Complications with peripherally inserted central catheters (PICCs) used in hospitalized patients and outpatients: a prospective cohort study

Our single-center prospective study describes an unselected cohort of patients, among the first to benefit from PICCs in our hospital in 2010. As is still the case in our institution, PICC were placed exclusively by trained radiologists, and not by dedicated teams at patients’ bedside as in other countries. This specific procedure can limit the generalizability of our data.

Since then, to our knowledge, no prospective follow-up study has evaluated PICC complications among hospitalized and outpatients, regardless of the type of medication infused and patients’ conditions.

In the absence of published guidelines for healthcare professionals using PICCs at the time of this study, our hospital’s Infection Control and Interventional Radiology Departments had established local recommendations for PICC care, including a leaflet for home care. However, all the healthcare professionals were not yet familiar with these best practice rules, which could in part explain the high complication rate, although similar to rates reported in other studies [9, 17–19].

Guidelines for PICC care are now better defined and protocolized [20]. Recent studies suggest lower incidence rates of PICC-related complications, probably due to several technological novelties, better respect of the maximal sterile barrier precautions and improvement of compliance with evidence-based recommendations regarding catheter management in selected populations [8, 11, 21]. Bertoglio et al. documented 15% of complications leading to catheter removal in cancer patients but still concluded that PICCs represent safe devices for chemotherapy delivery, in particular during the first months after insertion [22]. In pediatric outpatients receiving parenteral antibiotic therapy, Kovacich et al. reported that 8% of PICCs required removal due to a complication (4.6 per 1000 catheter-days), underlining the need to discuss the relevance of PICC insertion and maintenance in children [23]. These studies underscore the importance of the type of patients, infused therapies and best practice recommendations.

In our study, incidence of lumen occlusions was high (8.9%), leading to catheter removal in all cases. Recent studies showed occlusion rates of 2.4% and 6% among hospitalized patients [19, 22] and 4.5% and 7.4% among outpatients [16, 24]. Our multivariate analysis identified age??65 years and the presence of a pre-occlusive event as risk factors of lumen occlusion. In our study, catheter occlusion occurred on average 16 days after PICC placement, and was not associated with longer dwell times. We hypothesize that, because of the novelty of PICCs at the time of the study, healthcare professionals did not always follow instructions for the prevention of catheter obstruction and possibly did not heed the warning signs requiring timely prevention measures to be taken. French guidelines have since recommended systematic pulse flushing with saline after every use, heparin being used only as salvage therapy in some cases of lumen occlusion. Some published studies underline the role of nursing expertise in minimizing costs and complications and promote dedicated teams for safe PICC management [25, 26].

The second mechanical complication was accidental withdrawal of the catheter, which was as common as catheter occlusions (8.9%) and occurred mainly among hospitalized senior patients (mean age 70 years) with PICCs fixed on the skin by sutures. To prevent accidental removals, appropriate protection of the dressing is needful, particularly among elderly patients with behavioral disorders.

Concerning infections, we found a PICC-related BSI rate of 0.57 per 1000 PICC-days, which is lower than reported in the literature. For instance, Alenjo et al. reported an overall PICC-BSI rate of 3.13 per 1000 PICC-days in inpatients, higher in the ICU (4.79 per 1000 PICC-days) than in the non-ICU (2.78 per 1000 PICC-days) [27]; Chopra et al. also pointed out the ICU as risk factor for infectious complications with a PICC-BSI rate of 2.16 per 1000 PICC-days [28]. However, these differences can be in part explained by differing definitions of catheter-related BSIs between countries. Indeed, French studies that surveyed PICC-related complications among inpatients showed BSI rates comparable to ours [7, 9, 29, 30].

The prospective design of the study allowed us to register early local inflammatory signs in 25 PICCs (13% of the cohort), and to determine that less than half of these developed a confirmed infection. Moreover, 19 catheters were unnecessarily removed for suspected infections that were not confirmed. This underscores the need to apply rigorous diagnostic procedures for catheter-related infections (including differential blood cultures and insertion site swabbing), even if PICCs are seemingly easier to replace than CVCs.

We found a low incidence of symptomatic PICC-related venous thrombosis (1.6%; 0.57 per 1000 PICC-days). This result was similar to the incidence rate reported in the study of Kabsy et al. among oncologic patients (1.9%) but lower than in other published studies [5, 30, 31]. Turcotte et al. argued that, whereas the risk of infections related to CVCs and PICCs was similar, thrombotic complications were more frequent with PICCs and proposed a tailored approach in the choice of the most appropriate catheter [4].

We observed higher complications rates among hospitalized patients (14.38 per 1000 PICC-days) than in the outpatient settings (3.19 per 1000 PICC-days), with all the confirmed infections and 4/7 possible infections occurring in the inpatient settings. Smith et al. reported a 10-fold greater risk of PICC-BSI among hospitalized patients than outpatients and Chopra et al. demonstrated that PICCs were associated with a lower risk of infections (0.5%) that CVCs (2.1%) in outpatients [15, 32]. In our study, this can be explained by the differences between our in- and outpatient populations: the former had significantly more parenteral nutrition and daily catheter accesses, both known risk factors for catheter-related infections [11, 33]. Moreover, we might have underestimated the incidence of infectious complications, as PICC segments were not systematically cultured in the outpatient setting.