A COST ANALYSIS OF PERIPHERALLY INSERTED CENTRAL CATHETER IN … · 2013-10-10 · A cost analysis of peripherally inserted central catheter in paediatrics Zhaoxin Dong Master of
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A COST ANALYSIS OF PERIPHERALLY INSERTED
CENTRAL CATHETER IN PAEDIATRICS
by
Zhaoxin Dong
A thesis submitted in conformity with the requirements
for the degree of Master of Science in Health Services Research
Graduate Department Health Policy, Management, and Evaluation
1.5.1 Insertion and application of peripherally inserted central catheter ............................ 7
1.5.2 Comparison between peripherally inserted central catheter and peripheral intravenous lines ................................................................................................ 10
1.5.3 Peripherally inserted central catheter in pediatric patients...................................... 11
1.5.4 Complications related to peripherally inserted central catheters ............................. 12
1.5.5 Costs related to peripherally inserted central catheters ........................................... 22
2.5.3 Multivariate linear regression model ..................................................................... 52
2.5.4 Cost comparison between peripheral intravenous therapy and peripherally inserted central catheter ...................................................................................... 54
2.5.5 Regression diagnosis of the multivariate linear regression models ......................... 56
A PICC can be inserted at the bedside with a nurse-lead PICC team, or in an interventional
radiology (IR) suite with the interventional radiology team, consisting of an interventional
radiologist (IR), nurses (RN), technologists, and sometimes an anesthesiologist. The most
commonly used veins for access for a PICC are the basilic, brachial or cephalic veins, especially
the basilic vein which is considered as the access of first choice (Paulson & Miller, 2008). The
basilic vein has a lower incidence of phlebitis compared with other insertion sites (Mazzola et
al., 1999). Factors affecting successful insertion include patient’s age, vein size and condition,
patient edema, hypotension or dehydration as well as impaired skin integrity, all of which can
lead to an unsuccessful insertion. Level of experience of the nurses or interventional radiologists
is another factor that may affect a successful insertion. The success rate can improve from 55%
to 85% as the operator becomes more experienced (Evans & Lentsch, 1999). A PICC requires
careful maintenance in order to prevent complications after insertion, including dressing changes
and securement, daily flushing, and heparinization of the line (Gamulka, et al, 2005).
(2). Insertion procedure in the Hospital for Sick Children1
At the Hospital for Sick Children (SickKids), if a PICC is chosen for a patient, an IR will insert
the PICC in the interventional radiology department called the Image Guided Therapy (IGT).
SickKids is a paediatric, teaching and research hospital affiliated with the University of Toronto.
Before the insertion, informed consent is required. The duration of the procedure depends upon
many factors, including vein accessibility, size of veins, need for sedation. At SickKids most
PICCs (outside of the neonatal intensive care unit) are inserted by IRs. Different sizes of single
or double lumen PICCs, cuffed or uncuffed, are selected for individuals and inserted under a
1 An interventional radiologist provided this information
`
9
combination of sonographic and fluoroscopic guidance. A cuffed catheter is described as “a
catheter with a very short sleeve of dacron attached/adherent to its outer wall, which is
positioned close to the skin exit site in a short tunnel under the skin” while an uncuffed catheter
is “one that has no sleeve of Dacron, is usually not tunneled but placed directly into the vein”2.
A cuffed catheter is a commonly used access device to provide long term access for children
(Goldstein et al., 1997). The entire procedure requires a sterile environment. The use of a guide
wire facilitates the accurate location of the tip of the PICC under fluoroscopy and can direct the
PICC into the correct position. A fluoroscopic image (X-ray) is always taken at the end of the
procedure to document that the line tip is in the proper position. The children can be awake,
sedated, or under general anesthetic during the procedure, based on their procedure duration, age
and the nature of their illness. Most children have a PICC inserted under local anesthetic only.
After insertion, dressings are applied to cover the exposed catheter up to the hub. The site is
dressed and wrapped with a transparent dressing and the extension tube is used as the access to
the line for infusion of medication. The PICCs are later removed when no longer required. If the
PICC is cuffed and present longer than four weeks, the cuff becomes adherent to the skin and
requires local anesthetic and dissection of the cuff which is performed by IRs. If the PICC is
uncuffed or in situ for less than 4 weeks and not adherent to the skin, the PICC can be removed
by an RN.
(3). Peripherally inserted central catheter application
A PICC is most commonly indicated for patients who require IV therapy for a medium or long
term use such as chemotherapy, hyper-alimentation, repeated administration of blood or blood
2 An interventional radiologist gave this definition
`
10
products and venous blood sampling (Amesur et al., 2009). Patients, who require IV therapy but
have exhausted peripheral venous accesses because of frequent IVs, or have sclerotic and/or
thrombosed veins, may still benefit from PICC insertion (Islam et al., 2008). PICCs are also
suitable for patients who require home IV therapy or long term IV access (Weeks-Lozano H,
1991).
1.5.2 Comparison between peripherally inserted central catheter and peripheral intravenous lines
A peripheral intravenous line (PIV), also known as a peripheral cannula, is the most common
intravenous access method for short-term infusion (Lundgren et al., 1996). The cannula is
usually inserted through a peripheral vein, e.g. vein in the hand or arm. It is easy to insert but
requires frequent replacements every 48 to 72 hours in order to minimize the complications such
as extravasation and phlebitis (MOH nursing clinical practice guidelines, 2002). The risk of PIV
therapy includes infection, phlebitis, infiltration, and extravasation with possible issues such as
skin or tissue necrosis (Tully, et al., 1981).
Obtaining PIV access can be challenging in infants with tiny veins and poor cooperation, making
it difficult to provide IV therapies. Compared with a PIV which infuses directly in to a smaller
caliber vein, the central tip of a PICC is positioned in a central vein with larger diameter which
decreases the irritation from medications. Thus it is regarded as a safer approach to IV therapy
from vesicant drugs (Sol et al., 2007). It has been shown that patients have higher satisfactions
with PICCs as compared to PIVs (Polak et al., 1998; Schwengel et al., 2004). A recent study in a
tertiary hospital estimated that 96.8% of patients (mean age±SD, 67.0 ± 16.5) were satisfied with
a PICC as compared to 79.3% of patients with a PIV (Periard et al., 2008). The average number
`
11
of skin punctures was 1.36 per person in the PICC group for treatment compared to 8.25 per
person in PIV group (Periard et al., 2008). In the early 1990s, PICCs became a popular catheter
choice in both adult and pediatric settings because of higher rates of insertion success, better
satisfaction and smaller risk of catheter related complications than other catheters at that time.
PICCs had higher efficacy as an intravenous device because it only required 1.16 punctures on
average for a successful insertion, compared to 1.79 in PIV group in Periard’s study (Periard et
al., 2008). Moreover, a PICC enables an earlier hospital discharge with continuous treatment on
an outpatient basis with the assistance of a home care team. It was estimated that the cost saving
of a PICC with home care was $1,070 per day, compared to the cost of inpatient stay with
peripheral antibiotic therapy (Van Winkle et al., 2008). However, a PICC may not be ideal for
those patients with end-stage renal illness (Allen et al., 2000), serum creatinine level higher than
160 umol L -1 (Periard et al., 2008). Although the PICC may not be a suitable option for all
patients, it has become a viable alternative to subclavian lines, internal jugular lines or femoral
lines (Smith et al., 1998).
1.5.3 Peripherally inserted central catheter in pediatric patients
The increasing survival and better care of hospitalized children with complex medical conditions
means more patients requiring IV therapies (Pettit, 2009). PIVs are difficult to insert due to the
patient’s inability to cooperate, the small size of the veins, the pain from repeated IV punctures,
requirement of frequent changes, and the potential for infection from skin organisms (Stolfi et
al., 2009). Repeated IV punctures are accompanied with patients’ pain and labor cost (Stolfi et
al., 2009; Lago et al., 2008). PICCs can avoid multiple skin punctures, provide more reliable
venous access and increase patients’ satisfaction (Thiagarajan et al., 1997; Schwengel et al.,
2004). Thus a PICC is a preferable option for children requiring medium or long term access.
`
12
Moreover, as a PICC is inserted through a peripheral vein, risks of pneumothorax, hydrothorax,
and hemorrhage that may happen with central venous catheters inserted into major veins are
avoided (Levy et al., 2010). Therefore, since the 1970s, PICC lines have been increasingly used
as the preferable longer term venous access device option in paediatric patients, particularly for
patients in the neonatal intensive care unit (NICU) (Tully et al., 1981).
1.5.4 Complications related to peripherally inserted central catheters
Complications may lead to extended inpatient days, increased treatment costs and patient
discomfort (Webster et al, 2008; Ezingeard et al, 2009; Haddad et al, 2006; Chambers et al,
2002). PICCs have been popularized not only because of their ability to reduce hospitalized days
and costs but also because of their less frequent complication rate than PIV and other catheters
(Fairhall, 2008). Although PICCs have many advantages compared to PIVs and other catheters,
they are still associated with some problems of insertion or maintenance which may necessitate
premature removal of the catheters (Loughran et al., 1995). However, any procedure can carry
risk, which requires careful consideration of the tradeoff between benefits and risks. Careful
insertions and following the instructions for insertion and maintenance procedures of PICCs
reduce the risk of complications (RNAO, 2004). Nurses and Interventional Radiologists’ (IRs)
assessments of an individual patients potential risk factors include 1) site selection; 2) infection
prevention and control methods such as hand hygiene, skin antisepsis, antiseptic solution,
assessment of client risk factors, screening; 3) catheter material; 4) tip position; 5) dressing type,
frequency of dressing change, client tolerance; 6) type of securing device which was applied to
attach the catheter in order to prevent migration ; 7) flushing/locking (RNAO, 2004). To confirm
or diagnose a complication, over and above the clinical symptoms and signs, imaging
examinations such as doppler ultrasonography or venography are often required.
`
13
1.5.4.1 Definition of complications
(1). Catheter related infection (CRI)
Catheter related infection (CRI) has been defined several ways as: fever or elevated white blood
cell (WBC) or both; positive blood culture from the PICC, positive PICC tip culture, or positive
peripheral blood culture with no other source (Hampton et al., 1998). Infection is mainly
detected from lab examinations. A similar definition can be found in the study by Moureau et al.
“Catheter infection is identified through laboratory findings such as positive blood and catheter
cultures” (Moureau et al., 2002). Local CRI and catheter related blood-stream infection (CR-
BSI) belong to this category. If the organism grows from the proximal catheter segment,
accompanied with inflammation such as erythema, warmth, swelling or tenderness at the
insertion site, local CRI is presented (Pearson, 1996). If the organism grows from the catheter
segment and/or the blood of a patient is infected, CR-BSI can be diagnosed (Pearson, 1996). This
complication can occur in both adult and pediatric population.
(2). Phlebitis
Phlebitis is defined as a spectrum of clinical findings ranging from local inflammation at the
insertion site to a tender, erythematous, and palpable venous cord extending from the insertion
site and traveling along the arm” (Turcotte et al., 2006). It commonly presents as a local
inflammation of the vein accompanied with pain, erythema, edema, streak formation and/or
palpable cord (Hertzog &Waybill, 2008, Pearson, 1996). When phlebitis occurs, symptoms such
as redness, swelling, pain, skin warm to touch and a tender venous cord can be found in patients
which help make the diagnosis. Phlebitis is usually diagnosed by clinical signs and symptoms of
`
14
localized pain or sonographic examination instead of laboratory microbiological examination
(Hampton et al., 1998). This complication can occur in both adult and pediatric population.
(3). Occlusion
Occlusion usually results from either external or internal mechanical obstruction. It is defined as
“the inability to use the catheter for the assigned therapy due to obstruction” (Hampton et al.,
1998). Increased vigilance is required if external occlusion occurs as patients may face a higher
risk of pulmonary embolism from thrombotic obstruction in the vein. This is in contrast to
internal occlusion which usually results from clotted blood or drug precipitate within the lumen
of the line (Hampton et al., 1998). At the same time, excluding the extended hospitalized days,
occlusion may lead into higher costs, patient discomfort, or increase in the risk of catheter related
sepsis. Tissue plasminogen activator (TPA) is usually injected into the catheter as the first choice
to declot the internal occlusion. If the catheter remains occluded, replacement is then required
which can be attempted over a guide wire (Hoffer et al., 1999). This complication can occur in
both adult and pediatric population.
(4). Thrombosis
One of the typical causes of external occlusion is thrombosis which is defined as “the formation
of a blood clot attached to the exterior aspect of the catheter or to the venous wall in relation to
the catheter” (Turcotte et al., 2006). Upper extremity deep venous thrombosis (DVT) is a
specific type of thrombosis which presents as “a painful and swollen arm, or may be
asymptomatic” (Hertzog & Waybill, 2008). This complication can occur in both adult and
pediatric population.
`
15
(5). Malposition
The most suitable position of the central catheter tip is at the junction between the superior vena
cava and the right atrium (Connolly, et al, 2000). If the catheter tip is not in an appropriate
position, other complications such as thrombosis may be more likely to occur because of
malposition (Eastridge & Lefor, 1995). Malposition can happen at the time of insertion, or
develop later due to the change of intrathoracic pressure or catheter migration (Bowe-Geddes et
al., 2005). In Turcotte’s paper, catheter malposition was defined as accidental movement or
removal of the PICC (Turcotte et al., 2006). For example, a PICC could easily migrate outwards
during the process of dressing change or the patients’ accidental dislodgment the catheter with
activity. Confirmation of the catheter’s position on a chest X-ray is important, because
malposition can cause many complications such as venous thrombosis, phlebitis etc (Bowe-
Geddes et al., 2005). Careful catheter securement is important in order to prevent migration or
malposition. Many signs are used to discern malposition; for example, the external catheter’s
length will be increased; the neck or chest may possibly seem swollen; patient may feel pain or
discomfort during the infusion and/or there is no blood return (Paulson & Miller, 2008). When
malposition occurs, reposition is usually the first step to be considered. If reposition cannot solve
the problem, a new catheter is required (Hughes, 2006). This complication can occur in both
adult and pediatric population.
(6). Leakage/breakage
Leakage is defined as “infiltration of nonvesicant fluid into the tissue outside a vein” (Moureau
et al., 2002). The quality of the catheter or improper care can cause leakage/ breakage (Hughes,
2006). Leakage/breakage can be identified by clinical signs, fluoroscopy, or venography. For
example, if the dressing seems wet, or the line leaks during flushes, a leakage/breakage
`
16
complication should be considered. Damaged catheters can either be repaired using specific
repair tool kits, or exchanged for a new catheter (Hughes, 2006). This complication can occur in
both adult and pediatric population.
1.5.4.2 Classifications of complications
There are different classification systems of catheter related complications. Complications can be
classified into insertion complications and post-insertion complications. Pain, difficulty
advancing the catheter, damage to the catheter or veins, bleeding, nerve damage and embolism
can occur during insertion. In comparison, catheter occlusion, fracture or break, catheter related
bloodstream infection, thrombosis, phlebitis and edema may happen after insertion (Paulson &
Miller, 2008). Many steps are taken to prevent complications during insertion such as carefully
assessing the patient, optimizing comfort, avoiding excessive force. On the other hand,
procedures such as heparinization, flushing, aseptic technique should be taken in order to prevent
the post-insertion complications (Paulson & Miller, 2008).
Complications can also be classified as major or minor complications. Major complications are
defined as “an adverse event which a specific treatment, prolongation of hospitalization or re-
hospitalization is required”, for example, upper limb deep venous thrombosis (DVT) (Periard et
al., 2008). For those complications that do not require treatments or interventions, prolonged
hospitalization for more than 24 hours, or do not require re-hospitalization are considered minor
complications (Periard et al., 2008).
`
17
Additionally, complications can be divided into infectious or mechanical complications. For
infectious complications, according to the American Centers for Disease Control and
Prevention’s (CDC), there are three categories: exit site infection, catheter colonization, catheter-
related bloodstream infection (CR-BSI) (O'Grady et al, 2002). Mechanical complications include
catheter occlusion, catheter dislodgement or migration, and at the time of insertion they include
hemorrhage, vascular spasm, or arterial puncture. (Hampton et al., 1998)
Another classification approach is based on catheter dysfunction. Catheter dysfunction is defined
as inability to use the catheter normally. It can be divided into thrombotic and nonthrombotic
dysfunction. Thrombotic dysfunction is defined as “thrombus accumulation within a catheter
resulting in partial or complete blockage”. While nonthrombotic dysfunction is defined as
“inability to use the catheter as a result of causes other than thrombosis”. (Moureau et al., 2002).
1.5.4.3 Complication rates
Research reports that catheter complication rates are influenced by a variety of factors such as
the insertion mode, diameter of the PICC, lumen number, and patient immune status (Ng et al.,
1997, Grove et al., 2000). Complication rates are reported using the absolute incidence, or
incidence per 1,000 catheter days.
(1). Complication rate in the general population
It is estimated that 40% of PICCs have to be removed before therapy completion due to
complications (Turcotte et al., 2006). More specifically, 6% of PICCs are removed prematurely
due to phlebitis (Turcotte et al., 2006). There is a wide range of complication rates reported in
`
18
the literature. For example, the quoted incidence of thrombosis is as low as 2% or as high as 25%
(Ng et al., 1997, Periard et al., 2008). Similarly, the incidence of sepsis varies within a range of
0.4%-25.7%, or between 0.1-8 episodes of sepsis per 1,000 catheter days. The infection rate is
reported as approximately 7%, or 0.4–3.4/1,000 catheter days in these studies (Ng et al., 1997;
Cardella et al., 1996; Periard et al., 2008; Smith et al., 1998). Yamamoto’s study reveals that the
complication rate for leakage is 1.87% (Yamamoto et al, 2002). The incidence of DVT
associated with PICCs is reported to be much less than that associated with other VADs (Dubois
et al., 2007). However, another systematic review (Maki, et al., 2006) conclude that a range of
50,000 to 500,000 catheter related blood stream infections (CR-BSI) occur each year in USA, of
which PICCs have a rate of 0.5 per 1,000 catheter days while CVCs have a rate 2.7 per 1,000
catheter days.
(2). Complication rate in pediatrics
Compared to the general population, complication rates related to PICCs in pediatrics are
relatively low; only 30 per cent of PICCs are removed for complication (Racadio, et al., 2001;
Thiagarajan, et al., 1997). In Dubois’s study, infection and thrombosis are the main
complications and the complication rates are respectively 6% and 0.3%. In Racadio’s study, the
total complication rate was only 3.8%. 1.7% of the 1096 PICC lines became occluded; 1.5%
occurred phlebitis; 0.2% of the patients got infections (Racadio, et al., 2001). In Itzhak’s study,
twenty-six catheters of the 279 PICCs (9.3%) were dislodged accidentally, 13.6% of PICCs were
removed because of infection and 4.6 per cent were removed due to phlebitis. PICC associated
BSI rate was 4.3% (1.4 per 1,000 days). In another study by Thiagarajan, 7% of PICCs were
occluded and 8% are dislodged by accident. The incidence of catheter associated sepsis was 2%
and suspected infection accounts for 8%. However, in Crowley’s study, the incidence of
`
19
infection was 0.93 per 1,000 catheter dwell days (Crowley, et al., 1997). In another study, thirty-
eight PICCs were occluded and seven were accidentally dislodged among 269 PICCs.
Malposition was about 1.5% in Frey’s study as well (Frey, 1995).
1.5.4.4 The management of complications in relation to PICCs in pediatrics
Careful management can help reduce complication rates. Many approaches have been advocated
for catheter maintenance in order to reduce complications and improve patients’ outcomes. For
example, the “StatLock” securement device is used instead of tape or sutures to hold the line.
“Statlock” securement, made up of an adhesive-backed anchor pad with hinged clamps, is
reported safer and less time consuming than suture securement (Yamamoto et al., 2002; Held-
Warmkessel, 2001). Disinfecting the surface of the hub before access is strongly recommended
in order to prevent complications (Maki & Mermel, 1998). It has been reported that the longer
the PICC dwell time, the more complications a patient may encounter (Raad et al., 1993). It is
suggested to remove the catheter as soon as possible once it is no longer required. Keeping the
PICC dry helps reduce the risk of infection (Sanders, 2006, Periard et al., 2008). The use of
anticoagulation for prophylaxis in patients with a PICC tends to lower the rate of catheter-related
thrombosis (Paauw & Borders et al., 2008). Advantages of silicone catheters include greater
flexibility, decreased thrombogenicity, and decreased incidence of sepsis as compared with
polyethylene catheters. Verifying the central tip position of a PICC is essential to prevent
complications because improper position leads to many complications (Bowe-Geddes et al.,
2005). Recognized complications associated with incorrect tip position include: central venous
perforation; thrombosis and CVAD dysfunction (RNAO, 2004). In addition, securement devices
have been found to reduce the number of hospital days and complications (Sheppard et al. 1999
McMahon 2002) but require changing at least every seven days (CDC, 2002). Even if a
`
20
complication occurs, timely and well considered treatment can minimize the chance of PICC loss
and reduce the number of re-hospitalization days. Common treatment strategies for
complications are listed in table 2 (Paulson & Miller, 2008). Phlebitis’s symptoms can be
difficult to distinguish from infection. An incorrect diagnosis will lead to unnecessary antibiotic
treatment or unnecessary premature removal of the catheter.
Table 2 Common treatments for catheter-related complications (Paulson& Miller, 2008)
Complication Treatment methods
Pain Pacifier, containment, site numbing, sedation
Nerve damage Individual care
Occlusion tPA
Catheter fracture or break Repair kit
Infection cleanse the site with alcohol, oral or intravenous antibiotics
Malposition Reposition or remove
Deep Venous Thrombosis Chest X-ray, Doppler ultrasound, anticoagulant therapy, remove
Phlebitis Warm packs or remove
Edema Exercise or remove
If the catheter leaks and/or breaks, the dressing requires to be removed. The line requires to be
flushed carefully and at the same time, a repair kit used. If the repair fails, the catheter should be
replaced.
Redness at the exit site of a PICC can be indicative of a site infection. If there is no swelling and
pain at the site, the infection is considered in early stage. Cleansing the exit site with alcohol and
keeping the dressing dry are the proper way to manage the site. If patients have already
`
21
experienced pain and swelling, oral antibiotics are recommended for those patients. Intravenous
antibiotic therapy can be applied if the infection persists.
Measuring the external part of a PICC and comparing it with the original measurement at the
time of placement is one way to monitor for migration/malposition. If movement is detected, a
chest X-ray is required to review the tip position in order to choose appropriate treatment. If
malposition occurs, the first priority is to adjust the line into the proper position. If not possible,
the line may need to be removed or exchanged.
Patients with a PICC may develop a swollen arm and/or hand with bleeding, cyanosis, pain in the
arm or shoulder. If patients have these symptoms, 0.9% saline is used to flush the PICC, andif
the patients feel pain during the flushing, one possible reason is an internal catheter fracture. This
can be confirmed with linogram. If the patient does not feel any pain during the flush, deep
venous thrombosis (DVT) is considered. Chest X-ray and Doppler Ultrasound are required to
review the catheter and the vein. If thrombosis is confirmed, anticoagulant therapy(eg. Heparin)
is commenced to treat the thrombosis. TPA is considered as a very effective approach to dissolve
a thrombus; however, TPA is expensive and very potent with significant associated risks, and
used rarely.
If the patient reports pain in the shoulder, neck or chest, possible causes could be that, the tip pf
the catheter could have migrated into an improper location, requiring a chest X-ray to determine
its position. It is also possible that leakage/breakage may have occured. At this time, flushing the
catheter with saline and a linogram is required to determine the type of complication. As each
`
22
PICC is a valuable resource for the patient, removal is the last approach or option to be
considered.
1.5.5 Costs related to peripherally inserted central catheters
Previous studies have only focused on economic evaluation of PICCs in the general population.
In Schwengel’s study, the PICC insertion cost was measured by collecting the data of three
aspects of labor (anesthesiologist and phlebotomist time), equipment (PICC trays) and operating
room time costs. This study also showed that the total estimated insertion cost varied from US
$173.58 to US $440.70, which was more expensive than a PIV as the average cost for a PIV was
only US $108.49 (Schwengel et al., 2004). In Periard’s study, it was estimated that the insertion
cost related to a PICC was about US $690 per patient. More specifically, in that study, the author
also measured the material cost related to PICC insertion was US $210; angiography suite US
$265. The PICC maintenance cost was US $27 per patient per catheter for each insertion while
PIV maintenance cost was only US $18 per patient in 2006. From Periard’s study, we can also
know that as for the procedure time, nurses spent 4.1 hours per patient on average to manage the
PICC while PIV required 5.5 hours during the entire catheter dwell time. Therefore, for nurses’
salary, the estimate cost was US $165 in the PICC group; however, PIV was higher than PICC,
about US $219 (Periard, 2008). Smith’s study also calculated the PICC insertion cost which was
US $500 (Smith et al., 1998). Harattas’ study assessed a PICC insertion cost to be US $401,
which was close to Smith’s study (Horattas et al., 2001). Similarly, a PICC insertion costs CA
$270 in Murphy’s study. And the total cost of insertion, maintenance and managing
complications was CA $344 (Murphy et al., 2008). Cowl et al. used cost per day instead of total
cost to measure the PICC cost. In Cowl’s study, the PICC insertion and maintenance cost was
US $22.32+/-2.74 per day for hospitalized patients who required TPN therapy (Cowl et al.,
2000).
`
23
Complications are associated with additional costs over and above the total costs associated with
PICCs as they require specific examinations and treatments. For example, the treatment costs for
a blood-stream infection (BSI) ranged from US $10,000 up to US $35,000 (Donowitz et al.,
2001). Similarly, another study pointed out that BSI would extend hospital length of stay by an
additional 10 to 20 hospital days, with extra cost of US $4,000 to US $56,000 per episode (Maki,
et al., 2006).
Very few studies have focused on the break down costs associated with PICCs. Most of them
only calculated a portion of the health system costs such as the insertion costs or complication
treatment costs (Table 3). They did not consider other cost components such as patient/ family
productivity loss, travel costs, home care costs as we discussed above from the societal
perspective. Moreover, they did not stress the point whether PICCs yield cost-savings compared
to other catheters (e.g. PIV). Therefore, further research on costs related to PICCs is required. In
order to address the costs clearly, this study describes the average and/or median cost
components as well as the total cost in details. At the same time, factors that affect the total cost
of a PICC are also detected. Theoretical cost comparison between PICC and PIV was also
applied in this study.
`
24
Table 3 Cost information listed in the papers
1st author Published Year Cost items included Cost per total PICC
dwell time
Schwengel 2004 Total cost (labor, equipment and
operating room cost)
US $440.70
Periard 2008 Insertion cost
Material cost
Maintenance cost
US $690.00
US $210.00
US $27.00
Smith 1998 Insertion cost US $500.00
Murphy 2008 Insertion cost CA $270.00
Donowitz 2001 Complication treatment cost US$10,000- US
$35,000
Cowl 2000 Insertion and maintenance cost
per hospitalized day
US $22.32+/- 2.74
`
25
Chapter 2 Methods
This chapter describes the methods used for this study. It is divided into six sections. The first
section introduces the systematic review regarding PICC costs in pediatrics. The second section
provides an overview of the study design, including the study population, variables used in this
study. The third part presents details of measurements of different cost components. The fourth
section describes the statistical analysis used in this study followed by the fifth section of
sensitivity analysis adopted. Ethics are presented at the end of this chapter.
2.1 A systematic review of peripherally inserted central catheters
A systematic literature search was conducted using PUBMED (1946-2010), CINAHL (1981-
2010), Cochrane library (1995-2010), EMBASE (1948-2010) databases to capture all relevant
studies. At the same time, Health Technology Assessment (HTA) reports from Canadian Agency
for Drugs and Technologies in Health (CADTH), Centre for Reviews and Dissemination (CRD)
including DARE and NHS EED, Paediatric Economic Database Evaluation (PEDE) were also
searched for pertinent studies. The search period was limited to those databases’ coverage.
Medical Subject Headings (MeSH) terms were used in this study to retrieve relevant articles.
Terms used in this study were "Catheterization, Central Venous", "Costs and Cost Analysis",
"Infant", "Child, Preschool", "Child", "Adolescence", and "Pediatrics". This study aimed to
search articles related to "Catheterization, Central Venous or Peripheral" and "Costs and Cost
Analysis" in infant, child preschool, child, adolescent or pediatrics by restricting our search to
articles where "Catheterization, Central Venous or Peripheral" was a major subject. Therefore,
the search language was (MM “Catheterization, Central Venous” OR MM “Catheterization,
`
26
Peripheral”) AND (MH “Cost and Cost Analysis”) AND (MH “Infant” OR MH “Child,
Preschool” OR MH “Child” OR MH “Adolescence” OR MH “Pediatrics”). The following study
designs were included: practice guidelines, systematic reviews, meta-analyses, reviews,
randomized controlled trials, and controlled clinical trials. Types of articles such as letters,
editorials/commentaries, and lectures were excluded for this systematic review. All papers
identified as potentially relevant were initially assessed for inclusion by reviewing the titles and
abstracts by one master student. In order to be included, the articles had to meet the following
criteria: 1) The study population had to be limited to infants, child preschool, children or
adolescents; those who studied the whole population or the adults instead of pediatrics were
excluded; 2) The study had to focus on peripherally inserted central catheter; those studies
focusing on central venous catheters, midline catheters, or peripheral venous lines were excluded
from this study; 3) Cost must be reported as one of primary or secondary outcome measures;
those studies that did not provide actual cost numbers were excluded; 4) The study had to be
written in English. Inclusion and exclusion criteria were listed in the following Table 4 (Table
Based on the criteria discussed above, evidence was initially selected and reviewed based on
titles and abstracts. Studies that could not be excluded with certainty were retrieved and
reviewed in their entirety based on the inclusion/exclusion criteria described above (Table 5).
There were 134 papers reviewed in total which related to PICC cost. However, 69 papers of them
were excluded as duplicated citations; 3 papers were excluded because they were not written in
English; and 59 papers were not selected for this systematic review because the papers were not
relevant to PICCs, not pediatric population, or cost was not one of their primary or secondary
outcomes. Finally, only 3 papers were retained for analysis in terms of PICC cost in pediatrics.
(Schwengel et al., 2004; Moore et al., 2006; Van Winkle et al., 2008). A data extraction sheet
`
27
was also developed in order to collect the relevant data using Microsoft Excel®. This included
information about the 1st author’s name, year published, study design as well as key outcome
variables and their limitations. Results will be presented in Chapter 3.
Table 4 Inclusion and exclusion of systematic review
Inclusion Exclusion
Focus on pediatric population Studies based on adult population or the
whole population
Focus on PICCs Focus on other catheters such as central
venous catheters, peripheral intravenous
lines, midlines etc.
Cost must be one of the primary or
secondary outcomes
No actual cost numbers provided
Written in English Presented in other languages such as
French, Chinese, etc.
`
27
Table 5 Chart of included and excluded papers
Database Coverage MeSH terms No. of
papers
retrieved
Exclusion
reasons
No. of papers
retained for
analysis
PUBMED 1946-2010 (MM “Catheterization, Central Venous” OR
MM “Catheterization, Peripheral”) AND (MH
“Cost and Cost Analysis”) AND (MH
“Infant” OR “Child, Preschool” OR “Child”
OR “Adolescence” OR “Pediatrics”).
83 57 repeated
20 not relevant
3 not English
3
CINAHL(EBSCOhost) 1981-2010 Same as above 3 3 not relevant 0
The Cochrane Library 1995-2010 Same as above 9 3 repeated
6 not relevant
0
EMBASE(Ovid
MEDLINE (R) )
1948-2010 Same as above 14 1 repeated
13 not relevant
0
CADTH(HTA) 1990-2010 Same as above 0 0 0
DARE 1994-2010 Same as above 2 2 not relevant 0
NHS EED 1968-2010 Same as above 21 6 repeated
15 not relevant
0
HTA 1989-2010 Same as above 0 0 0
PEDE 1980-2010 Same as above 2 2 repeated 0
`
28
2.2 Overview of the cost analysis
2.2.1 Study population and the inclusion and exclusion criteria
A retrospective cohort of pediatric patients with PICCs inserted in IGT at SickKids between
January 1, 2008 and Dec. 31, 2008, was reviewed to determine the costs associated with having a
PICC. All PICC cases were followed until the PICC was removed, usually within two years,
from 2008 to 2009. The inclusion criteria were patients who underwent PICC insertion
procedures during the study period in the department of IGT in SickKids. Initially, 1,181 PICC
related procedures in 2008 were identified.
The exclusion criteria were as follows;
1) Patients who underwent other procedures such as catheter replacement, removal, or reposition
instead of a primary PICC insertion were excluded as they did not represent a new PICC
placement. Therefore, 564 patients were excluded due to this criterion.
2) Patients who did not reside in Ontario were excluded as these patients may have different
insurance plans from Ontario. Thus, 25 patients were excluded because of this.
3) Patients who transferred from or to other hospitals could not be followed due the unavailable
data; therefore, 12 patients were ruled out.
4) Some patients could not be followed due to missing data in the patients’ medical records. For
this reason, 19 patients were eliminated.
5) Patients who died in the PICU (21 patients) were excluded because relevant cost information
requires special application for data release.
6) Day surgery patients who had their PICC insertions with no subsequent hospital admission
were excluded because of limited data availability. Therefore, 16 patients were excluded because
of this reason.
`
29
Of 1,181 PICC procedures identified in 2008, 564 were excluded because they were not having a
primary insertion procedure. Twenty-five patients who did not reside in Ontario and 19 patients
who could not be followed due to missing information, 12 patients who transferred from or to
other hospitals, 21 patients who died in the PICU and 16 in day surgery patients were excluded
due to data limitation. Finally, 524 patients with 573 PICCs inserted were retained for analyses
(Figure 2).
Figure 2 Flow diagram of Inclusion and Exclusion Criteria:
Patients retained for analysis n= 524
Excluded n= 93
l Patients living outside of Ontario n= 25 l Patients died in PICU n=21 l Day surgery patients n=16 l Transferred from or to other hospitals n=12 l Patients lost to follow up n= 19
Patients included n= 617
All patients with PICC related procedures in 2008 n= 1181
Excluded because not primary insertion procedures n= 564
`
30
2.2.2 Data Collection
A search was run using the IGT-ESH database (a dedicated Interventional Radiology database in
the Image Guided Therapy department, IGT) to identify all patients who underwent a successful
PICC insertion in IGT during the study period (Appendix 3). The ESH database generated an
excel spreadsheet with information on patients age, sex, weight, date of procedure, time of
procedure, type of procedure, type of line used, operator information, sedation information,
procedural complications, procedural costs (broken down by equipment costs, labor costs and
material costs). Eligible cases were then cross referenced with the Vascular Access Database
which is an excel spreadsheet database managed and updated daily by the Vascular Access
Database administrator with information on all patients requiring vascular access services. For
the purpose of this study the database administrator pulled an excel spreadsheet from the
Vascular Access database during the study time-period which contained columns with the
patients age, date of birth, date of death (if applicable), sex, hospital in-patient unit, weight,
primary diagnosis, reason for insertion , line number (if this was the patients first line or had
previous lines), removal date, dwell days, type of catheter, manufacturer of the catheter, access
route, date of complication and type of complication (if applicable) and reason for removal.
Information of patients’ address, ward unit, lab examinations, emergency room visits, and
sedation methods were extracted from the database of the Electronic Patient Chart System (EPC)
using patients’ medical record numbers (MRN). One master student extracted these data and
entered them into excel spreadsheet. EPC was also used to retrieve those missing data if there
were any missing data in the excel sheet generated from ESH and Vascular Access Database, if
possible. The Picture Archiving and Communication Systems (PACS) of the Hospital for Sick
Children (SickKids) were utilized too to double check to make sure the reliability of the Vascular
`
31
Access Database. All of these four databases, ESH, PACS, EPC, Vascular Access Database,
were linked or cross-referenced by patient medical record number as personal identifiers. And all
the Excel spreadsheets were combined and merged into one excel sheet and imported to a SPSS
software (Statistical Package for the Social Sciences, version 17.0, SPSS Inc., Chicago, IL, USA)
for analysis. After import, patient personal information such as MRN and address was then
deleted for encryption purpose. This final dataset for this research was stored in the computer of
the student room. A password was set for this dataset file. Only the person who can access to the
computer with access account and know the password as well can open this dataset (Figure 4).
`
32
Figure 3 Flow chart of data collection process, encryption, analysis, and storage
2.2.3 Perspective of this study
The perspective taken in this cost analysis was that of the society. According to U.S. Public
Health Service Panel on Cost-Effectiveness in Health and Medicine (PCEHM)’s definition,
adopting a societal perspective means that all costs and types of resources of value to the entire
society should be considered no matter who paid the cost or who received the benefit. As said,
`
33
this perspective can offer the most comprehensive estimation of costs. In this study, all cost
components associated with the catheters were calculated.
2.2.4 Potential factors associated with peripheral inserted central catheters’ cost
Potential independent variables of costs in this study were selected based on Andersen and
Newman’s behavior model on health services utilization (Andersen & Newman, 1973) and/or
because those independent variables were identified from previous papers (Schwengel et al.,
2004; Periard, 2008; Smith et al., 1998; Horattas et al., 2001; Murphy et al., 2008; Cowl et al.,
2000). This model includes several types of variables classified as “predisposing”, “enabling”,
and “need for care”. “Predisposing factors” are defined as factors that wield effects by increasing
or decreasing a person’s motivation to undertake a behavior (Green & Kreuter, 2005). In
Andersen’s model, it means broadly to all factors that may predispose a person to need and/or
use a service. “Predisposing” includes demographic factors (e.g. age, sex), social structural
factors (e.g. education, occupation) and factors associated with health beliefs (e.g. attitudes,
values). “Need” factors means individualized perceived health status and function capacity such
as problems with daily activities, comorbid conditions, perceived health and mental status that
may influence a person’s utilization of a service. “Enabling” characteristics represents those
variables that may boost or impede individual health care services utilization such as financial
status, informal social supports and insurance condition. There are two types of enabling factors:
community enabling factors and personal/family enabling factors (Andersen & Newman, 1973;
Kempen & Suurmeijer, 1991).
In this study, variables such as age, sex, weight were selected as the predisposing factors for
analysis. In the “need” group, variables such as patients’ primary diagnosis, ward, catheter
`
34
insertion reason, access vein, type of anesthesia, catheter information including the size, type and
manufacturer, catheter dwell days, inpatient days, removal reason, complication during the
catheter dwell period were all included. Due to the data limitation, no financial status or
insurance information can be captured, there is no factor belonging to the “Enabling” category
(Table 6).
Table 6 Variables used in this study and their classification
classification Variables used in this study
predisposing Age, sex, weight,
need Patient primary diagnosis, ward, catheter insertion reason, access vein, type
of anesthesia, catheter size, catheter dwell days, inpatient days, line type,
line manufacturer, cuffed/uncuffed, lumen number, removal reason,
complication, home health care condition, insertion date, removal date,
admission date, discharge date
2.3 Measurement of cost components
To do an optimal economic evaluation, it is required to consider the resources consumed from
the four cost component perspectives: cost to the health sector, cost to other sectors, cost to the
patient/family, and cost attributed to productivity losses (Drummond et al., 2005). The cost to the
health sector is the direct medical cost which includes the items such as hospitalization,
physician visits, drugs, and so on. The cost to other sectors depends on their characteristics. Cost
to patient/family members is mainly deemed as out-of-pocket expenses such as traveling to the
hospital, copayments and other expenditures in the home. Opportunity costs, that is, the value of
the alternative choice of using those resources is the most suitable method of estimation (Liljas,
1998). Due to the illness, patients and their family members may be affected on their time at
`
35
work or productivity efficiency. The parental opportunity cost is measured by their time loss due
to take care of their children (Drummond et al., 2005).
2.3.1 Costs components associated with peripheral inserted central catheters
Previous studies have described the PICC as a preferable approach to vascular access but do not
provide a breakdown of the various cost components associated with having a PICC line in
children (Schwengel et al., 2004; Cheong et al., 2004). To our knowledge, there has been no
study focusing on detailed cost analysis in terms of different cost components related to PICCs.
In this study, both direct and indirect costs were included. “Direct cost” usually means the
resources consumed by the program when compared to other options, while “indirect cost”
usually denotes the time consumed by the patients or their family members (Drummond et al.,
2005). For the purpose of this study, direct costs include: the insertion cost, outpatient
management cost, complication cost, consultation cost, removal cost and travel cost. For the
indirect cost, parents’ productivity loss was considered by the measurement of time. Details were
listed as follows (Table 7). Regarding the insertion costs, we considered three aspects: material
costs, labor costs and equipment costs. In the event of a post insertion line complication, the line
might need to be repaired, exchanged, or repositioned, laboratory tests and other imaging
examinations might be required, patients may attend a clinic or visit an Emergency Room (ER);
sometimes they were given medications for treatments. All of these components were
incorporated into the cost component of a complication. Removal costs were recorded as well. A
PICC was removed by either a vascular access nurse (if uncuffed, or cuffed in situ for less than 4
weeks) or by an interventional radiologists (if cuffed and in situ for longer than 4 weeks) with
different associated costs. In addition, travel costs incurred with the patients and their families
was one of the cost components included. As patients who came to the hospital were usually
`
36
suffering from a severe disease, they seldom walked or used the public transportation to the
hospital. It was assumed that all patients traveled to SickKids by car or by plane. The cost of
home care nursing was calculated as the outpatient management cost. With respect to indirect
cost, parents’ productivity loss was considered during hospital days.
Dislodgement ER visit Exchange N/A Chest x ray N/A
Block Varies with different patients Flush or unblocking tPA x-ray
linogram
Dressing pack, sterile gloves, syringes, green
needle, Saline, etc
Malposition Varies with different patients Reposition in IGT N/A Chest x ray Vary with different patients
Note:1) N/A, not applicable;
2) Low molecular weight heparin (LMWH);
3) Emergency Room(ER);
4) Tissue plasminogen activator (tPA)
`
44
1. Emergency Room/Clinic visit costs
After discharge, costs of subsequent ER and clinic visits were usually not included as these were
primarily for the disease treatment rather than related to the PICC/PIV line related issues.
However, if the visit to the clinics or ER was related to a catheter complication, relevant costs
were considered. For the pediatric physician consultation cost in clinic, the fee for a general
assessment is $165.5 per episode according to the Ontario Health Insurance Plan (OHIP)
Schedule of Benefits (SoB) billing code A265. Patients do not need the ER service when they
stay in the hospital. But after discharge, patients may require ER services under emergent
situation; sometimes ambulance services may be also required. ER associated costs were drawn
from Guerriere et al.’s study, which was CAD $181 per visit in 2008 which fully allocated the
direct and an appropriate share of overhead costs associated with the treatments (Guerriere, et
al., 2010; Coyte, et al., 2001). An ambulance cost is assigned a $240.00 per visit by the
calculation of Ministry of Health and Long Term Care (MOHLTC). So each ER visit costs
$421.00 per visit (ER cost plus ambulance cost) in total if the ambulance is required.
2. Lab test costs
For laboratory test costs, the assumption is that only those patients with complications will
require additional laboratory tests. When referring to in-patients, the lab costs were included in
the ward cost. Three laboratory tests were assigned from OHIP SoB: bacteriology test,
biochemistry test and haematology test. The main lab tests were blood samples taken for
complete blood count (CBC), and blood cultures. For example, if the patient has a thrombosis,
CBC test is required; however, in the incidence of infection, they may need blood culture tests.
Labor, materials, supervision (LMS) units are the basis for OHIP billing by laboratories. They
are used to calculate the laboratory service prices. From OHIP SoB, we found that each CBC
`
45
was 16 LMS units (OHIP SoB code: L393); blood culture was 30 LMS units (OHIP SoB code:
L624). According to the MOHLTC’s rule, one LMS unit was 51.7 cents. Therefore multiplying
the LMS units with their unit values of 51.7 cents, we could assure the estimates of cost: $ 8.27
for per CBC test; $15.51 for per blood culture test.
3. Imaging examination costs
Interventional imaging costs were captured from ESH database such as linogram and venogram.
The cost of a linogram is $161.82 and venogram is $106.61. Chest X ray and ultrasound costs
were captured using unit costs assigned from OHIP SoB. The unit cost included three
components of cost: technology cost, professional fee and facility fee. The total cost drawn from
OHIP SoB for each Chest X ray (SoB code: X090) was $38.53 which included technology costs
$15.30, professional fee $6.75, and facility fee $16.48. Similarly, the total cost for each
ultrasound (OHIP SoB code: J207 or J507) was $62.11 with a technology cost of $22.60,
professional fee $16.35 and $22.16 facility fee.
4. Procedure costs
IGT procedure costs were captured from ESH system as well. Here is a table for some procedure
costs. One cost from a patient is shown as an example of the cost of each procedure. Different
patients faced different costs based on different situations; but usually those costs should be are
similar to the costs listed here. Some examples of procedure name and costs are listed in the
following table 10.
`
46
Table 10 Examples of procedure costs included
Procedure name Procedure cost
CVL PICC reposition $217.32
CVL PICC removal Nurse $38.56
Interventional Radiologist $143.12
CVL PICC exchange $1741.71
CVL Resuturing $22.76
CVL PICC repair by nurse $113.61
CVL PICC Flush $132.63
CVL PICC unblocking $223.47
Note: CVL - central venous line
5. Other relevant costs
Other relevant costs such as material costs used for treatment of complication while in IGT were
extracted from ESH system. As different patients use different quantity of materials or devices,
the costs will vary, examples of costs of materials or devices used are listed in the following
table 11.
Table 11 Examples of materials and tool’s costs
Examples of relevant stuff for complication treatment Costs
omnipaque 300 20 ml $38.07
syringe ll 10 cc $0.80
sodium chloride 0.9% inj 10ml $3.50
syringe tb slip tip 5cc $0.36
Syringe LL 3cc $0.59
Tegaderm 6X8.5cm $0.18
Catheter repair kit $109
Needle hypo 27G-1-1/4 $0.03
`
47
2-0 Prolene $2.25
Needle blunt fill 18GA *1.5 $0.08
Adhesive mastisol liquid $2.69
Blade sterile no.11 $0.36
Boundary Table cover $10.39
Soln Nacl 0.9% 250ml JB1322LP $1.00
Sharps Container $4.95
2.3.2.5 Removal Cost
PICCs can be removed either at the hospital or at home but must be removed by a trained
professional such as IRs, vascular access nurse, or community care nurse. Many factors can
influence the reason for line removal such as end of therapy, a complication, or a change in the
type of therapy required. If the PICC line is uncuffed or in situ for less than 4 weeks and the line
is not adherent to the skin, nurses can remove the line directly. Nurse removal costs, obtained
from ESH system, is $38.65 on average, including the material cost $11.80 and $16.76 nursing
time. When a cuffed line is removed after more than 4 weeks, by an IR, the cost for removal is
$165.11 on average, including the material cost $11.80, IR suite cost $75.00, and IR labor cost
$78.31.
2.3.2.6 Travel cost
Travel cost, defined as an out-of-pocket cost assumed by parents/family members, was included
from the societal perspective. It was assumed that most families arrived at the hospital by car.
However, because SickKids is a tertiary care centre, patients traveled from all over Ontario, at
times requiring air transportation methods. (Table 12).
`
48
Table 12 Travel approaches and their cost estimation
Distance to
the hospital
Travel way Live in
ward
Go home
daily
Cost estimation
<= 430km Auto (only on
admission and
discharge day)
Yes No Use reimbursement rate in SickKids for
driving to the hospital for research
purpose and parking fees.
>430km Airline Yes No Tango price for one adult plus one child
of the flight tickets from Air Canada,
West Jet and Porter Airlines
We assumed that if the driving distance was closer or equal than 430 kilometer (km), parents
would drive to the hospital first based on expert opinions. It was assumed that those families who
resided further than 430 km would travel by air to Toronto. It was also assumed that at least one
parent would stay on the ward with their child at all times. Assumption was that the parents
would drive to the hospital on the day of admission and drive back home after discharge. As
patients’ home addresses can be obtained from the database, the distance from their home to the
hospital can be measured by Google map. The kilometer reimbursement rate at SickKids for
driving to the hospital for research purpose was used to estimate the driving cost in this study,
which amounted to $0.35 cent per kilometer plus $11 per day for parking. Thus, the travel cost
for each visit was the reimbursement rate per km multiplied by the distance, plus parking fees.
The total travel cost was the cost per parking plus one round trip by car.
It was assumed that if the patient’s home was further than 430km from the hospital the patient
would travel by air to the hospital and stay with their child on the ward. . Prices were traced for
one year from June 15, 2010 until June 3, 2011 on Aircanada, West Jet and Porter Airlines.
Tango price (Lowest price in a day) of the flight ticket for one adult plus one child was searched
`
49
on these three airlines’ websites. In all cases, the nearest airport to the patient’s residence was
selected to determine air travel costs.
2.3.2.7 Home care cost
After discharge, patient’s PICCs are managed by the home care nurses. Home care nurses visit
daily to change dressings and perform general PICC maintenance such as flushing the line daily
and giving the medications. The home care days were measured from date of discharge until the
day of lines’ removal. To determine the cost of home-based professional services, the result of
the study of Guerriere et al. was used. The cost per home care visit for nursing and personal
supports (e.g. occupational therapy, physical therapy) was CA$88 per visit in 2008, including
23% of overhead cost (Guerriere et al., 2010). The payment rates for professional home care cost
in Guerriere’s study were obtained from home care agencies and an inflation factor was taken
into consideration as well. As this study was also aimed at Ontario home-based profession
services’ cost, we applied this cost for the home care costs too. Consequently, the home care
nursing cost was estimated at $88 per visit.
2.3.2.8 Indirect cost estimates
Productivity loss was calculated in this study to estimate the indirect cost. Parents/family
members’ absence from work or usual activities was an important index to measure this in direct
cost. Human capital cost was the most common method to calculate the cost of time lost. Most
people choose to use this approach as it is easier to apply and less expensive, though deficiencies
may exist (Liljas, 1998; Hodgson, 1983) Based on Howard’s research, during their
hospitalization, it was assumed at least one of their parents was present with the patient and
missed a full day of work (Hancock-Howard et al, 2010). So in this study we would have the
`
50
same assumption that every admitted child would have one parent to take care of them; the other
parent still retained their full time job. To value the productivity loss from labor market, average
hourly aged-based earnings calculated by Statistics Canada were assigned for this study.
According to the Statistics Canada’s calculation, the average hourly wage for people aged 25-50
years old was $23.87 during April 2008 to April 2009 (Statistics Canada 2007). The parent
income also includes the salary plus an additional 23% fringe benefit. Then it will be further
adjusted by 52/46 to account for vacations and holidays (Guerriere et al., 2010). The total
indirect cost was assigned as $23.87/hour multiplied by the inpatient days.
2.4 Cost components associated with PIVs
If a PICC had not been place, this cohort of patients would have to have their treatments given
through peripheral IV. In this scenario, the patient demographics, diagnosis, and duration of
therapy would be the same. However, they would not be able to have some of their treatment at
home. As PIVs and PICCs can in many instances be substituted for each other as an infusion
device, catheter dwell days were assumed to be the same in order to finish the same treatment
period. Therefore, in the PIV group, it was assumed that the total catheter dwell days would be
the same as the PICC group. Based on this assumption, the main difference for these two groups
would be inpatient days. PICC patients if stable from their disease can be discharged earlier with
the PICC in situ and receive their health care at home, whereas PIV patients cannot. They have
to stay in the hospitals for the entire infusion therapy because of the frequent exchanges of the
lines. Therefore, in the PIV group, inpatient days were assumed to be the date of admission to the
date of removal of the line and end the therapy. The inpatient days for the PIV group are
expected to be longer than the PICC group.
`
51
If the PIV dwell time is longer than 72 hours, the risk of complications such as thrombosis,
phlebitis is increased. Scheduled line replacement has been proposed to prevent such
complications (O’Grady et al, 2002). Therefore, the PIV was assumed to be replaced every three
days during the treatment period. For the PIV insertion, on average it requires two nurses 30
minutes to insert the IV (in house data from the IV team) based on the nurses’ expert opinion. It
was estimated at $50 in total for each PIV insertion, including the nursing time cost, the cost of
mexiline and other supplies used.
When compared to a PICC, management of the PIV is simpler, and does involve repairs or
repositioning the lines. Usually the PIV would simply be removed or replaced with a new line
inserted instead due to complications, assuming it is possible to place a new line successfully. In
contrast to PICCs, assessment time for a PIV is quite short or not necessary at all. Thus physician
and nurse assessment time costs were not considered in this study. If a complication occurs with
a PIV, removal of the PIV is usually the first resort. As removal of a PIV takes only a matter of
seconds, we assumed that the removal cost and related complication treatment cost were not
factored into the calculation. If intravenous therapy is still required after a PIV complication,
reinsertion a new PIV at a new site is sometimes all that is required, which would cost $50, at a
minimum.
For the travel cost, ward cost and cost of parents’ productivity loss, similar estimation
approaches to PICCs, the unit travel cost per day would be the same. The only difference is the
longer inpatient days for the PIV group. As we said before, patients with a PICC inserted can be
discharged earlier and receive home care instead of staying in the hospital; however, pediatric
`
52
patients with PIV inserted cannot usually go home. The whole infusion period must be finished
in hospital. Thus it is not necessary to consider the home care cost.
Four cost components were considered for the PIV group: insertion cost, travel cost, ward cost
and cost of parents’ productivity loss. Most of the estimation approach and assumptions were the
same as the PICC cost estimation except the insertion cost. The insertion cost for the PIV is
much lower than the PICC.
2.5.1 Statistical analysis software
Statistical Package for the Social Sciences, version 17.0 (SPSS Inc., Chicago, IL, USA) was used
to analyze data in our study to conduct analyses.
2.5.2 Descriptive analysis
Descriptive analyses on all pertinent variables included in this study were generated in this study.
For those continuous variables (e.g. age, weight, etc), mean, median, range and sum were
reported to describe the central tendency. However, for those categorical variables (e.g. sex,
reason for PICC, etc), frequency and percentage were used to measure the dispersion. To analyze
these variables, as patients may have more than one catheter during their therapy period, the case
unit was based on the catheter instead of the patient.
2.5.3 Multivariate linear regression model
A multivariable linear regression model was used to assess variables associated with the total
cost of a PICC. All cost components associated with a PICC were aggregated to get the total cost
of a PICC. The dependent variable was defined as the total cost associated with a PICC. All
`
53
pertinent variables as mentioned before were placed into this multivariable linear regression
model as independent variables to identify the determinants associated with the total cost of a
PICC: age, sex, weight, patient primary diagnosis, ward unit, distance to the hospital, catheter
insertion reason, access vein, type of anesthesia, catheter size, catheter dwell days, line type, line
manufacturer, cuffed/uncuffed, lumen number, removal reason, complication, death and home
health care condition.
The multivariate linear regression model is listed as follows:
Total cost of a PICC = a + b0D + b1X1+ b2X2…. + bnXn + c1 (D* X1) + c2 (D* X2)…. + cn (D*
Xn) +ξ.
In this model, “a” is the intercept in this regression model which is the cost spent on a PICC
under the circumstances when all pertinent independent variables equal zero. “ξ” is the error
term. “D” means the total catheter dwell days which is defined as the period from the date of
insertion to the date of removal of the line. ‘X1’, ‘X2’... and ‘Xn’ are the variables we mentioned
before that is used to explain the total cost. The series of bn are the unstandardized regression
coefficients that can be used to weight the independent variables. Interaction terms were also
considered in this model too. As this study mainly focuses on the catheter dwell days of the lines,
the interaction between the variable “catheter dwell days” and other variables such as age, sex,
weight, etc are taken into consideration. The series of D* Xn are the interaction terms and cn are
the relevant coefficients of those interaction terms. The magnitude of the difference in the R2
statistic of models with or without the interaction term could be used to assess the significance of
that interaction.
In this regression model, P values less than 0.05 was considered to be statistical significant. R2
`
54
and adjusted R2 were selected to assess how much the outcome can be explained by the selected
independent variables. The F statistic test was also tested to check whether the hypothesis that
each coefficient equals zero is rejected (P value <0.05). The T statistic of each coefficient was
applied to check each variable’s significance (P<0.05),
There are three major multiple regression procedures to do the regression analysis: simultaneous
entry of independent variables, stepwise and hierarchical regression. As this study aims to
determine the best subset of X’s to explain the dependent variable, stepwise regression is more
appropriate. That is, at each step, independent variables can be entered or removed by assessing
their importance. To further examine the relationship between the total cost of a PICC and
permanent independent variables, stepwise regression procedure is used to put those independent
variables in the model.
2.5.4 Cost comparison between peripheral intravenous therapy and peripherally inserted central
catheter
The insertion cost for a PICC is usually expected to be higher than a PIV; however, there is an
increasing tendency for PICCs to result in lower costs than a PIV as PICC patients can be
discharged earlier, thereby yielding savings in ward costs. As the catheter dwell days increase,
the total cost for PIV is expected to be eventually higher than those for a PICC. Consequently, if
the total cost of a PICC is initially greater than that for a PIV and if costs increase more rapidly
for a PIV than for a PICC, a point will be reached when both procedures entail similar costs. If
catheter dwell days were to increase further, a PICC would be associated with lower costs than a
PIV. For example, in Figure 3, the dashed curve line depicts the total costs associated with the
PICC while the solid line represents the total costs associated with a PIV. The dashed curve and
`
55
the solid curve can be divided into three segments for this graph, segment A, segment B and
segment C. Segment A means the insertion point. At this point, due to the insertion costs, the
cost curve will not start at the origin. Segment B starts after the insertion and ends before the
cross-point. It refers to short-term catheter-dwell-days and indicates that over this period total
costs associated with PICCs will be higher than those for a PIV because of the greater insertion
cost of the PICCs. The cross point D* (Figure 3) occurs where total costs for PIVs and PICCs are
the same. For shorter dwell-days, a PIV is less costly than a PIC; however, for longer dwell-days,
a PICC is less costly. Segment C starts after the cross point. Along this segment, PICCs cost less
than a PIV primarily because of the lower inpatient ward cost.
Figure 4 Theoretical model of capturing the breakeven dwell days when PIV and PICC have the
same total costs
To assess whether PICCs are cost saving, another multivariate linear model for a PIV was built.
The multivariate linear regression model is listed as follows:
`
56
Total cost of a PIV = e + f0D+ f1Y1+ f2Y2+…. + fnYn+ k1 (D*Y1) + k2 (D*Y2) +…. + kn (D*Yn)
+η.
Interaction terms were also considered in this model.
In this model, “e” is the intercept in this regression model which is the cost spent on a PIV under
the circumstances when all pertinent independent variables equal zero. “η” is the error term. “D”
means the total catheter dwell days which is defined as the period from the first date of insertion
to the last date of removal of the line when therapy is finished. ‘Y1’, ‘Y2’... and ‘Yn’ are the
variables we mentioned before to explain the total cost. The series of fn are the regression
coefficients that can be used to weight the independent variables. Interaction terms were also
considered in this model, the interaction between the variable “catheter dwell days” and other
variables such as age, sex, weight, etc are taken into consideration. The series of D* Yn are the
interaction terms and Kn are the relevant coefficients of those interaction terms. The magnitude
of the difference of R2 with or without the interaction term for this model was also used to assess
the significance of that interaction.
As we mentioned above, this study aims to assess the theoretical circumstances where a PICC
will become a cost saving catheter. That is, we analyze the data to find the breakeven dwell days
D* when the total cost of a PIV equals the total cost of a PICC. Before the breakeven dwell days
D*, total cost of a PICC is more expensive than a PIV, but after that point, the PICC will provide
cost savings.
2.5.5 Regression diagnosis of the multivariate linear regression models
As the best case scenario was that independent variables will significantly correlate with the
`
57
dependent variable but will have low correlations among themselves, correlations of these
independent variables were checked first. When there are correlations among the independent
variables to some degree, multicollinearity will occur. As it will make determining the
importance of a given explanatory factor difficult to identify, diagnosis of multicollinearity is
required. The tolerance and Variance Inflation Factor (VIF) are used to test the problem of
multicollinearity of the independent variables (Myers, 1990). If the value of tolerance is less than
0.2 or VIF is greater than 4, it suggests multi-collinearity; if the tolerance was less than 0.1 or
VIF is great than 10, it strongly indicates multi-collinearity. When multicollinearity occurs, the
variable with more clinical significance is considered to keep in the model; otherwise, priority
was given to the variables with more statistical significance (Myers, 1990).
To develop a good explanatory relation between independent and dependent variables, regression
diagnosis is required. Linear regression models have three primary characteristics: linearity,
homogeneity of variance, and normally distributed residuals. For multiple linear regressions,
multicollinearity is also required to be checked. Therefore, we will check the linearity,
multicollinearity, homogeneity of the variance and the normality.
Influential data are those points with different patterns of relationship between the independent
variables and the outcome, which can make a large difference in the result. There are two kinds
of influential data, outliers and leverage points. Outliers can bring large residual which may
indicate model misfit. It suggests a sample peculiarity or data entry error. Leverage points are
those extreme points which may cause changes in the standard errors of regression coefficients
estimates. To identify those potential influential data, Cook’s distance is used to measure how
much the residuals would change if the current case were deleted from the calculations. If
`
58
Cook’s distance is greater than 1, careful scrutiny is required. If the Cook’s distance is greater
than 4, potentially serious outliers may occur (Barnett, V. & Lewis, T. 1994).
Linearity between each explanatory factor and the dependent variable can be checked by the
scatter plot of the outcome against each explanatory factor. Homogeneity of variance means the
error variance should be constant. A well-fitted model should not have any pattern to the residual
plotted against the fitted value. Scatter plot is also used in this study to check the regression
assumption of homogeneity of variance (UCLA Academic Technology Services)
For linear regression, one of the assumptions is that the error term should be normally
distributed. There are many ways to check normality assumption. Normal Q-Q plot is applied in
this study to check the assumption of normality. If the normality is violated and the distribution
is skewed, transformations of the explanatory factor variable or dependent variable such as log
transforms are required to avoid abnormal distribution (Johnson & Kuby, 1999). Cost data are
usually skewed with a small number of very high costs but may not necessarily be treated as
outliers. Log-transformation of the dependent variable is often undertaken to ensure that the
assumption of normality of the residuals under the classical model (Rascati, et al, 2001)
2.6 Sensitivity analysis
Sensitivity analyses were conducted to test the robustness of the conclusion. One-way sensitivity
analyses were performed based on their plausible range or extreme values. There are two types
of uncertainty. One is parameter uncertainty; and the other is the structure uncertainty
(Guidelines for the economic evaluation of health technologies: Canada, 2006). Both structural
uncertainty and parameter uncertainty were analyzed in this study. Another regression with the
`
59
inpatient costs omitted was tested for structural uncertain purpose. For parameter uncertainty,
five variables were required to be tested in sensitivity analysis which was made based on
assumptions or professional opinions: nurse assessment time, parents’ time lost per day, home
care cost per day, travel cost per day, inpatient ward cost, were considered to alter in multilevel
conditions in terms of our theoretical model using plausible values for each parameter or
plausible alternatives for each assumption. Tornado diagram was used to present the one way
sensitivity analysis. The horizontal axis describes the total cost associated with PICCs per
patient, and the vertical axis is parameters analyzed. All of these variables were ordered from
widest to narrowest based on the parameter’s range. A dotted line was used to depict the base
case for each parameter. Bars were used to represent the range of each parameter in our analysis
(Guidelines for the economic evaluation of health technologies: Canada, 2006).
2.7 Ethics
This study was approved by the Research Ethics Review Board of the Hospital for Sick Children
(Appendix 5). As these data were already captured in the hospital’s databases, patients’ personal
information such as name, medical record number were deleted in this study, this research
involves only minimal risk and anonymous data collection. Thus informed consent from these
patients was not needed.
`
60
Chapter 3 Results
This chapter describes the systematic review and presents the findings of this study. The results
are shown in four sections. The first section describes the systematic review results. The
following two sections will give the descriptive analysis of variables associated with PICCs and
PIV. The fourth and fifth sections present the results of the multiple linear regression models.
Then the breakeven dwell days when a PICC will become cost saving is discussed. After that,
regression diagnosis is presented to check the assumptions of the regression model. Finally, one
way sensitivity analysis is presented to deduct uncertainty.
3.1 Systematic review results of peripherally inserted central catheter costs
There were 3 papers retained for review of the costs of PICCs in pediatrics. Based on these
papers, details about each paper’s cost result are listed as follows in table 13 (Table 13). The 1st
author’s names, year of study, study type, main cost items and limitations were extracted from
these three papers. Moore’ s study provided the total inpatient charge for a PICC ranging from
US$3,706.02 to US$28,792.16 with an average charge of US$14,209.81 for an average 4.11
hospital days in which the insertion cost ranged from US$1,363 to US$1,954 including the costs
of fluoroscopy, PICC placement, and insertion equipment cost. In addition, the outpatient
management cost for antibiotic therapy ranged from US$1,382 to US$1,889 for 14 days (Moore
et al., 2006). Van Winkle’s study estimated the average daily cost for a PICC at home was
US$115 while the average daily cost for a PICC as an inpatient was US$1,185. All costs for
medications, equipment, nursing, and outpatient physician visits were included as calculating the
PICC at home cost. All general billable costs were comprised in the inpatient cost estimation
(Van Winkle, et al., 2008). In Schwengel’s study, it mentioned that the lowest total cost for a
PICC per patient was US$173.58 and the highest total cost for a PICC per patient was
`
61
US$440.70. The cost estimations include the labor time (anesthesiologist & phlebotomist),
equipment (PICC trays & IV catheters), and operating room time costs (Schwengel et al., 2004).
However, there were flaws in these studies. First, all of these three papers’ sample size was
no more than one hundred, which may cause bias. Second, all papers did not provide details on
how the costing numbers were obtained. Finally, these three papers did not consider indirect
costs associated with PICCs. As the three papers calculated different cost items, meta-analysis is
not used in this review due to data heterogeneity.
`
62
Table 13 Systematic review for PICC cost results
1st author Year
published
Sample
size
Study
type
Cost items Limitations
Moore 2006 45 Retrospective 1. Inpatient hospital charges: average US$14,209.81, range
(US$3,706.02, US$28,792.16);
2. Direct PICC-associated costs: range (US$1,363,
US$1,954)
3. Outpatient at-home PICC antibiotic therapy cost: range
(US$1,382, US$1,889).
1. Small sample size
2. Rough direct costs
3. Lack of indirect cost
Schwengel 2004 96 Randomized
control trial
Insertion cost, range (US$173.58, US$440.70), underestimated insertion
cost
Van Winkle 2008 86 retrospective 1. Average daily cost for home health treatment US$115 per
patient
2. Average daily inpatient cost US$1,185 per patient
1. Small sample size
2. lack of indirect cost
`
63
3.2 Descriptive analysis of peripherally inserted central catheter
(1) Demographic variables
There were 524 patients who underwent 573 PICC insertions during this study period (Jan.1-
Dec.31, 2008) at IGT in SickKids. Among the 524 patients, 482 unique patients had one PICC
inserted during the study period; 35 patients had two PICCs insertions; while 7 patients had three
PICCs inserted during the study period. For our analysis, we use ‘per catheter’ as the basic unit
for analysis. Of these procedures, 53.9% (or 309) were performed on males and 46.1% (or 264)
performed on females (Table 14). The average age of the study population was 4.79 years old
with a median age of 1.10 years old. Similarly, the average weight was 19.00 kilogram (kg) with
a median weight of 9.30 kg. The average travel distance from the hospital was 81.66 kilometer
(km) with a median distance 42.60 km far away from the hospital (Table 17).
(2) Patient primary diagnosis, ward unit inpatient days and catheter dwell days
The most common primary diagnosis for these patients, using ICD-10 to classify, were digestive
problems, disease of circulatory systems, and disease of blood and certain disorders in immune
mechanism, which accounted for 19.3%, 11.2% and 11.2%, respectively. A patient’s ward unit
was analyzed by the wards they stayed in on the second last day of admission, prior to discharge.
Among the 573 patients, 84 patients stayed in NICU, which accounts for 14.7%; 53 patients
stayed in a medical ward, which accounted for 9.2%. 189 patients (33%) stayed in a surgical
ward and 243 patients (42.4%) stayed in a medical surgical ward. Four patients stayed in CCCU
(Table 14).
The total number of inpatient days for the newly inserted PICC was 20,186.00 days for 570 cases
during the study time period in 2008. Three patients’ data were missing either the admission date
`
64
or discharge date. The average hospital stay per patient was 37 days with the median days of 22
days. Similarly, the total catheter dwell days were calculated from the date of insertion to the
removal date of the line. The cumulative catheter dwell days were 29, 058 days with a mean of
51 days and median time of 27 days (Table 17).
(3) Insertion and removal information
The most common reason for PICC placement was medication administration and antibiotic
therapy (MEDS/ABX), which accounted for 50.3% of the total catheter insertions. The next most
common reason was administration of total parental nutrition (TPN). It was estimated that 21.1%
PICC lines were used for TPN plus medical therapies. An additional 16.9% were used for TPN
only treatment. A further 6.6% PICCs were used for chemotherapy treatment (Table 15).
Of the total 573 line insertions, 26.0% of catheters were inserted while patient was under general
anesthesia/sedation administered by an anesthesiologist (called GA). Only 3.5% were provided
with sedation administered by the IR team; most catheter insertions (70.3%) were provided with
local anesthesia with or without sucrose. Sixty five percent of PICCs were placed via the basilic
vein. The brachial vein was used in a 26.8% of cases. Other veins such as cephalic vein were
also used for access, but in fewer cases which only accounted for 5.9% in total (Table 15).
Among all these patients, 79.4% lines were removed because of end of therapy; 12.0% PICCs
were removed due to complications but without requiring or reinserting a new PICC; 6.1%
patients had their PICC removed due to a complication and had a new subsequent PICC inserted.
Both VANs and IRs could remove the catheters. As for the removal, 44.7% catheters were
removed by VANs while 46.6% were removed by IRs (Table 16).
`
65
(4) Line information
As for the type of PICC, 87.3% catheters were single lumen catheters and only 12.7% catheters
were double lumen catheters. Generally the patients were provided with size 3 French or size 4
French catheters, which accounted for 83.6% in total. 93.9% of catheters were cuffed catheters;
only 6.1% catheters were uncuffed catheters. Almost 81.5% catheters used in IGT were silicone
catheters supplied by COOK Medical. 7.5% of the catheters used in IGT were Power PICC
(BARD) (Table 15).
(5) Descriptive analysis of cost components (Unit: Canadian dollar)
With respect to the insertion, three components (material, labor, equipment costs) were analyzed
respectively. The mean of material, labor and equipment costs per case were $441.78, $647.60,
$338.69 with the median costs of $383.37, $541.50, $300.56, respectively. The mean total
insertion cost, including the three components, was $1,428.07 with a median of 1,280.79. As the
total catheter dwell days were 29,058 days, the total insertion cost can also be presented as
$28.16/day. At least one of the parents/family members in each family was assumed to take
leave from their job and stay with the child in the ward during admission. Therefore, parent
productivity loss was the main indirect cost component. The average productivity loss was
$4,711.43 with a median cost of $2,801.82 per case or $92.23/day (Table 17 & Table 18).
As for the complication, 199 catheters had a PICC related complications which accounted for
34.73% of the total number of the cases. If expressed in terms of catheter dwell days, it
calculates as 6.85 complications per 1,000 days. Different complications required nurse
consultation and the associated treatments in order to treat those complications. The
`
66
complication cost averaged over the entire study in the total 573 catheters. However, among 199
catheters with complication, 139 catheters occurred one complication; 43 catheters had two
complications; 11 catheters had three complications; and 6 catheters had four complications.
Some catheters may have more than one complication. There were 38 malpositions, 49
38 dislodgements, 15 thrombosis, 57 breakages or leakages, 76 blockages, and 9 other
complications such as swelling in total. The average complication treatment cost was $499.53
with a median cost $217.32 from the 199 lines. The total complication costs spent associated
with PICCs among the 524 patients was $117,390.19, or $4.03/day. When a complication
occurred, e.g. infection, a lab test was the primary way to detect and diagnose the problem. If we
looked at these 199 catheters, 61 cases required lab examinations because of the complication,
including CBC, blood culture, etc. The total laboratory cost was $1,624.15 for the 61 catheters,
with a mean cost $26.62 and median cost $28.95 per case. Meanwhile, in some circumstances,
imaging was also required. 102 patients underwent imaging examinations because of
complications. The total fee for imaging examinations, including ultrasound, linogram, and
venogram, was $8,525.40 with a mean cost of $83.58 and a median cost of $ 62.10 per case.
Thirty-four patients used the medical emergency services. $22,734.00 was spent on ER service
with the average cost $668.65, and a median cost of $421.00. There were different travel
methods for these patients, by car or by air. The total estimated travel cost incurred by parents
was $235,071.81 with a mean cost of $ 412.41 and a median cost of $256.03 per patient, or
$8.09/day. 248 patients went home earlier with their PICCs inserted, which accounts for 47.3%.
When the patients were at home, home care nurses will tend the patients, their lines and change
of their dressings and dealt with some minor complications. The average home care cost of a
PICC was $3,153.98 with a median cost of $ 176.00, or $61.94/day on average. The average
ward cost for total admission was $95,174.58 with a median cost of $53,911.00 or $1,778.50/day
`
67
during patients’ hospital stays. Adding up all the different cost components, we calculatedthe
total cost related to PICC. The average total cost per PICC (including the inpatient ward cost)
was $100,193.62 and the Median cost was $57,872.24. The minimum and maximum costs were
respectively $1653.38 and $171,199.01. There was $5.74*107 in total spent on these 573
catheters in this study, taken the catheter dwell days into consideration, the total cost associated
with a PICC per day was $1975.74. Total direct cost accounted for 94.31% of the total cost while
the indirect cost was 5.69% of the total cost associated with PICCs. In the direct cost category,
inpatient cost would influence the total costs significantly as it accounted for 86.90%.
Comparatively speaking, other cost components represent less, most of them less than 1.00%
(Table 17 & Table 18).
`
68
Table 14 Descriptive analysis of demographic characteristics and line information of PICCs
Note
*NICU- Neonatal Intensive Care Unit
**CCCU- Cardiac Care Unit
Variable Frequency Percent
Sex 573 Male 309 53.9 Female 264 46.1 Ward 573 Medical ward 53 9.2 Surgical ward 189 33.0 Medical surgical ward 243 42.4 NICU* 84 14.7 CCCU** 4 0.7 Type of line 573 Double 73 12.7 Single 500 87.3 Primary diagnosis 509 Disease of digestive system 128 19.3 Disease of circulatory system 74 11.2 Disease of blood and blood forming organs, certain disorders in immune mechanism
Others 50 8.8 Complication* 199 Block 59 or 2.03/1,000 days 0.30 Breakage 37 or 1.27/1,000 days 0.19 Dislodgement 22 or 0.76/1,000 days 0.11 Infection 29 or 1.00/1,000 days 0.15 Malposition 28 or 0.96/1,000 days 0.14 Thrombosis 7 or 0.24/1,000 days 0.03 Others 17 or 0.58/1,000 days 0.08 Death 522 Yes** 47 8.2 No 475 82.9
`
71
Table 17 Descriptive analysis of demographic variables and time variables of PICC (continuous variables)
Note: * The sample size of inpatient days and catheter dwell days are all 570, instead of 573. Three patients’ inpatient days and catheter
dwell days’ information were missed in the patient medical record.
Continuous variable Sample size Mean (per catheter)