**Bedside Procedure: Retained Central Venous Catheter**

**Bedside Procedure: Retained Central Venous Catheter**

DOI: 10.5772/intechopen.69748

Maureen E. Cheung, Logan T. Mellert and Michael S. Firstenberg Michael S. Firstenberg Additional information is available at the end of the chapter

Maureen E. Cheung, Logan T. Mellert and

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/intechopen.69748

#### **Abstract**

The placement of central venous catheters (CVC) is a common medical procedure and adjunct to current medical therapy. With millions of CVC placed yearly in the United States, complications occur. It is important to be aware of the potential immediate and long term complications associated with this procedure. In this chapter, a representative case of retained CVC guidewire is presented. Provider experience, appropriate patient selection and insertion technique can minimize the risk of immediate complication in most cases. A brief discussion of CVC complications with emphasis on guidewire malfunction is presented along with management and retrieval options.

**Keywords:** retained guidewire, central venous catheter, catheter complications, bedside procedures, J wire, guidewire fracture, intravascular foreign body

## **1. Introduction**

With more than 5 million central lines placed in the United States every year, complications are anticipated [1–4]. The most common immediate complications of central venous catheterization (CVC) include pneumothorax (incidence of 0.5–1.5%), arterial puncture (0.5–3.7%), and cardiac ectopy (0.9%) [1, 5, 6]. Retained guide wire is a recognized but rare complication during insertion of CVC. The estimated incidence is 0.05–0.1% during CVC insertion [7, 8]. Causes of retention include guide wire looping, entrapment, wedging within catheter and fracture during insertion [9–11]. While retained guide wire is a rare complication, it is entirely preventable except in cases of catastrophic equipment failure and is considered an unacceptable occurrence by physicians.

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2018 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

## **2. Vignette**

A 69 year old female presents with pneumonia and sepsis. She is hypotensive and unresponsive to multiple fluid challenges. The placement of a CVC is necessary for norepinephrine infusion therapy. She has a past medical history of hypertension, diabetes mellitus (type II), and chronic obstructive pulmonary disease. Her past surgical is significant for laparoscopic cholecystectomy and Cesarean section. She denies a history of head/neck surgery or prior central venous instrumentation.

After obtaining consent, a right internal jugular (IJ) CVC is placed using ultrasound guidance. After gaining venous access, the guide wire is threaded. It initially passes easily but resistance is met at approximately 10 cm. An attempt is made to withdraw the guide wire, however significant resistance is met. After multiple attempts to withdraw and advance the guide wire, there is a sudden change in resistance and the guide wire is easily advanced. The remainder of the procedure is completed using the modified Seldinger technique and a triple lumen catheter is placed. All ports easily draw blood and flush with saline (**Figure 1**).

A post procedure chest X-ray is obtained. The tip of the catheter terminates in the midsuperior vena cava (SVC) and a linear hyperdensity within the right atrium is noted. This is approximately 8 cm in length and consistent with retained guide wire. You inspect the procedural guide wire and note that the distal end does not have its characteristic "J" bend.

**Figure 1.** Chest X-ray revealing a retained guide wire (arrow).

## **3. Discussion**

**2. Vignette**

152 Vignettes in Patient Safety - Volume 2

central venous instrumentation.

**Figure 1.** Chest X-ray revealing a retained guide wire (arrow).

A 69 year old female presents with pneumonia and sepsis. She is hypotensive and unresponsive to multiple fluid challenges. The placement of a CVC is necessary for norepinephrine infusion therapy. She has a past medical history of hypertension, diabetes mellitus (type II), and chronic obstructive pulmonary disease. Her past surgical is significant for laparoscopic cholecystectomy and Cesarean section. She denies a history of head/neck surgery or prior

After obtaining consent, a right internal jugular (IJ) CVC is placed using ultrasound guidance. After gaining venous access, the guide wire is threaded. It initially passes easily but resistance is met at approximately 10 cm. An attempt is made to withdraw the guide wire, however significant resistance is met. After multiple attempts to withdraw and advance the guide wire, there is a sudden change in resistance and the guide wire is easily advanced. The remainder of the procedure is completed using the modified Seldinger technique and a triple lumen

A post procedure chest X-ray is obtained. The tip of the catheter terminates in the midsuperior vena cava (SVC) and a linear hyperdensity within the right atrium is noted. This is approximately 8 cm in length and consistent with retained guide wire. You inspect the procedural guide wire and note that the distal end does not have its characteristic "J" bend.

catheter is placed. All ports easily draw blood and flush with saline (**Figure 1**).

When inserting a CVC, it is essential that the provider is experienced and has been appropriately trained, supervised, and privileged to perform the procedure. If the clinical scenario allows, informed consent should be obtained. An appropriate insertion site should then be selected. The modified Seldinger technique is utilized for catheter placement. If an IJ or subclavian (SC) catheter is inserted, chest X-ray is performed to confirm placement location and ensure no pneumothorax has developed [2, 6, 12–16].

### **3.1. Indications and catheters**

CVC are an important component of medical care. They provide temporary short-term or long-term vascular access and are used for hematologic sampling and monitoring, longterm antibiotic administrations, hemodialysis, and the delivery of caustic therapeutics such as vasopressors, total parenteral nutrition (TPN), and chemotherapy. During insertion and when not in use, these catheters are flushed to maintain patency. Flushes are usually a sterile saline or heparin saline solution which is injected to fill the catheter. If an institution utilizes heparin saline for flushes, care should be exercised to identify patients with contraindications to heparin such as history of hypersensitivity reaction, and heparin-induced thrombocytopenia (HIT) [17–19]. There are many devices available for central venous access but they are generally composed of silicone or polyurethane which allows for relative stiffness at room temperature and a softened state at body temperature [6, 11]. CVC can be broadly categorized into non-tunneled, tunneled, and implantable with ports.

Of these, non-tunneled CVC are the most frequently inserted. They are typically placed at bedside for short-term therapy (1–4 weeks) and in patients with significant physiologic distress [12]. The most frequent entry sites are the internal jugular (IJ), subclavian (SC), and femoral veins. Peripherally inserted central catheters (PICC) are similar to non-tunneled CVC. They are non-tunneled catheters placed at bedside, although generally in a more elective fashion. They are inserted through a peripheral arm or leg vein and advanced until the tip lies in the SVC. They are small-bore catheters that usually enter the system in the brachial, antecubital or long saphenous veins. They have the benefit of comfortable positioning for the patient and have lower reported infection rates than IJ, SC, or femoral catheters. PICCs have a smaller total diameter with only 1–2 lumens and slower maximum infusion rates [20]. Midline catheters are placed in the same entry sites as PICCs but terminate in the axillary vein. While still considered a short-term device, PICC and midline catheters are often maintained longer than a non-tunneled CVC due to their lower infection rates. In the literature, there is emphasis that these should still be considered a short-term catheter and to cautiously maintain them for >4 weeks [20].

Tunneled CVC are placed in a procedure or operating room utilizing fluoroscopy for visualized insertion. These catheters are tunneled through subcutaneous tissue and contain a Dacron cuff which causes fixation. The combination of tunneling and cuffed fixation creates physical barriers to infection, reducing infection rates and allowing for long term use of these catheters. Implanted ports are also long-term devices placed in the operating room. They consist of a catheter and diaphragm which is accessed via cutaneous puncture. Inserted subcutaneously, ports are typically placed in upper anterior chest wall though an IJ or SC approach [12]. Once inserted, they have no externalized components allowing for the lowest infection rates amongst CVC [21]. For the purposes of this paper, we will briefly review relevant anatomy and insertion technique for non-tunneled IJ, SC, and femoral catheters. PICC and midline catheters are not addressed here as most hospitals have a dedicated team of providers who perform these insertions [22].

#### *3.1.1. IJ catheterization*

After selecting an appropriate access site, the care team should don hat and masks. The provider will then wash his/her hands and dress in sterile gown and gloves. The area is prepped and draped in normal sterile fashion with chlorhexidine solution. During the prep, the provider should examine the contents of the sterile central line insertion kit. Each hospital has its own standard CVC kit and it is important to be familiar with the equipment prior to proceeding. The catheter lumens are then flushed with sterile saline or heparin saline and compatibility of the needle, guidewire, and catheter should be confirmed. The contents of the kit should be organized in a fashion that allows for easy transition from one portion of the procedure to the next.

After sterile draping, the desired cannulation site is infiltrated with local anesthetic. If utilizing ultrasound, this should also be sterilely draped at this time. The provider will position themselves at the head of the bed with ultrasound in the nondominant hand. The patient is then placed in Trendelenburg position and the head rotated 30–45° away from the access side to allow access to the neck. Additional rotation of the head to the contralateral side does not aide the provider as it brings the sternocleidomastoid muscle into an anterior position over the IJ and increases the anterior-posterior overlap of the IJ with the carotid artery [2]. Therefore, only as much rotation as is needed to gain access to the neck is utilized.

Next, IJ and carotid artery are inspected sonographically. It is important to be familiar with regional anatomy and anatomic landmarks. For the IJ approach, the landmarks are the sternal and clavicular muscle heads of the sternocleidomastoid, and the clavicle. These borders create a triangle known as Sédillot's triangle. The carotid and IJ are identified slightly superior to the apex of the triangle. IJ is typically anterior/lateral and is easily compressed with ultrasound probe. For patients with difficult to identify IJ or complex anatomy, color doppler mode or duplex mode can be utilized for identification. Once identified, the large-caliber introducer needle is inserted through the skin at the apex of the triangle with approximately a 20–30° angle. While maintaining aspiration, it is advanced under direct visualization into the IJ. This is typically encountered 0.5 cm below the skin with the needle along an axis towards the ipsilateral nipple [2].

Once venous access is obtained, the US is set aside and the syringe is removed, taking care not to disturb the needle. With the J end of the guide-wire directed towards the midline, it is passed through the needle. It should thread smoothly and advance without resistance. If cardiac arrhythmia or ectopy is noted, immediately withdrawal the wire until it ceases. The needle is then removed, maintaining continuous control of the wire. A stab incision is made at the insertion site. A dilator is then passed over the wire, dilating the subcutaneous tissue. This is removed and the catheter is threaded over the wire. At this point, it is necessary to back feed the wire proximally until control is regained from the distal aspect of the catheter port. Once obtaining control of the wire, the catheter is inserted over the wire into the vessel. The wire is then removed and blood return is confirmed from each port with a following flush. Ultrasound can be utilized to confirm placement within the IJ and examine the lung fields for violation. The catheter is then secured in place and a sterile occlusive dressing is applied. Prior to removal of any equipment, a final sharps count and examination of the guidewire should be performed to ensure complete retrieval of guidewire with J-shaped tip and no sharps were misplaced. An upright chest X-ray is then obtained to confirm placement and verify no pneumo or hemothorax.

#### *3.1.2. SC catheterization*

catheters. Implanted ports are also long-term devices placed in the operating room. They consist of a catheter and diaphragm which is accessed via cutaneous puncture. Inserted subcutaneously, ports are typically placed in upper anterior chest wall though an IJ or SC approach [12]. Once inserted, they have no externalized components allowing for the lowest infection rates amongst CVC [21]. For the purposes of this paper, we will briefly review relevant anatomy and insertion technique for non-tunneled IJ, SC, and femoral catheters. PICC and midline catheters are not addressed here as most hospitals have a dedicated team of providers

After selecting an appropriate access site, the care team should don hat and masks. The provider will then wash his/her hands and dress in sterile gown and gloves. The area is prepped and draped in normal sterile fashion with chlorhexidine solution. During the prep, the provider should examine the contents of the sterile central line insertion kit. Each hospital has its own standard CVC kit and it is important to be familiar with the equipment prior to proceeding. The catheter lumens are then flushed with sterile saline or heparin saline and compatibility of the needle, guidewire, and catheter should be confirmed. The contents of the kit should be organized in a fashion that allows for easy transition from one portion of the procedure to

After sterile draping, the desired cannulation site is infiltrated with local anesthetic. If utilizing ultrasound, this should also be sterilely draped at this time. The provider will position themselves at the head of the bed with ultrasound in the nondominant hand. The patient is then placed in Trendelenburg position and the head rotated 30–45° away from the access side to allow access to the neck. Additional rotation of the head to the contralateral side does not aide the provider as it brings the sternocleidomastoid muscle into an anterior position over the IJ and increases the anterior-posterior overlap of the IJ with the carotid artery [2].

Next, IJ and carotid artery are inspected sonographically. It is important to be familiar with regional anatomy and anatomic landmarks. For the IJ approach, the landmarks are the sternal and clavicular muscle heads of the sternocleidomastoid, and the clavicle. These borders create a triangle known as Sédillot's triangle. The carotid and IJ are identified slightly superior to the apex of the triangle. IJ is typically anterior/lateral and is easily compressed with ultrasound probe. For patients with difficult to identify IJ or complex anatomy, color doppler mode or duplex mode can be utilized for identification. Once identified, the large-caliber introducer needle is inserted through the skin at the apex of the triangle with approximately a 20–30° angle. While maintaining aspiration, it is advanced under direct visualization into the IJ. This is typically encountered 0.5 cm below the skin with the needle along an axis towards the

Once venous access is obtained, the US is set aside and the syringe is removed, taking care not to disturb the needle. With the J end of the guide-wire directed towards the midline, it is passed through the needle. It should thread smoothly and advance without resistance. If cardiac arrhythmia or ectopy is noted, immediately withdrawal the wire until it ceases. The

Therefore, only as much rotation as is needed to gain access to the neck is utilized.

who perform these insertions [22].

*3.1.1. IJ catheterization*

154 Vignettes in Patient Safety - Volume 2

the next.

ipsilateral nipple [2].

A similar method is utilized for a SC approach [2, 12, 14–16]. A small shoulder roll may be placed between the shoulder blades to allow the shoulders to drop backwards. This maneuver exposes the necessary anatomy and brings the SC ventrally. If the shoulder roll is too large, the vein can collapse between the first rib and clavicle [1, 2]. Anatomic landmarks for SC access are the sternal notch, and the junction of the middle and medial third of the clavicle, near the deltopectoral groove. The provider will position him/herself on the side that is to be accessed. Following shoulder roll placement, the head is rotated slightly away from the side of insertion and the patient is placed into Trendelenburg position. The provider will place the nondominant index finger in the sternal notch and thumb on the clavicle above the deltopectoral groove. If accessing the right SC a right handed provider will turn their hips slightly towards the feet, if accessing the left their hips will be turned slightly towards the head; this allows comfortable positioning throughout the procedure. Local anesthetic is infiltrated and the access needle is inserted approximately 1–2 cm inferior and lateral to the junction of the middle and medial thirds of the clavicle. With continuous aspiration, the needle is directed medially and slightly cephalad in an axis towards the sternal notch and parallel to the floor. The needle may be "walked down" the clavicle if desired and passed beneath the clavicle with gentle downward pressure applied with the nondominant hand and advanced until venous blood is withdrawn. The syringe is then removed and the guidewire introduced with the J end directed caudally. The needle is removed; skin incised, and subcutaneus tract dilated. The catheter is then inserted over the guidewire again taking care to maintain continuous control of the wire throughout. The wire is removed, blood return verified and ports flushed. The catheter is secured and dressed. Post-procedural equipment verification is completed and an upright chest X-ray obtained.

#### *3.1.3. Femoral vein catheterization*

In the case of femoral vein catheterization, ultrasound can aid in anatomic visualization, however, the procedure is typically performed with anatomic landmarks only [2, 14]. The femoral triangle is identified with superior border of the inguinal ligament, medial border of the adductor longus muscle and the lateral border of the sartorius muscle. Within the triangle, the femoral vein is found medial to the common femoral artery, contained within the femoral sheath. It is important to remember that the inguinal ligament runs between the anterior superior iliac spine and the pubic tubercle and does not necessarily correspond to the "groin crease" [1, 2].

The most advantageous patient positioning is with the hip in a neutral or slightly abducted and externally rotated. Unlike the SC or IJ approach, the patient is placed supine or in slight reverse Trendelenburg position. The insertion site is identified by locating the arterial pulsation 1–2 cm below the inguinal ligament within the femoral triangle. The needle insertion site is approximately 1 cm medial to this maximal pulsation and the axis of insertion is cephalad and medially towards the umbilicus at a 45° angle from the skin. The femoral vein is typically encountered 2–4 cm below the skin and is accessed below the level of the inguinal ligament. The modified Seldinger technique is again utilized; post-procedure imaging is not typically indicated [2].

## *3.1.4. Other considerations*

Each insertion site has advantages and disadvantages; appropriate selection is affected by patient and clinical factors. The IJ can be accessed under direct visualization with ultrasound guidance and has a lower pneumothorax and hemothorax rate than the SC approach. However, it can be difficult in certain subsets of patients. Those with limited neck mobility, history of neck surgery, cervical collar, substantial subcutaneous tissue or significant cervical kyphosis can make IJ placement challenging [1, 14]. The SC has lower rates of arterial puncture than femoral or IJ locations and the lowest infection rate of the three. The femoral site has the advantage of no hemothoraces or pneumothoraces. It is a relatively safe and accessible location which is typically distant from other monitoring devices and can be placed without interruption of intubation or cardiopulmonary resuscitation. Femoral catheterization allows for free motion of the upper extremities and neck. Despite this, it has the highest rate of infection, limits ambulation and has the highest risk of associated thrombus formation [2, 15].

For proper positioning of SC and IJ CVC, the distance between insertion site and the SVC-atrial junction is vital to appropriate positioning and the avoidance of inducing arrhythmias. In an American based prospective study, fluoroscopy was utilized to determine this distance. It was found that the distance for right IJ insertion averaged 16 cm, right SC averaged 18.4 cm, left IJ averaged 19.1 cm, and left SC 21.2 cm [23]. A South Korean retrospective review of patients with CVC who underwent chest CT revealed slightly shorter distances. They reported a right IJ of 15 cm, right SC 14 cm, left IJ 18 cm, and left SC 17 cm [24]. These distances should be considered when selecting and inserting a CVC to ensure appropriate placement.

#### **3.2. Complications**

There are numerous case reports and case series documenting the potentially devastating complications of CVC. These include bleeding events, arterial puncture, infective sequela, cardiac conduction abnormalities, catheter malposition, thrombotic events, and mechanical device failure [1, 6, 25]. Identified risk factors for complications during CVC insertion are number of unsuccessful needle passes, inexperience of provider, body mass index >30 or <20, hypovolemia, large catheter size (specifically related to vascular complications), and previously failed catheterization attempts [1, 5, 25, 26]. When failed catheterization occurs, complications are reported as high as 28% [5]. Ultrasound assistance reduced immediate complications with the overall incidence decreased from 11.8 to 4–7% [5, 26]. In this section, the complications associated with CVC will be discussed briefly as well as advised action if they are encountered. The associated morbidity and mortality of these adverse events can be reduced with prompt recognition and appropriate action.

#### *3.2.1. Bleeding events*

the femoral vein is found medial to the common femoral artery, contained within the femoral sheath. It is important to remember that the inguinal ligament runs between the anterior superior iliac spine and the pubic tubercle and does not necessarily correspond to the "groin

The most advantageous patient positioning is with the hip in a neutral or slightly abducted and externally rotated. Unlike the SC or IJ approach, the patient is placed supine or in slight reverse Trendelenburg position. The insertion site is identified by locating the arterial pulsation 1–2 cm below the inguinal ligament within the femoral triangle. The needle insertion site is approximately 1 cm medial to this maximal pulsation and the axis of insertion is cephalad and medially towards the umbilicus at a 45° angle from the skin. The femoral vein is typically encountered 2–4 cm below the skin and is accessed below the level of the inguinal ligament. The modified Seldinger technique is again utilized; post-procedure imaging is not typically

Each insertion site has advantages and disadvantages; appropriate selection is affected by patient and clinical factors. The IJ can be accessed under direct visualization with ultrasound guidance and has a lower pneumothorax and hemothorax rate than the SC approach. However, it can be difficult in certain subsets of patients. Those with limited neck mobility, history of neck surgery, cervical collar, substantial subcutaneous tissue or significant cervical kyphosis can make IJ placement challenging [1, 14]. The SC has lower rates of arterial puncture than femoral or IJ locations and the lowest infection rate of the three. The femoral site has the advantage of no hemothoraces or pneumothoraces. It is a relatively safe and accessible location which is typically distant from other monitoring devices and can be placed without interruption of intubation or cardiopulmonary resuscitation. Femoral catheterization allows for free motion of the upper extremities and neck. Despite this, it has the highest rate of infection, limits ambulation and has the highest risk of associated thrombus formation [2, 15].

For proper positioning of SC and IJ CVC, the distance between insertion site and the SVC-atrial junction is vital to appropriate positioning and the avoidance of inducing arrhythmias. In an American based prospective study, fluoroscopy was utilized to determine this distance. It was found that the distance for right IJ insertion averaged 16 cm, right SC averaged 18.4 cm, left IJ averaged 19.1 cm, and left SC 21.2 cm [23]. A South Korean retrospective review of patients with CVC who underwent chest CT revealed slightly shorter distances. They reported a right IJ of 15 cm, right SC 14 cm, left IJ 18 cm, and left SC 17 cm [24]. These distances should be con-

There are numerous case reports and case series documenting the potentially devastating complications of CVC. These include bleeding events, arterial puncture, infective sequela, cardiac conduction abnormalities, catheter malposition, thrombotic events, and mechanical device failure [1, 6, 25]. Identified risk factors for complications during CVC insertion are number of unsuccessful needle passes, inexperience of provider, body mass index >30

sidered when selecting and inserting a CVC to ensure appropriate placement.

crease" [1, 2].

156 Vignettes in Patient Safety - Volume 2

indicated [2].

*3.1.4. Other considerations*

**3.2. Complications**

Bleeding events include hemorrhage, hematoma, and hemothorax from arterial or venous injury. While uncommon, innominate, aortic, SVC, and right ventricular perforation are reported in the literature [25, 27–30]. In those cases, improper use of the dilator and guidewire kinking were cited as the cause for injury [5, 25, 27, 28]. More common, is injury from puncture or cannulation of the carotid, subclavian, common femoral, or external iliac artery. The incidence of arterial injury from puncture (≤18 G) is reported to be 4.2–9.3% and 0.1–1.0% from cannulation (>7 Fr.) [25]. The majority of these injuries are identified at the time of occurrence with recognition of blood coloration, pulsatile flow, ultrasound visualization, or pressure measurements. If a needle puncture has occurred, it can typically be addressed with application of pressure. If the artery is cannulated, symptoms are reported in 30% of patients. In this subset, mortality reaches a rate of 20–40% [5]. Arterial catheter removal with direct pressure is associated with major complications in 47% of patients [31]. When removed by a surgical specialist or intervention radiologist under direct visualization with immediate surgical or endovascular intervention a 0% complication rate is reported [31]. This disparity in outcomes prompted the recommendation that if arterial cannulation is suspected, leave the catheter in place and seek immediate surgical consultation [25, 29, 31].

Arterial cannulation can lead to hemorrhage/hematoma, neurologic deficits, pseudoaneurysm, and AV fistula formation. Neurologic deficits are the result of either cerebral vascular ischemia or hematoma with nerve compression [5, 25, 31, 32]. Cerebral vascular ischemia from arterial CVC occurs when the inadvertent cannulation is not recognized and vasopressors are administered [5, 31]. Pseudoaneurysm and AV fistula formation can be an acute or delayed complication of arterial injury and discovery is reported years after catheterization [5]. The estimated incidence of AV fistula formation is 0.2% for IJ and 0.6% for SC catheterization attempts and can lead to symptomatic intracranial hypertension [5, 26]. Pseudoaneurysms are a recognized complication associated with arterial cannulation or rupture of a mediastinal vein during CVC placement [4, 33, 34]. The incidence of this rare complication is not defined in the literature. When a pseudoaneurysm does occur, treatment via endovascular stenting and open repair are described [4, 33]. Pseudoaneurysms can also occur with embolization following fracture of a catheter and in one case report resulted in a 5 × 4 cm pulmonary artery pseudoaneurysm which required surgical excision via lobectomy [33]. Both of these conditions require intervention. Various methods are described in the literature with coils, thrombin injections, manual compression, stenting, and open surgical approach utilized for repair [5, 26]. Once identified, the management of these complications should be undertaken with the aid of a vascular surgeon.

Localized hematomas with nerve compression of the brachial plexus or the sympathetic trunk (causing Horner syndrome) occurs in 4.7% of all CVC. It is caused by either arterial or venous hemorrhage [5, 26]. Other complications which can arise from hematomas include vocal cord paralysis, phrenic nerve injury, respiratory distress and airway obstruction [5, 26]. Venous hematomas can arise from multiple punctures, venous laceration, and attempted access at an inappropriate site. The utilization of ultrasound or fluoroscopy is advocated for prevention of these complications [1, 6, 25].

Catastrophic hemorrhage is a surgical emergency which requires prompt recognition and action. It occurs in the acute setting from puncture or perforation into regions of large potential space including the thoracic, abdominal, and retroperitoneal cavities [5, 28, 35, 36]. Unlike the mediastinum or neck which has relatively limited space, these cavities can accommodate a large amount of blood without clinical signs until hemodynamic instability is reached. Similar to arterial cannulation, if this occurs or is suspected, the catheter should be left in place and a vascular or cardiothoracic consultation immediately obtained. In this case, the catheter is partially occluding the tract and removal of the catheter blindly can lead to increased hemorrhage and death. If circumstances allow, imaging studies to define the path of the catheter are obtained to aid in planning for its safe removal [5, 25, 36].

#### *3.2.2. Central line-associated bloodstream infections (CLABSI)*

Infective sequela of CVC has become an area of interest in recent years, particularly as the Centers for Medicare and Medicaid Services (CMS) withdrew reimbursement for the treatment of hospital-acquired infections (HAI) including central line-associated bloodstream infections (CLABSI) [37]. The cause of CLABSI in non-tunneled CVC is attributed mainly to the migration of skin organisms at the insertion site through the cutaneous catheter tract and into the bloodstream with colonization of the external surface of the catheter. This modality of contamination is part of the rationale for tunneled CVC. Both non-tunneled and tunneled CVC contamination can occur via the catheter hub with intraluminal colonization of the catheters. Rarely, CVC can become seeded from other hematogenous infections [20, 38]. The infective organism encountered most frequently is Staphylococcus (37%), followed by Enterococcus [20, 22, 26, 38, 39]. The overall incidence of CLABSI is reported at 5.3 per 1000 catheter days with a CDC estimated cost per infection of \$16,550 and an attributed mortality of 18% [5, 26].

Higher rates of infection are noted based on entry site, emergent status of insertion, increased number of lumens, and type of CVC. With regards to location, PICC have the lowest infection rates of non-tunneled CVC at 1–2 per 1000 catheter days [20, 26, 38]. These are followed by SC (4 per 1000 catheter days) then IJ (8.6 per 1000 catheter days) and finally the femoral vein (15.3 per 1000 catheter days) [5, 20, 22, 26, 38]. In the case of elective non-tunneled CVC, infection incidence is reported at 1.1–3.35 per 1000 catheter days, significantly lower than the reported overall infection rates [5, 20, 38]. The rate for tunnelled CVC is 1.3 per 1000 catheter days and for implanted ports is 0–1 per 1000 catheter days [20, 26].

In recent years, a number of guidelines emerged to aid in the safe insertion and utilization of CVC. These were prompted by the Michigan Keystone project published in 2006 by Pronovost et al. which demonstrated significant reduction in CLABSI by implementing simple infectioncontrol practices [40]. These measures included maximum sterile barrier precautions, aseptic insertion technique, chlorhexidine skin preparation, transparent dressings, and removal of the catheter as soon as clinically possible [40–42]. It was also determined that the routine exchange of catheters for infection prevention was not necessary and may cause contamination [40, 42]. The increased interest and financial considerations have also led to device innovations for infection reduction. These include alcohol impregnated caps for the covering of hubs, and the development of antimicrobial-impregnated CVC with both internal and external surface impregnation [41, 42].

#### *3.2.3. Cardiac arrhythmias and ectopy*

Localized hematomas with nerve compression of the brachial plexus or the sympathetic trunk (causing Horner syndrome) occurs in 4.7% of all CVC. It is caused by either arterial or venous hemorrhage [5, 26]. Other complications which can arise from hematomas include vocal cord paralysis, phrenic nerve injury, respiratory distress and airway obstruction [5, 26]. Venous hematomas can arise from multiple punctures, venous laceration, and attempted access at an inappropriate site. The utilization of ultrasound or fluoroscopy is advocated for prevention of

Catastrophic hemorrhage is a surgical emergency which requires prompt recognition and action. It occurs in the acute setting from puncture or perforation into regions of large potential space including the thoracic, abdominal, and retroperitoneal cavities [5, 28, 35, 36]. Unlike the mediastinum or neck which has relatively limited space, these cavities can accommodate a large amount of blood without clinical signs until hemodynamic instability is reached. Similar to arterial cannulation, if this occurs or is suspected, the catheter should be left in place and a vascular or cardiothoracic consultation immediately obtained. In this case, the catheter is partially occluding the tract and removal of the catheter blindly can lead to increased hemorrhage and death. If circumstances allow, imaging studies to define the path of the catheter are

Infective sequela of CVC has become an area of interest in recent years, particularly as the Centers for Medicare and Medicaid Services (CMS) withdrew reimbursement for the treatment of hospital-acquired infections (HAI) including central line-associated bloodstream infections (CLABSI) [37]. The cause of CLABSI in non-tunneled CVC is attributed mainly to the migration of skin organisms at the insertion site through the cutaneous catheter tract and into the bloodstream with colonization of the external surface of the catheter. This modality of contamination is part of the rationale for tunneled CVC. Both non-tunneled and tunneled CVC contamination can occur via the catheter hub with intraluminal colonization of the catheters. Rarely, CVC can become seeded from other hematogenous infections [20, 38]. The infective organism encountered most frequently is Staphylococcus (37%), followed by Enterococcus [20, 22, 26, 38, 39]. The overall incidence of CLABSI is reported at 5.3 per 1000 catheter days with a CDC estimated cost per infection of \$16,550 and an attributed mortality

Higher rates of infection are noted based on entry site, emergent status of insertion, increased number of lumens, and type of CVC. With regards to location, PICC have the lowest infection rates of non-tunneled CVC at 1–2 per 1000 catheter days [20, 26, 38]. These are followed by SC (4 per 1000 catheter days) then IJ (8.6 per 1000 catheter days) and finally the femoral vein (15.3 per 1000 catheter days) [5, 20, 22, 26, 38]. In the case of elective non-tunneled CVC, infection incidence is reported at 1.1–3.35 per 1000 catheter days, significantly lower than the reported overall infection rates [5, 20, 38]. The rate for tunnelled CVC is 1.3 per 1000 catheter days and

In recent years, a number of guidelines emerged to aid in the safe insertion and utilization of CVC. These were prompted by the Michigan Keystone project published in 2006 by Pronovost

these complications [1, 6, 25].

158 Vignettes in Patient Safety - Volume 2

of 18% [5, 26].

obtained to aid in planning for its safe removal [5, 25, 36].

*3.2.2. Central line-associated bloodstream infections (CLABSI)*

for implanted ports is 0–1 per 1000 catheter days [20, 26].

Cardiac arrhythmias and ectopy are a recognized phenomenon during the placement of central venous devices and are frequently observed with incidence reaching 75% when the guidewire is advanced 25–32 cm from the IJ entry site [5]. Occurrence rates during insertion of CVC in adult patients are 41% for atrial arrhythmias, and 25% for ventricular ectopy [43]. A slightly lower rate is reported for pediatric patients with a 30% overall incidence of arrhythmias [44]. While complications from these are rarely reported in the literature, malignant/fatal arrhythmias have been described including complete heart block and sudden death. Typically these coincide with preexisting conduction abnormalities. In one such case report, a patient with a left bundle branch block was converted into a complete heart block during guidewire insertion [3]. The suggested mechanism was the superficial location of the right bundle branch making it vulnerable to guidewire trauma [3].

Continuous cardiac monitoring should be utilized throughout CVC placement. If any ectopy or arrhythmia is identified, the guidewire should be withdrawn until it resolves completely. It is imperative that the provider is aware of the guidewire length. Guidewire insertion for the placement of a CVC should never exceed 18 cm and insertion of 14 cm should be adequate for all insertion sites [23, 24]. Late onset of arrhythmia is also reported with an incidence of 0.9% in indwelling tunneled or implanted port CVC [5]. These are typically responsive to removal or replacement of CVC. Another cause of delayed arrhythmia which must be considered is mediastinal or cardiovascular perforation which is the result of catheter malposition [6].

#### *3.2.4. Catheter malposition and thrombosis*

Catheter malposition occurs in 3.3% of CVC insertions with the highest incidence with right SC entry (9.1%) and least frequently with right IJ (1.4%) [45]. Catheter malposition into the innominate, left internal mammary, azygous, hemiazygos, lateral thoracic, inferior thyroid, intercostal, and thymic veins have occurred [46, 47]. Incorrect position can result in inaccurate hemodynamic monitoring, thrombosis, and arrhythmias [6, 45–47]. While a rare event, perforation is possible. CVC perforation without tamponade occurs at a rate of 0.4–1% with an associated mortality of 12%; the rate of perforation with subsequent tamponade is 0.2% with a mortality of 81–90% [5, 6]. These perforations are attributed to catheter malposition [5, 6]. The exact definition of catheter malposition has evolved over the years but today is accepted as placement within the heart, an angle of incidence (the angle between the CVC tip and the wall of the vein) >40°, and placement into vessel other than the SVC or IVC [5].

The ideal catheter position for SC and IJ is with the tip in the SVC just above the right atrial junction. The accepted corresponding surface landmarks are the angle of Louis (the junction between the manubrium and sternum) or the right sternal border of the third intercostal space [1, 12, 15]. These can be used to estimate the appropriate length of catheter for insertion but final positioning should always be confirmed with a chest X-ray or fluoroscopy. The most reliable radiography landmark for placement verification is the CVC tip at the right tracheobronchial angle which ensures it lays ≥3 cm above the pericardial reflection [5]. This is accurate even when patient positioning gives the appearance of the CVC within the cardiac silhouette.

CVC thrombotic events are site, catheter, and patient dependent. In the case of non-tunneled CVC, site appears to be the major determining factor with the highest incidence occurring at the femoral site (21.5–29%) and lowest at the SC site (1.9%) [1, 6]. Catheter-related thrombosis in tunneled and indwelling port CVC has a reported incidence of 33–59% with a SVC obstruction rate of 0.1% [5]. Patients with malignancies are at particularly high risk for thrombus formation with an incidence of 41%; 15–30% of these patients will be symptomatic and 11% will experience an associated pulmonary embolism [5]. Anticoagulation and thromboprophylaxis was studied in this subset of patients without evidence of prevention or benefit [5]. Efforts to reduce the thrombogenicity of the catheter materials have been ongoing for >30 years [48–50]. This led to the refinement of materials with recognition that polyethylene catheters have a higher incidence of thrombus while silicone and polyurethane exhibit improved biocompatibility [49, 51]. Newer efforts have investigated the application of an athrombogenic layer, impregnation of catheter with medications including heparin and nitric oxide, as well as improved composition of catheter materials [51–53]. Due to the interaction between the catheter surface and hematogenous components, thrombus formation can occur at any point along the device. Morbidity from thrombus formation includes embolic events (particularly if the thrombus is associated with the tip of the catheter or >3 cm), infected thrombus, SVC occlusion, and ipsilateral edema [6, 54–57]. If a thrombus is identified and is <3 cm, the CVC can be safely removed without evidence of adverse events. However, if >3 cm there is an increased risk of embolic event upon line removal; anticoagulation or thrombolytic therapy is utilized to reduce the thrombus size but surgical removal is sometimes necessary [26]. In the event of thrombus identification, therapy should include symptomatic management, determination of the continued need to catheterization, monitoring for propagation and anticoagulation or thrombolytic therapy if indicated.

#### *3.2.5. Mechanical failure*

The mechanical failure of equipment is a potential complication of any device. Immediate mechanical failure is generally related to guidewire issues including retention from looping, entrapment, wedging within the catheter and fracture during insertion [9–11]. Cases of wire and catheter entrapment within inferior vena cava (IVC) filters, knotting with existing CVC, and incorporation with cardiac sutures are reported [26]. These have been managed with fluoroscopic endovascular procedures but at times require surgical intervention [11, 26]. When identified, immediate correction is preferred in stable patients. Known complications include catheter fragmentation with distal embolization and access site injury [26, 58]. The available literature on these cases is limited.

During insertion, guidewire fracture risk can be minimized by careful attention to associated resistance. If resistance is encountered, it may be from extraluminal placement, kinking, entrapment, or intraluminal stenosis [8, 25, 47, 58, 59]. When this occurs, remove the needle and guide wire en bloc (together), inspect immediately to verify complete removal, obtain new equipment and reinitiate procedure [12, 15, 16]. In this way, wire fracture and embolization is avoided. If you continue to encounter resistance, an alternative insertion site or fluoroscopic guidance should be considered.

Delayed mechanical failure is more commonly encountered with catheter fracture and embolization occurring in 0.5–3% of indwelling CVC with a morbidity rate of 71% and mortality of 30–38% [5]. Arrhythmia, cardiac arrest, pulmonary embolism with hemoptysis, perforation, and thrombosis are reported. Causes of delayed failure include breakage during catheter removal, entrapment, material properties of the catheter, and long-term mechanical fatigue [60, 61]. Material analysis from fractured catheters has shown an increased fracture risk with silicone catheters compared to polyurethane [60, 61]. Additionally, fatigue is particularly prevalent in SC catheters where mechanical shearing between the clavicle and first rib can occur [2, 5, 62]. This is referred to as pinch-off syndrome and is characterized by functional occlusion with postural changes. It will be reported by patients and staff as an inability to aspirate and difficultly flushing the catheter which is improved when the arm is raised. When pinch-off syndrome occurs, the repeated shear stress on the catheter will eventually cause fracture and embolization [2, 5, 62]. It is estimated that pinch-off syndrome is responsible for 41% of catheter embolic events and should be addressed immediately upon identification [2].

#### **3.3. Foreign body retrieval**

The ideal catheter position for SC and IJ is with the tip in the SVC just above the right atrial junction. The accepted corresponding surface landmarks are the angle of Louis (the junction between the manubrium and sternum) or the right sternal border of the third intercostal space [1, 12, 15]. These can be used to estimate the appropriate length of catheter for insertion but final positioning should always be confirmed with a chest X-ray or fluoroscopy. The most reliable radiography landmark for placement verification is the CVC tip at the right tracheobronchial angle which ensures it lays ≥3 cm above the pericardial reflection [5]. This is accurate even when patient positioning gives the appearance of the CVC within the cardiac silhouette. CVC thrombotic events are site, catheter, and patient dependent. In the case of non-tunneled CVC, site appears to be the major determining factor with the highest incidence occurring at the femoral site (21.5–29%) and lowest at the SC site (1.9%) [1, 6]. Catheter-related thrombosis in tunneled and indwelling port CVC has a reported incidence of 33–59% with a SVC obstruction rate of 0.1% [5]. Patients with malignancies are at particularly high risk for thrombus formation with an incidence of 41%; 15–30% of these patients will be symptomatic and 11% will experience an associated pulmonary embolism [5]. Anticoagulation and thromboprophylaxis was studied in this subset of patients without evidence of prevention or benefit [5]. Efforts to reduce the thrombogenicity of the catheter materials have been ongoing for >30 years [48–50]. This led to the refinement of materials with recognition that polyethylene catheters have a higher incidence of thrombus while silicone and polyurethane exhibit improved biocompatibility [49, 51]. Newer efforts have investigated the application of an athrombogenic layer, impregnation of catheter with medications including heparin and nitric oxide, as well as improved composition of catheter materials [51–53]. Due to the interaction between the catheter surface and hematogenous components, thrombus formation can occur at any point along the device. Morbidity from thrombus formation includes embolic events (particularly if the thrombus is associated with the tip of the catheter or >3 cm), infected thrombus, SVC occlusion, and ipsilateral edema [6, 54–57]. If a thrombus is identified and is <3 cm, the CVC can be safely removed without evidence of adverse events. However, if >3 cm there is an increased risk of embolic event upon line removal; anticoagulation or thrombolytic therapy is utilized to reduce the thrombus size but surgical removal is sometimes necessary [26]. In the event of thrombus identification, therapy should include symptomatic management, determination of the continued need to catheterization, monitoring for propagation and anticoagulation or thrombolytic therapy if indicated.

The mechanical failure of equipment is a potential complication of any device. Immediate mechanical failure is generally related to guidewire issues including retention from looping, entrapment, wedging within the catheter and fracture during insertion [9–11]. Cases of wire and catheter entrapment within inferior vena cava (IVC) filters, knotting with existing CVC, and incorporation with cardiac sutures are reported [26]. These have been managed with fluoroscopic endovascular procedures but at times require surgical intervention [11, 26]. When identified, immediate correction is preferred in stable patients. Known complications include catheter fragmentation with distal embolization and access site injury [26, 58]. The available literature on these cases is limited. During insertion, guidewire fracture risk can be minimized by careful attention to associated resistance. If resistance is encountered, it may be from extraluminal placement, kinking, entrapment, or

*3.2.5. Mechanical failure*

160 Vignettes in Patient Safety - Volume 2

Retained intravascular foreign bodies can occur during a variety of procedures. With the rapidly expanding scope of endovascular interventions, this complication will continue to be of clinical significance. In the case of CVC, guidewire retention during insertion is estimated at a rate of 0.03–0.1%, catheter retention during removal is estimated at 1.5%, and the overall incidence of retained foreign body due to CVC is reported between 0.3 and 1.5% [63–65]. Reports of CVC retention and recovery are presented in case reports, case series and retrospective reviews. These often pool the identification and recovery of all intravascular foreign objects. Embolization is common with identification in the venous system (46%), right heart (35%), and pulmonary arterial system (19%). The reported cases occur both immediately following procedures and with indwelling devices. In the case of procedure related retention, delayed identification of the foreign body is reported in 54% of cases, with the longest time to identification reported at 6 weeks [10]. Since post-procedure imaging often fails to demonstrate a retained object, the true incidence may be higher than reported. When an intravascular foreign object is identified, they are symptomatic in 5.6% of cases and an incidental finding on unrelated imaging in 37% of cases [64]. When identified, endovascular retrieval is the preferred method of recovery due to its minimally invasive approach [5, 9–11, 64, 66].

Endovascular retrieval methods have substantially improved since first reported by Thomas et al in 1964 [67]. Approaches for endovascular retrieval include fluoroscopic, CT, ultrasound, and rarely MRI guided retrieval [9]. Retrieval devices include a variety of loop snares, intravascular retrieval forceps, and retrieval baskets [9, 11, 64, 66]. A recovery rate of 86.6–94% is reported [64, 66]. One case series noted that two thirds of the non-retrieved items were related to CVC [64]. Interestingly, the rate of failed retrieval in case reports is 14.4% while only 3.7% in case series, suggesting publication bias is likely prevalent [66]. Proposed relative contraindications to recovery include small fragment size, difficult to access location, predicted potential for subsequent complication, patient's clinical status, and associated symptoms [64, 66].

Complication rates associated with retained intravascular foreign bodies vary substantially. In the pediatric literature, Chan et al described no complications in a case series of four retained CVC in pediatric patients with a median follow up of 7.5 months (range 1–53 months). Within the international literature, they noted an overall retention rate of 48.6% with no major complications reported [63]. When retention and embolism occurs in the adult literature, the rate of major complication is 71%. Major complications include persistent infection with sepsis, thrombosis, vascular occlusion, and migration into surrounding structures [63, 64, 66, 68–70]. Bacterial contamination is noted in 28% of these patients. Overall mortality associated with all intravascular foreign bodies ranges between 24 and 60% whereas the mortality rate specific to CVC related events is estimated to be much lower at 1.8% [64, 66, 71].

Attempts to identify complication risk factors and implement appropriate preventive measures prompted a recent study of intravenous retained surgical items (ivRSI) spanning 6 years [10]. In this multicenter study, 13 ivRSI were identified. Risk factors associated with ivRSI were unexpected procedural factors and equipment failure. Unexpected procedural factors were defined as blood loss >500 mL, technically difficult procedure, lack of familiarity with equipment, and difficult/emergent setting. Equipment failure included any documented malfunction/breakage of instrument, hardware, wire, or catheter during the procedure. While this study was specific to endovascular procedures, similar risk factors can be anticipated in bedside procedures (**Figure 2**).

**Figure 2.** CT-chest revealing a retained guidewire within right atrium (arrow). This was removed under local anesthetic with fluoroscopic guidance via an 8 Fr. right IJ sheath with a 2.5 cm loop snare and guiding catheter.

## **4. Conclusion**

to recovery include small fragment size, difficult to access location, predicted potential for sub-

Complication rates associated with retained intravascular foreign bodies vary substantially. In the pediatric literature, Chan et al described no complications in a case series of four retained CVC in pediatric patients with a median follow up of 7.5 months (range 1–53 months). Within the international literature, they noted an overall retention rate of 48.6% with no major complications reported [63]. When retention and embolism occurs in the adult literature, the rate of major complication is 71%. Major complications include persistent infection with sepsis, thrombosis, vascular occlusion, and migration into surrounding structures [63, 64, 66, 68–70]. Bacterial contamination is noted in 28% of these patients. Overall mortality associated with all intravascular foreign bodies ranges between 24 and 60% whereas the mortality rate specific to

Attempts to identify complication risk factors and implement appropriate preventive measures prompted a recent study of intravenous retained surgical items (ivRSI) spanning 6 years [10]. In this multicenter study, 13 ivRSI were identified. Risk factors associated with ivRSI were unexpected procedural factors and equipment failure. Unexpected procedural factors were defined as blood loss >500 mL, technically difficult procedure, lack of familiarity with equipment, and difficult/emergent setting. Equipment failure included any documented malfunction/breakage of instrument, hardware, wire, or catheter during the procedure. While this study was specific to endovascular procedures, similar risk factors can be anticipated in

**Figure 2.** CT-chest revealing a retained guidewire within right atrium (arrow). This was removed under local anesthetic

with fluoroscopic guidance via an 8 Fr. right IJ sheath with a 2.5 cm loop snare and guiding catheter.

sequent complication, patient's clinical status, and associated symptoms [64, 66].

CVC related events is estimated to be much lower at 1.8% [64, 66, 71].

bedside procedures (**Figure 2**).

162 Vignettes in Patient Safety - Volume 2

Central venous catheters represent an important clinical adjunct. Their utilization for the delivery of life saving therapies continues to expand. The complications discussed represent the most frequently reported and potentially devastating complications. With more than 15 million central venous catheter days yearly in the US, it is likely that providers will encounter complications [26]. Identified risk factors include number of unsuccessful needle passes, inexperience of provider, body mass index >30 or <20, hypovolemia, large catheter size and previously failed catheterization attempts [1, 5, 25, 26]. As a provider it is necessary to have a basic knowledge of complication management. Prompt action reduces the morbidity and mortality. In the case of a retained guide wire or embolized fragment, immediate retrieval is indicated.

## **Author details**

Maureen E. Cheung1 \*, Logan T. Mellert1 and Michael S. Firstenberg<sup>2</sup>

\*Address all correspondence to: cheungme@gmail.com

1 Department of Surgery, Western Reserve Hospital, Cuyahoga Falls, OH, United States

2 Department of Cardiovascular Surgery, Summa Health System, Akron, OH, United States

## **References**


[21] Coady K, Ali M, Sidloff D, Kenningham RR, Ahmed S. A comparison of infections and complications in central venous catheters in adults with solid tumours. The Journal of Vascular Access. 2015;**16**(1):38-41. DOI: 10.5301/jva.5000300

[8] Vannucci A, Jeffcoat A, Ifune C, Salinas C, Duncan JR, Wall M. Retained guidewires after intraoperative placement of central venous catheters. Anesthesia and Analgesia.

[9] Floridi C, Nocchi-Cardim L, De Chiara M, Ierardi AM, Carrafiello G. Intravascular foreign bodies: What the radiologist needs to know. Seminars in Ultrasound, CT, and MRI.

[10] Moffatt-Bruce SD, Ellison EC, Anderson HL, et al. Intravascular retained surgical items: A multicenter study of risk factors. The Journal of Surgical Research. 2012;**178**(1):519-

[11] Nayeemuddin M, Pherwani AD, Asquith JR. Imaging and management of complications of central venous catheters. Clinical Radiology. 2013;**68**(5):529-544. DOI: 10.1016/j.

[12] Brown P and Mitchell D. Central venous and dialysis access. In: Beard JD, Gaines PA, Loftus I, editors. Vascular & Endovascular Surgery. Edinburgh: Saunders Elsevier, 2014

[13] Seldinger SI. Catheter replacement of the needle in percutaneous arteriography; a new

[14] Graham AS, Ozment C, Tegtmeyer K, Lai S, Braner DAV. Central venous catheterization. The New England Journal of Medicine. 2007;**356**(21):e21. DOI: 10.1056/NEJMvcm055053

[15] Rhodes A, Grounds R, Bennett ED. Central venous catheter placement. In: Velasco JM.

[16] Parks RW. Practical procedures and patient investigation. In: Garden OJ. Principles &

[17] Zhong L, Wang H-L, Xu B, et al. Normal saline versus heparin for patency of central venous catheters in adult patients—A systematic review and meta-analysis. Critical Care.

[18] Pronovost PJ. Ensuring that guidelines help reduce patient harm. Journal of Oncology Practice/American Society of Clinical Oncology. 2013;**9**(4):e172-e173. DOI: 10.1200/

[19] Schiffer CA, Mangu PB, Wade JC, et al. Central venous catheter care for the patient with cancer: American Society of Clinical Oncology clinical practice guideline. Journal of

[20] Hughes JA, Cantwell CP, Waybill PN. Peripherally inserted central catheters and nontunneled central venous catheters. In: Mauro M, Murphy K, Thomson K, editors. Image-Guided Interventions: With Expert Consult; Expert Radiology Series. 2E ed. Edinburgh:

Clinical Oncology. 2013;**31**(10):1357-1370. DOI: 10.1200/JCO.2012.45.5733

Practice of Surgery. Edinburgh: Churchill Livingstone/Elsevier; 2014

2013;**117**(1):102-108. DOI: 10.1213/ANE.0b013e3182599179

2015;**36**(1):73-79. DOI: 10.1053/j.sult.2014.11.001

technique. Acta Radiologica. 1953;**39**(5):368-376

2017;**21**(5):1-9. DOI: 10.1186/s13054-016-1585-x

Saunders Elsevier, 2014; Chapter 118, 861-868.

Essential Surgical Procedures; 2016

523. DOI: 10.1016/j.jss.2012.02.053

crad.2012.10.013

164 Vignettes in Patient Safety - Volume 2

5th ed.; 2013

JOP.2012.000780


[46] Kumar USD, Shivananda S, Wali M. Malpositioned central venous catheters: A diagnostic dilemma. The Journal of the Association for Vascular Access. 2016;**21**(1):35-38. DOI: 10.1016/j.java.2015.11.002

[34] Hathi R, Patel S, Wilson P. An unusual case of pseudoaneurysm of the left brachiocephalic vein secondary to rupture during central venous catheterization treated with self-expanding, uncovered stent and coils. Clinical Radiology. 2011;**66**(1):89-92. DOI:

[35] Wallin D, Privette AR, Campbell AR, Tang JF. A case of unrecognized intrathoracic placement of a subclavian central venous catheter in a patient with large traumatic hemothorax. Case Reports in Emergency Medicine. 2015;**2015**:1-2. DOI: 10.1155/2015/382624 [36] Kuzniec S, Natal SRB, Werebe E de C, Wolosker N. Videothoracoscopic-guided management of a central vein perforation during hemodialysis catheter placement. Journal of

[37] Peasah S, McKay N, Harman J, Al-Amin M, Cook R. Medicare non-payment of hospital-acquired infections: Infection rates three years post implementation. Medicare &

[38] Hughes CJ, Ramsey-Stewart G, Storey DW. Percutaneous infraclavicular insertion of long-term central venous Hickman catheters. The Australian and New Zealand Journal of

[39] Miller LM, Clark E, Dipchand C, et al. Hemodialysis tunneled catheter-related infections. Canadian Journal of Kidney Health and Disease. 2016;**3**:1-11. DOI: 10.1177/

[40] Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheterrelated bloodstream infections in the ICU. The New England Journal of Medicine.

[41] Lorente L, Lecuona M, Jiménez A, et al. Chlorhexidine-silver sulfadiazine-impregnated venous catheters are efficient even at subclavian sites without tracheostomy. American Journal of Infection Control. 2016;**44**(12):1526-1529. DOI: 10.1016/j.ajic.2016.04.236 [42] Patel PA, Boehm S, Zhou Y, et al. Prospective observational study on central line-associated bloodstream infections and central venous catheter occlusions using a negative displacement connector with an alcohol disinfecting cap. American Journal of Infection

[43] Stuart R, Shikora S, Akerman P, et al. Incidence of arrhythmia with central venous catheter insertion and exchange. Journal of Parenteral and Enteral Nutrition. 1990;**14**(2):152-155.

[44] da Silva PSL, Waisberg J. Induction of life-threatening supraventricular tachycardia during central venous catheter placement: An unusual complication. Journal of Pediatric

[45] Pikwer A, Bååth L, Davidson B, Perstoft I, Akeson J. The incidence and risk of central venous catheter malpositioning: A prospective cohort study in 1619 patients. Anaesthesia

2006;**355**(26):2725-2732. DOI: 10.1056/NEJMoa061115

Control. 2017;**45**(2):115-120. DOI: 10.1016/j.ajic.2016.06.013

Surgery. 2010;**45**(8):e13-e16. DOI: 10.1016/j.jpedsurg.2010.05.013

DOI: 10.1177/0148607190014002152

and Intensive Care. 2008;**36**(1):30-37

Medicaid Research Review. 2013;**3**(3):E1-E16. DOI: 10.5600/mmrr.003.03.a08

Vascular Surgery. 2010;**52**(5):1354-1356. DOI: 10.1016/j.jvs.2010.05.109

10.1016/j.crad.2010.07.009

166 Vignettes in Patient Safety - Volume 2

Surgery. 1989;**59**(11):889-893

2054358116669129


[70] Reddy A, Stangl A, Radbill B. Retained catheter fragment from a fractured tunneled catheter—A rare and potentially lethal complication: Retained fractured catheter fragment. Seminars in Dialysis. 2010;**23**(5):536-539. DOI: 10.1111/j.1525-139X.2010.00756.x

[58] Hamad M, Rajan R, Kosai N, Sutton P, Das S, Harunarashid H. Retained fractured fragment of a central venous catheter: A minimally invasive approach to safe retrieval.

[59] Park K-H, Son J-W, Park W-J, et al. Characterization of the left atrial vortex flow by twodimensional transesophageal contrast echocardiography using particle image velocimetry. Ultrasound in Medicine & Biology. 2013;**39**(1):62-71. DOI: 10.1016/j.ultrasmedbio.2012.08.013

[60] Kojima S, Hiraki T, Gobara H, et al. Fracture of totally implanted central venous access devices: A propensity-score-matched comparison of risks for Groshong silicone versus polyurethane catheters. The Journal of Vascular Access. 2016;**17**(6):535-541. DOI:

[61] Braun U, Lorenz E, Weimann C, et al. Mechanic and surface properties of central-venous port catheters after removal: A comparison of polyurethane and silicon rubber materials. Journal of the Mechanical Behavior of Biomedical Materials. 2016;**64**:281-291. DOI:

[62] Viviani E, Giribono AM, Ferrara D, et al. Chest pain due to Pinch-off syndrome: Radiological findings and endovascular rescue. La Clinica Terapeutica. 2016;**167**(1):13-15

[63] Chan BKY, Rupasinghe SN, Hennessey I, Peart I, Baillie CT. Retained central venous lines (CVLs) after attempted removal: An 11-year series and literature review. Journal of

[64] Wolf F, Schernthaner RE, Dirisamer A, et al. Endovascular management of lost or misplaced intravascular objects: Experiences of 12 years. Cardiovascular and Interventional

[65] Song Y, Messerlian AK, Matevosian R. A potentially hazardous complication during central venous catheterization: Lost guidewire retained in the patient. Journal of Clinical

[66] Schechter MA, O'Brien PJ, Cox MW. Retrieval of iatrogenic intravascular foreign bodies. Journal of Vascular Surgery. 2013;**57**(1):276-281. DOI: 10.1016/j.jvs.2012.09.002

[67] Thomas J, Sinclair-Smith B, Bloomfield D, Davachi A. Non-surgical retrieval of a broken segment of steel spring guide from the right atrium and inferior vena cava. Circulation.

[68] Hassan A, Khalifa M, Al-Akraa M, Lord R, Davenport A. Six cases of retained central venous haemodialysis access catheters. Nephrology, Dialysis, Transplantation: Official Publication of the European Dialysis and Transplant Association—European Renal

[69] Jafferbhoy SF, Asquith JR, Jeeji R, Levine A, Menon M, Pherwani AD. A stuck haemodialysis central venous catheter: Not quite open and shut!. Journal of Surgical Case

Pediatric Surgery. 2013;**48**(9):1887-1891. DOI: 10.1016/j.jpedsurg.2013.01.050

Radiology. 2008;**31**(3):563-568. DOI: 10.1007/s00270-007-9201-8

Anesthesia. 2012;**24**(3):221-226. DOI: 10.1016/j.jclinane.2011.07.003

Association. 2006;**21**(7):2005-2008. DOI: 10.1093/ndt/gfl123

Reports. 2015;**2015**(4):rjv032-rjv032. DOI: 10.1093/jscr/rjv032

Ethiopian Journal of Health Sciences. 2016;**26**(1):85-88

10.5301/jva.5000606

168 Vignettes in Patient Safety - Volume 2

1964;**30**:106-108

10.1016/j.jmbbm.2016.08.002

[71] Tazzioli G, Gargaglia E, Vecchioni I, Papi S, Di Blasio P, Rossi R. Retained embolized fragment of totally implantable central venous catheter in right ventricle: It is really necessary to remove?. The Journal of Vascular Access. 2015;**16**(5):431-433. DOI: 10.5301/ jva.5000430

Provisional chapter

## **Psychometric Properties of the Hospital Survey on Patient Safety Culture (HSOPSC): Findings from Greece** Psychometric Properties of the Hospital Survey on

DOI: 10.5772/intechopen.69997

Patient Safety Culture (HSOPSC): Findings from Greece

Vasiliki Kapaki and Kyriakos Souliotis

Additional information is available at the end of the chapter Vasiliki Kapaki and Kyriakos Souliotis

http://dx.doi.org/10.5772/intechopen.69997 Additional information is available at the end of the chapter

#### Abstract

Background: Safety culture has been considered to be as one of the most crucial premises for the further development of patient safety in healthcare.

Objective: To study the psychometric properties of a translated Greek version of Hospital Survey on Patient Safety Culture (G-HSOPSC) of the Agency for Healthcare Research and Quality (AHRQ) in the Greek healthcare settings.

Methods: Factor analysis (FA) was performed to examine the applicability of the factor structure of the original questionnaire to the Greek data. In addition to the previously mentioned, internal consistency with Cronbach's coefficient alpha and construct validity was evaluated.

Results: Ten factors with 37 items were extracted by FA, with acceptable Cronbach's coefficients alpha and good construct validity. The factors jointly explained 62% of the variance in the responses. Five items were removed from the original version of the questionnaire. The composition of the factors was similar to that of the original questionnaire and five items moved to other factors. All the composites consisted of two to eight items.

Conclusions: The G-HSOPSC depicted sound psychometric properties for the evaluation of patient safety culture and therefore it is a reliable tool for use in research.

Keywords: hospital survey on patient safety culture, construct validity, reliability, internal consistency

© 2018 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

© The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and eproduction in any medium, provided the original work is properly cited.

## 1. Introduction

Safety culture has been deemed as one of the most significant premises for following improvement of patient safety in healthcare [1]. The term 'culture' is often substituted with 'climate' when questionnaire surveys are utilized to assess an organization's culture. The definition of 'safety culture' derives from the nuclear power industry and has been transferred to the field of the healthcare: 'the safety culture of an organization is the product of individual and group values, attitudes, perceptions, competencies and patterns of behaviour that determine the commitment to, and the style and proficiency of, an organization's safety management' [2]. Safety climate can be faced as the superficial features of the underlying safety culture [3]. It assesses workforce perceptions of procedures and behaviours in their work environment that point out the priority given to safety relative to other organizational goals [4, 5].

Therefore, individual and self-administered questionnaires allow measuring an organization's safety climate [6–9] while for assessing safety culture, other types of assessment (i.e. 'interviews, on-site observations, focus groups') are more suitable [10–12]. These questionnaires are distributed to a group of professionals that operate in the healthcare field having an aim to provide information on aspects of the organizational culture underlying active failures and latent conditions that have to be addressed by patient safety initiatives [13].

Most of the available tools were developed in the United States (US) but some researchers suggest that various US tools cannot be adapted to European context. For this exact reason, after translating a questionnaire into another language and applying it in a different setting, it is of crucial importance to validate it before extending its use to populations differentiating from the specific geographical and healthcare contexts for which it was initially developed. The psychometric techniques are commonly used in order to ensure potential users that tools will be a good predictor of safety events and provide actionable information [9].

## 2. Clinical vignette

M.G., a 75-year-old woman with stage four chronic kidney disease (CKD), hypertension and gout was admitted for a total knee replacement under the orthopedic team. According to the routine renal biochemistry results and following advice from nephrology group, she was on a low dose of an activated vitamin D analogue. She was also taking a diuretic, an angiotensin receptor blocker, aspirin, sodium bicarbonate and a statin drug. Serum calcium was not verified again during her admission. Even though discharge communication included the recently started medication with the advised dose, no advice was given to the general practitioner about the required rate of monitoring serum calcium and renal function post discharge and the patient was not duly informed of the necessity for this monitoring. At her home, the patient made a slow recovery from her operation and had limited ability to move around. Her son phoned for the surgery and requested a general practitioner to make a visit to his mother 3 weeks after discharge, as she looked very sick, was more and more confused and was not consuming food or water. The general practitioner arranges for the patient to be re-admitted

into the hospital. The last diagnosis was stage 2, acute kidney injury (AKI), second in importance, iatrogenic hypercalcaemia and dehydration.

#### 2.1. Key learning points

1. Introduction

172 Vignettes in Patient Safety - Volume 2

goals [4, 5].

2. Clinical vignette

Safety culture has been deemed as one of the most significant premises for following improvement of patient safety in healthcare [1]. The term 'culture' is often substituted with 'climate' when questionnaire surveys are utilized to assess an organization's culture. The definition of 'safety culture' derives from the nuclear power industry and has been transferred to the field of the healthcare: 'the safety culture of an organization is the product of individual and group values, attitudes, perceptions, competencies and patterns of behaviour that determine the commitment to, and the style and proficiency of, an organization's safety management' [2]. Safety climate can be faced as the superficial features of the underlying safety culture [3]. It assesses workforce perceptions of procedures and behaviours in their work environment that point out the priority given to safety relative to other organizational

Therefore, individual and self-administered questionnaires allow measuring an organization's safety climate [6–9] while for assessing safety culture, other types of assessment (i.e. 'interviews, on-site observations, focus groups') are more suitable [10–12]. These questionnaires are distributed to a group of professionals that operate in the healthcare field having an aim to provide information on aspects of the organizational culture underlying active failures and

Most of the available tools were developed in the United States (US) but some researchers suggest that various US tools cannot be adapted to European context. For this exact reason, after translating a questionnaire into another language and applying it in a different setting, it is of crucial importance to validate it before extending its use to populations differentiating from the specific geographical and healthcare contexts for which it was initially developed. The psychometric techniques are commonly used in order to ensure potential users that tools

M.G., a 75-year-old woman with stage four chronic kidney disease (CKD), hypertension and gout was admitted for a total knee replacement under the orthopedic team. According to the routine renal biochemistry results and following advice from nephrology group, she was on a low dose of an activated vitamin D analogue. She was also taking a diuretic, an angiotensin receptor blocker, aspirin, sodium bicarbonate and a statin drug. Serum calcium was not verified again during her admission. Even though discharge communication included the recently started medication with the advised dose, no advice was given to the general practitioner about the required rate of monitoring serum calcium and renal function post discharge and the patient was not duly informed of the necessity for this monitoring. At her home, the patient made a slow recovery from her operation and had limited ability to move around. Her son phoned for the surgery and requested a general practitioner to make a visit to his mother 3 weeks after discharge, as she looked very sick, was more and more confused and was not consuming food or water. The general practitioner arranges for the patient to be re-admitted

latent conditions that have to be addressed by patient safety initiatives [13].

will be a good predictor of safety events and provide actionable information [9].


## 3. Methods

### 3.1. HSOPSC measurement tool

The self-administered HSOPSC tool was developed by the US Agency for Healthcare Research and Quality (AHRQ). The HSOPSC tool assesses safety climate from the staff perspective and covers 7 unit-level composites (24 items) of safety climate, 3 hospital-level composites (11 items) and 4 outcome variables. Table 1 depicts the characteristics of the specific measurement tool [14].

HSOPSC was selected as the tool for testing for several reasons: (a) Organizations can use the tool to assess their patient safety culture, track changes over time and evaluate the impact of patient safety interventions [15]. (b) It had been designed for surveying all hospital personnel (clinical/non-clinical) [14]. (c) It was considered one of the few healthcare safety climate instruments for which initial psychometric results had been reported [6, 7]. (d) Benchmark statistics of HSOPSC can be retrieved from the internet [16]. (e) The questionnaire has been translated into 27 different languages and it is currently used in 59 countries [17]. To use the specific tool will allow for future international comparisons.

#### 3.2. Translation process: pre-test

Firstly, permission was obtained from the authors to use HSOPSC. It was translated into Greek language and then translated back into English by two independent researchers to ensure validity of the translation. In the translation process, it was stressed that the same meaning and 'strength' should be reproduced in the translation into the Greek language. In order to test if respondents understood the meaning of all items, HSOPSC was pilot tested in a group of 35 healthcare professionals which was not incorporated to the final sample. The overall Cronbach's alpha of the pre-test was 0.87.


Statistical analysis such as item analysis, exploratory factor analysis, confirmatory factor analysis and correlated composites scores across elements were performed to evaluate psychometric properties. It has a solid content validity and has been validated in all levels. FA resulted in 12 factors.

Table 1. Characteristics of the HSOPSC measurement tool.

#### 3.3. Sample

The study was carried out in 12 Greek hospitals over the period from May 2014 to November 2014. The participating hospitals included nine general hospitals, one of them is a teaching hospital, and three specialty hospitals (1 anticancer-oncology hospital, 1 psychiatric hospital and 1 cardiac surgery centre). The HSOPSC was originally designed for application to all hospital professionals [14]. However, the pre-test showed that items dealing with direct patient care could often not be answered by staff not involved directly in patient care (i.e. hospital managers, administrators). Consequently, the survey was returned by 820 participants (response rate = 59.6%), 10 questionnaires in which fewer than half the items were answered were also excluded. Finally, 810 questionnaires were retained for further analysis.

#### 3.4. Statistical analysis

Factor analysis (FA) clarifies the items which are in depth connected and allude in collaboration to a below composite (or factor). Therefore, the items are able to be lessened to the smallest potential number of understandings that as before make the largest potential part of the variance clear [18]. A FA was carried out (principal component analysis with varimax rotation) for the purpose of proving that the current scales/dimensions may be fairly employed within the Greek context. When proving the number of elements, the Eigen value (Eigen value > 1: Kaiser's criterion) was taken into consideration, in comparison with the range of explained variance, the shape of the screen plot and the future outcome of interpreting the elements. Kaiser's criterion is trustworthy in a specimen of more than 250 respondents and when the average communality adds up to or is larger than, 0.6. The figure of the screen plot supplies dependable knowledge when the sample is larger than 200 respondents [18]. The data fulfil the requirements.

The Kaiser-Meyer-Olkin (KMO) calculation of sampling appropriateness was ascertained. This value is able to fluctuate from 0 to 1. A value near 1 points out that there is just any diffusion in the correlation pattern, empowered trustworthy and unique elements by FA [18]. The KMO score was 0.9, not close to Kaiser's standard of 0.5.

Additionally, the writers confirmed whether the inter-item correlations were adequate, by a test of the correlation matrix. Queries are a member of the common underlying composite, which will be related as they calculate the identical feature of patient safety culture. Objectives that are not related, or correlate with only a few other variables, are not compatible with FA [18]. Bartlett's test of sphericity illustrated that the inter-item correlations were adequate: (x<sup>2</sup> = 12,190, df = 861, p < 0.001).

Last but not least, the writers confirmed whether the contrary existed: too much connection between the items. According to an ideal, each feature of patient safety culture exclusively is responsible for the patient safety culture. An important connection between two items signifies that patient safety culture aspects cross each other to a comprehensive range. The amount overlapped in the answer patterns is about 50% when a connection is 0.7 [18]. No connections surpassed the specific boundary score. The pre-analyses depict that the data could be employed for FA.

The construct validity was accomplished by determining scale scores for each factor (after any essential opposite coding) and next measuring Pearson correlation coefficients (r) between the scale scores. The construct validity of each factor is revealed in scale scores that are reasonably connected. Despite this, strong correlations (r > 0.7) would point out that factors calculate the identical concept and the above factors may be joined and/or a few objectives could be taken out. Also, connections of the scale scores were measured with the outcome variable 'Patient safety grade'. No connections were measured with the other outcome variable, 'Number of events reported', due to the shortage of variability and distorted type of the specific item (40.1% of the respondents pointed out not to have reported any events during the past 12 months and 35% had reported only one or two events).

3.3. Sample

3.4. Statistical analysis

The study was carried out in 12 Greek hospitals over the period from May 2014 to November 2014. The participating hospitals included nine general hospitals, one of them is a teaching hospital, and three specialty hospitals (1 anticancer-oncology hospital, 1 psychiatric hospital and 1 cardiac surgery centre). The HSOPSC was originally designed for application to all hospital professionals [14]. However, the pre-test showed that items dealing with direct patient care could often not be answered by staff not involved directly in patient care (i.e. hospital managers, administrators). Consequently, the survey was returned by 820 participants (response rate = 59.6%), 10 questionnaires in which fewer than half the items were answered

Factor analysis (FA) clarifies the items which are in depth connected and allude in collaboration to a below composite (or factor). Therefore, the items are able to be lessened to the smallest potential number of understandings that as before make the largest potential part of the variance

were also excluded. Finally, 810 questionnaires were retained for further analysis.

Characteristics HSOPSC measurement tool Writers and date of development Sorra and Nieva, 2004

Scale On a 5-point Likert scale

Dimensions/elements 1. Communication openness

Psychometric evaluation 1. Sufficient psychometric properties

Objective To empower hospitals to evaluate their patient safety culture

4. Staffing

3. No punitive response to error

6. Teamwork within units 7. Teamwork across hospital units

3. Tested on large specimen

5. Hospital management support for patient safety

8. Organizational learning—continuous improvement 9. Feedback and communication about error 10. Hospital handoffs and transitions 11. Overall perception of patient safety 12. Frequency of event reporting 13. Overall patient safety grade

14. Number of events reported in the past 12 months

2. Cronbach's alpha range from 0.63 to 0.84

Statistical analysis such as item analysis, exploratory factor analysis, confirmatory factor analysis and correlated composites scores across elements were performed to evaluate psychometric properties. It has a solid content validity and has

2. Supervisor/manager expectations and actions promoting patient safety

Country USA

Setting Hospital Staff Health care staff

been validated in all levels. FA resulted in 12 factors.

Table 1. Characteristics of the HSOPSC measurement tool.

Number of items 44

174 Vignettes in Patient Safety - Volume 2

Cronbach's alpha was determined to examine the internal consistency of composites. It is expressed as a number between 0 and 1. In case that separate items are considered to calculate the identical concept, the internal consistency (reliability) should be greater than or equivalent to 0.6 [18]. To the reason that the form with questions composed of in a positive and negative way phrased items, the negative ones were made an entry in first reason, due to ensure that a higher score every time signifies a more affirmative reply. Statistical analysis was carried out using the IBM SPSS 21.0 (IBM Corp. Released 2012. IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY: IBM Corp.).

## 4. Results

#### 4.1. Sample

Most respondents were nursing staff (45.7%), followed by physicians (25.4%), nurse/unit assistants (17.9%) and physical/occupational/speech therapists (3.8%). These percentages give a reasonable reflection of the real distribution of disciplines at the units (Table 2).


Psychometric Properties of the Hospital Survey on Patient Safety Culture (HSOPSC): Findings from Greece http://dx.doi.org/10.5772/intechopen.69997 177


Table 2. Respondents—hospital characteristics.

4. Results

176 Vignettes in Patient Safety - Volume 2

4.1. Sample

Most respondents were nursing staff (45.7%), followed by physicians (25.4%), nurse/unit assistants (17.9%) and physical/occupational/speech therapists (3.8%). These percentages give

Characteristics Category N (%) Hospital type General hospital 501 (61.9)

Hospital size (beds) 100–250 642 (79.3)

Work area/unit Many different hospital units/no specific unit 166 (20.5)

Staff position Resident physicians 110 (13.6)

Location of hospital Central hospitals 9 (75)

Anticancer-oncology hospital 110 (13.6) Psychiatric hospital 132 (16.3) Cardiac surgery centre 67 (8.3)

400 or more 168 (20.7)

Medicine (non-surgical) 166 (20.5) Surgery 204 (25.2) Emergency department 12 (1.5) Intensive care unit (any type) 49 (6) Laboratory 38 (4.7) Psychiatry/mental health 117 (14.4) Rehabilitation 11 (1.4) Pharmacy 1 (0.1) Social services department 19 (2.3) Other 27 (3.4)

Specialist physicians 95 (11.8) Nurses (university training) 49 (6.1) Nurses (technological education institute training) 320 (39.6) Nurse assistants 136 (16.8) Unit assistants 9 (1.1) Physical/occupational/speech therapists 31 (3.8) Psychologists 5 (0.6) Welfare workers 26 (3.2) Pharmacy staff 1 (0.1) Other 26 (3.2)

Peripheral hospitals 3 (25)

a reasonable reflection of the real distribution of disciplines at the units (Table 2).

#### 4.2. FA: internal consistency

Ten factors were drawn by FA with 37 items. All the items of 'Hospital handoffs and transitions' (F3r, F5r, F7r, F11r) blended into the factor 'Teamwork across hospital units'. Two of the items of 'Feedback and communication about errors' (C3, C5) from the US version blended into the factor 'Communication openness'. A new factor originated, which comprised four items from the original questionnaire (B3r, B4r, A7r, A10r). The factors of 'Non-punitive response to error', 'Hospital management support for patient safety' and 'Frequency of event reporting' from the American study remained stable to the G-HSOPSC. The overall Cronbach's coefficient alpha for the G-HSOPSC was 0.91. Seven out of 10 factors in the G-HSOPSC had Cronbach's coefficients alpha > 0.70 and three factors had values between 0.60 and 0.70, which indicate fairly good internal consistency of the Greek version of the questionnaire (Table 3).



a The codes in items' column refer to the sections in the questionnaire and the numbers of the questions.

\* Some of the items of the American factors 'Feedback and communication about errors', 'Hospital handoffs and transitions' and 'Overall perceptions for safety' assimilated to other factors and other items removed from the questionnaire.

Table 3. Cronbach's alpha and characteristics of the factors after factor analysis.

Five items (A11, A15, A17r, A18, C1) did not have a sufficient factor loading on any of the factors (all loadings < 0.50) and were eliminated. Table 4 gives the mean scores with standard deviations and factor loadings per item. The factors jointly explained 62% of the variance in the responses (Table 4).

#### 4.3. Construct validity: inter-correlations

4.2. FA: internal consistency

178 Vignettes in Patient Safety - Volume 2

Composite Itemsa Cronbach's

Unit-level 1. Supervisor/ manager expectations and actions promoting

safety

2. Organizational learning continuous improvement

3. Teamwork within units

4. Communication openness

5. Feedback and communication about errors

6. Non-punitive response to error

Hospital-level 8. Hospital management support for patient safety

7. Staffing A2, A5r, A7r,

A14r

Ten factors were drawn by FA with 37 items. All the items of 'Hospital handoffs and transitions' (F3r, F5r, F7r, F11r) blended into the factor 'Teamwork across hospital units'. Two of the items of 'Feedback and communication about errors' (C3, C5) from the US version blended into the factor 'Communication openness'. A new factor originated, which comprised four items from the original questionnaire (B3r, B4r, A7r, A10r). The factors of 'Non-punitive response to error', 'Hospital management support for patient safety' and 'Frequency of event reporting' from the American study remained stable to the G-HSOPSC. The overall Cronbach's coefficient alpha for the G-HSOPSC was 0.91. Seven out of 10 factors in the G-HSOPSC had Cronbach's coefficients alpha > 0.70 and three factors had values between 0.60 and 0.70, which indicate fairly good internal consistency of the Greek version of the questionnaire (Table 3).

> Cronbach's α Greek data

Composite Itemsa Cronbach's α

B1, B2 0.84

A9, A13 0.60

A1, A3, A4, A6 0.80

0.77

0.71

0.60

0.79

C2, C4 C6r, C3 C5

A8r, A12r A16r

A14r

F1, F8 F9r

supervisor/manager expectations and actions promoting

units—continuous improvement

communication openness about errors

response to error

management support for patient safety

0.63 0.51 6. Sufficient staffing A2, A5r

C1, C3, C5 0.78 0.74 \* \*\*

safety

learning

HSOPSC factor analysis G-HSOPSC factor analysis

α American data

B1, B2, B3r, B4r 0.75 0.70 1. Competent

A6, A9, A13 0.76 0.49 2. Organizational

A1, A3, A4, A11 0.83 0.61 3. Teamwork within

C2, C4, C6r 0.72 0.62 4. Feedback and

A8r, A12r, A16r 0.79 0.71 5. Non-punitive

F1, F8, F9r 0.83 0.79 7. Hospital

For each of the 10 factors, scale scores were calculated by obtaining the mean of the item scores within one factor for every respondent. Immediately after, the mono-item outcome variable 'Patient safety grade' has been determined with the connections of the scales. The factors were anticipated to be related in a positive way with the specific outcome measure. Every one of connections with 'Patient safety grade' was important. With the 'Teamwork across hospital units and handoffs & transitions', the most significant correlation of this outcome was measured (r = 0.49). Moreover, correlations between the scale scores were calculated. The highest correlation was between 'Hospital management support for patient safety' and 'Teamwork across hospital units and handoffs & transitions' (r = 0.52) but no correlation was exceptionally high (Table 5).


Item Factors

180 Vignettes in Patient Safety - Volume 2

3.62 0.95 0.823

3.67 0.91 0.805

3.12 0.94 0.772

3.29 0.91 0.597

3.52 0.91 0.802

3.71 0.89 0.711

3.50 0.89 0.778

3.98 0.75 0.618

3.78 0.92 0.695

3.75 0.95 0.645

2.77 0.96 0.687

3.66 0.89 0.626

3.62 0.99 0.604

2.35 0.94 0.753

2.68 1.00 0.699

B1. My supervisor/manager says a good word when he/she sees a job done according to established patient safety

B2. My supervisor/manager seriously considers staff suggestions for improving

A9. Mistakes have led to positive changes here

A13. After we make changes to improve patient safety, we evaluate their effectiveness

A1. People support one another

A3. When a lot of work needs to be done quickly, we work together as a team to get the

A4. In this unit, people treat each other with respect

A6. We are actively doing things to improve patient

C2. Staff will freely speak up if they see something that may negatively affect patient care

C3. We are informed about errors that happen in this unit

C4. Staff feel free to question the decisions or actions of those

C5. In this unit, we discuss ways to prevent errors from

C6r. Staff are afraid to ask questions when something does not seem right (reverse

A8r. Staff feel as if their mistakes are held against them

A12r. When an event is reported, it feels like the person is being written up, not the problem (reverse worded)

(reverse worded)

with more authority

happening again

worded)

procedures

patient safety

in this unit

work done

safety

Mean SD 1 2 3 4 5 6 7 8 9 10



Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a a Rotation converged in seven iterations.

Table 4. Mean scores and factor loadings of the items regarding patient safety culture.

Psychometric Properties of the Hospital Survey on Patient Safety Culture (HSOPSC): Findings from Greece http://dx.doi.org/10.5772/intechopen.69997 183


Correlations

Item Factors

D1. When a mistake is made, but is caught and corrected before affecting the patient, how often is this reported?

182 Vignettes in Patient Safety - Volume 2

D2. When a mistake is made, but has no potential to harm the patient, how often is this

D3. When a mistake is made that could harm the patient, but does not, how often is this

A7r. We use more agency/ temporary staff than is best for

A10r. It is just by chance that more serious mistakes do not happen around here.

B3r. Whenever pressure builds up, my supervisor/manager wants us to work faster, even if it means taking shortcuts. (reverse worded)

B4r. My supervisor/manager overlooks patient safety problems that happen over and over. (reverse worded)

A11. When one area in this unit gets really busy, others help

A15. Patient safety is never sacrificed to get more work

A17r. We have patient safety problems in this unit. (reverse

A18. Our procedures and systems are good at preventing errors from happening.

C1. We are given feedback about changes put into place based on event reports.

Rotation converged in seven iterations.

reported?

reported?

patient care.

out.

done.

worded)

a

Mean SD 1 2 3 4 5 6 7 8 9 10

3.41 1.11 0.788

3.05 1.13 0.881

3.17 1.19 0.808

3.44 1.02 0.571

3.24 1.12 0.505

3.41 1.01 0.596

3.91 0.94 0.656

2.30 1.11 0.29

4.12 0.81 0.41

3.07 0.97 0.46

3.06 0.98 0.44

3.24 1.07 0.49

Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a

Table 4. Mean scores and factor loadings of the items regarding patient safety culture.


Table 5. Mean factor scores, correlations with patient safety grade and inter-correlations of the 10 composites.

## 5. Discussion

Correlations

Factor

Patient

1

 2

 3

 4

 5

 6

 7

 8

 9

 10

184 Vignettes in Patient Safety - Volume 2

safety

grade

<0.001

810

 793

 788

 794

 786

 784

 797

 797

 775

 797

 756

 0.003 <0.001 0.009 <0.001 <0.001

 <0.001 <0.001 0.019 <0.001

Sig. (two-tailed)

N Mean (SD)

Pearson r Sig. (two-tailed)

N Mean (SD)

Pearson r Sig. (two-tailed)

N Mean (SD)

Pearson r Sig. (two-tailed)

N Mean (SD)

Pearson r Sig. (two-tailed)

N

> NA: non applicable.

\*Correlation

\*\*Correlation

Table 5. Mean factor scores, correlations

 is significant at the 0.01 level

 is significant at the 0.05 level

(two-tailed).

(two-tailed).

 with patient safety grade and

inter-correlations

 of the 10 composites.

0.38

<0.001

810

 767

 759

 763

 757

 755

 756

 767

 748

 767

 767

 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001

 0.31\*\*

0.26\*\*

0.39\*\*

0.33\*\*

0.29\*\*

0.21\*\*

0.31\*\*

0.42\*\*

0.26\*\* 1

 3.49 (0.70)

10. Adequate procedures

safety

 and systems for

0.30

<0.001

810

 806

 800

 806

 798

 795

 797

 810

 787

 810

 767

 <0.001 <0.001 <0.001 <0.001 <0.001 0.019 0.002 <0.001

 0.16\*\*

0.17\*\*

0.14\*\*

0.36\*\*

0.13\*\*

0.08\*

0.11\*\*

0.15\*\* 1

 0.26\*\*

 <0.001

 3.20 (0.97)

9. Frequency of event reporting

0.49

<0.001

810

 784

 780

 783

 777

 772

 775

 787

 787

 787

 748

 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001

 0.27\*\*

0.37\*\*

0.39\*\*

0.39\*\*

0.29\*\*

0.30\*\*

0.52\*\* 1

 0.15\*\*

 <0.001 <0.001

0.42\*\*

 3.13 (0.69)

8. Teamwork handoffs & transitions

 across hospital units and

0.44

<0.001

810

 806

 800

 806

 798

 795

 797

 810

 787

 810

 767

 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001

 0.29\*\*

0.37\*\*

0.29\*\*

0.25\*\*

0.22\*\*

0.30\*\* 1

 0.52\*\*

 <0.001 0.002 <0.001

0.11\*\*

0.31\*\*

 2.86 (0.87)

7. Hospital patient safety

management

 support for

Cultural and healthcare differences in terms of context between US and Greece set obvious that reproduction of HSOPSC would be meaningful in Greek hospital settings. The available evidence from studies which were conducted in European and non-European countries—such as Norway [19]; Sweden [20]; Slovenia [21]; the West Bank [22]; Iran [23]; Scotland [24]; the United Kingdom [25]; the Netherlands [26]; Norway [15]; Switzerland [26] and Belgium [27]—suggests that the HSOPSC developed based on the original US version should be cautiously adjusted to other healthcare contexts. In Switzerland, for instance [26], the use of agency staff in nursing is currently relatively uncommon. Moreover, the role of hospital management and the way it is organized presents differences between hospital types and national or regional regulations. Consequently, taking into account the relative published studies, the number of composites varied between 8 and 15 and included 27 to 50 items.

This is the first study which was conducted in Greece which reports the structure as well as the psychometric properties of G-HSOPSC in accordance with the guidelines of the AHRQ. Despite the fact that our results are aligned with the original version, some adaptations were demanded so that the Greek context is fitted correctly. A 10-factor model with 37 items performed better than the original one in the sample of the 12 Greek hospitals. The main difference was that the composite 'Teamwork across hospital units' merged with 'Hospital handoffs and transitions' and 'Communication openness' merged with 'Feedback and communication about error' except an item (C1). The studies [21, 28, 29] showed the same conflations. The items B3r and B4r, A7, A10r loaded slightly more on a new composite which was named 'Adequate procedures and systems for safety' instead of 'Supervisor/Manager expectations & actions promoting safety', 'Staffing', 'Overall perceptions for safety', respectively. Last but not least, the item A6 loaded slightly more on 'Teamwork within units' instead of 'Organizational learning—continuous improvement' which renamed the first one as 'Teamwork within units continuous improvement'.

Finally five items (A11, A15, A17r, A18, C1) of the original questionnaire were removed. Three of them (A11, A15, C1) have been eliminated from the Arabic, Dutch and French version, respectively too [22, 26, 28]. Ten underlying factors offered 62% of the variance of the items. The originally proposed 12 safety culture composites had explained 64.5% of the variance in the US version [14] and 57.1% and 59.8% in the Dutch adaptation and German version, respectively [26, 29].

If the factor structures of the various applications of the HSOPSC in Europe are compared to the original pilot tested US version, most of the composites presented similar patterns in the Cronbach's alpha. The internal consistency of G-HSOPSC ranged between 0.60 and 0.88 with lowest Cronbach´s alpha values for 'Organizational learning' and 'Sufficient staffing' (both α = 0.60). These findings have also been presented in other studies [25, 26, 29, 30]. As far as the present study is concerned, our belief is that these composites and items should be kept since they signify important aspects of patient safety and as such shape a useful foundation for improvement work.

Correlations among the 10 safety culture composites varied from 0.08 to 0.52 (p < 0.01). These correlations are deemed satisfactory and do not indicate problematic associations among dimensions. 'Patient Safety Grade' showed its highest correlations with 'Teamwork across hospital units and handoffs & transitions' (r = 0.49). 'Frequency of events reported' has actually only a small interrelationship with the other safety culture sub-dimensions (the highest with 'Feedback and communication openness about error', r = 0.36). The above results underline the crucial role of the hospital procedures in developing a cooperative and communication openness environment that cultivates free process of evaluation about the adverse events, sharing data about the errors that take place, discussing the way to prevent adverse events and reporting the identified errors. As data indicate an aftermath of that environment will lead to a frequency of event report and improved patient safety grade [31]. Finally, the highest intercorrelation was between 'Hospital management support for patient safety' and 'Teamwork across hospital units and handoffs & transitions' (r = 0.52). Considering that both composites share some attention towards transference of important patient care information, this outcome was not considered as surprising; although these composites share a common meaning, they were not integrated into one concept.

#### 5.1. Strengths and limitations of the study

The main strength of the study is the heterogeneity of the selected healthcare facilities. The sample was opted from different types of hospitals in order to capture a more comprehensive view of perceptions towards patient safety culture because the studies which have been published show that the patient safety culture composites may vary among different types of healthcare settings [32].On the other hand, the study has some limitations. Firstly, selection bias might have occurred as hospitals were selected on a voluntary basis and as head nurses were responsible for distributing the questionnaires. It is possible that head nurses chose not to include some healthcare professionals. Secondly, the relatively lower internal consistency of some scales (i.e. organizational learning, sufficient staffing) than that of the original AHRQ data consist another cause. Further studies are needed to investigate the possible association between certain composites and their items. Thirdly, the difficulty of achieving high response rates among hospital professionals, which was thought to be the most practical challenge after conducting this study.

## 6. Conclusion

The G-HSOPSC is suitable for clinical and research purposes and allows clinicians and researchers to make cross-national comparisons. Healthcare managers could benefit from using the G-HSOPSC for benchmarking when improving hospital patient safety culture in general and at the same time to obtain knowledge about specific areas of improvement (i.e. shift-working, staffing and over-occupancy). Examination of patient safety culture differences between staff groups and factors affecting patient safety culture is also a term of need in order to obtain knowledge of areas in order to take action to improve safety.

## Acknowledgements

Correlations among the 10 safety culture composites varied from 0.08 to 0.52 (p < 0.01). These correlations are deemed satisfactory and do not indicate problematic associations among dimensions. 'Patient Safety Grade' showed its highest correlations with 'Teamwork across hospital units and handoffs & transitions' (r = 0.49). 'Frequency of events reported' has actually only a small interrelationship with the other safety culture sub-dimensions (the highest with 'Feedback and communication openness about error', r = 0.36). The above results underline the crucial role of the hospital procedures in developing a cooperative and communication openness environment that cultivates free process of evaluation about the adverse events, sharing data about the errors that take place, discussing the way to prevent adverse events and reporting the identified errors. As data indicate an aftermath of that environment will lead to a frequency of event report and improved patient safety grade [31]. Finally, the highest intercorrelation was between 'Hospital management support for patient safety' and 'Teamwork across hospital units and handoffs & transitions' (r = 0.52). Considering that both composites share some attention towards transference of important patient care information, this outcome was not considered as surprising; although these composites share a common meaning, they

The main strength of the study is the heterogeneity of the selected healthcare facilities. The sample was opted from different types of hospitals in order to capture a more comprehensive view of perceptions towards patient safety culture because the studies which have been published show that the patient safety culture composites may vary among different types of healthcare settings [32].On the other hand, the study has some limitations. Firstly, selection bias might have occurred as hospitals were selected on a voluntary basis and as head nurses were responsible for distributing the questionnaires. It is possible that head nurses chose not to include some healthcare professionals. Secondly, the relatively lower internal consistency of some scales (i.e. organizational learning, sufficient staffing) than that of the original AHRQ data consist another cause. Further studies are needed to investigate the possible association between certain composites and their items. Thirdly, the difficulty of achieving high response rates among hospital professionals, which was thought to be the most practical challenge after

The G-HSOPSC is suitable for clinical and research purposes and allows clinicians and researchers to make cross-national comparisons. Healthcare managers could benefit from using the G-HSOPSC for benchmarking when improving hospital patient safety culture in general and at the same time to obtain knowledge about specific areas of improvement (i.e. shift-working, staffing and over-occupancy). Examination of patient safety culture differences between staff groups and factors affecting patient safety culture is also a term of need in order

to obtain knowledge of areas in order to take action to improve safety.

were not integrated into one concept.

186 Vignettes in Patient Safety - Volume 2

conducting this study.

6. Conclusion

5.1. Strengths and limitations of the study

The authors would like to extend their gratitude and appreciation to the administration of the 12 hospitals and to their staff who devoted time to complete the survey. Furthermore, the authors did not receive any kind of funding for the research and they have no competing interests.

## Ethical approval

The research complied with every one of the dilemmas concerned with ethics. The privacy and secrecy or anonymity of employees were made certain.

## Abbreviations


## Author details

Vasiliki Kapaki<sup>1</sup> \* and Kyriakos Souliotis<sup>2</sup>


## References

[1] Institute of Medicine Committee on Quality of Health Care in America. In: Kohn LT, Corrigan JM, Donaldson MS, editors. To Err is Human: Building a Safer Health System. Washington (DC): National Academies Press (US); 2000. Executive Summary. Available from: https://www.ncbi.nlm.nih.gov/books/NBK225179/


[19] Cappelen K, et al. Psychometric properties of the nursing home survey on patient safety culture in Norwegian nursing homes. BMC Health Services Research. 2016;16(1):446

[2] International Atomic Energy Agency Safety Culture and I.S.A. Group. Safety Series 75-

[3] Flin R, et al. Measuring safety climate: Identifying the common features. Safety Science.

[4] Zohar D. Safety climate in industrial organizations: Theoretical and applied implications.

[5] Zohar D, et al. Healthcare climate: A framework for measuring and improving patient

[6] Colla JB, et al. Measuring patient safety climate: A review of surveys. Quality and Safety

[7] Flin R, et al. Measuring safety climate in health care. Quality & Safety in Health Care.

[8] Nieva VF, Sorra J. Safety culture assessment: A tool for improving patient safety in healthcare organizations. Quality and Safety in Health Care. 2003;12(Suppl 2):ii17-23 [9] Singla AK, et al. Assessing patient safety culture: A review and synthesis of the measure-

[10] Guldenmund FW. The nature of safety culture: A review of theory and research. Safety

[12] Hutchinson A, et al. Use of a safety climate questionnaire in UK health care: Factor structure, reliability and usability. Quality & Safety in Health Care. 2006;15(5):347-353

[14] Sorra JS, Nieva VF. Hospital survey on patient safety culture. Prepared by Westat, under Contract No. 290-96-0004. in AHRQ publication no. 04-0041. USA: AHRQ; 2004

[15] Olsen E. Reliability and Validity of the Hospital Survey on Patient Safety Culture at a Norwegian Hospital, in Quality and Safety Improvement Research: Methods and Research Practice, National School of Public Health. Lisbon: International Quality Improvement

[16] AHRQ. Comparing your Results: Preliminary Benchmarks: Hospital Survey on Patient

[17] Agency for Healthcare Research and Quality (AHRQ). International Use of the Surveys on Patient Safety Culture. Agency for Healthcare Research and Quality, Rockville, MD; 2014. Available from: http://www.ahrq.gov/professionals/quality-patient-safety/patientsa

[18] Field AP. Discovering Statistics Using SPSS for Windows: Advanced Techniques for the

INSAG-4. Vienna: International Atomic Energy Agency; 1991

Journal of Applied Psychology. 1980;65(1):96-102

safety. Critical Care Medicine. 2007;35(5):1312-1317

ment tools. Journal of Patient Safety. 2006;2:105-115

Research Network (QIRN); 2008. pp. 173-186

Beginner. London: Sage Publications; 2000

Safety Culture. USA: AHRQ; 2006

fetyculture/pscintusers.html

[11] Hale AR. Culture's confusions. Safety Science. 2000;34(1-3):1-14

[13] Reason J. Human Error. New York, NY: Cambridge University Press; 1990

in Health Care. 2005;14(5):364-366

Science. 2000;34(1-3):215-257

2000:34(1-3):177-192

188 Vignettes in Patient Safety - Volume 2

2006;15(2):109-115


## *Edited by Michael S. Firstenberg and Stanislaw P. Stawicki*

Over the past two decades, the healthcare community increasingly recognized the importance and the impact of medical errors on patient safety and clinical outcomes. Medical and surgical errors continue to contribute to unnecessary and potentially preventable morbidity and/or mortality, affecting both ambulatory and hospital settings. The spectrum of contributing variables—ranging from minor errors that subsequently escalate to poor communication to lapses in appropriate protocols and processes (just to name a few)—is extensive, and solutions are only recently being described. As such, there is a growing body of research and experiences that can help provide an organized framework—based upon the best practices and evidencebased medical principles—for hospitals and clinics to foster patient safety culture and to develop institutional patient safety champions. Based upon the tremendous interest in the first volume of our Vignettes in Patient Safety series, this second volume follows a similar vignettebased model. Each chapter outlines a realistic case scenario designed to closely approximate experiences and clinical patterns that medical and surgical practitioners can easily relate to. Vignette presentations are then followed by an evidence-based overview of pertinent patient safety literature, relevant clinical evidence, and the formulation of preventive strategies and potential solutions that may be applicable to each corresponding scenario. Throughout the Vignettes in Patient Safety cycle, emphasis is placed on the identification and remediation of team-based and organizational factors associated with patient safety events. The second volume of the Vignettes in Patient Safety begins with an overview of recent high-impact studies in the area of patient safety. Subsequent chapters discuss a broad range of topics, including retained surgical items, wrong site procedures, disruptive healthcare workers, interhospital transfers, risks of emergency department overcrowding, dangers of inadequate handoff communication, and the association between provider fatigue and medical errors. By outlining some of the current best practices, structured experiences, and evidencebased recommendations, the authors and editors hope to provide our readers with new and significant insights into making healthcare safer for patients around the world.

Vignettes in Patient Safety - Volume 2

Vignettes in Patient Safety

Volume 2

*Edited by Michael S. Firstenberg and Stanislaw P. Stawicki*

Photo by wernerimages / iStock