To determine current practice in the delivery of parenteral nutrition (PN) in Australian hospitals.
A cross-sectional mail survey.
Acute-care adult hospitals with greater than 200 beds in Australia.
A total of 67 hospitals (65.7% response rate).
Surveys were posted to hospitals. A reminder letter with a second copy of the survey was posted 3 weeks later to non-respondents.
Twenty-seven (40.3%) of the hospitals have a PN team and 50 (74.6%) have a hospital protocol for PN delivery. An inaccessible or non-functional gastrointestinal tract is the most common indicator for commencing PN. Fat infusion is calculated by 24 (38.7%) respondents with a mean (s.d.) maximum amount of fat provided of 2.0 (0.7) g/kg/day. Over half (n=35) reported calculating carbohydrate infusion at a maximum amount of 5.4 (1.0) mg/kg/min. Two-thirds (n=41) reported commencing PN at a rate of 50% or less of goal rate. Blood glucose levels (BGL) were monitored at least once per day by the majority of respondents (n=56, 83.6%). Insulin infusion was commenced at varying BGL. Most respondents (n=40, 59.7%) reported ceasing PN when at least half of the patient's requirements are being met either orally or enterally. A number of practice guidelines were identified and the results of the survey were compared with these guidelines.
Where there are clear practice guidelines, current practice appears to be in line with these recommendations, however, where evidence is lacking, practice is varied.
Parenteral nutrition (PN) is used for feeding patients when feeding via the enteral route is not possible. The choice of parenteral versus enteral feeding remains a controversial decision (Woodcock et al., 2001), particularly owing to the high cost and complications associated with PN (Dudrick et al., 1972; Tay et al., 2002). Complications include catheter-related sepsis, hyperglycaemia and hepatic steatosis. Regardless, PN is clearly indicated for use in particular patients, where a clinical benefit is anticipated, which would outweigh any harmful effects of restricting nutrition support (Wolfe and Mathiesen, 1997; Duggan et al., 2002).
Previous audits of PN administration within hospitals have indicated: a tendency to prescribe PN to malnourished patients regardless of severity of illness or gut function (Ioannides-Demos et al., 1995); provision of varied glucose loads (Ioannides-Demos et al., 1995) and increased short-term PN use (less than 5 days) before introduction of practice guidelines (Duggan et al., 2002).
Database and Internet searches indicated that there are a number of practice guidelines available for PN, which have been endorsed by a national body (British Association for Parenteral and Enteral Nutrition Working Party, 1996; British Society of Gastroenterology, 1996; Australasian Society for Parenteral and Enteral Nutrition, 1999; American Gastroenterological Association, 2001; ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002; The EAST Practice Management Guidelines Workgroup, 2003). These guidelines tend to provide recommendations that are too general and do not focus on practical aspects relating to the everyday delivery of PN (ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002), or appear to be based on expert opinion rather than higher levels of evidence (The EAST Practice Management Guidelines Workgroup, 2003).
The current American Society for Parenteral and Enteral Nutrition (ASPEN) (2002) guidelines refer to the nutrition care process, which encompasses patient screening, assessment, development of nutrition care plan, implementation, monitoring and termination. To be of practical use, guidelines for nutrition support and intervention should address each stage within the nutrition care process and provide specific guidelines, so that practice is standardized across practitioners.
Wolfe and Mathiesen (1997) highlight the problems with randomized controlled trials (RCTs) in nutrition support, which result in practice guidelines commonly based on lower levels of evidence rather than the ideal Level I or II evidence. A limitation of RCTs is often the exclusion of the group that is thought to most likely benefit from nutritional support (e.g. severely malnourished patients) owing to ethical reasons preventing randomization to a ‘control’ group or withholding nutrition support.
To evaluate the delivery of PN to hospitalized patients, it is important to first define current practice and identify where current practice differs from known evidence either in practice guidelines or from the current literature. This study therefore aimed to determine current practices in the delivery of PN in Australian hospitals.
Acute-care adult hospitals across Australia with greater than 200 beds were identified from The Australian Hospitals Directory (2000). Rehabilitation or repatriation only, maternity only and children's only hospitals were excluded from the study. One hundred and three hospitals were identified as the study population. Health professionals responsible for the delivery of PN in the hospital were the target sample, with the intention for only one response per hospital.
A survey consisting of 26 questions covering each stage of the nutrition care process was developed. The questions included indications for PN, prescription of PN including type of solutions used, monitoring of complications, cessation of PN and follow-up after cessation, as well as information on the existence of PN teams. Information was also collected on the hospital – type of hospital (publicly or privately funded), geographical location (metropolitan or non-metropolitan) and number of beds. Questions were primarily multiple-choice with some short response questions. The survey was piloted for content and face validity on a convenience sample, from five hospitals, including metropolitan and regional hospitals.
A cross-sectional survey was conducted. Surveys were posted to 103 acute-care adult hospitals across Australia. The covering letter explained the intention was for the survey to be completed by a health professional responsible for the delivery of PN in the hospital. A reminder letter, with a second copy of the survey, was posted 3 weeks after the initial posting to hospitals that did not respond.
Statistical analyses were carried out using SPSS for Windows (Version 11.0.1, 2001, SPSS Inc., Chicago, IL, USA). Categorical variables are presented as counts (percentages). Continuous variables were normally distributed and are presented as means (standard deviation, s.d.). Two-sample analysis using Fisher's exact test was performed to determine if respondents were similar to non-respondents with respect to hospital type and location.
Analyses conducted were primarily descriptive to address the aim of the study. Bivariate comparisons of responses from hospitals with and without PN teams were also conducted using Fisher's exact tests and independent sample t-tests. Statistical significance was set at the conventional 95% level (two-tailed).
A response rate of 65.7% (n=67) was achieved. Respondents and non-respondents did not differ with respect to location (metropolitan or non-metropolitan, Fishers exact test P=0.183); however, a significantly greater proportion of respondents were from publicly funded hospitals than expected (Fishers exact test, P<0.001). Characteristics of respondents are shown in Table 1. Dietitians (89.6%) followed by doctors (7.5%) and nurses and pharmacists (3.0%) completed the survey.
Total parenteral nutrition within the hospital
Less than half of the respondents reported having a hospital PN team (n=27). PN teams generally consisted of at least four members of a multidisciplinary team (n=18, 66.7%) including a doctor (n=27, 100%), dietitian (n=26, 96.3%), pharmacist (n=23, 85.2%) and nurse (n=20, 74.1%). Smaller hospitals (less than 300 beds) appeared to be less likely to report having a hospital PN team (Fishers exact test, P=0.10). PN delivery in hospitals without PN teams was primarily managed by intensivists (n=22, 55%). Only half of the hospital PN teams (n=14) managed the delivery of PN within the intensive care unit. Most hospitals (n=50, 78.1%) have a hospital protocol for the delivery of PN.
With respect to the delivery of PN, only minor differences were observed between hospitals with PN teams and those without. Areas where differences were identified are indicated below.
Patient screening and assessment
Reasons for commencing PN are indicated in Table 2. Hospitals with a PN team indicated that in over half of the cases (n=15, 55.6%) the team determined patients' suitability for PN, whereas for a quarter of the hospitals (n=7) determination of suitability was a combined decision between the team and the referring medical unit. Almost all hospitals (n=66, 98.4%) assessed biochemical parameters before the commencement of PN. Commonly assessed biochemical parameters are shown in Table 2.
Development and implementation of nutrition care plan
Table 3 shows responses for items relating to the development and implementation of the nutrition care plan. Over half of the respondents reported calculating the maximum amount of carbohydrate to be infused for individual patients, with fewer respondents calculating the maximum amount of fat infusion (Table 3). The mean (s.d.) maximum amount of carbohydrate infusion reported by respondents was 5.4 (1.0) mg/kg/min, ranging from 3.5 to 7.0 mg/kg/min. For fat infusion, the mean (s.d.) maximum amount provided was 2.0 (0.7) g/kg/day, ranging from 1.0 to 3.5 g/kg/day.
There is considerable variation in the type of PN solution and lipid emulsion used (Table 3).
Lipid emulsions are most commonly infused continuously over 24 h (n=45, 67.2%). Almost 10% (n=6) provided the lipid emulsion over a period of less than 12 h, whereas 16.4% (n=11) provided the lipid emulsion over a period of between 12 and 24 h.
A number of precautionary methods were reported for avoiding refeeding syndrome (Table 3). Hospitals with PN teams appeared to be more likely to report using intravenous thiamine as a precautionary method for refeeding syndrome (n=15, 55.6%) compared to hospitals without the teams (n=12, 33.3%), although this difference was not statistically significant (Fishers exact test, P=0.122).
Responses to questions relating to patient monitoring are shown in Table 4. The majority of respondents indicated monitoring the following: fluid balance at least once per day (n=56, 83.6%); blood glucose levels (BGL) at least once per day (n=56, 83.6%); temperature at least once per day (n=45, 67.2%); weight once to twice a week (n=41, 61.2%); pulse rate at least once per day (n=40, 59.7%) and respiratory rate at least once per day (n=34, 50.7%). There is considerable variation in the reported BGL at which insulin infusion is commenced. A third of respondents (n=22, 32.9%) reported uncertainty in the maximum BGL, relying on the medical team to recommend commencement of insulin infusion.
Termination of therapy
Most respondents (n=40, 63.5%) reported ceasing PN when at least half of a patient's requirements are being met either orally or enterally. A small number reported ceasing PN as soon as the patient commences on oral diet or enteral feeds (n=9, 14.3%) or when the referring medical team makes the decision (n=8, 12.7%). When ceasing PN 69.4% (n=43) reported decreasing the volume of PN gradually, whereas 16.1% (n=10) stopped PN immediately. Almost two-thirds (n=40) of hospitals have a follow-up protocol for patients following the termination of PN. Specific details on the follow-up protocols, however, were not collected.
PN is indicated when nutritional support is required and enteral feeding is not possible or inadequate. The high expense and risk of complications associated with PN delivery necessitates the need to ensure that current practice is in line with practice guidelines and/or the evidence-based literature. This study aimed to determine current practices in the delivery of PN in Australian hospitals. The following discussion considers the results in the context of current practice guidelines or the recent evidence-based literature, where available.
Owing to the difficulty in conducting RCTs in nutrition support and the often critical state of patients receiving PN, very few guidelines and research exist that are based on high levels of evidence. As such, unless otherwise stated, evidence from practice guidelines and literature is based on expert opinion only.
Guidelines on the suitability for PN recommend that PN should be indicated for patients with non-functional or inaccessible gastrointestinal tract (GIT) and those who cannot be nourished by oral diets or enteral nutrition alone (Level III evidence) (British Society of Gastroenterology, 1996; ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002). Similar evidence suggests that an inadequate oral intake over a period of 7–14 days and if oral intake is expected to be inadequate over the same time frame are appropriate indicators for commencing PN (British Society of Gastroenterology, 1996; American Gastroenterological Association, 2001; ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002). Based on the results of this study, current practice appears to be in line with these guidelines as the most common reported indicator for commencing PN was due to a non-functional or inaccessible GIT.
Guidelines for patient assessment are based on expert opinion only. Current opinion recommends both anthropometric and biochemical assessment be performed before commencement of PN. Biochemical assessment should include full blood count, liver function tests, urea and electrolytes, including calcium, magnesium and phosphate (British Association for Parenteral and Enteral Nutrition Working Party, 1996). In severely malnourished, stressed or unstable patients, micronutrient status should also be assessed (British Association for Parenteral and Enteral Nutrition Working Party, 1996).
The results of this survey indicated that liver function tests, electrolytes and BGL are commonly assessed before commencement of PN in current practice. Micronutrients do not appear to be commonly assessed. Assessment of anthropometric measures was not addressed as an item in the survey.
Development of nutrition care plan
The ASPEN guidelines (2002) recommend determining individualized nutrient requirements for patients, based on patient assessment (Level III evidence). In this survey, the majority of hospitals reported individualizing the nutrition prescription and regimen to individual patients' requirements. The survey did not ask about methods for estimating individual patients' energy requirements; however, recent studies have indicated that practice varies considerably (Reeves and Capra, 2003).
There is limited evidence regarding the maximum infusion rate of carbohydrate or fat in PN solutions. As hyperglycaemia (defined as BGL greater than 12 mmol/l (200 mg/dl)) is the most common complication associated with PN administration (Ioannides-Demos et al., 1995; Rosmarin et al., 1996; Tay et al., 2002), the maximum infusion rate at which hyperglycaemia can be avoided is an important clinical question. The British Association for Parenteral and Enteral Nutrition (BAPEN) (1996) recommend a maximum glucose infusion of 4 mg/kg/min. Guenst and Nelson (1994) found that 73% of critically ill patients who received glucose infusion at rates greater than 4 mg/kg/min had respiratory quotients greater than 1, indicating net fat synthesis. In a retrospective audit of patients receiving PN, almost half of the patients receiving a PN glucose infusion rate of >5 mg/kg/min developed hyperglycaemia, whereas no patients receiving <4 mg/kg/min developed hyperglycaemia (Level III-3 evidence) (Rosmarin et al., 1996). Patients who were sensitive to glucose loads (e.g. history of diabetes mellitus or glucose intolerance; corticosteroid therapy) were excluded from the study, therefore even lower levels of glucose infusion may be necessary (ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002).
Most likely owing to a lack of clear guidelines or sufficient evidence, our study indicated that current practice is associated with varied rates of glucose infusion. The mean rate of 5.4 mg/kg/min is higher than current opinion recommendations, with reported glucose infusions of up to 7.0 mg/kg/min.
A similar lack of evidence exists for the recommended maximum lipid infusion rate. BAPEN (1996) recommend not exceeding a lipid infusion of 50 mg/kg/h (1.2 g/kg/day). Two retrospective studies in home PN patients have indicated that risk of hepatic dysfunction may be associated with lipid infusions of greater than 1 g/kg/day (Level III evidence) (Chan et al., 1999; Cavicchi et al., 2000). This survey indicated that lipid infusion is commonly provided above this recommendation in practice.
Implementation of nutrition care plan
Very few guidelines exist regarding the implementation of the nutrition care plan for patients receiving PN. Guidelines recommend commencing PN below estimated requirements in severely depleted or critically ill patients (British Association for Parenteral and Enteral Nutrition Working Party, 1996) and commencing with a low glucose infusion rate for patients with diabetes or glucose intolerance (ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002). Results of this survey indicated that the majority of hospitals initiate PN at a rate below estimated requirements and increase PN at a slower rate for patients at risk of refeeding syndrome. Practices used as precautionary methods for patients at risk of refeeding syndrome, such as replacing abnormal electrolytes, are supported by the literature (Solomon and Kirby, 1990; Crook et al., 2001).
Different lipid emulsions based on different concentrations of triglyceride, fatty acids and phospholipid are available (Garcia de Lorenzo et al., 2003). A number of early studies suggested that high triglyceride emulsions suppress immunological function, thereby increasing patients' risk of infection and complications (Fischer et al., 1980; Fraser et al., 1983). Other studies, however, did not confirm these findings (Monson et al., 1986; Palmblad, 1991). Moreover, recent studies have supported the use of lipid emulsions as they provide essential fatty acids, energy substrate and a means to reduce carbon dioxide production (Druml et al., 1998).
Although it is well accepted that lipid emulsions should be provided as a component of PN solutions, there are no guidelines or recommendations as to the concentration of lipid emulsion that should be provided. A recent study found that lipid emulsions of high triglyceride concentration (30% intralipid) resulted in fewer disturbances in lipid metabolism in critically ill septic and trauma patients, compared to lipid emulsions of lower triglyceride concentration (20% intralipid) (Level II evidence) (Garcia de Lorenzo et al., 2003). The authors suggest that this may be due to the lower phospholipid/triglyceride ratio in the former emulsion, with both lipid emulsions showing similar results with respect to tolerance and safety (Garcia de Lorenzo et al., 2003). Kalfarentzos et al (1998) observed similar results.
In our survey, we found that most respondents reported using lipid emulsions of moderate level of triglyceride concentration (20% lipid emulsion), whereas few reported using lower triglyceride concentration lipid emulsions (10% lipid emulsion). Although most respondents reported providing lipid emulsions over a 24-h period, over a quarter of respondents reported a shorter time period. Research has suggested that when lipid emulsion is provided intermittently, glucose infusion rates may be increased to provide the calories needed, thereby increasing the risk of hyperglycaemia (Rosmarin et al., 1996).
Although there are guidelines regarding monitoring of specific parameters during PN administration (British Association for Parenteral and Enteral Nutrition Working Party, 1996; British Society of Gastroenterology, 1996; ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002), the guidelines do not address the precise frequency of monitoring or the level at which intervention is required. The ASPEN guidelines (2002) provide Level II evidence to recommend periodical monitoring of liver function tests (LFTs), so that hepatobiliary complications can be identified early and intervention to alter PN prescription may prevent significant liver complications.
There is also Level III evidence to support frequent monitoring of serum electrolytes (sodium, potassium, chloride and bicarbonate) upon initiation of PN and until stabilized (British Society of Gastroenterology, 1996; ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002); close monitoring of phosphorus, magnesium and glucose levels upon initiation of PN in patients at risk of refeeding syndrome (British Association for Parenteral and Enteral Nutrition Working Party, 1996; British Society of Gastroenterology, 1996; ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002); and frequent monitoring of BGL upon initiation of PN, after changes in insulin dose and until stabilized (ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002). Recommendations also include monitoring weight, anthropometry, temperature, fluid balance, triglycerides and trace elements and vitamins (British Association for Parenteral and Enteral Nutrition Working Party, 1996; British Society of Gastroenterology, 1996; ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002).
The results of this study indicated that specific parameters are monitored following commencement of PN; however, the frequency of monitoring varied among respondents, possibly reflecting the lack of detailed guidelines in this area.
There are no clear guidelines on the management of elevated LFTs in patients receiving long-term PN. A review of total PN-associated liver disease suggested methods for minimizing hepatic dysfunction, which include avoiding excess carbohydrate (use <40 kcal dextrose/kg/day) and excess lipid (use <1 g/kg/day) infusions and also cycling PN (Buchman, 2002). Pharmacological means (e.g. metronidaxole, ursodeoxycholic acid, choline) have also been tried as treatment (Buchman, 2002). Druml et al (1998) found that lipid emulsions of approximately 0.6 g/kg provided over 2 h were efficiently metabolized in critically ill patients with combined hepatic failure and sepsis. Overall, management of patients with elevated LFTs in this study appeared to be unclear with respondents reporting varied practice.
Guidelines regarding the frequency of blood glucose monitoring were not identified. Hyperglycaemia is associated with hyperosmolarity, which may lead to a number of undesirable consequences including osmotic diuresis (depleting potassium, sodium and phosphate), non-ketotic coma, increased rate of infection and impaired leukocyte function (Rossini, 1976; Marhoffer et al., 1992). The ASPEN guidelines (2002) recommend maintaining BGL between 5.6 and 11.1 mmol/l (100–200 mg/dl) in patients with diabetes mellitus and between 5.0 and 6.7 mmol/l (90–120 mg/dl) range during pregnancy. A prospective randomized controlled study in critically ill patients found a reduction in morbidity and mortality with intensive insulin therapy to maintain BGL between 4.5 and 6.1 mmol/l (80–110 mg/dl) compared to conventional treatment (Level II evidence) (Van den Berghe et al., 2001). This survey found that there is large variation and uncertainty in practice regarding the BGL at which insulin infusion should be commenced. This may reflect the lack of clear guidelines in this area.
Termination of therapy
The ASPEN guidelines (2002) recommend ceasing PN and providing total nutrition support via enteral or oral feeding, when more than half the patient's requirements are being met and tolerated by oral or enteral feeding. Infusion of PN should reduce as enteral feeding increases (ASPEN Board of Directors and The Clinical Guidelines Task Force, 2002). This study identified that for the majority current reported practice is in line with these recommendations. There are no guidelines regarding the follow-up of patients following cessation of PN. In current practice, however, most respondents reported a follow-up protocol.
This study was not without limitations. The majority of respondents were dietitians and therefore the results may not reflect current practice where dietitians are not involved in the delivery of PN. Another limitation is that the questions are related only to general PN practice and administration and did not address any issues relating to specific disease states or complications.
Ideally, we would have liked to assess current hospital practice against a clear standard or guideline based on the evidence. However, in many of these areas guidelines do not exist or evidence is inconsistent. Areas where there are clear guidelines include indications for PN, assessment of biochemical parameters before commencement and individualization of the regimen. This survey found that for the majority of hospitals, current practice appeared to be consistent with the current guidelines. Areas such as calculation of macronutrient infusion rates, monitoring and management of complications where evidence is limited or conflicting have the greatest variation in reported practice.
This study has indicated that where there are clear guidelines, whether based on sufficient evidence or opinion only, there appears to be consensus in current practice in line with these recommendations. There are important stages within the nutrition care process for patients receiving PN where guidelines and evidence is lacking. Consequently, current practice appears to vary in relation to these stages. In particular, practice in relation to the BGL at which insulin infusion is commenced varies greatly. Adherence to practice guidelines can improve the effectiveness of practice, both by decreasing risk of complications and improving patients' outcomes as well as resulting in significant cost savings (Duggan et al., 2002).
ASPEN Board of Directors and The Clinical Guidelines Task Force (2002). Guidelines for the use of parenteral and enteral nutrition in adult and pediatric patients. JPEN 26, 1S–138S.
American Gastroenterological Association (2001). American Gastroenterological Association medical position statement: parenteral nutrition. Gastroenterology 121, 966–969.
Australasian Society for Parenteral and Enteral Nutrition (1999). AuSPEN Guidelines for Intravenous Trace Elements and Vitamins. AuSPEN: Melbourne.
British Association for Parenteral and Enteral Nutrition Working Party (1996). Current Perspectives on Parenteral Nutrition in Adults. BAPEN: Berks.
British Society of Gastroenterology (1996). Guidelines for Artificial Nutrition Support. British Society of Gastroenterology: London.
Buchman A (2002). Total parenteral nutrition-associated liver disease. JPEN 26, S43–S48.
Cavicchi M, Beau P, Crenn P, Degott C, Messing B (2000). Prevalence of liver disease and contributing factors in patients receiving home parenteral nutrition for permanent intestinal failure. Ann Intern Med 132, 525–532.
Chan S, McCowen KC, Bistrian BR, Thibault A, Keane-Ellison M, Forse RA et al. (1999). Incidence, prognosis, and etiology of end-stage liver disease in patients receiving home total parenteral nutrition. Surgery 126, 28–34.
Crook MA, Hally V, Panteli JV (2001). The importance of the refeeding syndrome. Nutrition 17, 632–637.
Druml W, Fischer M, Ratheiser K (1998). Use of intravenous lipids in critically ill patients with sepsis without and with hepatic failure. JPEN 22, 217–223.
Dudrick SJ, Macfadyen Jr BV, Van Buren CT, Ruberg RL, Maynard AT (1972). Parenteral hyperalimentation. Metabolic problems and solutions. Ann Surg 176, 259–264.
Duggan C, Rizzo C, Cooper A, Klavon S, Fuchs V, Gura K et al. (2002). Effectiveness of a clinical practice guideline for parenteral nutrition: a 5-year follow-up study in a pediatric teaching hospital. JPEN 26, 377–381.
Fischer GW, Hunter KW, Wilson SR, Mease AD (1980). Diminished bacterial defences with intralipid. Lancet 2, 819–820.
Fraser I, Neoptolemos J, Woods P (1983). The effect of intralipid on human lymphocyte and monocyte function. Clin Nutr 2, 37–40.
Garcia de Lorenzo A, Lopez-Martinez J, Planas M, Chacon P, Montejo JC, Bonet A et al. (2003). Safety and metabolic tolerance of a concentrated long-chain triglyceride lipid emulsion in critically ill septic and trauma patients. JPEN 27, 208–215.
Guenst JM, Nelson LD (1994). Predictors of total parenteral nutrition-induced lipogenesis. Chest 105, 553–559.
Ioannides-Demos LL, Liolios L, Topliss DJ, McLean AJ (1995). A prospective audit of total parenteral nutrition at a major teaching hospital. Med J Aust 163, 233–237.
Kalfarentzos F, Kokkinis K, Leukaditi K, Maroulis J, Onoufriou A, Alexopoulos K (1998). Comparison between two fat emulsions: intralipid 30% vs intralipid 10% in critically ill patients. Clin Nutr 17, 31–34.
Marhoffer W, Stein M, Maeser E, Federlin K (1992). Impairment of polymorphonuclear leukocyte function and metabolic control of diabetes. Diabetes Care 15, 256–260.
Monson JFT, Ramsden CW, MacFie J, Brennan TG, Guillou PJ (1986). Immunorestorative effect of lipid emulsions during total parenteral nutrition. Br J Surg 73, 843–846.
Palmblad J (1991). Intravenous lipid emulsions and host defense – a critical review. Clin Nutr 10, 303–308.
Reeves MM, Capra S (2003). Variation in the application of methods used for predicting energy requirements in acutely ill adult patients: a survey of practice. Eur J Clin Nutr 57, 1530–1535.
Rosmarin DK, Wardlaw GM, Mirtallo J (1996). Hyperglycemia associated with high, continuous infusion rates of total parenteral nutrition dextrose. Nutr Clin Pract 11, 151–156.
Rossini AA (1976). Why control blood glucose levels? Arch Surg 111, 229–233.
Solomon SM, Kirby DF (1990). The refeeding syndrome: a review. JPEN 14, 90–97.
Tay SM, Ip-Yam PC, Lim BL, Chan YW (2002). Audit of total parenteral nutrition in adult surgical intensive care. Ann Acad Med Singapore 31, 487–492.
The Australian Hospitals Directory (2000): ATA Professional Services Pty Ltd: North Sydney.
The EAST Practice Management Guidelines Workgroup (2003). Practice Management Guidelines for Nutritional Support of the Trauma Patient. Eastern Association for the Surgery of Trauma (EAST): Winston-Salem.
Van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M et al. (2001). Intensive insulin therapy in critically ill patients. N Engl J Med 345, 1359–1367.
Wolfe BM, Mathiesen KA (1997). Clinical practice guidelines in nutrition support: can they be based on randomized clinical trials? JPEN 21, 1–6.
Woodcock NP, Zeigler D, Palmer D, Buckley P, Mitchell CJ, MacFie J (2001). Enteral versus parenteral nutrition: a pragmatic study. Nutrition 17, 1–12.
We thank all the respondents who willingly completed the surveys. We also thank Dr Peter Kruger and Dr Andrew Pascoe for assistance with proof-reading and editing.
Guarantor: A Ali.
Contributors: ABA initiated and designed the study, carried out data collection and assisted with the interpretation and discussion of results, and writing of the manuscript. CCK initiated the study and assisted in the design of the study and interpretation and discussion of results. MMR carried out data analysis and was responsible for the writing of the manuscript.
About this article
Journal of Parenteral and Enteral Nutrition (2017)
Japanese Guidelines for Nutrition Support Therapy in the Adult and Pediatric Critically Ill Patients
Nihon Shuchu Chiryo Igakukai zasshi (2016)
Clinical Nutrition (2009)