This study analyzed 522 patients, and after multivariate logistic regression analysis and multiple regression analysis, there was no statistically significant association between the needle type and failure of initial PIVC insertion and between the needle type and the number of procedures, respectively.
We did not find the efficacy of the newly developed needles in the current study; however, this may be a matter of the population at risk. While a previous study reported 24% failure of initial PIVC insertion, the rate of failure of initial PIVC insertion in our study was much low, at approximately 15%.[5] Easy cases for insertion have factors such as palpable or visible veins, and the veins of approximately 90% of the patients included in the current study were palpable or visible; this percentage is considered high. Therefore, the patients included in the current study represented a population in which PIVC insertion was easy, and the failure rate was low. Difficult intravenous access (DIVA) is a global concept. Interventions for DIVA are the focus of the future worldwide; for example, ultrasound-guided insertion has been effective for DIVA.[12, 13] In previous studies, various factors were thought to affect patients with DIVA: 1) no palpable or visible veins, 2) previous history of difficult venous catheterisation, 3) patient age < 4 years, and 4) tissue health (e.g. renal failure, diabetes, edema, and cachexia).[12, 13] With regard to the first factor, in our study, the veins of approximately 90% of the included patients were palpable or visible, and PIVC insertion was easy. Moreover, regarding the fourth factor, the patients included in our study were older adults and had skin that was not highly elastic. A previous study reported that PIVC insertion into highly elastic tissue is difficult because of needle flexion.[14] Moreover, previous studies have reported that the skin of older adults is less elastic than that of younger individuals owing to the decrease in collagen fibers, and insertion resistance is expected to decrease as elasticity decreases.[9, 15] Therefore, the older patients included in our study had reduced skin elasticity, and the needle did not flex; thus, PIVC insertion may not have been difficult. In contrast, to reduce resistance during insertion and facilitate PIVC insertion, the bevel angle of the newly developed needles is 20°.[9] Comparing the newly developed and existing needles when inserting a vinyl chloride tube, the penetration forces at the needle and catheter tips of the newly developed needles were significantly lower than those of the existing needles, and the success rate of the initial PIVC insertion was improved.[9] The newly developed needles may be more effective than the existing needles, particularly for patients with DIVA with large vascular and skin elasticity. Patients whose vessels were palpable or visible and older patients included in our study were unlikely to benefit from the ease of insertion of the newly developed needles.
We did not find a statistically significant association between the use of newly developed needles and the failure of initial PIVC insertion in multivariate logistic regression analysis. However, given the advantages of the structure of newly developed needles, they may be more effective than the existing needles, and patients with DIVA are likely to benefit from the ease of insertion, particularly patients whose vessels are not palpable or visible. If studies were conducted on patients with DIVA, the results may differ from those of our study.
This study had several limitations. First, a significant difference between the newly developed and existing needles was not detected owing to the low power of this study. While a previous study reported 24% failure of initial PIVC insertion, the rate of failure of initial PIVC insertion in our study was much low, at approximately 15%.[5] Furthermore, the absolute risk reduction associated with the use of newly developed needles is calculated to be 29 patients per 1,000 (2.9%), converting the adjusted OR in our study to a risk ratio.[16] Given a 1:1 allocation, it was determined that the enrolment of 4,800 patients would provide 80% power at a 2-sided α error level of 0.05 to detect statistically significant differences between the two needles. However, the power was calculated to be 12% for the 522 patients in the current study. Therefore, a significant difference between the newly developed and existing needles was not detected because of the insufficient number of patients. Nevertheless, in the results of the subgroup analysis as an interaction analysis, there was a statistically significant difference in insertion by physicians excluding residents, with individuals with ≥ 6 years of nursing experience as a reference (e-Table 4 in Supplemental File 2). This significant difference may have been detected as an accidental error because of the insufficient sample size, and the interaction of the person performing the insertion may no longer be applicable if a sufficient number of patients were included. Second, the difference between the two needles in our study may have been detected if different subjective outcomes were measured. The bevel angle of the newly developed needles was 20°, and the penetration forces at the needle and catheter tips for the newly developed needles were low.[9] A previous study indicated that a decrease in skin resistance during insertion may reduce pain, and it is thought that the number of pain spots to be stimulated decreases because a low resistance reduces the extent of skin deformation during PIVC insertion.[17] The newly developed needles may be less painful than the existing needles because of their low penetration forces and resistance during PIVC insertion.[9, 17] Therefore, differences between the newly developed and existing needles could have been detected if studies had been conducted considering different outcomes, such as pain scoring. Finally, external validity may be low because there were only a few patients with DIVA in our study. In a previous study that evaluated the risk of DIVA on a three-level scale of low, medium, and high, approximately 30% of the included patients were judged to be at medium and high risk.[18] The items used to classify risk into three levels in the previous study were palpable and visible veins. However, in our study, the veins of approximately 10% of the included patients were not palpable or visible, and this percentage is considered low compared with that in a previous study.[18] Moreover, BMI is thought to affect DIVA.13 The median BMI in our study was 22.0 kg/m2. Given that the mean BMI in developed countries is approximately 25–30 kg/m2, the patients included in our study would be considered thin compared with those in other countries.[19] Therefore, the results of our study may not be applicable to patients with DIVA whose veins are not palpable or visible and whose BMI is high.