We carried out a post hoc analysis of CLEAN 3 database to assess the link between PIVC insertion site and its infectious risk. The value of the CLEAN 3 database is that it is recent and includes almost 1000 PIVCs with few missing data. Moreover, we used research staff to ensure high quality data collection and we sent over 85% of PIVC tips to the laboratory for culture. We used catheter colonization instead of PIVC-related BSI as it is by far a much more common event and is regularly used as a surrogate of PIVC-related BSI because colonization usually precedes BSI [8]. Using PIVC-related BSI would have required inclusion of tens of thousands of PIVC, which is difficult to achieve with the collection of large amounts of data and the sending of PIVC tips for culture.

In our study, insertion of the PIVCs at the wrist or cubital fossa increased the risk of colonization and/or local infection, as well as the risk of positive PIVC tip culture. The choice of the best PIVC insertion site provided conflicting results in the literature. In a secondary analysis involving 12 prospective studies and almost 12,000 PIVCs, insertion of PIVCs at the wrist or cubital fossa were associated with increased non-infectious complications leading to PIVC failure (i.e., infiltration, occlusion and dislodgment) [6]. Unfortunately, the association between PIVC insertion site and the risk of infectious complications were not investigated. In a retrospective study of 24 cases of PIVC-related BSI in adult patients, PIVC involved were more frequently inserted in the forearm and arm and less frequently inserted in the back of hand [5]. As this was a retrospective study, the authors included only cases for which all diagnostic criteria were met. Thus, the actual number of cases was probably underestimated, which could have influenced the impact of insertion site choice on PIVC-related BSI occurrence. In addition, the insertion sites for catheters without BSI were not recorded. Thus, it was not possible to establish a link between insertion site and infectious risk. Finally, in a large prospective cohort study involving more than 400,000 PIVC, hand insertion reduced the risk of PIVC-related BSI (HR [95% CI], 0.42 [0.18–0.98], p = 0.046) compared with proximal insertion sites [9].

We believe that insertion sites close to the joints could lead to PIVC dislodgment, thus damaging the endothelium of the vein and enabling bacteria from the insertion site to penetrate the body. These two components increase the risk of phlebitis and infectious as well as noninfectious complications. Moreover, the joints compromise the hold of the polyurethane dressing. Dressing disruption is a well-known major risk factor of infectious complications associated with vascular catheters [10].

In our study, the number of attempts for PIVC placement did not increase the risk of colonization and/or local infection, as well as the risk of positive PIVC tip culture. Few studies have examined this issue. A multicenter observational study of 5,300 PIVCs reported that more than two puncture attempts increased the number of catheter failures (HR [95% CI], 1.48 [1.19–1.84], p < 0.001), although this study did not specifically look at infectious complications. We believe that unlike with central venous catheters, where the same site is frequently punctured in the event of insertion failure, PIVC insertion failure requires the operator to change the insertion site, which may explain why the number of attempts is not correlated with risk of infectious complications.

Our study has several limitations. Firstly, this is a post-hoc analysis of a single-center study, which may compromise the external validity of the results. However, the large number of patients included and the wide range of medical conditions presented makes it possible to explore a representative sample of the general population. Secondly, only patients visiting our emergency department were included. PIVC inserted in emergency departments are at greater risk of infectious complications. However, only experienced nurses took part in the study, guidelines to prevent PIVC-related BSI were rigorously applied and PIVC inserted urgently were excluded. Thirdly, the study was not randomized according to insertion site. In emergency departments, PIVCs are mostly inserted in the cubital fossa or forearm, as these veins are easy to puncture and of large diameter. This enables insertion of larger-diameter PIVCs, which are more effective when vascular filling, blood products or contrast media administration are required. However, we did multivariate analyses taking into account all covariates of interest to identify independent factors associated with PIVC-related infectious complications.

The choice of the insertion site for a PIVC depends on a variety of factors, including the quality of the patient's venous network, the diameter of the catheter to be inserted, patient comfort and the risk of infectious and non-infectious complications. Our study suggests that the wrist and cubital fossa should be avoided whenever possible to reduce the risk of infectious complications. Prevention measures should consider the insertion site to reduce the risk of severe infections associated with PIVC.