Effects of Calorie and Fluid Intake on Adverse Events During Hemodialysis Jennifer Strong, MS, RD, LMNT, * Marlys Burgett, MS, RD, LMNT, * Mary Lou Buss, MS, LCSW, if Michelle Carver, RN, BSN,~ Shalini Kwankin, RN, BSN,.~ and Dalene Walker, RN, BAN, CNNII

Purpose: To investigate the incidence of adverse events during hemodialysis treatments as a function of calories and fluid intake.

Methods: The study period was August 3-26, 1999. Hemodialysis visits were studied. Twenty-three patients receiving hemodialysis during the 2nd shift on Tuesday, Thursday, and Saturday were studied. A total of 166 hemodialysis patient visits were studied. Data collected included: amount of fluid and food consumed, blood pressure levels, and mannitol use during each hemodialysis treatment; and any symptoms that occurred either during or after the dialysis treatment (hypotension, nausea, vomiting, diarrhea, cramping, and access problems).

Results: Using regression analysis, calories and fluids were strong predictors of both hypotension (P = .003) and mannitol use (P = .000), but not of cramping or access problems. Patients were 3 times more likely to have hypotension if taking any fluids (P = .011). Patients consuming >200 calories were 2 times as likely to have hypotension (P = .058). Patients were 5 times more likely to use mannitol if taking any fluids (P = .005). Mannitol use increased significantly (P = .001) with those patients consuming >200 calories.

Conclusion: Patients who ate more than 200 calories and consumed more than 200 mL of fluid during hemodialysis had an increased incident of hypotensive events and increased use of mannitol. © 2001 by the National Kidney Foundation, Inc.

M ANY HEMODIALYSIS patients experi- ence hypotension, diarrhea, nausea, vom-

iting, and/or cramping during hemodialysis treat- ments. It has been our observation that patients complaining of the aforementioned side effects ingest food and/or fluid during the dialysis treat- ment or have high intradialytic fluid weight gains (>3 kg). It has been shown that food intake during hemodialysis treatments causes hypoten- sion by accelerating the fall in blood pressure caused by decreased systemic vascular resis- tance.1 4 Many authors have recommended that meals should not be consumed during hemodial- ysis based on the correlation between food intake

From Dialysis Center of Lincoln, Lincoln, NE. *Renal Dietitian. "~RenaI Social Worker. ~Education Coordinator. 5Chary, e Nurse. IIQuality Assurance Coordinator. Address reprint requests to Jennifer Stron2, MS, RD, L M N T ,

Dialysis Center of Lincoln, 7910 " 0 " St, Lincoln, N E 68510. © 2001 by the National Kidney Foundation, Nc. 1051-2276/01/1102-0007535.00/0 doi: 10.1053/jren. 2001.22490

and hypotensive episodes. 2 4 Currently, the pol- icy in our facility is to allow all patients to eat and/or drink during the dialysis treatment.

Despite the research, many units still allow patients to consume meals during hemodialysis treatments. Meal intake during hemodialysis al- lows for meal consistency. Consistent timing of meals can be particularly important in the diabetic hemodialysis patient. Patients who spend 30 min- utes or more traveling to/from the dialysis unit may also benefit from meal intake during dialysis. Allowing meal intake during hemodialysis can increase protein and calorie intake in patients who are malnourished. Protein-energy malnutri- tion is a common complication of hemodialysis and is significantly predictive of morbidity and mortality, s And finally, for many patients, it can be a pleasurable way to pass the time spent on dialysis.

A random, observational study was proposed to investigate the incidence of adverse events during hemodialysis as a function of calorie and fluid intake. By identifying correlations between food/fluid intake and hypotension, diarrhea, nau- sea, vomiting, and/or cramping, the patients at

.Journal of Renal Nutrition, Vol 11, No 2 (April), 2001: pp 97-100 07

98 S T R O N G E T A L

risk could be identified and improvements in dialysis therapy can be made.

Patients and Methods This investigation was a random, observational

study of data collected from hemodialysis visits at the Dialysis Center of Lincoln, Lincoln, NE. Data collected was studied by visit, not by individual patient (166 collective visits were studied from 23 different patients receiving hemodialysis). Con- sent forms to collect patient data were obtained from patients before initiation of the study.

During the month of August 1999, 166 hemo- dialysis patient visits were studied at the Dialysis Center of Lincoln located in Lincoln, NE. The patient visits consisted of 23 patients receiving hemodialysis during the second shift (11:00 AM- 4:00 PM) of the Tuesday, Thursday, Saturday schedule. Patients did not have to meet any selection criteria other than dialyzing on the sec- ond shift. All patients on the second shift were asked to participate; those who refused were not included in the data collection.

The patient visits included patients who un- derwent dialysis 3 times per week for approxi- mately 4 hours each treatment (average length of treatment is 4.01 hours). The average patient age was 71. The average length of time on hemodi- alysis was 2.7 years. The causes of end-stage renal disease for the patients were atheroembolism (3), hypertension (10), and Type 2 Diabetes Mellitus (10). The majority of the patients were female (14).

Patient data was collected during each visit to the center on a standard flowsheet. The flowsheet was kept on the patient clipboard during the entire hemodialysis treatment. The primary care nurse recorded the following: patient name, date, all food and fluid consumed during dialysis, blood sugar results (if diabetic), mannitol use, problems with access, and any adverse events that occurred either during or after the dialysis treatment, in- cluding: hypotension (systolic pressure less than 90 m m Hg), nausea, vomiting, diarrhea, and/or cramping. The dietitians analyzed the flowsheet and determined actual kilocalories (kcal) and amount of fluid (mL) consumed at each visit. Fluid was only considered from beverages, pud- ding, gelatin, and soups. The fluid content of other foods was not calculated. The kilocalories consumed were calculated using the American

Diabetes Exchange List figures and Bowes and Church's Food Values of Portions Commonly Used. 6 Patient visits with incomplete flowsheets were not analyzed. A total of 70/236 flowsheets had either incomplete data regarding food intake during the visit or had no data at all. Only 6 patients had more than 2 visits with incomplete data. O f these 6 patients, one individual had 4 occasions of cramping; one individual had 2 oc- casions ofhypotension; and one individual had 2 occasions of cramping and 1 occasion of hypo- tension during the study.

Analysis of Data The statistical program SPSS (SPSS Inc, Chi-

cago, IL) was used to calculate all analytical tests. Frequency statistics, and percentages were used to determine the number of patient visits; kilocalo- rie and fluid consumption; incidence ofhypoten- sion, nausea, vomiting, diarrhea, cramping; blood sugar results; mannitol use; access problems; high and low calorie intake; and high and low fluid intake. Cross tabulation statistics were used to develop contingency tables for each variable (hy- potension, nausea, vomiting, diarrhea, cramping; blood sugar results; mannitol use; access prob- lems) between 2 categories (high defined as more than 200 kcal or 200 mL and low defined as less than 200 kcal or 200 mL) for both kilocalorie and fluid consumption. The high and low categories were retrospectively defined based on statistical significance. Chi-square tests were used for com- parison of the categorical variables. Multiple re- gression analysis was used to determine the cor- relation between kilocalorie and fluid intake (constant predictors) and each dependent variable (hypotension, cramping; mannitol use; and/or access problems).

Results A total of 166 visits were included in this study.

Hypotension occurred in 32 visits, cramping and mannitol both occurred in 13 visits, and access problems occurred during 7 patient visits. Nau- sea, vomiting, and diarrhea did not occur in any of the patient visits and were not included as variables for the regression analysis.

For comparison, both kilocalorie and fluid consumption were divided into "high" and "low" intake categories. High intake was defined

E F F E C T S OF C A L O R I E A N D FLUID O N H E M O D I A L Y S I S 99

as consuming more than 200 kcal and /o r 200 mL fluid at a single visit. Low intake was defined as consuming less than 200 kcal and /or 200 mL fluid at a single visit. The majority o f patients (61.4% o f all patient visits) consumed more than 200 kcal at each visit (individuals consuming more than 200 kcal averaged 490 _+ 183 kcal per visit v those consuming less than 200 kcal aver- aged 28 _+ 57 kcal per visit). The majority o f patients (75.9% of all patient visits) limited their fluid intake to less than 200 mL per visit (indi- viduals consuming less than 200 mL averaged only 81 mL _+ 69 mL per visit v those who consumed more than 200 mL averaged 278 mL _+ 79 mL per visit).

The incidence o f hypotension was twice as c o m m o n when kilocalories consumed per visit reached 200 or more with a significant difference in use o f mannitol (P < .001) at the higher kilocalorie level. Mannitol was not needed on any visits when the kilocalories consumed was kept below 200. There was not a significant difference in occurrence o f cramping or problems with access between high and low kilocalorie levels.

Patients were 3 times more likely to experience hypotension if taking any fluids and were 5 times more likely to use mannitol with fluid intake. The incidence o fhypotens ion was twice as com- mon when fluid consumed per visit reached 200 mL or more with a significant difference in use o f mannitol (P < .014) between the high and low fluid categories. There was not a significant dif- ference in occurrence o f cramping or problems with access between high and low fluid levels.

Both fluid intake and kilocalorie consumption were strong predictors o f both hypotension (P < .003) and mannitol use (P < .000). Nei ther fluid intake nor kilocalorie consumption were statisti- cally significant predictors o f either cramping or problems with access.

Limitations There are several limitations that must be con-

sidered when evaluating this research project. The study was purely observational and was based on a number o f collective hemodialysis visits rather than individual patients. The study design did not take into consideration the other variables that might have an effect on blood pressure and was not controlled for these other variables. The

data collected was dependent on accuracy and standardization among the nurses completing the flowsheets for food and fluid intake. Because o f these limitations, it is important to point out the results are not conclusive in determining "cause and effect."

Discussion Patients who consume food and /o r fluid dur-

ing hemodialysis are at an increased risk o f suf- fering from a hypotensive episode and subse- quently use mannitol to treat the episode. Patients who consume more than 200 kcal or 200 mL fluid are more likely to have a hypotensive event and use more mannitol than patients who limit their intake to less than 200 kcal or 200 mL fluid. The patients who limited food and/or fluid were more likely to bring a snack as opposed to a meal. Examples o f snacks (less than 200 kcal) com- monly consumed during dialysis include: ½ sand- wich, and ½ cup fruit; ½ cup carrots, 6 crackers, and 1 tbsp peanut butter; or 1 hard boiled egg, ½ bagel with ½ tbsp margarine.

A primary goal, as a result o f this study, is patient education. It is essential to help our pa- tients understand that the physical complications o f dialysis may be directly related to high inter- dialytic fluid gains; as well as fluid and /o r calorie intake o f 200 mL/200 kcal or more consumed during their dialysis treatment. The multidisci- plinary team will provide consistent education and reinforcement o f educational goals related to fluid and calorie intake to improve patient out- comes during and away from dialysis.

W e currently have a few patients who volun- tarily choose to not eat or drink during dialysis because they state they feel better when they abstain from consumption during the treatment. They are self-actualized and realize their behavior directly impacts their health and well-being dur- ing dialysis. The majority o f patients feel it is a burden to schedule meals around dialysis (diabetic patients and those who have lengthy travel times to and f rom the unit). A few patients believe we are trying to control their behavior by wi thhold- ing food and drink during dialysis.

After much discussion, we have decided to individualize the diet/fluid restriction order per patient. W e believe it is not necessary to restrict patients f rom eating and /o r drinking during di- alysis as long as they show stability doing so. A

1 O0 S T R O N G E T A L

multidiscipl inary effort wil l be used to identify patients at risk. Intensive patient educat ion will be ini t iated on "at risk" patients wi th close m o n - i tor ing by the team. A " N o eat a n d / o r drink during dialysis" order will be wr i t ten by the physician when , despite intensive educat ion, the patient experiences hypotens ion or o ther physical complicat ions dur ing dialysis.

Conclusion As part o f our mission at the Dialysis Cen te r o f

Lincoln, w e strive to provide high quality, p ro - gressive dialysis therapy, wh ich contributes to the quality o f life exper ienced by our patients. W e strive to provide our patients wi th the highest quality o f care and p romo te o p t i m u m health as a hemodialysis patient. W e at tempt to provide this care in the most efficient manner possible. W e feel to accomplish our mission each patient must have an individual care plan, inc luding diet.

The diet prescript ion must be individual ized to meet the best quality ou tcome for each patient. O n e diet is not best for all patients and as w e have seen wi th this research study, not all patients are adversely affected by food a n d / o r fluid consump- t ion during dialysis. Therefore , it may be in the

patients ' best interest to restrict food a n d / o r fluid consumpt ion during dialysis when they consis- tent ly experience hypotension. It is also in the best interest o f our patients to make all at tempts to el iminate the cause o f hypotens ion before bl indly treating the p rob lem wi th medicat ion.

Acknowledgment The authors thank June Smith for statistical guidance.

References 1. Shibagaki Y, Takaichi K: Significant reduction of the

large-vessel blood volume by food intake during hemodialysis. Clin Nephrol 49:49-54, 1998

2. Barakat MM, Nawab ZM, Yu AW, et al: Hemodynamic effects ofintradialytic food ingestion and the effects of caffeine. J Am Soc Nephrol 3:1813-1818, 1993

3. Zoccali C, Mallamaci F, Cicarelli M, et al: Postprandial alterations in arterial pressure control during hemodialysis in uremic patients. Clin Nephrol 31:323-326, 1989

4. Sherman RA, Torres F, Cody RP: Postprandial blood pressure changes during hemodialysis. Am J Kidney Dis 12:37- 39, 1988

5. Kopple JD: Therapeutic approaches to malnutrition in chronic dialysis patients: The different modalities of nutritional support. Am J Kidney Dis 33:180-188, 1999

6. Pennington J: Bowes and Church's Food Values of Por- tions Commonly Used (ed 17). Philadelphia, PA, Lippincott, 1998