Snapshot of transurethral resection of bladder tumours in the United Kingdom Audit (STUKA)

Authors


Correspondence: Timothy O'Brien, The Urology Centre, Guy's Hospital, Great Maze Pond, London, SE1 9RT, UK.

e-mail: tim.obrien@gstt.nhs.uk

Abstract

Objectives

  • To determine the quality of transurethral resection of bladder tumour (TURBT) in the UK.
  • To evaluate the utility of a novel ‘snapshot’ methodology in carrying out national audits.

Patients and Methods

  • Every consultant Urologist in the UK was asked to contribute details of their first patient with a new bladder cancer treated with TURBT after midnight of 31st January 2010.
  • Responses were received from 192 consultants.

Results

  • The median (range) time from referral to first Urology appointment was 11 (0–161) days, and the median (range) time from first appointment to TURBT was 27 (1–588) days.
  • In all, 12 (6.3%) patients underwent photodynamic diagnosis-assisted TURBT and 119 patients (61%) received a dose of Mitomycin C after TURBT.
  • The rate of major complications was low, with five incidences (2.6%) of bladder perforation.
  • There was no record of muscle present in resected specimens in 40 cases (20.8%) and resection was considered incomplete in 26 cases (13.5%). In all, 31 patients (16.1%) underwent early re-resection with residual tumour or carcinoma in situ detected in 17 cases, although no tumour was upstaged.
  • Of the 37 patients classified with intermediate-risk non-muscle-invasive bladder cancer (NMIBC), there were nine recurrences (24.3%) at 3 months, and 13 recurrences (35.1%) at 1 year. Newly presenting MIBC managed with currently available treatments has a high mortality rate of 33.3% at 1 year.

Conclusions

  • The quality of TURBT in the UK is high.
  • Areas for improvement include the timeliness of diagnosis and treatment, and improved care of patients with intermediate-risk NMIBC and MIBC.
  • The ‘snapshot’ methodology is promising but widening participation is a priority.
Abbreviations
CIS

carcinoma in situ

DOH

Department of Health

EORTC

European Organisation for Research and Treatment of Cancer

NCEPOD

the National Confidential Enquiry into Patient Outcome and Death

(N)MIBC

(non-)muscle-invasive bladder cancer PDD, photodynamic diagnosis

TURBT

transurethral resection of bladder tumour

Introduction

Improving the quality of care delivered to patients must be at the heart of every urologist's practice. Patients, and the purchasers and regulators of healthcare need to have the confidence that whenever and wherever patients present, they will receive high quality care. Central to this process is understanding the limitations and strengths of the care currently being delivered. In this regard the scientific literature may be of relatively limited value, as it is so often dominated by single-centre studies from a few centres of research excellence [1]. The care delivered and the results achieved in centres of excellence may be very different from that delivered to the population at large. Furthermore, surgeons and surgical conferences tend to focus on complex procedures rather than the much more commonly performed ‘bread and butter’ procedures, variations in the quality of which would impact on far more patients.

In the UK >10 000 patients per year are diagnosed with a new bladder cancer [2], with virtually all patients undergoing an initial transurethral resection of bladder tumour (TURBT). It is well recognised that non-muscle-invasive bladder cancer (NMIBC) has a very high recurrence rate of >50% after initial TURBT [3]. Two reasons advanced for this are incomplete initial resection, e.g. residual disease not recurrence, and implantation of floating tumour cells [4]. Studies of early re-resection suggest that over half of recurrences may be due to the presence of residual tumour [5]; and studies of perioperative chemotherapy show a 30% reduction in early recurrence with ‘single shot’ postoperative chemotherapy [6]. Both of these factors impact on the quality of bladder cancer care and both can be influenced by the clinician at or immediately after TURBT.

Studies from Europe suggest that the quality of TURBT may vary widely [7]. Despite the importance of a high quality initial TURBT, to date there has been no national study to evaluate performance. One approach would be to attempt prospectively to audit every initial TURBT carried out across the UK over a 12-month period. Whilst clearly of interest, this would be a massive undertaking, would be very expensive, time consuming and in units without research support probably doomed to fail through ‘research fatigue’. As a research methodology it would also not lend itself easily to being repeated at a later date to judge whether standards were improving or deteriorating. Might there be another way? One very successful approach to national audit in the UK has been that used by the National Confidential Enquiry into Patient Outcome and Death (NCEPOD) to investigate perioperative deaths in 1997 [8]. Every surgical consultant was approached and asked to contribute details of just one operation performed by them within a specified time frame. This method carries the attraction of drawing the sample from a very widely dispersed group of surgeons whilst not over-burdening each surgeon with an undeliverable administrative task. It has proved very successful in improving standards, particularly in emergency care.

In the present study, we have adopted a modified NCEPOD methodology in designing a national study to determine the quality of the initial management of bladder cancer by TURBT across the UK by attempting to study just one TURBT from every urologist in the country. Our aims were to determine:

  • A ‘snapshot’ of the current demographics of patients with newly presenting bladder cancer.
  • The quality of TURBT using conventional surgical and pathological parameters.
  • Areas for improvement in administrative processes experienced by patients with bladder cancer.
  • Recurrence rates at 3 and 12 months after initial TURBT, and 90-day mortality rate.
  • Priorities for further research and training.
  • Evaluate the utility of this novel ‘national snapshot’ methodology.

It was felt that the metrics in the study should reflect the domains of quality as set out by the Institute of Medicine in their landmark publication ‘Crossing the Quality Chasm’ [9]; viz, care that is effective, safe, efficient, timely, equitable, and patient-centred. Specific surgical metrics were determined by evaluation of contemporary review articles on bladder cancer, and by review of guidelines issued by the various urological associations [10-13]. These include the timeliness of surgery, TURBT complication rate, the appropriate use of adjuvant intravesical therapy after TURBT, the presence of muscle tissue in resected specimens, the performance and results of early re-resections, and recurrence and mortality rates.

Patients and Methods

In all, 798 urology consultants in the UK were approached by e-mail in June 2011, inviting them to retrospectively identify the first patient with newly presenting bladder cancer undergoing TURBT under their care at an NHS institution after midnight of 31 January 2010. Data was entered into an online proforma via the BAUS data and audit system (http://audits.baus.org.uk). The study was launched on 13 June 2011, and urologists were given until 31 August 2011 to enter their data. A copy of the study proforma is appended (Appendix S1).

Results

Of the 798 urologists approached, we received 192 responses (24%). In all, 174 of these (91%) urologists were existing users of the BAUS registry. Overall, 82% of data fields were completed.

Patient Demographics

Patient demographics and reasons for presentation are shown in Table 1.

Table 1. Patient demographics
VariableValue
Gender, n (%): 
Male149 (77.6)
Female43 (22.4)
Mean (range) age, years73 (27–93)
N (%): 
ECOG status 
085 (44.3)
157 (29.7)
234 (17.7)
37 (3.6)
Presentation: 
Visible haematuria155 (80.7)
Non-visible haematuria (NVH)9 (4.7)
LUTS3 (1.6)
LUTS and NVH8 (4.2)
UTI2 (1.0)
Incidental finding8 (4.2)
Other2 (1.0)
No response5 (2.6)

Quality of Administrative and Clinical Processes

The median (range) length of time from referral to first urology appointment was 11 (0–161) days. The median (range) length of time from first urology appointment to TURBT was 27 (1–588) days. In all, 82 patients (42.7%) were seen in a dedicated haematuria clinic, 49 (25.5%) in a ‘one-stop’ clinic, 43 (22.4%) in a general Urology clinic, and one (0.5%) in a specialist bladder cancer clinic. A range of investigations were used, including flexible cystoscopy (164 patients, 85.4%), urinary tract ultrasound (130, 67.7%), CT urogram (74, 38.5%), IVU (36, 18.8%), MRI of the pelvis (eight, 4.2%), urinary cytology (33, 17.2%) and other urinary markers (one, 0.5%). Of the 164 patients who were investigated with a flexible cystoscopy, 114 (69.5%) had their cystoscopy on the same day as their first urology appointment. Of the remaining patients who had their flexible cystoscopies at a separate appointment, this took place at a median (range) of 15 (1–90) days after their first appointment.

In all, 53.6% of TURBTs were performed by a consultant urologist, 34.4% by a specialty registrar, 8.9% by non-consultant permanent staff, and the remainder not specified. Most procedures were performed under standard white light; with only 12 patients (6.3%) undergoing photodynamic diagnosis (PDD)-assisted TURBT. In all, 119 patients (61%) were given Mitomycin C after TURBT. Of these, 32 patients (26.9%) received Mitomycin C <2 h after surgery, 48 (40.3%) received it within 2–6 h after surgery, and 39 (32.8%) received it 6–24 h after surgery.

Most procedures (75%) did not have any complications. Where complications did occur, they were mostly (20%) minor complications (Clavien grades 1 and 2) [14]. There were seven (3.6%) grade III complications, and one (0.5%) grade IV complication. There were five incidences (2.6%) of bladder perforation, three of which were small extraperitoneal perforations, one was a large extraperitoneal perforation, and one was intraperitoneal.

In 152 cases (79.2%), muscle was present in the resected specimens. Muscle was absent in 31 cases (16.1%), and no response was recorded in the remaining nine (4.7%). Resection was considered full and complete in 159 (82.8%) cases, partial and incomplete in 26 (13.5%) cases, and not recorded in seven (3.6%) cases.

Cases where there was either no muscle in resected specimens or where resection was incomplete were further classified according to the European Organisation for Research and Treatment of Cancer (EORTC) risk stratification system [15]. The subsequent management plans for these patients are shown in Figs 1 and 2.

Figure 1.

Subsequent management for cases where muscle was absent in resected specimens.

Figure 2.

Subsequent management for cases where resection was incomplete.

In all, 31 patients (16.1%) underwent early re-resection. Table 2 shows the indications given for re-resection. Residual tumour was found in 14 cases (45.2%), while three cases (9.7%) had carcinoma in situ (CIS) only on re-resection (Table 3). No tumour was upstaged on repeat resection.

Table 2. Indications for re-resection
Reasons for re-resectionN (%)
Second resection (high-risk histology)20 (64.5)
Restaging resection (no muscle in specimen)6 (19.4)
Repeat resection (incomplete first resection)3 (9.7)
Other2 (6.5)
Table 3. Positive findings on re-resection
Initial histologyHistology on re-resection
G3pT2+G1pTx
G3pT1 + CISG3pT1 + CIS
G3pT1 + CISG2pT1 + CIS
G3pT1 + CISpTa + CIS
G3pT1 + CISCIS only
G3pT1G3pT1
G3pT1G3pT1
G3pT1G3pTa + CIS
G3pT1G3pTa
G3pT1G2pTa
G3pT1CIS only
G3pT1CIS only
High-grade pT1High grade
G2pT1G3pTa
G2pT1G2pT1
G1pTaG2pTa
G1pTaG1pTa

Further Management

In all, 148 patients (77.1%) were diagnosed with NMIBC, and 30 patients (15.6%) with MIBC. We had insufficient data to determine the diagnosis for 14 patients. The subsequent management plans for patients diagnosed with NMIBC and MIBC are shown in Tables 4 and 5, respectively.

Table 4. Further management of NMIBC stratified by EORTC risk
EORTC risk groupManagementN (%)
Low, n = 61Follow-up cystoscopy at 3 months57 (93.4)
Intravesical chemotherapy3 (4.9)
Early resection1 (1.6)
Intermediate, n = 37Follow-up cystoscopy at 3 months30 (81.1)
Intravesical chemotherapy6 (16.2)
Other1 (2.7)
High, n = 50BCG31 (62)
Follow-up cystoscopy at 3 months6 (12)
Cystectomy4 (8)
Radiotherapy2 (4)
Intravesical chemotherapy1 (2)
EMDA Mitomycin C + BCG1 (2)
Other5 (10)
Table 5. Further management of MIBC
Management of MIBC (n = 30)N (%)
Cystectomy +/– neoadjuvant chemotherapy12 (40)
Radiotherapy +/– neoadjuvant chemotherapy9 (30)
Early re-resection1 (3)
Other8 (27)

Outcomes

The rate of recurrence at 3 and 12 months, and the incidence of death at 12 months are shown in Table 6, stratified according to EORTC risk group.

Table 6. Outcomes stratified by EORTC risk
EORTC risk groupRecurrence (3 months), n (%)Recurrence (12 months), n (%)Death (12 months), n (%)
Low, n = 616 (9.8)11 (18.0)0
Intermediate, n = 379 (24.3)13 (35.1)0
High, n = 5010 (20)14 (28)1 (2.0)
MIBC, n = 3010 (33.3)

Discussion

Despite the importance of the initial TURBT in both staging and determining outcomes in newly presenting bladder cancer, to our knowledge this is the first study to evaluate its quality across a nation. The quality of the surgical procedure itself in the UK appears to be very high as evidenced by low complication rates, a very high percentage of cases with muscle in the specimen, e.g. correct staging; zero cases upstaged on early re-resection; and low recurrence rates for low-risk disease. Nevertheless, the study has identified several areas where the care of patients with bladder cancer could be significantly improved.

Timeliness is an important determinant of quality. Issues relating to delays in care in the NHS can dominate the political agenda in the UK. The Department of Health (DOH) stipulates that all patients referred by their GP for suspected cancer should be seen by a specialist ≤2 weeks of referral, and that the wait from diagnosis to first definitive treatment should be no longer than 31 days. In the present study, whilst the median time from referral to first urology appointment was 11 days, only 64% of patients were seen within the DOH target time of 2 weeks after referral and only 60.9% of patients met the 31-day DOH target for time to first treatment (e.g. TURBT) from diagnosis. ‘One-stop’ haematuria clinics have been in operation in the UK for >25 years but only 70% of patients had a diagnostic flexible cystoscopy on the day of the initial consultation.

In the present study, the quality of the surgical procedure appears very high. In all, 53.6% of procedures were performed by a consultant urologist, and a consultant was present at the operation in 71.9% of cases where the main operating surgeon was not a consultant. It has previously been shown that surgeon seniority is independently associated with a lower rate of recurrence after TURBT [16]. The frequent presence of a consultant urologist in theatre while a junior urologist is operating also reflects a good level of support and teaching provided for trainees. Major complications after TURBT were rare with a bladder perforation rate of 2.6%, similar to the frequency of 1 in 50 quoted by BAUS [17]. Most patients (61%) received a single instillation of Mitomycin C ≤24 h of surgery and in those instances where Mitomycin C was not given, reasonable contraindications were present in over two-thirds of cases (invasive or high-grade tumour, concerns about bladder perforation, haematuria, patient refusal etc.). In only four cases did respondents state that their local policy was not to give Mitomycin C after TURBT.

The presence of detrusor muscle in resected specimens is a frequently used surrogate marker for the quality of TURBT. In the present survey, 20.8% of cases had no record of muscle in the resected specimens, which is a lower rate than that quoted by several large volume studies [5, 18, 19]. Furthermore, most cases (75%) where muscle was recorded to be absent turned out to be low- or intermediate-risk tumours where the risk of understaging is less likely to be significant. In those cases where muscle was absent and the tumour was high risk, all patients went on to have appropriate further management, either with re-resection, or definitive treatment. Resection was thought to be incomplete in 26 cases (13.5%) but many of these were invasive tumours, suggesting that the reason for incomplete resection was a reflection of the extent of the tumour rather than surgical ability. In all, 18 of these patients (69.2%) went on to have either early re-resection or definitive treatment, and of the eight remaining patients, five had died ≤12 months after diagonsis, indicating the advanced nature of the disease at diagnosis in these patients.

Early re-resection was performed in 31 patients (16.1%), and of these, there was residual tumour or CIS in 17 cases (54.8%). No tumours were upstaged at early re-resection, which is again significantly lower than the frequency described in large series of early re-resections [20], and could be taken to indicate a high quality of the initial TURBT.

A notable finding of the present study was the high rate of recurrence in patients with intermediate-risk tumours (24.3% at 3 months and 37.8% at 1 year). It would appear that adjuvant treatment for disease in this category is underused. Of 37 patients diagnosed with intermediate-risk NMIBC, only six (16.2%) went on to have adjuvant treatment with intravesical chemotherapy. This clearly does not adhere to European Association of Urology (EAU) guidelines, which recommend that patients at intermediate risk of recurrence or progression should be managed with a course of chemotherapy or at least 1 year of intravesical BCG instillations [7]. One reason may be that urologists in the UK rarely use EORTC risk tables to formally calculate risk of recurrence in individual patients.

Interestingly, PDD-assisted TURBT was used in only 6% of patients. PDD with hexaminolevulinate has been licensed for use in Europe for 7 years and urological associations across Europe including BAUS have all issued guidelines recommending the use of PDD at different stages of the management of NMIBC [21], but UK surgeons clearly remain sceptical despite studies showing improved diagnostic assessment of bladder tumours with PDD [22].

The ‘snapshot’ methodology used in the present clinical study appears to be a promising tool for use in national surgical audits. It has the advantages of providing a very up to date perspective on surgical performance across a wide population, is quick and cheap to carry out, and should be easy to repeat. One limitation of the present study is the apparently relatively low response rate of 24%. It is possible that the only surgeons who responded to the study were those with an interest in the disease or who acknowledge the power of national surgical audit. A positive bias could also have been introduced if we make the assumption that better organised units with more favourable outcomes are more likely to contribute data to studies of this type, and also by surgeons submitting their own data. However, this is the first study of its type and the challenge now is to widen participation. Only when data is included from every surgeon in the UK performing the operation can the ‘snapshot’ be considered complete. The reasons for non-participation are not completely clear, but all non-responders were contacted by email to discover what their reasons were. Responses were received from 77 of the 506 consultants contacted (15.2%). Of these, 26 stated that they do not perform TURBT, 24 did not have time to respond, seven did not have access to the BAUS registry, and the remaining 20 gave various other reasons, mostly that they did not remember receiving the invitation to participate, or were unable to access case notes. In addition, over 40 consultant urologists practising in June 2011 had not been appointed to consultant positions at the time of the TURBT analysed, e.g. February 2010, so could not submit data.

In conclusion, the demographics of patients undergoing TURBT in the UK for new bladder cancer are comparable to those previously published. The quality of surgery offered to patients in the UK is high but this impression is dependent on a strategy of self-reporting by the surgeon and ideally would need external validation. Notable areas for improving quality of bladder cancer care are the same-day availability of flexible cystoscopy whenever a diagnosis of bladder cancer is suspected; reductions in delays to resection from diagnosis; improved care of patients with intermediate-risk disease and treatments to improve the dismal prognosis of many patients with muscle-invasive disease. This feasibility study of ‘snapshot’ methodology is a promising tool for future national audits but widening participation is a priority. The ‘snapshot’ can only be considered truly representative when data is available from every surgeon in the UK.

Acknowledgements

We would like to acknowledge and thank all Urologists who took time to contribute data to this study. (All names attached in Appendix S2.)

Conflict of Interest

None declared.

Ancillary