Introduction

Hospital-at-home (HaH) programs enable the treatment of infectious diseases through outpatient parenteral antimicrobial therapy (OPAT), which largely corresponds to prolonged courses (>14 days)¹,². This is why an adequate vascular access device (VAD) is necessary to minimize the rate of catheter-associated complications²-⁴.

International guidelines for selecting the most appropriate VAD consider various factors, such as compatibility with peripheral veins, staff expertise in insertion and maintenance, the condition of the patient's superficial venous network, the presence of chronic kidney disease (stage 3b, 4, or 5), and the duration of treatment²,⁵-⁷. For prolonged OPAT or when solutions are not suitable for peripheral administration, the guidelines recommend using a peripheral inserted central catheter (PICC). PICCs provide central venous access, can be inserted by nurses, offer patient satisfaction, and allow for easy care, making them suitable for home-based use⁵,⁷,⁸.

It is common to use the short peripheral catheter (SPC) in hospitalized patients, with some registries indicating it as the most utilized VAD⁹. Although this catheter is less expensive and easier to insert, complication rates of up to 59% have been reported, often leading to catheter failure¹⁰. These complications can be major, such as deep vein thrombosis (DVT) and catheter-associated bloodstream infection (CRBSI), as well as minor, including phlebitis, local infections, occlusion, extravasation, infiltration, and displacement or accidental removal⁴,¹⁰. The necessity for a new catheter insertion increases the risk of further failures, impacts the patient's venous network, causes pain, reduces patient satisfaction levels, prolongs care, and results in additional costs¹⁰.

Available literature reports a lower rate of minor complications and reduced use of catheters and venipunctures during treatment with PICC compared to SPC, resulting in a lower rate of catheter failure with PICCs¹¹,¹²; however, this benefit is offset by the higher cost of consumables associated with PICC use¹³-¹⁶. As for major complications, there is a higher tendency to report them with PICCs in the in-hospital setting. Nevertheless, in the outpatient setting, there appears to be a lower frequency of CRBSI (0.04 per 1000 catheter-days), as well as reduced rates of infections and thrombosis⁸,¹³,¹⁷,¹⁸.

In Colombia, some studies have assessed the complication rates associated with VADs¹⁹, including evidence confirming a lower frequency of CRBSI in HaH settings when using PICCs²⁰. However, no studies evaluate the cost-effectiveness of using PICCs versus SPCs for prolonged OPAT regarding the complications associated with each device. This study aims to support more informed decision-making on allocating resources for VAD use.

Materials and Methods

An economic evaluation study was conducted to assess the relationship between safety (i.e., the occurrence of complications) and the costs associated with two VAD options for prolonged OPAT in adults receiving HaH care. The methodology followed the Colombian Health Technology Manual (IETS, for the Spanish acronym) guidelines for economic evaluation studies²¹ and adhered to the recommendations of the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) 2022²².

The time horizon for the analysis corresponded to the duration of antimicrobial therapy (a minimum of 14 days), considering that VAD-related complications occur during the catheter's dwell time. The third-party payer's cost perspective was considered, evaluating direct medical costs associated with the use of the devices, required supplies, as well as the cost of any complications that occur. A discount rate was not applied, as the time horizon was less than 12 months. Approval from the institutional ethics board was obtained, and no patient data were used. No funding was received for this study.

Outcomes

Complications were classified as minor (phlebitis, occlusion, extravasation, dislodgement, and superficial vein thrombosis [SVT]) and major (DVT, CRBSI)¹³,²³. To determine their frequencies, a comprehensive literature search was performed in the PubMed, Embase, CINAHL, LILACS, and Scopus databases without year restrictions. The search was limited to studies published in English or Spanish and included comparative, analytical, observational, and cross-sectional studies involving adult populations in home and/or in-patient settings, with exposure to PICC and/or SPC as the VAD. Studies involving pediatric populations, intensive care unit hospitalization, case reports, case series, and the use of VADs for chemotherapy administration were excluded. For study selection, an independent screening of the documents according to title and abstract was performed by two authors, with a third reviewer resolving any disagreement. The Rayyan - Intelligent Systematic Review tool was used²⁴, and a full- text review of the articles was subsequently conducted. Data extraction was then performed using a database designed in Microsoft Excel to capture the highest quality of evidence for each complication.

Subsequently, focus group meetings with experts from disciplines involved in the management of device-related complications were held. The aim was to contrast the findings from the literature with the reality and experience of each group and to compile the inputs, diagnostic methods, and therapeutic methods necessary for managing each complication. The distribution was as follows: for minor complications, a focus group was conducted with three experts in vascular access nursing and interventional radiology; for thrombotic complications, information was gathered from the institution's anticoagulation clinic group; and for infectious complications, meetings were held with experts in infectious diseases.

Cost evaluation

Based on the information gathered from the focus groups, all diagnostic and therapeutic tools for using each VAD (insertion, maintenance, and complications) were grouped. A unit cost was assigned to each of these inputs using information from the institution’s purchasing department and the institution's charge rate conditions, obtaining the charge rates from an insurer. The costs of inserting a new device were not included, as the device is not always removed in the event of complications, and even when it is removed, a new device is not always inserted. All costs were estimated in 2022 Colombian pesos (COP) and converted to U.S. dollars using the average exchange rate for that year (1 USD= 4,255 COP).

Modeling and analysis

A decision tree model was designed using the TreeAge Pro2015 software tool (TreeAge Software, Inc., Williamstown, MA). For each VAD, branches were constructed, including probability nodes with the frequencies of major and minor complications, along with the associated costs. The results were measured using the incremental cost-effectiveness ratio (ICER) (PICC cost - SPC cost / complication avoided with PICC - SPC). A univariate sensitivity analysis was performed using a tornado diagram, and a multivariate sensitivity analysis was conducted using Monte Carlo simulation. The dataset is available in Mendeley Data²⁵.

Results

Literature search for complications

The initial search identified 1053 references. Three were obtained from additional sources, including expert recommendations, reference lists, and gray literature. The final selection process included 18 articles. Of these, 2 evaluated PICC-related CRBSI¹⁷,²⁶, 3 evaluated PICC-associated DVT²⁷-²⁹, 2 explored local infection and phlebitis³⁰,³¹, and 5 various outcomes related to PICC use⁸,¹¹-¹³,³². Regarding SPCs, 1 article evaluated the presence of CRBSI33, and 6 evaluated the other outcomes¹⁰,¹³,³⁴-³⁷. For the selection of the source of the reference case for each outcome, the article with the best available quality of evidence and a reported range of frequencies was chosen and evaluated. Overall, the frequency of major complications for each catheter is similar, while minor complications were more frequent with SPCs than PICCs ( Table 1).

Focus groups

The list of inputs obtained through the focus groups is summarized in Supplementary Table 1. Regarding minor complications, the nursing focus group agreed with the frequencies found in the literature search. Considering thrombotic complications (DVT and SVT), the anticoagulation clinic group agreed with the frequencies reported by the reference case, and, finally, the infectious disease group supported the results found for CRBSI.

Costs, modeling, and analysis

The total cost of using each VAD was calculated in COP according to the person responsible for insertion (nurse or interventional radiologist), maintenance, and the occurrence of each complication. The highest costs were associated with PICC insertion by interventional radiology ($281) and the cost of major complications with the use of both VADs (DVT: $207, CRBSI: $123.3) (Table S2).

The reference case for PICC assumed 50.00% insertion by interventional radiology and 50.00% by nursing staff, a catheter price of $74.7, a minimum treatment duration of 15 days, and the complication frequencies outlined in Table 1. For SPC, the reference case included nursing-led insertion, a catheter price listed in Table S1, the same treatment duration as PICC, and the complication percentages reported in Table 1. In the decision tree model (Figure 1), the frequency of major complications was similar for both VADs (3.30% for PICC, 3.45% for SPC), which limited the analysis to minor complications, whose costs varied depending on the difference in the occurrence of each outcome. SVT and phlebitis associated with SPC use were the costliest minor complications (Table S2). In the reference case, the ICER per minor complication avoided with the use of PICC was $199. The cost-effectiveness scatterplot illustrating various catheter price scenarios is shown in Figure 2.

PICC: Peripheral inserted central catheter

Figure 1 Decision tree model  

PICC: Peripheral inserted central catheter

Figure 2 Cost-effectiveness Scatterplot  

Sensitivity analysis

In the univariate sensitivity analysis of minor complications for PICC, the variables with the most significant influence on the overall cost were those related to catheter insertion costs. For SPC, the key factors affecting costs included insertion expenses, phlebitis-related costs, and the duration of hospital stay. These findings are illustrated in the tornado diagram (Figure 3a-b).

Considering the catheter cost and the individual responsible for insertion, in the two-way sensitivity analysis, the most unfavorable scenario is PICC insertion in 100,00% of cases by interventional radiologists at the current catheter price, resulting in an ICER of $323.40. Conversely, with 100,00% nurse-led PICC insertion at the standard catheter price, the ICER to avoid a minor complication is $74.60. In more realistic scenarios, with 90,00% of nurse-led PICC insertions and a 30,00% reduction in PICC price, the ICER would be $69,40; it would be $49,2 with a 50,00% price reduction and $24,3 if 100,00% of catheters were inserted by nurses at that last price.

Figure 3 Univariate sensitivity analysis. Tornado graph for PICC (3a) and SPC (3b)  

Discussion

In the literature review, the frequency of major complications is comparable between both VADs when the patient is at home, as the reported incidence of CRBSI and DVT is lower than in hospitalized patients, in some cohorts being close to 0%⁸,¹¹,¹⁷,³⁰ and reaching a frequency similar to that reported with SPC use³³.

The economic evaluation conducted in this study suggests that PICC is more effective than SPC in reducing the occurrence of minor complications, except for accidental withdrawal, which showed similar rates for both VADs. SVT and phlebitis were identified as the most common and costly complications associated with SPC use. Furthermore, an incremental cost of $199 was documented to avoid a minor complication with PICC. The univariate sensitivity analysis showed the influence of higher costs of PICC insertion and materials, aligning with the findings of Dychter et al.; however, complications were not considered in this review³⁸. Likewise, a study by Periard et al.¹³ assessing the cost-effectiveness of both VADs showed higher costs associated with PICC due to the insertion materials. However, unlike our study, they evaluated efficacy based on the number of catheters and venipunctures needed for intravenous treatment. In contrast, a review of cost-effectiveness studies of PICC and SPC in the pediatric population found favorable outcomes for PICC use in short-term parenteral nutrition³⁹. This highlights that, to the best of our knowledge, no economic evaluation studies have considered hidden post-insertion costs of VADs, such as maintenance and complications, in the adult population.

Catheter failure rates are reported at 54-59% for SPC¹⁰,⁴⁰ and 20% for PICC in two cohorts of patients undergoing long-term antibiotic treatment¹¹,³¹. It is widely recognized that the risk of subsequent failures increases with each catheter failure due to factors like phlebitis, infiltration, extravasation, vascular quality, and history of previous venous punctures⁴⁰. While our study did not account for this factor, given the complexity of cost and measurement implications, the univariate sensitivity analysis revealed that factors with the greatest economic impact for SPC included insertion cost, phlebitis, and length of hospital stay. These factors may align with a higher risk of subsequent failures and complications, given the longer exposure time. Considering the failure rate of PICCs, their use may be advantageous when evaluating total costs, especially in prolonged treatments, supporting international guidelines recommending PICC use for treatments exceeding 14 days²-⁵.

Other potential hidden costs associated with extended treatments, such as transportation expenses for home care personnel, patient satisfaction, and venous depletion due to catheter failure, were not factored in this study due to challenges in cost assessment¹⁰. A randomized study comparing SPC and PICC use in treatments lasting more than 5 days showed higher levels of patient satisfaction with PICCs (96.8% vs. 79.3%)¹³. Despite the limitations in assessing patient satisfaction costs, whether this factor could indicate an overall benefit in prolonged parenteral treatment scenarios remains uncertain.

Complementarily, we emphasize the two-way sensitivity analysis, which helped to determine the impact of VAD price variation and the person responsible for insertion on the modification of the ICER. With a higher proportion of insertions performed by nurses, the ICER tended to decrease, as did the cost of the catheter. In a realistic scenario where 100,00% of PICCs are inserted by nurses and the catheter price decreases by 50,00%, the ICER would drop to $24,3. This possibility is supported by a study showing a 93% success rate for PICCs inserted by vascular access nurses, with a 42% increase in costs (p<0. 01) when PICCs are inserted by radiologists, and higher patient satisfaction in the nursing group⁴¹. Another study in pediatric patients also demonstrated lower prices of nurse-led PICC insertion⁴². This highlights the role of nurse-led vascular access teams (VATs), recommended by the Infusion Nurses Society⁶ as a strategy to improve healthcare quality, seize the window of opportunity, reduce venous depletion, and potentially lower total cost, as shown by Cortés et al.⁴³. These findings suggest that the increased institutional demand for PICC use, in line with VAD guidelines, could facilitate price negotiations, potentially leading to a reduction in ICER.

Several opportunities for future research have been identified, particularly the evaluation of the costs associated with peripheral catheter failure and the potential for missed antibiotic doses in Hospital-at-Home programs. This would complement the hidden costs not accounted for in this study and highlight the importance of selecting the most appropriate vascular access device despite its initial cost. Another area worth investigating, though challenging to quantify in terms of cost, is the comparison of patient and caregiver satisfaction in the context of hospital-based home care services, specifically when comparing SPCs and PICCs.

It is hoped that these findings will support similar institutions in selecting the most appropriate VAD for prolonged OPAT to minimize minor complications, catheter failure, and venous depletion. Similarly, promoting the development of vascular access teams that offer cost advantages during VAD insertion and ultimately lead to reduced overall patient care expenses is recommended. Nevertheless, each organization within the system should assess the value they place on preventing minor complications and subsequent catheter failure.

Conclusions

When comparing PICCs and SPCs for prolonged OPAT in HaH settings, this study suggests that PICCs are more effective in preventing minor complications and subsequent catheter failure. However, they remain more costly due to insertion and material expenses.

Nurse-led PICC insertion and implementation of VATs within healthcare institutions could lower overall costs.

With the growing need for PICCs in prolonged therapies, a reduction in incremental costs could be attained, promoting the adoption of this device for the benefit of patients and OPAT services.