Parenteral Nutrition With Standard Solutions

2 downloads 0 Views 177KB Size Report
of the nutrition support interest group of the Dietitians' Association ... ally 3 times a day in ICU patients. ... 3% use individually compounded PN solutions on a daily .... Potassium (3.5-5.0 mmol/L). 3.5. 2.5. 5. 6. Magnesium (0.70-0.95 mmol/L).
The Art and Science of Infusion Nursing Suzie Ferrie, MNutrDiet, AdvAPD (Australia), CNSC Sharon Carey, MNutrDiet, APD Rachelle Ryan, MNutrDiet, APD Charbel Sandroussi, MBBS(Hons), MMSc, FRACS Lynn Jones, MNNP, MPH, BHSc David Storey, MBBS, FRACS Nicole Segaert, BPharm, DipHospPharm

Parenteral Nutrition With Standard Solutions: Not the Best Solution for Everyone? A Retrospective Audit of 300 Patients ABSTRACT An observational retrospective study audited the incidence of adverse events in 300 consecutive inpatients receiving a single, premade total nutrient admixture. No patient experienced critically high triglycerides; 16% of patients had a metabolic adverse event, including raised bilirubin, urea, creatinine, or liver enzymes. Line sepsis occurred on 30 occasions representing 0.67 infections per 1000 catheter days. Mortality was significantly

higher in dialysis, nonsurgical, and intensive care unit patients. The use of a standard formulation for all parenterally nourished patients does not lead to an unacceptable incidence or severity of metabolic complications; however, it did not meet the protein requirements of surgical or critically ill patients. (See Abstract Video, Supplemental Digital Content 1, http://links.lww.com/JIN/A62). Key words: parenteral nutrition, standard solutions, total nutrient admixture

Author Affiliations: Royal Prince Alfred Hospital, Sydney, Australia (Mss Ferrie, Carey, Ryan, Jones, and Segaert; Drs Storey and Sandroussi); and University of Sydney, Sydney, Australia (Mss Ferrie and Carey). Suzie Ferrie, MNutrDiet, AdvAPD (Australia), CNSC, has been the critical care dietitian in the intensive care unit at Sydney’s Royal Prince Alfred Hospital since 2001. She is also the national convenor of the nutrition support interest group of the Dietitians’ Association of Australia. Sharon Carey, MNutrDiet, APD, is the deputy manager in the Department of Nutrition and Dietetics at Royal Prince Alfred Hospital, the senior dietitian in the upper gastrointestinal unit, and a consultant to the hospital’s intestinal failure clinic. She was awarded a Churchill Fellowship to conduct an international study of dietetic management of intestinal failure. Rachelle Ryan, MNutrDiet, APD, is the clinical educator in the Department of Nutrition and Dietetics at Royal Prince Alfred Hospital and a dietitian in the upper gastrointestinal unit. Charbel Sandroussi, MBBS(Hons), MMSc, FRACS, is a surgeon in the Royal Prince Alfred Hospital’s Department of Gastrointestinal

and Liver Services. He also heads the hospital’s parenteral nutrition service. Lynn Jones, MNNP, MPH, BHSc, is a nurse practitioner in parenteral nutrition at Royal Prince Alfred Hospital, where she oversees the management of parenteral nutrition for inpatients, as well the hospital’s home parenteral nutrition and intestinal failure service. David Storey, MBBS, FRACS, is the head of upper gastrointestinal surgery at Royal Prince Alfred Hospital. Retired from his leadership role in the hospital’s parenteral nutrition service, he continues to provide consultation and support. Nicole Segaert, BPharm, DipHospPharm, is the pharmacist for the parenteral nutrition service at Royal Prince Alfred Hospital. Corresponding Author: Suzie Ferrie, MNutrDiet, AdvAPD (Australia), CNSC, Royal Prince Alfred Hospital, Department of Nutrition and Dietetics, Missenden Road, Camperdown, NSW 2050, Australia ([email protected]). The authors of this article have no conflicts of interest to disclose. Supplemental digital content is available for this article. Direct URL citations appear in the printed text, and links to the digital files are provided in the HTML text of this article on the journal’s Web site (http://journals.lww.com/journalofinfusionnursing).

DOI: 10.1097/NAN.0000000000000077

424 Copyright © 2014 Infusion Nurses Society

Journal of Infusion Nursing

Copyright © 2014 Infusion Nurses Society. Unauthorized reproduction of this article is prohibited.

JIN-D-13-00048.indd 424

31/10/14 7:23 PM

T

he use of standard parenteral nutrition (PN) solutions has been a topic of recent discussion. In Australasia, however, such solutions are usual practice. A recent survey1 of all hospitals in Australia and New Zealand with adult intensive care units (ICUs) revealed that only 3% use individually compounded PN solutions on a daily basis. The remainder mostly or solely use standard solutions, either commercially ready-made products in multichamber bags or a limited range of standard solutions routinely made for the hospital or the health care district. There are a dozen or so different ready-made products available in Australia and New Zealand. Most of these are all-in-1, total nutrient admixtures, used without an in-line filter. In Australia and New Zealand, filters are used in neonatal PN (which is usually individually compounded) but are not commonly used in adult PN, perhaps because of the widespread use of standard products in which the proportions of potential precipitants are fixed at known safe levels. Usual practice is for the dietitian to estimate the nutritional requirements of an individual patient and then select the most appropriate product from the hospital’s available range of ready-made admixtures. The target administration rate is calculated to meet the patient’s estimated energy requirement; the patient then receives the corresponding fraction of a standard bag, with the remainder being discarded after a 24-hour hang time. In this respect, nutrition management is similar for all nutrition support patients, whether they are receiving PN, enteral nutrition, or a combination of both. The study came about in 1 hospital when only 1 standardized commercial premade PN admixture was used for more than 12 months because of unique, facility-specific factors. This extreme example created an opportunity to investigate the implications of standardized PN solutions. The study aimed to audit medical records and relevant databases to document the experiences of 300 consecutive patients at 1 hospital who received the same standard PN product, recording adverse events and seeking to identify whether they disproportionately affected any particular patient group. The study was approved by the hospital’s ethics review committee.

MATERIALS AND METHODS Three hundred consecutive patients requiring PN were given the same standard 3-in-1 total nutrient admixture, Oliclinomel N7-1000E (Baxter Healthcare, Lessines, Belgium). Details of the product’s nutrient composition are shown in Table 1. All patients were managed according to the hospital’s PN protocols, with tunneled venous access devices used for all patients requiring long-term PN, mandatory change of short-term central venous

VOLUME 37

|

NUMBER 6

|

NOVEMBER/DECEMBER 2014

TABLE 1

Composition of the Parenteral Nutrition Admixture Used in the Study Per Liter Energy (kCal)

1200

Amino acids (g)

40

Total nitrogen (g)

6.6

Lipid: 80% olive oil/20% soybean oil (g)

40

Sodium (mmol)

32

Potassium (mmol)

24

Magnesium (mmol)

2.2

Calcium (mmol)

2

Phosphate, including phosphate content of lipid emulsion (mmol)

10

Acetate (mmol)

57

Chloride (mmol)

48

pH

5.5-6.5

Osmolality (mOsm/kg)

1760

catheters (CVCs) every 7 days, infusion sets changed only by central venous access-accredited registered nurses, and careful swabbing of the line at every occasion of connecting or disconnecting it. No in-line filters were used. All patients received standard micronutrient additives providing the daily amounts of vitamins and trace elements recommended by the Australasian Society for Parenteral and Enteral Nutrition (AuSPEN).2 Blood glucose levels were monitored at least twice a day in stable ward patients and at least every 4 hours in the ICU. Most ICU patients received intravenous insulin infusions following current protocols for moderate glucose control.3 Laboratory parameters were measured at least 3 times a week in stable ward patients and generally 3 times a day in ICU patients. Retrospective data collection was conducted to identify the frequency and type of any adverse events occurring during the administration of standard PN, as measured by mortality, septic events, and “unacceptable” metabolic parameters. (Where possible, unacceptable metabolic parameters were defined according to established guidelines. Where these were not available, a structured consensus process, described below, was used.) All available values were collected for each indicator. Additionally, analyses were performed aimed at identifying patient characteristics at initiation of PN that are predictors of the adverse events in order to

Copyright © 2014 Infusion Nurses Society

425

Copyright © 2014 Infusion Nurses Society. Unauthorized reproduction of this article is prohibited.

JIN-D-13-00048.indd 425

31/10/14 7:23 PM

investigate whether any particular patient group was disadvantaged by receiving standardized PN. Data collection was conducted by the Surgical Outcomes Research Centre independent of the nutrition support team. To develop consensus about “acceptable” and “unacceptable” ranges for key metabolic parameters where no established guidelines were available, a modified Delphi approach was used with an objective facilitator from the research center involving personnel from the multidisciplinary nutrition support team at the hospital. The team included gastrointestinal surgeons, gastroenterologists, intensivists, dietitians, pharmacists, and specialist nutrition nurses. A convenience sample participated in the Delphi process: 2 gastrointestinal surgeons, 1 gastroenterologist, 1 intensivist, 3 dietitians, 1 pharmacist, and 2 specialist PN nurses. A list of the metabolic parameters of interest to this study was distributed to the clinicians, who then were asked independently to identify the measurement ranges they would consider “acceptable” and “unacceptably low” or “unacceptably high” during the administration of PN. If applicable, “unacceptable” ranges could be further classified as “a level of concern, to monitor closely” or “a critical level requiring immediate intervention.” The survey results were collated by the data collection team into an anonymous summary. This was then disclosed to the nutrition support team for review. Negotiation about the nominated ranges was then conducted as a group until a consensus regarding the agreed “acceptable” and “unacceptable” ranges was obtained. Each “PN occasion” was entered separately into the study database. A PN occasion was defined as the continuous administration of PN with any interruptions being no more than 5 days. That is, if an interruption occurred that was longer than 5 days, the next occurrence of PN administration was recorded as a new PN occasion. Major events in the patient’s admission that reflected a change in the indication for PN—including major surgical procedures or admission to the ICU— were also considered to define a new PN occasion. Using the agreed acceptable and unacceptable ranges derived from guidelines or developed in the Delphi process, all medical records and electronic databases were audited to record the dates on which any unacceptable parameters occurred for each patient. All available values were collected. The following metabolic parameters were assessed in this manner: serum sodium, potassium, creatinine, urea, alkaline phosphatase, gamma glutamyl transferase, aspartate transaminase, alanine transaminase, bilirubin, glucose, triglycerides, magnesium, phosphate, zinc, selenium, and copper. At the beginning and conclusion of each episode of PN, weight and subjective global assessment status were also recorded. Complications during the administration of PN were also noted, including

426 Copyright © 2014 Infusion Nurses Society

any septic events; identification of a positive blood culture; source of infection, if known; intravenous access complications, such as line-related thrombosis; and mortality. Line-related sepsis was defined as clinical sepsis with a positive blood culture and positive CVC tip.

STATISTICAL ANALYSIS Continuous variables were grouped in tertiles or quintiles, whichever was clinically meaningful for each variable, for comparison with the outcome of interest—that is, adverse events. Univariate and multivariate (logistic regression) analyses were conducted to determine whether there were any associations between the demographic and clinical characteristics of the patients and any adverse outcomes. Unadjusted logistic regression analyses were first conducted to identify univariate associations between each of the measured variables and the outcome of an adverse event. Variables with a P value < .25 in the unadjusted analyses were selected for inclusion in the multivariate model. Because some patients had more than 1 PN occasion, data at the level of individual occasions were clustered by patient, so generalized estimating equations (GEEs) were used to account for this cluster effect. Multivariate logistic regression analyses were conducted to identify the independent predictors of the outcome of an adverse event. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Further analyses were conducted to model the number of days of an unacceptable metabolic parameter and the number of days from commencement of PN to the measurement of the first unacceptable metabolic parameter in order to predict not only the occurrence of an adverse event but also the time at which the adverse event would likely occur. Univariate and multivariate analyses were also undertaken to identify the patients most at risk for adverse outcomes and to identify any patient groups for which it may be more appropriate to administer individually tailored PN.

RESULTS Adverse metabolic events were defined using guidelines and a modified Delphi approach. See Table 2 for details of the ranges classified as “unacceptably low” or “unacceptably high,” either “a level of concern, to monitor closely” or “a critical level requiring immediate intervention.” Three hundred patients had full data available, for a total of 400 occasions of PN; see Table 3 for demographic details of these patients. This table includes the indication for commencement of PN where this was available in the patient’s medical record. For many patients, the reason for commencing PN was not clearly

Journal of Infusion Nursing

Copyright © 2014 Infusion Nurses Society. Unauthorized reproduction of this article is prohibited.

JIN-D-13-00048.indd 426

31/10/14 7:23 PM

TABLE 2

Definitions of Adverse Biochemical Events From Delphi Process Unacceptable Low Level

Unacceptable High Level

Parameter (Normal Range)

Concerning

Critical

Concerning

Critical

Triglycerides (0.2-2.0 mmol/L)





4

8

Bilirubin (1-21 μmol/L)





30

60

ALP (30-130 U/L)





200

400

GGT (1-35 U/L)





200

400

ALT (5-55 U/L)





150

300

AST (5-55 U/L)





150

300

Urea (3-8 mmol/L)





10

20

Creatinine (50-90 μmol/L)





120

200

Glucose (3.0-7.7 mmol/L)

3.3

3.0

10

16

Sodium (135-145 mmol/L)

135

125

145

150

Potassium (3.5-5.0 mmol/L)

3.5

2.5

5

6

Magnesium (0.70-0.95 mmol/L)

0.65

0.45

1

1.6

Phosphate (0.8-1.5 mmol/L)

0.7

0.4

1.5

2.5

Copper (12-22 μmol/L)

12

6

22

30

0.75

0.4

1.35

1.7

8

5

20

30

Selenium (0.75-1.35 μmol/L) Zinc (10-18 μmol/L)

Abbreviations: ALP, alkaline phosphatase; ALT, alanine transaminase; AST, aspartate transaminase; GGT, gamma glutamyl transferase.

documented. In 38% of PN occasions, the documented reason was “patient needed to be nil by mouth” with no further detail. Most patients had 1 occasion of PN; median duration was 8 days, and the interquartile range (IQR) was 6 to 14 days. Six patients had more than 4 PN occasions (all were home PN patients with multiple hospital admissions during the study period), and an additional 47 patients had 2 to 4 occasions. In general, patient weight did not change significantly over the period of a PN occasion (mean 70.23 [standard deviation 17.56] kg at the beginning of PN and 69.98 [16.72] kg at the end of a PN occasion). Mean daily infused volume of PN admixture was 1592 (289) mL, or about 65 mL per hour, providing mean energy 7993 (1451) kilojoules/1911 (347) calories and mean protein 64 (12) g per day. For ICU and surgical patients, protein intakes from PN were on average 12.4 (11.5) grams per day less than the dietitian’s documented protein daily goal. For other patients, the mean protein intake met the dietitian’s documented goal. Additional nutrition intake came from enteral or oral nutrition at some point during all but 13 of the 400 PN occasions recorded, with patients receiving oral nutrition at some point during 332 (83%) of PN occasions

VOLUME 37

|

NUMBER 6

|

NOVEMBER/DECEMBER 2014

and enteral nutrition during 68 (17.0%) occasions. Propofol infusions were used during 60 of the PN occasions in ICU, lasting for an average of 2.36 (1.50) days. Univariate modeling was undertaken to identify specific patient characteristics that were significantly associated with increased risk of the adverse events studied. Adjustment was made for individual patients having multiple occasions of PN (cluster effect). The significance of univariate associations between patient factors and each of the adverse events is shown in Table 4. Metabolic adverse events (occurrences of “critical” high or low parameters) are shown in Table 5. The majority occurred on day 3 or 4 of PN, with the exception of low serum zinc level, which occurred at median 6 days (IQR 5-16 days) after starting PN. Eight patients experienced hyperglycemia, and, of these, 3 were managed with a transient decrease in PN infusion rate, 2 were managed by adding or increasing insulin, and 3 were monitored for resolution of glucose levels without intervention. Eleven patients experienced hypoglycemia, and all were managed by administering extra carbohydrate intravenously or orally. Electrolyte supplementation was used during 200 (50%) of PN occasions, mostly potassium (200 occasions), magnesium (153 occasions), and phosphate (140 occasions).

Copyright © 2014 Infusion Nurses Society

427

Copyright © 2014 Infusion Nurses Society. Unauthorized reproduction of this article is prohibited.

JIN-D-13-00048.indd 427

31/10/14 7:23 PM

TABLE 3

Patient Demographics Demographics

n

%