unit, and ⅲ) voluntary participation in this study. The exclusion criteria were as follows: ⅰ) nurse directors who did not participate in PIVC insertion, or ⅱ) outpatient nurses. Data collection was conducted from January 2018 to October 2019. A questionnaire was provided to voluntary participants who understood the purpose and procedure of this study, and a self-reported questionnaire was retrieved by the researcher. The questionnaire included nine items of general characteristics and 13 items of the PIVC-IP preliminary tool. The response scale comprised a 5-point Likert scale (5=completely, 4=often, 3=partially, 2=rarely, 1=never).
4. Sample size
The sample size for EFA can be calculated as 5–10 cases per item (Netemeyer et al., 2003;Waltz et al., 2010). Thus, at least 80– 160 participants for a 16-pooled item were needed to confirm the construct of the items. We distributed the questionnaire to 400 nurses to include clinical nurses with various clinical experiences; 365 data were used for analysis, excluding insufficient response data (more than half of the missing responses).
5. Statistical analysis
The collected data were statistically processed using IBM SPSS Statistical 26. Descriptive statistics were used to analyze the general characteristics of the participants. I-CVI and S-CVI were calculated for content validity. In item analysis, mean, standard deviation (SD), item-total correlation, and Cronbach's alpha if an item is deleted were calculated. EFA was performed by principal component factor analysis with varimax rotation. The Cronbach's alpha coefficient was calculated to confirm reliability. Missing data were replaced by mean values, and statistical significance was set at p<.05. Strengthening the Reporting of Observational studies in Epidemiology (STROBE) reporting guideline was used to confirm the quality of this cross-sectional study (see Appendix 1).
6. Ethical considerations
Informed consent from the participants was obtained. This study was approved by the Institutional Review Board of the University Hospital in D city (IRB No. CNUH2017- 06-004). The researcher did not directly contact participants for confidentiality, privacy, and withdrawal from the study at any time and for any reason.
Ⅲ. RESULTS
1. General characteristics
The general characteristics of the participants are presented in Table 1. A total of 365 nurses participated in this study, of whom 99.2% were women. The mean age was 29.76 ± 6.01 years, and 28.2% were married. Among the participants, university graduates accounted for the most, at 65.8%, and 92.3% were staff nurses; the mean nursing career period was 81.60 ± 70.50 months. Approximately 96% of the participants had experience participating in education related to infection control while working as a nurse, and the mean number of participating educations was 4.59 ± 3.89 times.
2. Item generation and content validity
Reviewing the guidelines from the INS (Gorski et al., 2016;Gorski, 2017), RCN (Denton et al., 2016), KHNA (Gu et al., 2012), and an existing tool (Schuster et al., 2016), generated 16 pool items that are most commonly applied. These included items about hand hygiene, preparation and disinfection of the intravenous insertion site before insertion, performance immediately after insertion, exchange-related activities, and patient education. For information on the difference between domestic and foreign guidelines, refer to domestic guidelines. After completing the questionnaire, 19 experts participated and scored a 16-pooled item to confirm the content validity, the mean I-CVI for items was .89, and S-CVI/Ave was .90. Based on the results of analyzing the degree of agreement for the items of expert group, three items were deleted. The three items were ‘Do not apply sterile tape directly to the IV catheterization site,’ ‘Replace all IV sets every 72 to 96 hours, except blood or lipid solutions,’ and ‘Change the IV catheter every 72 to 96 hours,’ and the I-CVI was .76, .71, and .59, respectively.
3. Item analysis
The item with the highest mean score was Q11, ‘Replace immediately if IV set is contaminated’ (4.78 ± 0.48), the lowest item was Q13, ‘Regular assessment of nurses' IV-related knowledge and compliance with guidelines’ (4.26 ± 0.78), and the total mean score was 4.51 ± 0.40. The item-total correlation coefficients for 13 items ranged from .43 to .63; thus, no items were deleted (Table 2).
4. Construct Validity: EFA
The KMO value was meritorious at .89, and Bartlett's test (χ2=1483.03, p<.001) was significant, suggesting that the data were appropriate for performing an EFA. As a result of EFA, a scree plot of eigenvalues revealed that the elbow-point is a second factor. After orthogonal rotation, the eigenvalues of factors 1 and 2 were 5.10 and 1.17, respectively. The 2-factor solution was preferred and explained 48.2% of the variance (Table 3).
Factor 1 was the point of care (POC), while factor 2 was called the follow-up care (FUC). The Q6 and Q7 questions showed that both POC and FUC had a factor loading value of .4 or higher. Considering the value of factor loading, theoretical guidelines, and clinical validity, Q6 and Q7 questions were determined as POC factors. Thus, POC factor consisted of eight questions (Q1, Q2, Q3, Q4, Q5, Q6, Q7, and Q8), with explained variance of 39.19%. The FUC factor had five questions (Q9, Q10, Q11, Q12, and Q13), which explained variance of 9.01%.
5. Concurrent validity
The positive correlation between the total items of PIVC-IP and nursing career period (r=.27, p<.001) was significant. The POC (r=.26, p<.001) and FUC (r=.21, p <.001) factors were positively correlated with the nursing career period, respectively (Table 4).
6. Reliability
Cronbach’s alpha coefficient for POC and FUC were .84 and .71, respectively, and the overall PIVC-IP was .87 (Table 3).
Ⅳ. DISCUSSION
In this study, a simple and convenient PIVC-IP tool was developed so that nurses could evaluate the performance of PIVC insertion in clinical settings. The 13-item PIVC-IP was valid and reliable for measuring nurses’ PIVC insertion performance.
The PIVC-IP was developed to reflect prior guidelines and expert opinions to increase usability and feasibility. First, items of the PIVC-IP were generated based on INS’s The Infusion Therapy Standards of Practice (Gorski et al., 2016;Gorski, 2017), RCN’s Standards for Infusion Therapy (Denton et al., 2016), KHNA’s Evidence-Based Clinical Nursing Practice Guideline-Intravenous Infusion (Gu et al., 2012), and Short Peripheral Intravenous Catheter Insertion Skills Checklist (Schuster et al., 2016). The guidelines include both general and special management. For this item development, we focused on general preparation such as selection and identification of PIVC insertion site and equipment, hand hygiene, and skin disinfection, and its management such as securement and infusion set exchange, and monitoring patients’ discomfort. In the process of listening to the opinions of experts, the opinion that it is desirable to compose the items suitable for general situations so that the use of the tools can be broadened was reflected. For example, items for special cases such as the use of special disinfectants (e.g., povidone iodine) for skin disinfection and the use of joint stabilization device to immobilize a venous catheter, were not included in the tool.
In the content validity test, three items with an I-CVI lower than .78 were dropped, although Polit et al. (2007) view that an S-CVI/Ave of .90 or higher was desirable was met. Experts disagreed with these items as the prior guidelines were revised after the generation of preliminary items. At the time of initial item development, the KHNA guidelines (Gu et al., 2012) recommended that the peripheral intravenous catheter or infusion set be replaced every 72 to 96 hours. However, the guidelines have since revised and the 2021 INS guidelines (Gorski et al., 2021) state that replace continuous administration sets used to administer solutions other than lipid, blood, or blood products no more frequently than every 96 hours but at least every 7 days. ‘Replace all IV sets every 72 to 96 hours, except blood or lipid solutions’ and ‘Change the IV catheter every 72 to 96 hours’ were deleted. In addition, the ‘Do not apply sterile tape directly to the IV catheterization site’ item is interpreted as not obtaining agree because it is ambiguous compared to the guideline ‘attach to the IV catheter hub, not directly to the IV insertion point.’ Eventually, the PIVC-IP consisted of 13 items for general PIVC insertion performance and patient assessment immediately after insertion. In the item analysis, the range of item-total correlation coefficients for 13 items met the criteria and no items were deleted.
In construct validation, the PIVC-IP tool was identified as two factors named ‘point of care’ and ‘follow-up care’: POC items at the moment of PIVC insertion and FUC items reflecting management strategy and education immediately after insertion. The POC includes eight items ranging from pre-insertion preparation to catheter fixation. However, compared with the Short Peripheral Intravenous Catheter Insertion Skills Checklist (Schuster et al., 2016), the detailed process of catheterization such as ‘apply tourniquet’ and ‘remove needle’ were not included in POC factor of this tool. This is because the PIVC-IP tool was designed to check the overall process rather than to evaluate each technique individually. The FUC factor has five items, including IV catheter and infusion set exchange, and monitoring the patient's discomfort.
The final PIVC-IP tool was determined to be a 2-factor solution and accounted for nearly 50% of the variance. And for any one factor to be significant, it met the criterion that at least 5% of the total variance was attributable to that factor (Hair, Black, Babin, Anderson, & Tatham, 2006). Various extraction approaches and rules exist for determining the number of factors, but these are not absolute criteria (Waltz et al., 2010). Kim et al. (2015) stated that for researchers to determine the number of factors, they should consider the conceptual meaning of the items constituting the factors, not simply statistical criteria. Therefore, the authors decided that the 2-factor solution was the most appropriate considering the theoretical guidelines and clinical validity of the items.
For concurrent validation, the correlation coefficient between the PIVC-IP scores and the nursing career period was calculated, and a significant positive correlation was confirmed. Although this study did not reach Nunnally and Bernstein (1994)'s suggestion that the correlation coefficient of .3 to .6 is appropriate, it cannot be said to be a low correlation. In previous studies, it has been reported that the older the nurse's age and experience, the lower the PIVC insertion failure rate (Jacobson & Winslow, 2005;Marsh et al., 2021), but it is interpreted that this is not consistent with a better following of the guidelines. Therefore, in future studies, it is necessary to consider education on performance according to the guidelines as a variable. In addition, although there is currently no legal vascular access expert system in Korea, it is expected that a known group validity evaluation between experts and general nurses will be possible when a legal system is established in the future.
Despite the small number of items, the internal consistency of the PIVC-IP tool can be interpreted as very good according to the criteria of DeVellis and Thorpe (2021). As a result, the developed PIVC-IP has been confirmed for validity and reliability and has secured practicality. The final items of the PIVC-IP tool (see Appendix 2, 3) do not violate the most recent guidelines, the 2021 INS Infusion therapy standards of practice.
The PIVC-IP tool is valid and reliable, but there are some limitations in terms of feasibility. First, the PIVC-IP can use only the general PIVC insertion situation. It can be of limited use if the patient is receiving chemotherapy or special drug administration through infusion therapy. Second, PIVC-IP use as a protocol for nursing student education can be limited. This is because the PIVC-IP tool does not contain items to evaluate the venipuncture process in detail. It is recommended to use in combination with the PIVC Insertion Skill 28 item checklist when used for student evaluation in a simulation situation. Third, management of complications due to PIVC was not included in this tool, thus we suggest the development of a tool for managing complications. Since the management of complications of PIVC is detailed and extensive, items for the management of complications should be developed separately. Finally, in psychometric evaluation, data were collected at a tertiary hospital by convenient extraction, and the sample size was insufficient to perform confirmatory factor analysis. In future studies, confirming that the developed tool constitutes an appropriate factor structure by expanding the sampling and sample size is necessary.
Nevertheless, the advantage of the PIVC-IP tool is that it can be actively utilized quickly and easily in clinical settings where nurses perform PIVC insertion. With this tool, it takes less than 5 minutes to see how a peripheral intravenous catheterization is performed. The PIVC insertion technique is an essential clinical skill in the nursing field, and it must be acquired through continuous education and training based on guidelines. The high competency of nurses will also contribute to patient comfort and safety. The PIVC-IP tool, which can be quickly and easily evaluated, will be useful for the education and training of new nurses and the self-reflection of experienced nurses. As a structured tool, PIVC-IP can be used not only for self-evaluation but also for observational evaluation. Response bias may occur when the PIVC-IP measurement tool is used as a self-evaluation. Therefore, it is necessary to explain the purpose of using the measurement tool to the nurse and to specify that it should be objectively evaluated immediately after PIVC-IP. This tool organized the questions so they could be applied in the most common situations. Therefore, it can be widely used even in environments with relatively poor equipment.
From a research point of view, it will enable the accumulation of more valid and consistent research results while using the developed tools. It may also contribute to further studies by enabling them to measure the performance of peripheral intravenous catheterization and develop intervention programs that can improve performance skills and capabilities.
Ⅴ. CONCLUSIONS
Intravenous therapy is an invasive intervention with a risk of potential complications performed by nurses. Therefore, monitoring and evaluation of intravenous catheterization is essential for improving patient care and safety. The development of the PIVC-IP tool was a study to help nurses safely perform peripheral intravenous catheterization in a clinical setting. The PIVC-IP tool was a valid and reliable measurement tool composed of POC and FUC. The response format of this tool is a 5-point Likert scale (scores from 1 to 5), with higher scores indicating greater performance for PIVC insertion, with a total score ranging from 13 to 65. In addition, it is concisely composed of 13 items and can be used quickly and easily. It will be useful in reviewing the procedure of performing peripheral intravenous catheterization in most common clinical settings.
