So, next we'll welcome Dr. Justin DiVanzo to the podium, and he's completed a randomized trial looking at the complication rates of PICC lines and central venous lines. Hello, I'm Justin DiVanzo. I'm one of the residents at Hershey, and I'm going to talk to you about a randomized trial that we did in our neuro-ICU looking at PICC lines and central lines and all cause complications of those two lines when placed by members of our study team. No disclosures. So we know that central venous access is an important part of the care of critically ill patients, and we know there are two ways that this can be accomplished, through traditional central venous catheters and through peripherally inserted central catheters. The risk profile for each of these lines is very different. We know that PICC lines have been found to have an association with large vein thromboses, and central venous catheters have had some insertional-related complications that are not necessarily noted with PICC lines. Some of the previous data that we looked at found these to be equivocal with each other, while others found that one was superior to the other. In addition to that, we felt that patients with critical illness secondary to neurologic disease represented a very unique population. A lot of them have decreased mobility, they have depressed mental status, and an increased risk of hemiplegia. So we posed the question, which type of line poses the least risk to critically ill patients with neurologic disease, either PICC lines or central venous catheters? So this was a randomized controlled trial for which our randomization was carried out using a computer-generated randomization sequence. The allocation was concealed prior to randomization, and all of the lines were placed by study team members who were available at all times. So we had 24-7 call coverage for placing these lines within the study. The inclusion criteria for this study included being admitted to the neuroscience ICU and being deemed to require central access by the ICU team. Our exclusion criteria included non-English speaking, an emergent condition in which informed consent could not be obtained. That was a fairly rare reason for exclusion since we were available at all times. Previous large vein thrombosis anywhere in the body, an existing central catheter, or pre-existing septicemia. Our study endpoints included failure to insert, removal of device, discharge from the hospital, or death. And the primary outcome that we looked at were all-cause complications with a specific focus on infection, large vein thrombosis, and insertional complications. All of our PICC lines were placed by study team members using ultrasound guidance. Our central venous catheters were placed either in the internal jugular using ultrasound guidance or in the subclavian with traditional anatomic landmarks. It was the provider discretion with regards to location of placement and when to deem failure of insertion. So if the patient randomized a PICC line, they could choose either arm, right or left, regardless of hemiplegia. And they could choose brachial, basilic, or cephalic veins. And same for central venous catheters, right or left side, internal jugular, or subclavian were all acceptable places of placement. And the final position of all of our lines were confirmed with the chest X-ray. We did a pre-study power calculation based on a previous observational study. For the primary outcome that we were looking at, we expected a rate of 11% for PICC lines and 4% for central lines. And using our alpha of 0.05 and a power of 0.8, we determined that we would need 181 patients in each arm. And we planned for interim analyses at 50 and 150 patients. This was our flow chart. Starting in July of 2015 to December of 2017, 580 patients were admitted to our ICU, 428 of those were excluded from our study, the large majority of which were just deemed not to need central access. We then randomized 25 to each PICC versus central line. An interim analysis was completed at that point and the trial was continued. We then proceeded to randomize another 47 patients to PICC line and 55 to central lines where we went to our second interim analysis and the study was stopped. These were our demographics. So ultimately we had 72 patients admitted to the PICC line portion of our trial and 80 to the central line portion. You can see that the numbers throughout this are very similar and we had a wide variety of diagnoses for which we placed central venous catheters in the ICU. As far as complications go, we divided these into classical complications and then the more odd complications. So we had four complications in PICC lines, all of which were large vein thromboses and we had one complication in our central venous catheter group which was insertional trauma. And you can see there was no statistical significant difference between those two groups. When we put everything together, we ended up having a total of 14 complications with PICC lines, the majority of which were failures of insertion due to either patient anatomy or other situational surroundings, and two early removals which were the patient pulled the line out unintentionally. And then we had 10 total complications with central venous catheters, also with five failures of insertion, a couple mechanical failures and tip malpositions, and one early removal. Again, even including all of these, we noted no statistical significant difference. And then overall mortality between the two groups, not necessarily secondary to the line, was also not statistically significant. We also looked at this by diagnoses to see if there was a group of patients that we could single out as one line being better than the other. And we also found here that there was no statistical significant difference between all of these different diagnoses that we see in our ICU routinely. We also looked at ultrasound for our central venous catheters to see if that made a difference with regards to failure of insertion or classical complication. As you can see, we had two failures of insertion with ultrasound with the central venous catheter and three with anatomic guidance, which turned out to not be statistically significant. And one classic complication with ultrasound guidance, which was a pneumothorax, and zero with an anatomic guidance, which was also not statistically significant. Finally, we looked at our first 50 patients and then the last 102 patients to see if there was any change in our complication rates as time goes on. And you can see that the percentages are fairly similar. All complications, 18% with the first 50 group and 14% with the last 102. And same goes for classical complications, 4% in the first 50 and 3% in the last 102, all of which found not to be statistically significant. So this was the largest randomized study to assess for all complications in patients receiving PICS or central lines in a neuroscience ICU setting. We found no difference between PICS and central lines with regards to all complications and classic complications, including large vein thromboses, insertional trauma, and infection. All of our lines were placed by study team members, which allowed these to be placed at all hours and in a semi-emergent fashion if necessary. We did not do any screening protocols for our complications, so we didn't routinely draw blood cultures or routinely get ultrasound. We let all of that to the ICU team, so when they felt it was necessary to do that, they pursued cultures or an ultrasound. We included other complications associated with central axis, which are not often seen in other studies, including failure of insertion, tip malposition, catheter malfunction, et cetera. The DSMB stopped our study early based on the interim analysis at 150 patients. And the reason we were stopped is based on our current complication rate, we were unlikely to show the pre-study expected difference, even in rolling to our pre-specified power calculation. This was secondary to our overall low rate of complications, and we noted that the complication rates were the same between our 0 to 50 patients and our patients in 51 to 152, so we didn't feel as though continuing on at that point was worthwhile. The limitations to this study, this is a non-blinded study, which is an unalterable source of bias. We obviously know what catheters we put in, so it would be hard to blind that any further. And all of our lines were placed by experienced providers, so oftentimes PICs were placed by vascular access teams. The low complication rate could at least partially be accounted for by the experience of the study team. We became very facile with putting these lines in quickly and efficiently. So in conclusion, PICs and central lines have similar complication rates when placed by experts in a neuroscience ICU. We feel this provides level two evidence, and clinical experience and judgment should be used to choose the appropriate line for each patient. Thank you very much. Thank you for your talk.

complication rates

PICC lines

central venous catheters

neuro-ICU

experienced providers