Oral Nutrition After Regular Radical Cystectomy

Design

A prospective interventional non-randomized before-after trial was conducted. Eligible patients admitted from 01/02/2009 were consecutively assigned to the control group until the predefined sample size was reached. After enrollment and follow-up of control patients had been finished, all newly admitted eligible patients were consecutively assigned to the interventional group.

In addition to the before-after study, a sample of eligible patients was collected from March 2011 to March 2013 (follow-up group) to examine the post-study impact of the ONP.

Patients

All adult patients admitted to the 48-bed urological ward in the 1850 bed University Hospitals Leuven and treated with elective and regular RC, whether or not for underlying cancer, were eligible for inclusion. Three different diversions are performed: an ileal conduit, an orthothopic neobladder (N-pouch)12 or a continent cutaneous diversion. The surgical procedure was considered regular if the patient returned after the procedure to the urological ward and was not transferred to the intensive care unit. If the latter was deemed necessary by the treating surgeon, the patient would be subjected to the intensive care nutrition protocol. Therefore, only regular RC patients were included.

Exclusion criteria were transfer to the intensive care unit, preoperative contra-indications for EN and discharge to a rehabilitation center. Contra-indications for EN were defined as intestinal obstruction, malabsorption, multiple high-output fistula or intestinal ischemia8. Uretero-ileal or ileo-ileal anastomoses, resulting from RC, were not considered as a contra-indication for EN9,10. Discharge to a rehabilitation center could surpass one of the criteria for discharge, i.e. the ability of independent mobilization, and could therefore influence the postoperative LOS.

Care practices other than nutrition management

Preoperatively, all patients received bowel preparation with a 2 liter osmotic laxative (macrogol 3350 100 g, sodiumsulphate 7.5 g, NaCl 2.69 g, KCl 1.02 g, sodium salt of ascorbic acid 5.9 g and ascorbic acid 4.7 g). Before incision, a loading dose of 0.05 ml/cm epidural analgesics (levobupivacaine 2,5 mg/ml and sufentanil 1 mcg/ml) was administered. During surgery, a central venous catheter, a nasogastric or percutaneous gastrostomy tube and ureteral stents were inserted. The nasogastric or percutaneous gastrostomy tube was removed on the first or second postoperative day after restoration of transit (defined as a residual gastric volume of less than 50 ml/24h). Acetaminophen, non-steroidal anti-inflammatory drugs and patient-controlled epidural analgesia (PCEA; levobupivacaine 2,5 mg/ml and sufentanil 1 mcg/ml) were available for postoperative acute pain management. Intraoperative antimicrobial prophylaxis consisted of cefazoline and metronidazole. Thromboprophylaxis consisted of enoxaparin. Anti-emetics and other prokinetic drugs were not routinely administered but only prescribed when needed. All treating surgeons were fully trained at the start of the study.

After a minimum of 10 days, ureteral stents were removed if the following 3 criteria were met: restoration of transit, presence of bowel movements and absence of fever (defined as body temperature <38,3 °C). Prophylactic levofloxacin 500 mg once daily was administered after stent removal, until the first consultation. Patients were discharged if the following 4 criteria were fulfilled: 1) removal of all drains and stents, 2) the absence of fever (body temperature < 38,3 °C), 3) the ability to tolerate solid food and 4) the ability of independent mobilization.

The abovementioned care practices (anesthesia modalities, surgical techniques, doses of narcotics, bowel stimulants and anti-nauseants, and mobilization practices) did not alter during the course of the study.

Nutrition management

In the control group, PN was initiated immediately after surgery and continued until the patient was able to tolerate solid food. PN infusion therapy consisted of Olimel® (Baxter S.A., Lessines, Belgium) N7E 1000 ml, 1500 ml or 2000 ml; a parenteral solution with a total energy of 1100 kcal/1000 ml and containing polyamino-acids (43.75g/1000 ml), glucose (140g/1000ml), lipids (40g/1000ml) and electrolytes. Cernevit® (Baxter S.A., Lessines, Belgium), a multivitamin powder for injection, and Addamel® (Fresenius Kabi, Schelle, Belgium), a concentrate for injection containing trace elements, were added daily to the PN. The amount of PN administered depended on the non-protein requirements, calculated by 30 ± 10% kcal/kg ideal body weight13. Extra fluids, up to two liter per day, were given intravenously, at the discretion of the treating physician.

In the interventional group, the ONP was implemented (Table 1). Oral intake was increased progressively with oral fluids and easily digestible food, independent of bowel movements. The corresponding energy content from the meals and oral fluids were calculated. Fortimel Jucy®, 200 ml containing 300 kcal, was used as the formulary energy sip. Extra fluids, up to two liter per day, were given intravenously, at the discretion of the treating physician. Nurses verified daily whether the patient was able to tolerate the ONP. If the patient tolerated the ONP well, the oral intake was considered equal in terms of calories as the corresponding oral meal in the ONP. From the sixth day, the patient was allowed to eat at will. Only if oral intake remained insufficient after 5 days, which was left to the opinion of the treating physician, PN could be initiated in this group.

Data collection

The following baseline demographic data were retrieved from medical files: patient characteristics (gender, age, weight, length, body mass index, age-adjusted Charlson comorbidity index14, indication for RC (oncologic or neurogenic) and bladder cancer staging (TNM classification15)), surgical aspects (type of urinary diversion, attending surgeon, intra-operative blood loss and duration of procedure), postoperative ileus (POI) and day of nasogastric tube removal. The age-adjusted Charlson comorbidity index is a method for classifying comorbid conditions. As the increase in elderly population and concurrent increase in oncologic disease (and thus in bladder cancer) emphasized the importance of understanding the interaction between age an comorbid illness on life expectancy, the age -adjusted version was used. POI was defined as the inability to tolerate solid food, the need to place a nasogastric tube in suction, the need to stop oral intake due to abdominal distention (evaluated by the treating clinician), nausea or emesis.

Endpoints

The primary endpoint was LOS from surgery to discharge.

Secondary endpoints were the number of patients with successful implementation of the ONP, as well as number of administered PN infusion bags per patient, time to resumption of full diet, number of patients with one or more postoperative complications (POC), number of POCs per patient, type and severity grade of POCs and the incidence of catheter related bloodstream infections (CRBSIs). The implementation of the ONP was considered successful if the protocol could be completely applied and no PN was needed. The type and severity of POCs were classified using the Clavien-Dindo classification16. POCs were classified into 11 categories and stratified by severity. CRBSI, a common PN-related complication6, was defined as bacteremia or fungemia in a patient having an intravascular device and more than one positive blood culture result obtained from a peripheral vein, clinical manifestations of infection (e.g., fever, chills, and/or hypotension) and no apparent other source for blood stream infection17.

Post-study impact

An additional patient cohort was assembled, using the same inclusion and exclusion criteria as described above. The LOS and number of PN infusion bags after surgery were determined and compared to the results obtained from the intervention group.

Analysis of PN associated costs

The avoided PN cost per patient were calculated. For every group, the total number of PN infusion bags was divided by the number of patients and multiplied by € 60, which was the average actual cost in euros of one bag of Olimel® in Belgium at the time of the study. This resulted in a direct and comparable PN infusion bag related cost per patient per group.

Statistical analysis and sample size calculation

Statistical analysis was performed by L-Biostat, Catholic University Leuven, using SPSS package (SPSS Statistics 20.0 for Windows). Chi-square and Fisher's Exact tests were used for categorical data. The unpaired T-test or the Wilcoxon rank sum test was used for normal and non-normal distributed continuous variables respectively. Baseline demographic data and the outcome measures were considered statistically significant if p-values were < 0.05. The Bonferroni correction was used to account for multiple comparisons (6 secondary outcome measures). Therefore, secondary outcome parameters were considered statistically significant if p-values were < 0.008. To asses factors influencing LOS, general linear model statistics - univariate analysis followed by ANOVA - were carried out.

Preliminary data from our institution showed that all RC patients receiving PN directly after surgery were discharged after a mean of 19.3 ± 5.6 days. We expected to reduce our LOS with 3 days from 19 to 16 days by implementing the ONP. Assuming that variances for both groups are comparable, 88 (2x44) patients were needed, giving at least 80 % power (α=0.05, one-tailed) to reject the null hypothesis defined as no difference in postoperative LOS between the two groups.

The study was approved by the hospital's Ethics Committee and registered on clinical trials.gov n° NCT01777126. No informed consent was obtained as the introduction of the ONP was part of continuous improvement of standard of care in our hospital.