This was a prospective, cohort, pilot study to assess two quality of life and patient satisfaction assessment tools. Ethics approval for this study was obtained from the Institution’s Biomedical Ethics Board.
2.1. Subject Selection
All subjects with newly diagnosed breast or colorectal cancers who were seen at a major provincial Cancer Center were invited to participate in the study. Subjects were enrolled between January 2016 and December 2018. Patients with a coagulopathy, an active infection, or those who were not candidates for chemotherapy were excluded. Locally, it is empirically accepted that if intravenous treatment is to be provided for less than 6 months, a PICC is an appropriate device. Alternatively, if treatment is expected to be for longer than 6 months, a port is implanted.
The referring clinicians from the Cancer Center evaluated candidates for this project based on their initial patient contact interview, physical, and laboratory assessment. Once the duration of the required intravenous chemotherapy was determined, based on the patient’s disease status, a request for the desired intravenous device was sent to the Medical Imaging Department. Consent for involvement in the research project was acquired by a research nurse.
2.2. Central Venous Catheter Devices
All venous access devices for this project were inserted by an interventional radiologist from the Medical Imaging Department. Consent for the venous access device was acquired by the interventional radiologist involved.
The PICC implanted for this project was a 5F, dual lumen, Angiodynamics, BioFlow, Power PICC (Angiodynamics Canada Inc., Oakville, ON, Canada). The non-power injectable arm port was a 5F, single lumen, Cook Vital Mini Titanium port (Cook Medical LLC, Bloomington, IN, USA), while the power injectable device was a 6F, single lumen, Bard Titanium Slim power port (Bard Canada, Inc. Oakville, ON, Canada).
All venous access devices were implanted into the arm, between the antecubital fossa and the axilla, using ultrasound or arm venography for venous access. The vein selected for catheter insertion was preferentially the basilic vein. The preferential arm for PICC and port insertion for trial subjects was their non-dominant arm if possible. For women with breast cancer, the contralateral arm was used for device insertion. The venous access procedures were performed using a sterile technique and local anesthetic. Patients with colorectal cancer often require frequent intravenous contrast enhanced computed tomography (CT) examinations during their care; therefore, a power injectable venous access device was implanted for this group to enable the rapid injection of intravenous contrast agents for follow-up CT scans. In general, patients with early stage breast cancer do not require frequent contrast enhanced CT scans; therefore, this group of subjects received the non-power injectable port device. The current PICC, used for all subjects, was power-injectable. Technical details for these device implantation procedures have been previously described and were followed for this project [5
]. As per institutional guidelines, PICCs were flushed once weekly, whereas ports were flushed every 4 weeks when not in use.
2.3. Subject Surveys and Clinical Follow-Up
The European Organization for Research and Treatment of Cancer General Quality of Life assessment (EORTC QLQ-C30) targeted towards patients with cancer was administered along with a locally developed Quality of Life Assessment Venous Device (QLAVD) survey comprising 30 items [4
] (Table S1
The QLAVD was modeled after a survey tool created by Marcy et al. entitled, “Questionnaire for Acceptance of and Satisfaction with Implanted Central Venous Catheter (QASSIC)”. The QASSIC was statistically validated and found reliable by Marcy et al. However, this was a French language survey [7
]. The QLAVD was created by a local working group of oncologists, oncology nurses, and interventional radiologists using the QASSIC, after it was translated locally into English, revising some of original questions, adding several new questions, and finalizing the structure of this new survey tool. The initial deployment of the QLAVD was completed by Burbridge et al. [4
] Subsequently, Liu et al. translated the QLAVD into Chinese and performed a statistical assessment of reliability and validity. Liu et al. found that the QLAVD demonstrated good content validity, had internal consistency, and had a high degree of reliability and stability [8
The subjects for our pilot study were asked to complete both surveys, at four different times, within one month after device insertion (baseline), then at three months, six months, and twelve months after venous access device insertion. Completion of these surveys continued until the removal of their venous access device for any reason, termination of treatment with the venous access device, or subject death (whichever came first). These surveys were administered personally to each subject by a research assistant.
The subjects were followed clinically for the detection of venous access device complications during the course of their infusion therapy, as summarized from the previous literature [9
]. Imaging for all subjects was collected in a solitary provincial Picture Archive and Communication in Medicine System (PACS). Therefore, any venous access device complication that required imaging was easily available and was reviewed.
2.4. Quality of Life Constructs
The EORTC QLQ-C30 comprises 30 items, 24 of which are aggregated into 9 multi-item scales: 5 functioning scales (physical, role, cognitive, emotional, and social); 3 symptom scales (fatigue, pain, and nausea and/or vomiting); and 1 global health-status scale. The remaining six single items assess symptoms of dyspnea, appetite loss, sleep disturbance, constipation, diarrhea, and financial impact. All of the scales and single-item measures range in score from 0 to 100. The questions in these constructs were scored individually and summated, and a mean score was generated. The principle for scoring the scales was the same for all patients. Firstly, the average of the items that contributed to the scale was estimated to create the raw score; then, a linear transformation was performed to standardize the raw data [15
]. The mean scores were compared to assess potential differences in responses between the PICC and port groups.
A similar strategy was employed for the QLAVD survey and the questions in this survey were pooled into larger question blocks. The QLAVD survey questionnaires were aggregated into psychosocial, treatment, satisfaction, and procedural constructs. Subjects’ pain experience at the time of device implantation (procedural construct) was assessed using a single question: “how painful was it for you to have your device implanted?” Subjects were asked to respond based on a scale from 1 (minimal discomfort) to 10 (severe discomfort). The psychosocial construct was calculated by summing up 18 questions relating to subjects’ feelings, beliefs, and experiences of having the device implanted in relation to daily living, psychosocial, emotional, and physical activities. The higher the score, the more negative the subject’s experience. The satisfaction construct consisted of three questions related to satisfaction with the venous access device. The treatment construct was calculated by summing up 8 questions relating to the subjects’ feelings, beliefs, and experiences in relation to the treatment process. The higher the score, the more negative the subject’s experience.
2.5. Statistical Analysis
Descriptive analysis was performed to summarize the data. Mean and standard deviation estimates of each construct (procedural, psychological and treatment) were calculated separately and compared between patients with a PICC or a port for each of the four time points using Student’s t-tests. To investigate the changes in mean scores for different constructs between the PICC and port groups across the four time points, mixed effects models were used, taking into account the interactions between the groups (type of device used) and the time points (4 survey times). Due to the high drop-out rates at times 3 and 4, mixed effects models were only run for the first two surveys (1 month and 3 month), taking into account the interaction between groups and times, controlling for other covariates such as gender, whether the subject was left- or right-handed, and the arm used for device implantation. Statistical analyses were performed with the Statistical Package for the Social Sciences (SPSS) Version 26 (SPSS Inc., Armonk, NY, USA; IBM Corp., Endicott, NY, USA). Statistical significance was set by an alpha level of 0.05.
It is understandable that subjects with the port experience more pain at the time of device insertion and during chemotherapy infusion. Port insertion requires a skin incision and the creation of a subcutaneous pocket in which the device resides. Skin closure is also required to facilitate wound healing. It is a minor surgical procedure with risks of hemorrhage, infection, and pain.
Infusion of chemotherapy for those with a port requires the direct insertion of a non-coring infusion needle through the skin into the inner chamber. The pain is similar to that involved for phlebotomy. However, accessing the port is much more predictable and reliable in the hands of trained personnel, resulting in fewer needle punctures than for those patients who do not have a port or PICC.
Given the findings associated with port device pain, this could be considered an important issue to raise with patients when discussing the type of venous access device required for infusion therapy. If patients have a severe needle phobia, a PICC may be more appropriate.
Ports had very little impact on how the subjects dress, because they are associated with a small upper arm incision and a small protrusion of the skin on the upper arm. There are no components of the port that hang from the arm or are external to the skin. Hence, there is no need to wrap or fixate the port. The PICC, on the other hand, requires great care in covering the device with a clean dressing and a fixation device that protects the PICC from inadvertent extraction and infection. Additionally, some patients may choose to dress in a fashion that protects the PICC from inadvertent damage by it becoming caught on something in one’s surroundings and becoming inadvertently extracted. Clothing choices are definitely not as limited with the port as with the PICC.
Difficulty with showering or bathing with the PICC is related to the need to keep the PICC insertion site clean and dry. This is not an issue with a healed port incision because patients are able to swim, bathe, shower, and participate in unrestricted water activities once the insertion site incision has healed. If patients are avid water activity enthusiasts, this is an important differential functionality between the two devices. Overall, the level of activity related to work, sports, and social interactions should also be taken into consideration.
Having people negatively comment on the presence of the device is a result of the fixation of the device, the site dressing, and the external limbs with the PICC being readily visible, which is not the case for the port due to its subcutaneous location. This is unavoidable due to the differences in device design. Usually, the only way to avoid this is to dress in a fashion that covers the implantation site and prevent others from seeing the location of the device. The mixed effects model did not show significant changes in mean scores for different constructs including mean pain and psychological scores (improvement or worsening) between the PICC and port groups from baseline to three months. This suggests that patients with a PICC line continue to experience less pain during device access, but a high negative psychological effect compared to subjects with a port.
The complication profiles for subjects with a PICC or port seem comparable, with the exception that the power injectable port does seem to be associated with a slightly higher rate of venous thrombosis, although this was not found to be statistically significant in this cohort. Additionally, the detection of venous thrombosis was based on clinical signs and symptoms; this may not be a reliable method of thrombosis detection because there is a tendency for asymptomatic venous thrombosis to develop in this patient group [16
]. Given that power injectable venous devices are usually associated with a larger catheter diameter, which has been proven to increase the risk of venous thrombosis, the role that power injection may play in the treatment and follow-up of patients should be balanced against the potential risk for venous thrombosis [5
Although our study showed that both groups had similar complication rates, other studies have shown that PICCs compared to ports tend to be associated with a high complication rate [17
]. For example, Patel et al. prospectively examined CVC-related complications in 70 patients with a solid tumor and reported that, overall, 6% of the patients with a port compared to 20% patients with a PICC line experienced major complications [17
]. A systematic review of the literature showed that PICC compared to the other central venous lines including a port was associated with an increased risk for DVT [20
]. In the present study, the equivalent complication rates related to the device could be a random occurrence. Furthermore, complications leading to PICC removal may have been underestimated if the device was removed and the reason for removal was not recorded in the medical record. The rate of non-occlusive and occlusive venous thrombosis detected using ultrasound was outside the expected norm, which may very well be related to catheter material and catheter size. When the project was initiated, a 6F power port was used for arm implantation and power injection. We have transitioned to a smaller catheter diameter with a different manufacturer. Finally, there are risks of contrast extravasation with simple peripheral intravenous catheters, PICCs and ports [22
]. The ports are designed and rated for high-volume contrast, being injected under increased pressure. Contrast extravasation is a nuisance complication that fortunately rarely occurs and usually does not have dire consequences, especially now that contrast is low-osmolality and non-ionic in chemical composition. It is important that ports must be properly accessed by trained healthcare professionals.
Long-term comparison of these two different venous access devices is limited by the expectation that PICCs will be employed for intravenous treatment for a shorter time frame, i.e., usually less than six months. This results in a significant drop-out effect for data capture for those subjects with a PICC that may have confounded complication rates in favor of PICC lines. In addition, it is much easier to have a PICC removed, and clinicians are more inclined to simply remove a PICC and have a new one implanted if there are any potential complications, regardless of the severity of the complication. Hence, one of the limitations of our study is the lack of longer-term follow-up and detailed information on duration and reasons of the removal of PICC lines.
An additional limitation of our pilot study is the small cohort size. However, the cohort was found to be statistically comparable to the EORTC QOL survey results, suggesting that there were no substantial differences in the overall group make-up that would adversely affect the results of the QLAVD survey. The groups were balanced, with 50 subjects in the PICC group and 51 subjects in the port group. However, no attempt was made to balance the group according to the stage of their malignancy. This could be a potential factor to consider for future research. Our results do not suggest that quality of life was diminished between the two major groups, i.e., PICC vs. port. Nevertheless, we do not know if the frequency of line care, such as venous catheter flushing, has adverse effects on the QoL of patients with a venous catheter who are not on active treatment. Lastly, this study was performed in subjects with breast and colorectal cancers, and findings may not be generalizable to patients with other solid organ and hematological cancers.
In conclusion, this pilot study suggests that the decision to employ a venous access device for the treatment of a malignancy is a multi-factorial one. The duration of treatment cannot be the only factor taken into consideration. Determination of the degree of needle phobia, levels of sporting and social activities, ability of the patient to swim, shower, and bathe, and the need to alter one’s dressing habits should also be considered. The need for a power-injectable device is also potentially clouded by the possibility of an increased risk of venous thrombosis, which should be taken into account.