Introduction of an enhanced recovery protocol for radical cystectomy


Nimalan Arumainayagam, Clinical Research Registrar, Bristol Urological Institute, Southmead Hospital, Bristol, BS10 5NB, UK.



To describe and assess an enhanced recovery protocol (ERP) for the peri-operative management of patients undergoing radical cystectomy (RC), which was started at our institution on 1 October 2005, as RC is associated with increased morbidity and longer inpatient stays than other major urological procedures.


An ERP was introduced in our institution that focused on reduced bowel preparation, and standardized feeding and analgesic regimens. In all, 112 consecutive patients were compared, i.e. 56 before implementing the ERP and 56 since introducing the ERP. The primary outcome measures were duration of total inpatient stay and interval from surgery to discharge, and the morbidity and mortality. Data were analysed retrospectively from cancer network and hospital records.


The demographics of the two groups showed no significant difference in age, gender distribution, American Society of Anesthesiologists grade, or type of urinary diversion. Re-admission, mortality and morbidity rates showed no statistically significant difference between the groups. The median (interquartile range) duration of hospital stay was 17 (15–23) days in the no-ERP group, and 13 (11–17) days in the ERP group (significantly different, P < 0.001, Wilcoxon rank-sum test). The median duration of recovery after RC was 15 (13–21) days in the no-ERP group and 12 (10–15) days in the ERP group (significantly different, P = 0.001, Wilcoxon rank-sum test).


The introduction of an ERP was associated with significantly reduced hospital stay, with no deleterious effect on morbidity or mortality.


enhanced recovery protocol


radical cystectomy


deep vein thrombosis.


Despite improvements in perioperative care, radical cystectomy (RC) is still associated with greater morbidity and prolonged inpatient stay after surgery than other urological procedures. In a recent European series the reported overall complication rate was 27.3% and the mean (sd) inpatient stay was 17.4 (4.7) days [1]. Bowel complications, and particularly paralytic ileus, are amongst the commonest problems.

Enhanced recovery protocols (ERPs) have been used successfully in patients undergoing colorectal surgery in the UK, with resulting reduced inpatient hospital stay [2]. Such protocols were initially pioneered in Denmark and have been shown to prevent a reduction in lean body mass, pulmonary function, oxygenation and cardiovascular response to exercise in such patients [3]. Furthermore, the use of routine preoperative mechanical bowel preparation in left-sided colorectal surgery has been deemed unnecessary and might indeed be harmful [4]. Bowel preparation has subsequently been shown to confer no advantage in patients undergoing RC [5].

Traditional management after surgery has involved resting the gastrointestinal tract until it regains its function. However, studies have confirmed the safety of early postoperative feeding in elective colorectal surgery [6], and such findings have been reported in patients undergoing RC [7]. Based on these reports and such evidence, we developed an ERP for patients undergoing RC and implemented it in our unit, and analysed the effect of this protocol.


After an exhaustive literature search and in consultation with surgical, anaesthetic and nursing staff, an ERP was designed, shown in the Appendix. The key features of this ERP involved stopping the use of mechanical bowel preparation before RC, which allowed patients to be admitted to hospital on the day before RC (as opposed to 2 days before), and the implementation of early enteral feeding and mobilization as tolerated soon after RC.

The ERP was introduced in our institution on 1 October 2005; we compared a series of 112 consecutive patients, i.e. 56 before implementing the ERP and 56 afterward. All three surgeons who conducted RCs at our centre during this period agreed with and were involved in the formulation of the ERP and its subsequent implementation.

Data were analysed retrospectively from the cancer network and hospital records. The primary outcome measures were the duration of total inpatient stay and the interval from surgery to discharge, and the complications, re-admissions and mortality.


The demographics of the two groups are outlined in Table 1, below; there was no statistical difference in any of the variables assessed (using the Pearson chi-square test for gender, American Society of Anesthesiologists grade, type of urinary diversion, previous major abdominal surgery, transfusion; a two-sample t-test for age; and Fisher’s exact test for radiotherapy or chemotherapy before RC). Nor was there a significant difference (Pearson chi-square test) in the incidence of minor and major complications, including superficial wound dehiscence, wound infection, intra-abdominal bleed, pelvic collection, urine leak, mechanical bowel obstruction, fistulae, cardiac arrest, Clostridium difficile diarrhoea, atrial fibrillation, line infections and deep vein thrombosis (DVT) (Table 1).

Table 1. 
The demographics, complications, reasons for re-operation, reasons for re-admission and the outcome
VariableBefore ERPAfter ERP
  1. MI, myocardial infarct; AF, atrial fibrillation; ARF, acute renal failure. *Vaginal pouch fistula, required surgery to repair; †all enterocutaneous, healed with conservative management.

No. of patients5656
Mean age, years65.965.9
Ileal conduit4547
Neobladder 11 9
ASA grade
 1 8 11
 3 9 5
No. patients
 Transfused peri-op2419
 Radiotherapy before RC 3 2
 Chemotherapy before RC 2 2
 Previous major abdominal surgery1316
Complications, n
 Superficial wound dehiscence 8 2
 Wound infection 0 1
 Intra-abdominal collection 2 1
 Urine leak 1 1
 Bowel obstruction 0 1
 Intra-abdominal bleeding 0 2
 Fistula 1* 3
 Diarrhoea (C. difficile–ve) 3 1
 Diarrhoea (C. difficile+ve) 3 2
 Pyelonephritis 2 0
 DVT 1 1
 Peripheral/central line infection 0 2
 Lower respiratory tract infection 4 3
 Other medical (MI, AF, ARF) 4 2
Reason for return to theatre
 Urine leak 2 1
 Postoperative bleeding 0 2
 Repair of vaginal pouch fistula 1 0
 Mechanical bowel obstruction 0 1
 Reason for re-admission
 Pelvic abscess (drained by TRUS) 0 1
 Blocked catheter 0 1
 Acute renal failure (pre-renal) 0 1
 Vomiting (settled) 2 0
 Wound infection 1 0
 Superficial wound dehiscence 2 0
 DVT 1 0
  Time to defecation 6 6
  Total stay1713
  Stay after RC1512
 Return to theatre 3 4
 Deaths within 28 days (days) 1 (7) 1 (20)
 Re-admissions within 28 days 5 3

The number of patients needing a re-operation was comparable (three before ERP and four after ERP). There was no significant difference in the re-admission rate (Fisher’s exact test). The reasons for re-operation and re-admission are also outlined in Table 1.

There was one death after RC in each group, from acute myocardial infarction, at 7 days (before ERP group), and at 21 days (ERP group), which were both before discharge; these patients were not included in the subsequent analysis.

The time to first defecation was equivalent, with a median duration of 6 days in each group. The median (interquartile range) duration of hospital stay was 17 (15–23) days before ERP and 13 (11–17) days after the ERP (statistically significant, P < 0.001, Wilcoxon rank-sum test). The median duration of recovery after RC was 15 (13–21) days before ERP and 12 (10–15) days after the ERP (statistically significant, P = 0.001, Wilcoxon rank-sum test). These results are summarized in Table 1 and shown as box plots in Fig. 1.

Figure 1.

Box plots of a, total hospital stay and b, hospital stay after surgery.


Our experience of implementing an ERP for the perioperative management of RC resulted in a significantly reduced total and postoperative hospital stay, with no effect on morbidity or mortality. Comparative data are difficult to find, as many studies are based in the USA, where the criteria for discharge from hospital differ. Such studies in the USA showed the benefit of using a clinical care pathway in reducing postoperative hospital stay, with a potential median hospital stay of 7 days, and with 74% of patients discharged by 8 days after surgery [8]. Pruthi et al.[7] reduced their mean time to discharge after RC from 10 to 5.2 days by using a care plan. They continued to use bowel preparation before surgery and routine nasogastric tube insertion immediately after RC (removed on day 1). In addition, their feeding regimen was less aggressive (with restricted fluids instituted on day 2, unrestricted fluids on day 3, and regular diet on day 4), than in our ERP.

Although recovery times from these studies appear to be much shorter, the results should be considered in context, because their service is based on the health system model in the USA. The present study is based on the model of healthcare delivery in the UK and thus direct comparison with the previous results is difficult.

There might be many reasons for the present results. First, the reduced bowel preparation might be a factor. Patients before the ERP received two sachets of bowel-cleansing solution (e.g. sodium picosulphate), which required admission 2 days before RC, rather than 1 day in the ERP group. This would partly account for the difference in total stay. Mechanical bowel preparation is also known to have detrimental effects, e.g. metabolic acidosis, hyponatraemia, hypokalaemia, and hypochloraemia [9,10]. It might have therefore also affected the recovery times after RC. Removing the epidural on day 3 might also contribute to resolving postoperative ileus, and hence recovery. However, we found that resolving such ileus does not seem to be a factor in the lower postoperative stay in the present study. Other explanations include the possibility that establishing an ERP ensures that factors such as nutrition and mobilization are repeatedly addressed. The use of nutritional supplements in the peri-operative period might also be of benefit.

There might be more consistency in routine postoperative management decisions. The presence of an ERP might also ensure that decisions influencing recovery are not delayed over periods such as weekends, when management might be decided by a surgeon or trainee who might not be part of the patients’ usual team.

Complications of RC will always result in a longer inpatient hospital stay, as can be seen by the outlying scatter points in the box plot (Fig. 1), and such a protocol will not expedite or improve recovery after RC in this subset of patients. However, an ERP improves the recovery in those patients who do not have major complications.

In conclusion, introducing an ERP for RC was associated with a significantly reduced total hospital stay and postoperative stay, with no deleterious effect on morbidity or mortality.


None declared.


The ERP focused on reduced bowel preparation and standardized feeding and analgesic regimens. All patients received that same prophylaxis for DVT as used before the ERP. This comprised 20 mg s.c. enoxaparin sodium, low molecular weight heparin, given on the evening before surgery and each evening thereafter. Patients also wore compression stockings for the duration of their inpatient stay, and had intermittent mechanical calf compression during RC. The protocol consisted of two Fortijuices/Fortisips (Nutricia Limited, Trowbridge, Wilts, UK) daily in addition to:

Day before RC

Normal breakfast

Admit to hospital

Unrestricted clear fluids

Refer to dietician

Stoma therapist to see patient

Assess social circumstances and refer if needed

Day of RC

Clear fluids up to 4 h before RC, then nil by mouth

Restart clear fluids as tolerated when in recovery

Start food chart

Epidural analgesia in-situ.

After RC – Day 1

Free fluids as tolerated

Female patients, remove vaginal pack

Mobilise and refer to physiotherapist

Metoclopramide regularly

Ranitidine three times daily i.v. or twice daily orally

Remove drain if draining <50 mL in 24 h

Flush 20 mL into neobladder, 2-hourly for 12 h and then 4-hourly

Day 2

Light diet as tolerated

Mobilise and encourage self care (catheter care/flushing in neobladders, and stoma bag emptying in patients with a conduit)

Day 3 and 4

Remove epidural on day 3

Continue to mobilise and encourage self care

Light diet as tolerated

Start planning for discharge

Continue to mobilise and encourage self-care

Day 5, 6 and 7

Dietician to assess nutritional requirements on day 5

If a patient is not eating or drinking after 5–6 days but with bowel activity, then start nasogastric feeding. If there is no bowel activity then start total parenteral nutrition.

Day 8

Stents out (no stentogram); patient to stay at least 24 h after stent removal

Day 10

Remove clips

Day 11–14

Continue as previous and schedule for return to home