RESEARCH/PRACTICE REPORTS
PHARMACIST INTERVENTIONS IMPROVE FLUID BALANCE IN FLUID-RESTRICTED PATIENTS REQUIRING PARENTERAL NUTRITION Joyce E. Broyles, Rex O. Brown, Kathryn L. Vehe, RobertJ. Nally, and R. Wayne Luther
ABSTRACf: Many intensivecare unit (ICU) patientsrequire parenteral nutrition(PN) and fluid restriction, makingdelivery of adequate nutritiondifficult. We studiedthe effectsof pharmacistinterventions on fluidbalancein fluid-restricted ICU patientsrequiringPN. Twenty patientswererandomizedto the treatmentgroup (dextrose70% injection [D70Wj plus 15%amino acids for PN, 25-mL piggybacks, selecteddrugs added to the PN solution)or the control group (D70W plus 10% aminoacids, 50- or IOO-mL piggybacks). Each group contained 10patientsand they werenot significantly differentfor age, gender, weight, hospitaldays, and serum albuminconcentration. The duration(9.3 ± 1.2 vs. 9.7 ±2.4 d) and doses ofPN (29 ±6.8 vs. 28.7 ±6.9 kcal/kg/d; 1.1 ±0.3 vs. 1.1 ±0.4 g/kg/d protein)were similar between treatment and controlgroups. Mean fluidintake(3112 ± 1146 vs. 3498± IIII mUd), fluidbalance(146± 1581 vs. 708 ± 1402mUd), and cumulative fluid balance(1358 vs. 6867 mL) wereall significantly lowerin the treatmentgroup. Mean fluidoutput was similarbetweenthe two groups. Pharmacist interventionscan significantly decrease intake and result in a better fluidbalance in fluid-restricted ICU patientswho require PN. D1CP Ann Pharmacother 1991;25: 119-22. MANY PATIENTS HOSPITALIZED in intensive care units (lCUs)
require fluid restriction because they require administration of multiple intravenous drugs, blood and blood products, or JOYCE E. BROYLES, Phann.D., at the time of this study was a Clinical Pharmacy Resident, Regional Medical Centerat Memphis; sheis nowa Clinical Pharmacist. MethodistHospital, Memphis, TN, and an Assistant Professor of Clinical Pharmacy, University of Tennessee; REX O. BROWN, Phann.D., is an Associate Professor of Clinical Pharmacy. University ofIennessee, 26S. DunlapSt., Memphis, TN 38163; anda NutritionSupportPharmacist, Regional Medical Centerat Memphis, University ofTennessee Medical Center; KATHRYN L. VEHE, Phann.D., at the time of this study was a Research Fellow, Department of Clinical Pharmacy, University of Tennessee; she is now a Clinical Pharmacist, BarnesTeaching Hospital,St. Louis,MO;ROBERTJ. NOLLY, M.S., is the Directorof Pharmacy,Universityof TennesseeMedical Center, and an Assistant Professor of HealthScienceAdministration, University of Tennessee; and R. WAYNE LUTHER, M.D., is an AssociateProfessorof Surgery, UniversityofTennessee. Reprints: Rex0. Brown,Phann.D. This study was funded by the American Societyof Hospital Pharmacists' Research and Education Foundation (R07-3630-36).
specialized nutrition support. Also, they may have disease states that require fluid restriction, such as congestive heart failure, acute renal failure, adult respiratory distress syndrome, acute respiratory failure, head injury associated with elevated intracranial pressure, the syndrome of inappropriate antidiuretic hormone secretion, and advanced cirrhosis with ascites. Failure to restrict intake in these patients can result in prolonged mechanical ventilation, cerebral edema, worsened ascites, and severe hyponatremia with seizures. Dasta and Armstrong reported an average cumulative fluid gain of 11.6 kg in 181 patients admitted to a surgical intensive care unit.' Another recent study reported that patients who increased their preoperative weight by greater than ten percent during the perioperative period had a significantly higher dependence on pressor agents and colloids when compared with a group that increased their weight by less than ten percent. 2 Fluid restriction may compromise administration of optimal medical therapy to ICU patients (e.g. , amount of fluid allowed for nutrition support). Broyles et al. reported that using concentrated nutritional substrates for parenteral nutrition (PN) in ICU patients requiring fluid restriction reduced PN intake by 187 mLld; however, the effect on overall fluid balance was insignificant. The authors suggested that a more comprehensive approach to fluid intake would be needed to impact fluid balance in this patient population. 3 Pharmacists, with their thorough knowledge of drugs, should be helpful in designing fluid restriction regimens for ICU patients. Possible interventions include using concentrated macronutrients for PN solutions and concentrating piggybacks for intravenous drug administration. Other potential ways that a pharmacist may design fluid-restricted regimens for patients include using multiple-dose containers, changing the dose of medications to less-frequent intervals (e.g., vancomycin administered every 12 versus
DICP, The Annals ofPharmacotherapy • Downloaded from aop.sagepub.com at CORNELL UNIV on September 19, 2016
1991 February, Volume 25 •
119
every 6 hours), and giving medications intravenous push rather than via piggybacks. Other interventions include adding selected intravenous drugs to PN solutions, discontinuing maintenance intravenous solutions, and converting medications to the oral route if possible. This study was conducted to determine if some of the above pharmacist interventions could reduce fluid intake and improve fluid balance in fluid-restricted ICU patients who required PN.
Methods Consecutive ICU patients who required fluid restriction and were referredto a nutritionsupportservicefor PN werestudied. Patientswere considered in need of fluid restrictionif (I) they had a disease state that traditionally required fluid restriction (e.g., congestive heart failure, adult respiratory distress syndrome, syndrome of inappropriate antidiuretichormonesecretion); (2)theyhadanelevatedcentralvenouspressure (> 12 mmHg) or pulmonary capillary wedge pressure (> 12 mm Hg); or (3) they received a volumeof intravenous medications (>3 medications), bloodproducts,or otherfluids thatwouldmakethe administration of the appropriate volume of PN difficult to achieve. The patients werestudiedfor a minimumof 7 and a maximumof 14days. Patients who had severe liver dysfunction (total bilirubin >8 mgldL or hepaticencephalopathy>grade I), thermalinjury, acute or chronicrenal failure,or a protein-losing enteropathy wereexcludedfromthisprotocol. Demographic data such as age, weight,and genderwererecordedfor each patient. The patients' maximum temperature for the previous 24 hours, serum creatinine, blood urea nitrogen, and serum albuminconcentrationswere measuredat study entrance and exit. Nitrogenbalance was measuredon day 2 of the studyand repeatedon day 6. Patients were assigned to one of two treatmentgroups by a random numberstable. GroupI (control) received10%crystallineaminoacidsas Travasol (Clintec Nutrition Company, Deerfield, IL), dextrose70% in water(mOW), and 20% lipidemulsionas Intralipid(Clintec)forPN and standardSO- or 1000mL piggybackbagsfor intravenous drug administration. Group II (treatment) received IS% crystalline amino acids as Novamine (Clintec), mow, and 20% lipidemulsionfor PN and 2S-mL piggyback bags whenever possible. Treatment-group patients also receivedselectedmedications (e.g., histamine,antagonists) in their PN formula, and continuous-infusion medications (e.g., pressor agents, aminophylline) wereconcentrated in smallervolumesif possible.Medications were added to PN solutions according to previously published guidelines.' Standards for minimum dilution of other intravenous medications weretaken from publishedguidelines."? PN solutions were prepared using the Automix Plus High Speed Compounder (Clintec). Dextrose and amino acids were admixed with individualized electrolytes andstandardvitaminsand traceelementsin 1L containers. The intravenous lipid was administered as a SOO-mL 20% emulsiontwo times a week to preventessential fatty acid deficiency or infusedcontinuously as a calorie source (2So--S00 mUd). The PN formulaswerepreparedto containthe same numberof nonprotein calories and gramsof proteinin each unit (Table I). All patients received PN by the central venous route after infraclavicular placement of a triple-lumen catheter. All PN solutions and intravenous fluids were administered by Travenol Flo-gard model 6200 infusion pumps (Baxter International, Deerfield, IL). The goal of specialized nutrition support in these patients was attainment of nitrogen equilibriumas measured by nitrogen balance (nitrogen out was calculated by adding 4 g to the grams of urea nitrogenexcretedduring a 24-
Table 2. Demographic Data of the Patients at Study Entry
Patients (n) Gender(M/F) Age (y) Weight (kg) Maximum temperature (OC) Hospital days* Bloodurea nitrogen (mmoIlL) Creatinine (IJomoI/L) Albumin (g/L)
CONTROL
TREATMENT
10
10 8/2
515 55.6± 15.4 68.6± 18.7 38.1±0.6 9.9±8.7 8.6±5.5 124±53 24±8
46.1 ± 18.6 71.3± 16.3 38.1±0.5 8.2 ± 8.4 9.4±6.1 115±53 25±5
P
0.23 0.74 0.93 0.66 0.75 0.68 0.74
*Before parenteral nutritionwas started.
Table 3. Major Diagnoses in the Patients Treatment group Surgical SIP exploratory laparotomy with ileus (2) severe pancreatitis (2) SIP liver transplant closed head injury with ileus intraabdominal abscesses SIP coronary artery bypass with ileus gunshot wound to the abdomen Medical respiratory failure with ileus
Control Group Surgical SIP exploratory laparotomy with ileus (2) necrotizing pancreatitis SIP liver transplant (2) closed head injury with ileus intraabdominal abscesses ovarian cancer Medical respiratory failure with ileus sicklecell crisis with sepsisand ileus
SIP = status post.
hour urine collection). The PN formulas were initiatedat 25-40 mUh and advanced as toleratedtoa proteingoalofl-I.S g/kg/d and an energy goal of 30-35 kcal/kgld. fluid intake and output were recorded daily for the durationof the study period. fluid intake was defined as the volume administered as PN, other intravenous solutions, piggybacks, blood products, and oral! enteral intake. Output was defined as urine excretion and fluid losses from nasogastric suction, abdominal drains, chest tubes, ostomies and rectum (if a rectal tube was placed). fluid balancewas calculateddaily from the difference between fluid intake and output. Cumulative fluid balance was recorded as the sum of the daily fluid balances for each patientduring the study. fluid balancemeasurements began during the firstfull day of PN administration, which was usuallyday 2 of nutrition support. Use of human albuminas an adjunct to PN therapyfor severe hypoalbuminemia «2S g/L) was recorded, as was all diuretic therapy the patients' received. All data are expressed as the mean± one standard deviation. Demographic, nutrition support administration, and cumulative fluid balance data werecomparedbetweengroups using the Student'sunpaired r-test. fluid intake, output, and balance(repeatedmeasures) between the two groups were compared by ANOVA with Scheffe's test." This protocol was approved by a university institutional review board and informed consentwasdeemed unnecessary.
Results Table I. Parenteral Nutrition Solutions
Dextrose 70% (rnl.) Amino acids* (mL) Nonprotein energy (kcal/unit) Protein (g/unit) Nitrogen (g/unit)
CONTROL
TREATMENT
500 500 1190 50 8.2
500 335 1190 50 8.0
*Control patients received 10% amino acids and treatment patients received 15% amino acids.
120
•
D1CP. The Annals ofPharmacotherapy
•
Of 24 patients who were entered into the study, 20 (10 in each group) completed at least seven days of PN and were included for analysis. Four patients (two in each group) were excluded because they received less than seven days of PN. The 20 patients received 190 days of PN during the study. Patient demographic data appear in Table 2 and suggest that the two groups were well matched. One treatment patient and one control patient died during the study, and three other control patients and one treatment patient died after the study. Table 3 summarizes the patients' major
1991 February, Volume 25
Downloaded from aop.sagepub.com at CORNELL UNIV on September 19, 2016
Research/Practice
diagnoses upon inclusion into the study. Although many diagnoses are represented in both study groups, the groups appear well matched. Nutrition support data are summarized in Table 4. Both control and treatment groups were given PN for a similar number of days. Energy intake, percent of energy given as lipid, protein intake, and nitrogen balances were similar between the two groups during the study. The targeted energy goal was slightly above the actual energy received because we included all days that the patients received some PN in our fluid balance calculations. This usually included one to two days to attain the desired rate and the days the patients were weaned from PN if they finished their therapy between days 7 and 14. Table 5 summarizes the fluid balance data. There were 97 fluid balances in the control group and 93 in the treatment group. Treatment patients had a significantly lower fluid intake, fluid balance, and cumulative fluid balance when compared with control patients (Figure 1). Fluid output was not significantly different between the two groups. Six patients in each group received human albumin supplementation during PN administration. Human albumin 25 g was added to each liter of PN in all cases. The mean cumulative doses of human albumin received were also similar between groups (225 g in control patients vs. 265 g in treatment patients). Five treatment patients received a mean of 1500 mL and seven control patients a mean of 1380 mL of blood products during the study period. Six control patients and eight treatment patients received intravenous furosemide during the study (control group 1850 mg, treatment group 770 mg). One control patient received 1240 mg of furosemide. The amount of furosemide given to the patients was not significantly different between groups.
Discussion Pharmacist intervention to decrease fluid intake resulted in significantly lower fluid intake, fluid balance, and cumulative fluid balance in a group of ICU patients who required fluid restriction and PN. As fluid output was similar between the two groups, the difference in fluid balance appears to be due to the decrease in fluid intake, which re-
Table 4. Nutrition Support Data
Length of parenteral nutrition (d) Nonprotein energy (kcal/kg/d) Percent of energy as lipid (%) Protein (glkg/d) Nitrogen balance on day 2 (g/d) day 6 (g/d)
CONTROL
TREATMENT
P
9.7±2.4 28.7±6.9 16 1.1±0.4 -6.4±9.4 -1.0±6.9
9.3± 1.2 29.0±6.8 II 1.1±0.3 -5.6±3.4 -0.8±4.9
0.64 0.92 0.84 0.84 0.96
Table 5. Fluid Balance Data
Intake (mUd) Output (mUd) Balance (mUd) Cumulative balance (mL) *p
PHARMACIST INTERVENTIONS IMPROVE FLUID BALANCE IN FLUID-RESTRICTED PATIENTS REQUIRING PARENTERAL NUTRITION Joyce E. Broyles, Rex O. Brown, Kathryn L. Vehe, RobertJ. Nally, and R. Wayne Luther
ABSTRACf: Many intensivecare unit (ICU) patientsrequire parenteral nutrition(PN) and fluid restriction, makingdelivery of adequate nutritiondifficult. We studiedthe effectsof pharmacistinterventions on fluidbalancein fluid-restricted ICU patientsrequiringPN. Twenty patientswererandomizedto the treatmentgroup (dextrose70% injection [D70Wj plus 15%amino acids for PN, 25-mL piggybacks, selecteddrugs added to the PN solution)or the control group (D70W plus 10% aminoacids, 50- or IOO-mL piggybacks). Each group contained 10patientsand they werenot significantly differentfor age, gender, weight, hospitaldays, and serum albuminconcentration. The duration(9.3 ± 1.2 vs. 9.7 ±2.4 d) and doses ofPN (29 ±6.8 vs. 28.7 ±6.9 kcal/kg/d; 1.1 ±0.3 vs. 1.1 ±0.4 g/kg/d protein)were similar between treatment and controlgroups. Mean fluidintake(3112 ± 1146 vs. 3498± IIII mUd), fluidbalance(146± 1581 vs. 708 ± 1402mUd), and cumulative fluid balance(1358 vs. 6867 mL) wereall significantly lowerin the treatmentgroup. Mean fluidoutput was similarbetweenthe two groups. Pharmacist interventionscan significantly decrease intake and result in a better fluidbalance in fluid-restricted ICU patientswho require PN. D1CP Ann Pharmacother 1991;25: 119-22. MANY PATIENTS HOSPITALIZED in intensive care units (lCUs)
require fluid restriction because they require administration of multiple intravenous drugs, blood and blood products, or JOYCE E. BROYLES, Phann.D., at the time of this study was a Clinical Pharmacy Resident, Regional Medical Centerat Memphis; sheis nowa Clinical Pharmacist. MethodistHospital, Memphis, TN, and an Assistant Professor of Clinical Pharmacy, University of Tennessee; REX O. BROWN, Phann.D., is an Associate Professor of Clinical Pharmacy. University ofIennessee, 26S. DunlapSt., Memphis, TN 38163; anda NutritionSupportPharmacist, Regional Medical Centerat Memphis, University ofTennessee Medical Center; KATHRYN L. VEHE, Phann.D., at the time of this study was a Research Fellow, Department of Clinical Pharmacy, University of Tennessee; she is now a Clinical Pharmacist, BarnesTeaching Hospital,St. Louis,MO;ROBERTJ. NOLLY, M.S., is the Directorof Pharmacy,Universityof TennesseeMedical Center, and an Assistant Professor of HealthScienceAdministration, University of Tennessee; and R. WAYNE LUTHER, M.D., is an AssociateProfessorof Surgery, UniversityofTennessee. Reprints: Rex0. Brown,Phann.D. This study was funded by the American Societyof Hospital Pharmacists' Research and Education Foundation (R07-3630-36).
specialized nutrition support. Also, they may have disease states that require fluid restriction, such as congestive heart failure, acute renal failure, adult respiratory distress syndrome, acute respiratory failure, head injury associated with elevated intracranial pressure, the syndrome of inappropriate antidiuretic hormone secretion, and advanced cirrhosis with ascites. Failure to restrict intake in these patients can result in prolonged mechanical ventilation, cerebral edema, worsened ascites, and severe hyponatremia with seizures. Dasta and Armstrong reported an average cumulative fluid gain of 11.6 kg in 181 patients admitted to a surgical intensive care unit.' Another recent study reported that patients who increased their preoperative weight by greater than ten percent during the perioperative period had a significantly higher dependence on pressor agents and colloids when compared with a group that increased their weight by less than ten percent. 2 Fluid restriction may compromise administration of optimal medical therapy to ICU patients (e.g. , amount of fluid allowed for nutrition support). Broyles et al. reported that using concentrated nutritional substrates for parenteral nutrition (PN) in ICU patients requiring fluid restriction reduced PN intake by 187 mLld; however, the effect on overall fluid balance was insignificant. The authors suggested that a more comprehensive approach to fluid intake would be needed to impact fluid balance in this patient population. 3 Pharmacists, with their thorough knowledge of drugs, should be helpful in designing fluid restriction regimens for ICU patients. Possible interventions include using concentrated macronutrients for PN solutions and concentrating piggybacks for intravenous drug administration. Other potential ways that a pharmacist may design fluid-restricted regimens for patients include using multiple-dose containers, changing the dose of medications to less-frequent intervals (e.g., vancomycin administered every 12 versus
DICP, The Annals ofPharmacotherapy • Downloaded from aop.sagepub.com at CORNELL UNIV on September 19, 2016
1991 February, Volume 25 •
119
every 6 hours), and giving medications intravenous push rather than via piggybacks. Other interventions include adding selected intravenous drugs to PN solutions, discontinuing maintenance intravenous solutions, and converting medications to the oral route if possible. This study was conducted to determine if some of the above pharmacist interventions could reduce fluid intake and improve fluid balance in fluid-restricted ICU patients who required PN.
Methods Consecutive ICU patients who required fluid restriction and were referredto a nutritionsupportservicefor PN werestudied. Patientswere considered in need of fluid restrictionif (I) they had a disease state that traditionally required fluid restriction (e.g., congestive heart failure, adult respiratory distress syndrome, syndrome of inappropriate antidiuretichormonesecretion); (2)theyhadanelevatedcentralvenouspressure (> 12 mmHg) or pulmonary capillary wedge pressure (> 12 mm Hg); or (3) they received a volumeof intravenous medications (>3 medications), bloodproducts,or otherfluids thatwouldmakethe administration of the appropriate volume of PN difficult to achieve. The patients werestudiedfor a minimumof 7 and a maximumof 14days. Patients who had severe liver dysfunction (total bilirubin >8 mgldL or hepaticencephalopathy>grade I), thermalinjury, acute or chronicrenal failure,or a protein-losing enteropathy wereexcludedfromthisprotocol. Demographic data such as age, weight,and genderwererecordedfor each patient. The patients' maximum temperature for the previous 24 hours, serum creatinine, blood urea nitrogen, and serum albuminconcentrationswere measuredat study entrance and exit. Nitrogenbalance was measuredon day 2 of the studyand repeatedon day 6. Patients were assigned to one of two treatmentgroups by a random numberstable. GroupI (control) received10%crystallineaminoacidsas Travasol (Clintec Nutrition Company, Deerfield, IL), dextrose70% in water(mOW), and 20% lipidemulsionas Intralipid(Clintec)forPN and standardSO- or 1000mL piggybackbagsfor intravenous drug administration. Group II (treatment) received IS% crystalline amino acids as Novamine (Clintec), mow, and 20% lipidemulsionfor PN and 2S-mL piggyback bags whenever possible. Treatment-group patients also receivedselectedmedications (e.g., histamine,antagonists) in their PN formula, and continuous-infusion medications (e.g., pressor agents, aminophylline) wereconcentrated in smallervolumesif possible.Medications were added to PN solutions according to previously published guidelines.' Standards for minimum dilution of other intravenous medications weretaken from publishedguidelines."? PN solutions were prepared using the Automix Plus High Speed Compounder (Clintec). Dextrose and amino acids were admixed with individualized electrolytes andstandardvitaminsand traceelementsin 1L containers. The intravenous lipid was administered as a SOO-mL 20% emulsiontwo times a week to preventessential fatty acid deficiency or infusedcontinuously as a calorie source (2So--S00 mUd). The PN formulaswerepreparedto containthe same numberof nonprotein calories and gramsof proteinin each unit (Table I). All patients received PN by the central venous route after infraclavicular placement of a triple-lumen catheter. All PN solutions and intravenous fluids were administered by Travenol Flo-gard model 6200 infusion pumps (Baxter International, Deerfield, IL). The goal of specialized nutrition support in these patients was attainment of nitrogen equilibriumas measured by nitrogen balance (nitrogen out was calculated by adding 4 g to the grams of urea nitrogenexcretedduring a 24-
Table 2. Demographic Data of the Patients at Study Entry
Patients (n) Gender(M/F) Age (y) Weight (kg) Maximum temperature (OC) Hospital days* Bloodurea nitrogen (mmoIlL) Creatinine (IJomoI/L) Albumin (g/L)
CONTROL
TREATMENT
10
10 8/2
515 55.6± 15.4 68.6± 18.7 38.1±0.6 9.9±8.7 8.6±5.5 124±53 24±8
46.1 ± 18.6 71.3± 16.3 38.1±0.5 8.2 ± 8.4 9.4±6.1 115±53 25±5
P
0.23 0.74 0.93 0.66 0.75 0.68 0.74
*Before parenteral nutritionwas started.
Table 3. Major Diagnoses in the Patients Treatment group Surgical SIP exploratory laparotomy with ileus (2) severe pancreatitis (2) SIP liver transplant closed head injury with ileus intraabdominal abscesses SIP coronary artery bypass with ileus gunshot wound to the abdomen Medical respiratory failure with ileus
Control Group Surgical SIP exploratory laparotomy with ileus (2) necrotizing pancreatitis SIP liver transplant (2) closed head injury with ileus intraabdominal abscesses ovarian cancer Medical respiratory failure with ileus sicklecell crisis with sepsisand ileus
SIP = status post.
hour urine collection). The PN formulas were initiatedat 25-40 mUh and advanced as toleratedtoa proteingoalofl-I.S g/kg/d and an energy goal of 30-35 kcal/kgld. fluid intake and output were recorded daily for the durationof the study period. fluid intake was defined as the volume administered as PN, other intravenous solutions, piggybacks, blood products, and oral! enteral intake. Output was defined as urine excretion and fluid losses from nasogastric suction, abdominal drains, chest tubes, ostomies and rectum (if a rectal tube was placed). fluid balancewas calculateddaily from the difference between fluid intake and output. Cumulative fluid balance was recorded as the sum of the daily fluid balances for each patientduring the study. fluid balancemeasurements began during the firstfull day of PN administration, which was usuallyday 2 of nutrition support. Use of human albuminas an adjunct to PN therapyfor severe hypoalbuminemia «2S g/L) was recorded, as was all diuretic therapy the patients' received. All data are expressed as the mean± one standard deviation. Demographic, nutrition support administration, and cumulative fluid balance data werecomparedbetweengroups using the Student'sunpaired r-test. fluid intake, output, and balance(repeatedmeasures) between the two groups were compared by ANOVA with Scheffe's test." This protocol was approved by a university institutional review board and informed consentwasdeemed unnecessary.
Results Table I. Parenteral Nutrition Solutions
Dextrose 70% (rnl.) Amino acids* (mL) Nonprotein energy (kcal/unit) Protein (g/unit) Nitrogen (g/unit)
CONTROL
TREATMENT
500 500 1190 50 8.2
500 335 1190 50 8.0
*Control patients received 10% amino acids and treatment patients received 15% amino acids.
120
•
D1CP. The Annals ofPharmacotherapy
•
Of 24 patients who were entered into the study, 20 (10 in each group) completed at least seven days of PN and were included for analysis. Four patients (two in each group) were excluded because they received less than seven days of PN. The 20 patients received 190 days of PN during the study. Patient demographic data appear in Table 2 and suggest that the two groups were well matched. One treatment patient and one control patient died during the study, and three other control patients and one treatment patient died after the study. Table 3 summarizes the patients' major
1991 February, Volume 25
Downloaded from aop.sagepub.com at CORNELL UNIV on September 19, 2016
Research/Practice
diagnoses upon inclusion into the study. Although many diagnoses are represented in both study groups, the groups appear well matched. Nutrition support data are summarized in Table 4. Both control and treatment groups were given PN for a similar number of days. Energy intake, percent of energy given as lipid, protein intake, and nitrogen balances were similar between the two groups during the study. The targeted energy goal was slightly above the actual energy received because we included all days that the patients received some PN in our fluid balance calculations. This usually included one to two days to attain the desired rate and the days the patients were weaned from PN if they finished their therapy between days 7 and 14. Table 5 summarizes the fluid balance data. There were 97 fluid balances in the control group and 93 in the treatment group. Treatment patients had a significantly lower fluid intake, fluid balance, and cumulative fluid balance when compared with control patients (Figure 1). Fluid output was not significantly different between the two groups. Six patients in each group received human albumin supplementation during PN administration. Human albumin 25 g was added to each liter of PN in all cases. The mean cumulative doses of human albumin received were also similar between groups (225 g in control patients vs. 265 g in treatment patients). Five treatment patients received a mean of 1500 mL and seven control patients a mean of 1380 mL of blood products during the study period. Six control patients and eight treatment patients received intravenous furosemide during the study (control group 1850 mg, treatment group 770 mg). One control patient received 1240 mg of furosemide. The amount of furosemide given to the patients was not significantly different between groups.
Discussion Pharmacist intervention to decrease fluid intake resulted in significantly lower fluid intake, fluid balance, and cumulative fluid balance in a group of ICU patients who required fluid restriction and PN. As fluid output was similar between the two groups, the difference in fluid balance appears to be due to the decrease in fluid intake, which re-
Table 4. Nutrition Support Data
Length of parenteral nutrition (d) Nonprotein energy (kcal/kg/d) Percent of energy as lipid (%) Protein (glkg/d) Nitrogen balance on day 2 (g/d) day 6 (g/d)
CONTROL
TREATMENT
P
9.7±2.4 28.7±6.9 16 1.1±0.4 -6.4±9.4 -1.0±6.9
9.3± 1.2 29.0±6.8 II 1.1±0.3 -5.6±3.4 -0.8±4.9
0.64 0.92 0.84 0.84 0.96
Table 5. Fluid Balance Data
Intake (mUd) Output (mUd) Balance (mUd) Cumulative balance (mL) *p
