Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

Total parenteral nutrition: how to reduce the risks

  • Comment

VOL: 98, ISSUE: 22, PAGE NO: 40

Lorraine Bosonnet, RGN, is Macmillan nurse, upper gastrointestinal tract, Royal Liverpool and Broadgreen University Hospitals NHS Trust

Total parenteral nutrition involves the administration of a complex mix of glucose, lipids, amino acids, electrolytes, vitamins and trace elements via the circulatory system when the body is unable to tolerate nutrition by any other method. It is often used to prepare undernourished patients for major surgery and is used in gastrointestinal disease, such as ischaemic bowel disease and fistula. It involves long-term venous access using a catheter, which has a range of associated complications.

Before 1998, the standard of care given to TPN lines at the Royal Liverpool Hospital was fragmented and not based on evidence. There was a nutritional team, but it was unclear who had responsibility for line insertion and care was haphazard. No nursing or medical team had direct responsibility for patients receiving TPN and this was reflected in an infection rate of up to 20%. Indications that clinical practice was not of a consistently high standard gave cause for concern.

In an attempt to reduce the incidence of complications, the management of TPN lines was examined and problem areas identified. It was apparent that new evidence-based guidelines and ongoing staff education on all aspects of intravenous feeding were required. Two visits to a regional nutritional unit and a literature review clarified which aspects of care needed to be addressed.

This article reports on the evidence-base from which written guidelines and educational packages were developed, and an audit that examined the effectiveness of revised practice. The guidelines and education packages concur with standards outlined by the British Association for Parenteral and Enteral Nutrition.

Early audit results facilitated the opening of a designated unit on one of the surgical wards for patients requiring TPN. All nurses were given an opportunity for supervised work in the unit, to facilitate education and assessment, before they were deemed competent to care for feeding lines. A practice routine was also introduced for nurses.

Insertion of line

The complication rate and development of catheter sepsis are inversely proportional to operator experience (Armstrong et al, 1986). Line insertion should therefore be undertaken by a specialist registrar or consultant connected with the nutritional team in the operating theatre using single-lumen Hickman catheters: multi-lumen catheters increase the risk of infection (Early et al, 1990; McCarthy et al, 1987). Specific long-term catheters have a Dacron cuff about 30cm from the hub. These provide a barrier to ascending organisms and stabilise the catheter.

The subclavian veins are the preferred insertion sites for feeding lines as they allow easier handling. The exit site is easily accessible with the line securely attached to the patient’s skin. Jugular vein access makes care of the exit site and catheter hub difficult (Elliot et al, 1994).

The insertion site is a major source for catheter colonisation (Dhagastani et al, 1996). This and the surrounding area should be cleansed with chlorhexidine 0.5% in alcohol 70% (Hydrex) before applying sterile drapes (Elliot et al, 1994). The line should be tunnelled under the skin, as this is associated with reduced infection rates (Keohane et al, 1983). Once the line is in situ and blood has been aspirated via the lumen, it should be flushed with a heparin sodium solution (10 units/ml) (Hepsal) to ensure patency, then sutured in place.

There is conflicting evidence on whether application of an antiseptic to the insertion site reduces bacterial contamination (Conly et al, 1989; Maki et al, 1991). To facilitate viewing of the site, an occlusive transparent dressing should be used, as this will reduce the number of times the site is disturbed.

The catheter hub has been implicated as an important portal of entry for micro-organisms causing catheter-related sepsis. Salzman and Ruben (1993) report that cleaning with 70% ethanol dramatically reduces microbial contamination. Chlorhexidine in alcohol has also proved an effective skin disinfectant (Maki et al, 1991; Baranowski, 1993). It is recommended that catheter hubs are cleaned with a solution of chlorhexidine 0.5% in alcohol 70%.

Once the connection has been cleaned and a needle-less IV administration device (Bionector) applied, a gauze dressing must be wrapped around the connection site and secured with tape to reduce the risk of contamination. Placing a shield around the catheter hub to ensure it does not come into contact with skin reduces the incidence of catheter-related sepsis from 39% to 8% (Stotter et al, 1987). The line can then be coiled on the patient’s chest until the position is confirmed and feeding is ready to start.

The use of prophylactic antibiotics does not seem to prevent catheter-related sepsis (Dhagastani et al, 1996), indeed, it may encourage the emergence of resistance (Elliot et al, 1994).

A chest X-ray will exclude other complications relating to catheter insertion, such as pneumothorax and haemothorax, and confirm correct catheter placement.

Improved aseptic technique

Effective infection control is essential. The number of staff handling tubing, changing giving sets, and manipulating catheters and entry sites should be kept to a minimum (Elliott et al, 1994). When a specialist team provides meticulous care to TPN lines, infection rates fall from 25-33% to 4% (Faubian et al, 1986).

Infection: prevention and early detection

The meticulous care of intravenous lines plays a large part in the prevention of infection (Dhagastani et al, 1996). Causes of infection in feeding lines include:

- Wrong type of line with multiple-entry ports;

- Poor siting or incorrect insertion technique;

- Incorrect use of lines, including administration of other fluids or drugs through them.

The reduced infection rate in the unit did not prevent lines being removed prematurely. Despite evidence suggesting that systemic and localised infection of central lines may produce a low-grade pyrexia (Elliott et al, 1994), catheters were being removed inappropriately when patients developed a pyrexia. Clear guidelines to help identify and treat suspected catheter-related infection were evidently required.

Other causes of pyrexia should be excluded before the line is removed for culture. When there is a system of strict aseptic care, the line should be considered a possible source of infection, not a probable cause. Post-operative pyrexia may remain unexplained and subside spontaneously.

Routine administration of antibiotics in pyrexial patients should be discouraged (there is no evidence that they are effective), while the incidence of antibiotic-related complications and emergence of resistant organisms is significant.

Identifying infection

Localised infections can occur at the exit site or along the track of a tunnelled line. Inflammation without exudate does not always signify infection. If pus is present, a diagnosis of infection can be made (Salzman and Rubin, 1995).

Systemic infections are more difficult to recognise. Low-grade pyrexia, which does not respond to antibiotics, or leukocytosis with hypotension may be evident. There are usually no definite signs of local infection and no other obvious source of the septicaemia (Salzman and Rubin, 1995). Systemic infection can be confirmed when the same organism is isolated from peripheral blood cultures and one of the following:

- Central-line blood cultures;

- Exudate at the catheter site;

- Culture from the catheter hub.

Any patient who presents with pyrexia should have a full septic screen. Blood cultures must be obtained from the central line and a peripheral source. Swabs should be obtained from the exit site, catheter hub and any other possible source, such as a wound or drain site.

Exit-site infections can usually be treated with systemic antibiotics and local care without removing the catheter. Tunnelled infections are usually more difficult to treat and require intravenous antibiotics. These can progress rapidly and the catheter may have to be removed (Salzman and Rubin, 1995).

When systemic infection is a possibility, feeding should be stopped and the line locked with heparin sodium solution (Hepsal) (Sizer et al, 1996). Intravenous fluids should be administered through a peripheral intravenous cannula to ensure hydration while the feeding line is not in use. The result of the blood culture should be known within 24 hours. If the patient’s condition permits, antibiotic therapy should be withheld until an organism is isolated.

Gaillard (1990) showed that the internal lumen of a colonised catheter could be sterilised with an antibiotic to which the organism is sensitive. It was proposed that instilling the appropriate antibiotic using a lock technique for two hours would kill the bacteria. However, this has not been confirmed in human studies. Krzywda et al’s study (1995) of human subjects found that instilling a high concentration of an antimicrobial agent into a lumen using a lock technique for 12 hours was effective (except against candida infections), after which the catheter was flushed with saline to clear the line.

If the pyrexia subsides and the patient’s condition is stable, feeding should be resumed within 48 hours. Antibiotic therapy should be continued through an additional peripheral IV cannula for the recommended period. The dedicated TPN lumen should not be used for drug access once feeding has resumed (Kruse and Nipurn, 1993).

Early recognition of symptoms relating to pyrexia and line infection can prevent unnecessary line removal. Early detection of an infected line will ensure that correct treatment is started as soon as possible and infected lines can be salvaged. It is essential to remove the line in cases of septic shock related to an infected catheter.

Blocked lines: prevention and treatment

When parenteral feeding is occluded, the pump and giving set should be checked. There should be no kinks in the line and no external pressure on it. Asking the patient to change position by standing or moving their arm up and down above their head may resume the flow if the catheter tip is pressed against the vessel wall.

A chest X-ray should be obtained to ensure the catheter is not kinked or twisted. If the catheter is still occluded, this may be the result of fibrin sheath formation, blood clots, lipid deposits or the precipitation of medications or minerals and electrolytes (Holcombe et al, 1992; Johnston et al, 1992).

Fibrin deposits and blood clots

Fibrin sheath formation may occur over several days or within 24 hours of catheter placement (Wickham et al, 1992). Line occlusion can occur after blood sampling or blood product administration as deposits from the blood residue can form along the lumen (Baranowski, 1993). Careless flushing can lead to aspiration of blood into the catheter tip and subsequent clotting (Riella and Scribner, 1976).

Urokinase is a safe agent for clearing central lines thought to be occluded with blood clots (Baranowski, 1993; British Committee for Standards in Haematology, 1997; Wickham et al, 1992), and is effective in 92% of cases (Bagnall-Reeb and Ruccionne, 1993). A small dose is required, so systemic fibrinolysis is unlikely to occur (BCSH, 1997). Patients with a history of active bleeding, blood dyscrasias and other relevant history must be treated with caution.

The procedure involves instilling 5,000 units per ml equal to the catheter volume (Baranowski, 1993) gently into the catheter using a to-and-fro movement to maximise mixing in the lumen. A 10ml syringe must be used - a smaller size will increase the pressure inside the lumen and catheter fracture may occur (Wickham et al, 1992; Riella and Scribner, 1976).

The mean time to catheter clearance is 27.4 minutes (Lawson et al, 1982). It is recommended that the solution remain in the line for a minimum of 30 minutes and up to two hours. If flow does not resume after the solution is withdrawn, a further 10,000 units should be instilled and left for up to six hours, possibly overnight (Bagnall-Reeb and Ruccione, 1993; Holcombe et al, 1992).

Lipid formation

Lipid deposits can form a white or creamy coloured waxy substance along the internal lumen of the line (Baranowski, 1993; Holcombe et al, 1992). It is not known why this occurs. Gradual increasing resistance over two to three days results in complete blockage (Johnston et al, 1992). In such cases, a 4ml solution of 70% ethanol should be inserted gently into the lumen and locked in place for one hour (Pennington and Pithie, 1987) to unblock the catheter. When cyclical feeding is taking place, daily flushing of the line with 10ml of 20% ethanol solution helps to prevent the build-up of the sludge that leads to occlusion (Johnston et al, 1992). If it is not known whether occlusion is caused by lipid or clots, it is recommended that urokinase be tried first (Holcombe et al, 1992).

If occlusion is caused by lipids, the urokinase can completely block the line; however, this can be reversed by an ethanol lock (Pennington and Pithie, 1987).

Precipitation of drugs or minerals

Precipitation is more likely if TPN is used several days after preparation and administered with other drugs (Wickham et al, 1992). If a precipitation of medication or minerals is suspected, the pharmacy department’s advice should be sought to ensure the correct solution is used to unblock the line. Hydrochloric acid decreases the pH content, while sodium bicarbonate increases it. Both can effectively clear blockages caused by certain antibiotics and heparin.

Hydrochloric acid has also been used successfully to dissolve insoluble calcium phosphate precipitate (Baranowski, 1993; Wickham et al 1992). When it is not known which way to alter the pH balance, both methods can be attempted before resorting to removing the catheter (Holcombe et al, 1992). As the volume of an occluded catheter is not known, extreme force must not be used.


Many authors recommend regular flushing to maintain line patency and prevent build-up of substances along the lumen of the catheter. Heparinised saline is an acceptable solution; the volume should be twice that of the lumen of the line (Baranowski, 1993). Solutions of 10 units per ml (Cottee, 1995), and 50 units per ml (BCSH, 1997) have been shown to be successful.

Before and after the heparinised solution, a saline flush of at least 10ml should be inserted if the line is being used for drug therapy and blood sampling or infusion. Holcombe (1992) suggests the line should be flushed routinely before each line change with 10ml of 0.9% saline. The frequency varies widely from six-hourly to weekly.

It is vital to use a positive-pressure flushing technique to minimise the risk of blood reflux into the catheter, thus preventing the formation of blood clots (Goodwin and Carlson, 1993). The positive pressure is achieved by clamping the catheter just before the flush is finished, for example when the plunger of the syringe is moving forward (Riella and Scribner, 1976).

Audit results

Data was collected on 235 episodes of feeding between May 1998 and August 2001 where the guidelines were used. Early results were promising: infection and complication rates were greatly reduced if the guidelines were adhered to. Problems identified relating to blocked lines led to another literature review and a change in policy after 12 months. The data produced was used to highlight the need for a designated feeding unit and a nutritional nurse specialist, which have been set up.

By August 2001, 235 lines had been inserted, 77 peripherally and 158 using central line catheters.

There were no complications in 188 (80%) of the lines. Ten (4%) were removed owing to infection, a significant reduction from the previous rate of up to 20%. None of the catheters placed and cared for according to guidelines (using a tunnelled single-lumen Hickman catheter) became infected.

Eight (3%) catheters had problems with flow delivery. Blocked catheter episodes were associated with long-term feeding and thought to be the cause of lipid build-up rather than thrombus formation. Three (1%) catheters were removed because the lines were being fractured by incorrect flushing technique, and nine peripherally inserted lines (4%) were removed due to phlebitis. Fifteen (7%) were removed for other reasons, which included the catheters falling out, being pulled out or problems with the manufacture of the catheter itself.


The Davies Parenteral Nutrition Unit has been open for three years and has cut complication rates. Ongoing staff education will continue, including an induction day for house officers, and data will be presented regularly at audit meetings. Nurses from other areas will have the chance to spend time in the unit to ensure expertise is shared. The audit has proved that inserting a correct line and treating it appropriately can practically eliminate infections. The team recognises the need for further work in this area and the audit is ongoing.

  • Comment

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Links may be included in your comments but HTML is not permitted.