Adoption, consistency, and usability of NICE vs EAU guidelines in low‐ and intermediate‐risk NMIBC: a cross‐sectional survey of UK urologists

Abstract

Objectives

To evaluate current adoption of guidelines for low‐ and intermediate‐risk non–muscle‐invasive bladder cancer (NMIBC) in the UK and to assess clinicians' ability to apply their chosen guideline consistently to real‐world clinical scenarios.

Patients and Methods

This study is a cross‐sectional survey distributed to clinicians involved in the management of NMIBC in the UK. Respondents reported their unit‐level and personal guideline adoption (National Institute for Health and Care Excellence [NICE] vs European Association of Urology [EAU]) and completed three case‐based real‐life scenarios derived from NMIBC cases with discordant risk stratification between NICE and EAU guidelines. Cases were randomly selected from a local database using a computer‐generated random number function in Microsoft Excel. Concordance between respondents' decisions and their stated guideline was assessed for risk stratification and surveillance duration.

Results

A total of 174 eligible responses were analysed. Most units (38%) followed NICE guidelines, but individual clinicians most frequently reported that they were following EAU guidelines (39%). Those who followed NICE guidelines demonstrated significantly higher compliance with the duration of follow‐up as recommended in the guidelines than did those who followed EAU guidelines (78.9% vs 49.2%, P < 0.001). Deviations from the guidelines for those who followed NICE were predominantly in duration of surveillance, which was longer than recommended. Conversely, those following EAU guidelines frequently opted for shorter‐than‐recommended follow‐up (P < 0.001). Consistent guideline‐based risk stratification was achieved in 77.1% of those following NICE guidelines and 34.1% of those following EAU guidelines (P < 0.001). Acceptable recurrence risk at discharge did not differ between guideline groups (P = 0.278).

Conclusions

Substantial variation exists in NMIBC guideline adherence and real‐world application in the UK. NICE guidance was associated with greater consistency and utility compared with EAU guidance for low‐ and intermediate‐risk disease. These findings highlight the need for clearer, unified recommendations to promote equitable care and rational use of surveillance resources.

Abbreviations

BCG

Bacillus Calmette–Guérin

CI

confidence interval

EAU

European Association of Urology

EORTC

European Organisation for Research and Treatment of Cancer

HG

high grade

LG

low grade

NHS

National Health Service (UK)

NICE

National Institute for Health and Care Excellence

NMIBC

Non‐muscle Invasive Bladder Cancer

WHO

World Health Organization

Introduction

Non–muscle‐invasive bladder cancer (NMIBC) represents a major healthcare challenge. It accounts for almost three‐quarters of new bladder cancer diagnoses, and – although overall prognosis is generally favourable – patients face a substantial lifetime risk of recurrence and in some cases progression. This necessitates regular, long‐term cystoscopy surveillance, making NMIBC one of the most expensive malignancies to manage on a per‐patient basis. Despite promising diagnostic accuracy, urinary biomarkers such as EpiCheck are not yet ready to replace routine cystoscopy in NMIBC surveillance because of heterogeneous evidence, limited real‐world validation, and lack of guideline endorsement for standalone use [1, 2].

The risk of recurrence and progression in NMIBC can be estimated using clinical, endoscopic, and histopathological factors derived from the European Organisation for Research and Treatment of Cancer (EORTC) trials. These factors underpin risk stratification systems that categorise NMIBC into low‐, intermediate‐, and high‐risk groups across different clinical guidelines [1]. Evidence supporting the definition and management of high‐risk disease is relatively robust, resulting in greater concordance between guidelines in both risk classification and recommended treatment and surveillance strategies. In contrast, the evidence base for low‐ and intermediate‐risk disease is less strong, with recommendations largely informed by expert opinion and retrospective data. This has led to considerable variation between guidelines in both risk stratification criteria and surveillance protocols [3, 4].

Multiple national and international guidelines for NMIBC management are currently available. In the UK, the most commonly adopted guidelines are those produced by the European Association of Urology (EAU) and the National Institute for Health and Care Excellence (NICE) [5]. Notable discrepancies exist between these guidelines in the definition and management of low‐ and intermediate‐risk disease. For example, a patient with recurrent low‐grade pTa disease occurring more than 1 year after the primary tumour ablation would be classified as low risk in NICE guidance and surveillance recommendations would be for 1 year followed by discharge. However, the same patient would be considered intermediate risk in EAU guidelines, to be treated with adjuvant intravesical therapy and followed up for up to 10 years before discharge. Nevertheless, overestimation of intermediate‐risk NMIBC may drive unnecessary BCG use which is a EAU‐recommended treatment for intermediate‐risk patients; this is an important concern in the setting of the global BCG shortage [6]. Consequently, identical patients may receive markedly different management depending on which guideline is applied. Similar discrepancies arise from differential weighting of tumour multiplicity, tumour size, World Health Organization (WHO) grade, and patient age, all of which can significantly alter risk classification and subsequent management recommendations (Table 1).

Table 1.

Key differences in application of various factors in risk stratification between National Institute for Health and Care Excellence (NICE) and European Association of Urology (EAU) guidelines.

Stratification variable	NICE	EAU
Recurrence	Depending on timing of recurrence (< or >12 months), it can be low or intermediate risk*	All recurrent cases are intermediate risk*
Multiplicity	All multiple tumours are intermediate risk (at least)*	Depending on presence of other risk factors, † it can be low or intermediate risk*
Size	All >3 cm tumours are intermediate risk*	Depending on the presence of other risk factors, † it can be low or intermediate risk*
Grading	NICE applies hybrid grading system of WHO 1973 and WHO 2004: G1: Low risk LG/G2: Low risk HG/G2: Intermediate risk G3: High risk	EAU applies WHO 1973 and WHO 2004 grading systems separately WHO 1973: G1: Low risk G2: Intermediate risk G3: High risk WHO 2004: LG: low risk HG: high risk
Patient age	Not considered	If patient is >70 years old, this can upgrade their risk category, depending on the presence of other risk factors.

^*In absence of high‐risk features; T1, CIS, HG/G3, and variant histology.

^† Risk factors included in EAU: age >70 years, tumour size >3 cm, and multiplicity.

Digital decision‐support tools, such as the EAU online NMIBC risk calculator and the BLATUR NMIBC Navigator, aim to translate complex guideline recommendations into structured clinical pathways and may reduce misinterpretation. However, despite their availability, uptake in routine clinical practice remains limited and has not led to meaningful standardisation of care [7, 8].

Variations in guideline adoption between institutions and individual clinicians within a single healthcare system risk undermining the primary objectives of clinical guidelines: the delivery of consistent, evidence‐based care. Against this background, we conducted a UK national survey to describe current patterns of NMIBC guideline adoption and to assess clinicians' ability to apply their chosen guideline consistently to real‐world clinical scenarios. This study focuses on low‐ and intermediate‐risk NMIBC, where maximum discrepancies exist between guidelines and where variation in practice may have the most significant impact on patient care.

Methods

Study Design and Survey Structure

A cross‐sectional, questionnaire‐based survey was conducted to evaluate the adoption and application of guidelines in the management of low‐ and intermediate‐risk NMIBC among UK clinicians. The survey consisted of two sections. Section 1 collected participant characteristics, including demographics, grade, employment setting, and bladder cancer clinical practice. Section 2 assessed the real‐world application of the respondent's self‐reported preferred guideline using case‐based clinical scenarios (Appendix S1).

Case Selection and Risk Stratification

Cases that formed the basis of the scenario‐based assessment within the survey were identified from an institutional transurethral resection of bladder tumour (TURBT) database. All cases had undergone independent risk stratification according to both NICE and EAU. EAU risk stratification was performed using both WHO 2004/2016 and WHO 1973 grading systems, employing the official EAU online risk stratification calculator [9, 10]. Cases eligible for inclusion were those in which risk categorisation differed depending on the guideline applied. Three cases were randomly selected using a computer‐generated random number function (RAND()) in Microsoft Excel (Fig. 1). Cases included in the survey are outlined in (Table 2).

Fig. 1.

The case‐selection process for cases used as real‐life scenarios included in survey questions 7–9.

Table 2.

History of cases randomly selected and included in the survey, and their risk category according to National Institute for Health and Care Excellence (NICE) and European Association of Urology (EAU) guidelines.

Case no. & summary	Risk category as per various guidance
	NICE	EAU‐1973	EAU‐2004
Case (1): 62‐year‐old patient with a single low‐grade (G1) pTa lesion measuring 1 cm, with a history of a similar lesion resected 7 years ago	Low	Intermediate	Intermediate
Case (2): 68‐year‐old patient with multiple low‐grade (G1) pTa lesions, each smaller than <1 cm, and no previous history of bladder cancer	Intermediate	Low	Low
Case (3): 81‐year‐old patient with a single low‐grade (G2) pTa lesion measuring 1 cm, no previous history of bladder cancer	Low	Intermediate	Low

Validation and Dissemination

Prior to national dissemination, the survey was piloted locally within our unit to assess clarity, relevance, and ease of completion. Minor revisions to wording and layout were made in response to pilot feedback.

The final survey was distributed nationally using professional WhatsApp groups (Meta Platforms, Inc., Menlo Park, CA, USA) and the official social media platforms of the British Association of Urological Surgeons and the British Society of Urology Trainees on X (X Corp., Bastrop, TX, USA). Participation was restricted to UK‐based clinicians involved in NMIBC management. UK National Health Service (NHS) email verification was required for enrolment, ensuring that responses were obtained from practising UK clinicians only.

Outcomes

The primary outcomes were self‐reported adoption rate of NMIBC clinical guidelines (NICE vs EAU) and consistency and alignment of clinical decisions with the respondent's stated preferred guideline when managing case‐based scenarios.

Declaration

Artificial intelligence (ChatGPT, OpenAI, USA) was used to check grammar and refine the language. All authors reviewed and take full responsibility for the final content.

Statistical Analysis

Descriptive statistics were used to summarise participant characteristics and guideline adoption patterns, reported as frequencies and proportions. Case‐based decision accuracy was defined as concordance between the respondent's management decisions and the recommendations of their self‐reported guideline. Comparisons between NICE‐ and EAU‐adopters were performed using the chi‐squared test for categorical variables, with Fisher's exact test applied where expected cell counts were <5. Ordered categorical variables were compared using the Mann–Whitney U test. Effect sizes are reported as absolute differences in proportions with 95% confidence intervals (CIs). All tests were two‐sided, with statistical significance defined as P < 0.05. Analyses were performed using IBM SPSS Statistics version 29 (IBM Corp., Armonk, NY, USA).

Results

Survey Response and Participants

A total of 182 responses were received, of which eight were excluded as ineligible because they were not confirmed as practicing in the UK, leaving 174 respondents for analysis. A true response rate could not be calculated because of the absence of a reliable denominator.

Respondents were geographically distributed across England (n = 159 [89%]), Scotland (n = 7 [4%]), Wales (n = 6 [3.4%]), and Northern Ireland (n = 2 [1.1%]) (Table 3). Participants consisted of 100 consultants (57%), 70 trainees (41%), and four clinical nurse specialists (2%). Table 4 shows the total responses for the survey questions.

Table 3.

number and percentage of participants from various regions within England.

Respondents' regional distribution in England	Number (%)
London	32 (18.4)
Southwest of England	30 (17.2)
Northwest of England	26 (14.9)
East and West Midlands	22 (12.6)
East of England	14 (8.0)
Yorkshire and Humber	11(6.3)
Kent, Surrey, and Sussex	9 (5.2)
Other (Northeast of England, South Central England, Not specified)	15 (8.6)

Table 4.

Questions included in the survey and distribution of responses.

Question (t: Total responses)	Responses n (%)
Q1: What is your current role?	Consultant	Trainee	CNS
(t: 174)	100 (58)	70 (40)	4 (2)
Q2: What is your practice region?	England	Wales	Scotland	Northern Ireland
(t: 174)	159 (91)	6 (4)	7 (4)	2 (1)
Q3: Do you have agreed guidance on management of NMIBC within your unit?	NICE	EAU	Local protocol	Not aware of Any
(t: 174)	66 (38)	41 (23.5)	41 (23.5)	26 (15)
Q4: Which guidance do you personally use for NMIBC surveillance?	NICE	EAU	Local protocol	Other
(t: 173)	64 (37)	67 (39)	27 (15)	15 (9)
Q5: How long do you follow up low‐risk NMIBC for?	1 year	2 years	3 years	5 years	More
(t: 172)	88 (51)	12 (7)	8 (4.7)	63 (36.6)	1 (0.6)
Q6: How long do you follow‐up intermediate‐risk NMIBC for?	2 years	3 years	4 years	5 years	More
(t: 172)	4 (2.3)	3 (1.7)	1 (0.6)	117 (68.1)	47 (27.3)
Q7: According to your guideline what is the risk category of 62‐year‐old with a single low‐grade (G1) pTa lesion 1 cm, with similar lesion 7 years ago?	Low	Intermediate	High
(t: 174)	107 (61.5)	67 (38.5)	0 (0)
Q8: According to your guideline what is the risk category of 68‐year‐old with multiple low‐grade (G1) pTa lesions <1 cm, no previous history?	Low	Intermediate	High
(t: 174)	35 (20)	137 (79)	2 (1)
Q9: According to your guideline what is the risk category of 81‐year‐old with a single low‐grade (G2) pTa lesion 1 cm, no previous history?	Low	Int	High
(t: 174)	118 (67.8)	53 (30.5)	3 (1.7)
Q10: What recurrence risk is acceptable at time of discharge?	<1	<5	<10	<20	<30
(t: 172)	8 (4.7)	93 (54)	54 (31.3)	12 (7)	5 (3)
Q11: What is your NHS email address?

EAU, European Association of Urology; NHS, National Health Service; NICE, National Institute for Health and Care Excellence; NMIBC, non–muscle‐invasive bladder cancer.

Guideline Availability and Adoption

Participants reported that their units most frequently recommended NICE guidelines (38% [66]), followed by EAU guidelines (24% [41]) and locally developed protocols (24% [41]). In total, 15% (n = 26) reported that there was no agreed NMIBC guideline or protocol within their unit.

Individual clinician practice did not always align with unit recommendations. When asked which guideline they personally adopted in clinical practice, 39% (n = 67) reported following EAU guidelines and 37% (n = 64) followed NICE guidelines. The remainder followed local protocols (15% [26]) or other guidance (9% [15]).

Adherence to unit‐recommended guidance varied significantly by guideline type. Clinicians working in units that recommended EAU or NICE guidelines were more likely to adhere to their unit's recommendation than those working in units with locally developed protocols (EAU: 35/41 [85.4%]; NICE: 55/66 [83.3%]; local protocols: 26/41 [63.4%]; P = 0.023).

Duration of Surveillance

When asked about surveillance duration for low‐ and intermediate‐risk NMIBC, NICE followers made decisions consistent with NICE recommendations in 101 of 128 responses (78.9%), compared with 65 of 132 responses (49.2%) among EAU followers. This difference was statistically significant (absolute difference 29.7%; 95% CI 18.1–40.1, P < 0.001).

Deviations among NICE followers predominantly reflected longer surveillance (23/128), largely in low‐risk disease (13/64), with minimal shortening (4/128). In contrast, EAU followers rarely extended follow‐up (1/132) but frequently adopted shorter surveillance schedules (66/132), affecting both low‐risk (27/66) and intermediate‐risk (39/66) NMIBC. The direction of deviation differed significantly between NICE and EAU followers (P < 0.001), driven by a markedly higher rate of shortened surveillance among EAU followers.

Case‐Based Guideline Application

Responses to the three clinical scenarios were assessed for concordance with the guideline each participant reported adopting. Among NICE followers, 192 case responses were analysed for all three cases. Consistent risk stratification across all three cases, in accordance with NICE guidance, was achieved in 77.1% of responses (n = 148).

Among EAU followers, adherence to guidelines were assessed separately for cases 1 and 2 given consistent risk category irrespective of the WHO grading used; 132 responses were available. Guideline‐consistent risk stratification was observed in 34.1% of responses (n = 45), which was significantly lower than that observed among NICE followers (77.1%) (absolute difference 43.0%; 95% CI 32.4–52.2; P < 0.001).

For case 3, where both WHO 2004/2016 and WHO 1973 grading systems were referenced, out of 64 responses by EAU followers, risk stratification answers were in keeping with WHO 2004 and WHO 1973 in 56% (n = 36) and 44% (n = 28), respectively, when both grading systems were referenced in the scenario.

Variations Adjusted to Participants' Experience

Whenever a departmental recommendation was present, compliance was similar between consultants and trainees; 68/87 (78.2%) consultants and 44/58 (75.9%) trainees adhered to their unit's recommended guideline, with no significant difference between groups (P = 0.90). Likewise, across a total of 345 and 298 responses from consultants and trainees, respectively, correct response rates were comparable between both groups (205 [59.4%] vs 175 [58.7%], P = 0.92), irrespective of the guideline referenced.

Attitudes to Acceptable Recurrence Risk

To assess whether threshold for recurrence risk influenced guideline adoption, respondents were asked to select the maximum acceptable recurrence risk at discharge using ordered categories (0%, <5%, <10%, <20%, <30%). The median response was <5% for both NICE and EAU followers, with no significant difference in distributions between groups as shown in Table 5 (Mann–Whitney U test, P = 0.278).

Table 5.

Distribution of responses for acceptable risk of recurrence at time of discharge among followers of various guidance.

	0%	<5%	<10%	<20%	<30%
NICE followers	4 (6.3%)	37 (58.7%)	17 (27.0%)	5 (7.9%)	0 (0.0%)
EAU followers	1 (1.5%)	36 (54.5%)	26 (39.4%)	2 (3.0%)	1 (1.5%)
Followers of neither	3 (7.1%)	19 (45.2%)	11 (26.2%)	5 (11.9%)	4 (9.5%)
Total	8 (4.7%)	92 (53.8%)	54 (31.6%)	12 (7.0%)	5 (2.9%)

EAU, European Association of Urology; NICE, National Institute for Health and Care Excellence.

Discussion

This national survey demonstrates substantial variation in the adoption and real‐world application of NMIBC surveillance guidelines in the UK, particularly for low‐ and intermediate‐risk disease. By using case‐based scenarios derived from real patients in whom guideline discordance is known to occur, this study shows that differences between guidelines translate directly into inconsistent decision‐making, even among clinicians following the same guideline. This is particularly relevant given that low‐ and intermediate‐risk NMIBC represents the majority (>70%) of bladder cancer diagnoses [1].

Our results show wide variation of practice across the country. At a departmental level, most units reported recommending either NICE or EAU guidance; however, a substantial proportion (24%) relied on locally developed protocols, and 15% had no agreed guidance at all. Clinicians working in units using local protocols were significantly less likely to follow departmental recommendations than were those recommending recognised national/international guidelines, suggesting that locally produced guidance may be less effective at standardising care (84% vs 63%, P = 0.023). No previous studies have assessed clinicians’ compliance with their departmental recommendations. In a publicly funded healthcare system, such variation is significant, as it may contribute to inequities in patient management depending not only on where but also by whom care is delivered.

At the individual clinician level, adoption of NICE and EAU guidance was broadly similar, yet it did not translate into consistent application. Clinicians who followed NICE guidance were significantly more likely to classify risk and recommend surveillance durations in line with their chosen guideline than those following EAU guidance (77% vs 34%, p = <0.001). This finding is important because it suggests that compliance can be affected by guideline structure, clarity, and applicability in everyday practice. Simpler categorisation and clearer surveillance endpoints may facilitate more consistent decision‐making. One potential source of confusion in EAU guidelines is referencing two grading systems that are applied differently to risk stratification. In our institutional dataset, one‐third of cases were classified differently by the EAU‐NMIBC‐risk‐calculator depending on the grading system used (WHO 1973 or WHO 2004/2016), and this was reflected in the discordant responses from clinicians to the case scenario, with 56% choosing the risk category according to WHO 2004/2016 despite the EAU guideline recommendation to use WHO 1973 if both grading systems are available [9].

Decisions about the duration of surveillance further reinforce this interpretation. NICE followers were significantly more likely to recommend guideline‐consistent follow‐up durations than were EAU followers, despite similar opinions about the recurrence risk at the point of discharge between groups. The greater consistency observed among NICE followers might reflect previous assessments of NMIBC guidelines using the AGREE‐II framework, which identified applicability as a major weakness of many international guidelines, with NICE guidelines scoring higher because of their attention to implementation, cost, and service delivery considerations [11]. Our findings provide real‐world support for these assessments, suggesting that guidelines designed with practical application in mind may reduce variation and unnecessary intervention.

Previous international European and Asian surveys showed lower compliance in applying guidelines for low‐ and intermediate‐risk NMIBC criteria than for high‐risk disease, with inclination to overtreat when compared with high‐risk patients [5, 12]. Mori et al. conducted a systematic review including 15 studies and more than 10,000 patients and showed low worldwide compliance with NMIBC guidelines. That was particularly significant, with more than 50% non‐adherence rates when it came to NMIBC surveillance because of limited access, high costs, discomfort, and morbidities associated with repeated cystoscopies [13]. However, unlike earlier studies, this survey demonstrates that guideline complexity translates into inconsistent decision‐making when clinicians face real clinical scenarios. This distinction is important because it suggests that awareness of guidelines alone is insufficient to ensure consistent care.

Although digital tools such as the EAU risk calculator aim to standardise care, their uptake remains limited and has not reduced variability in practice. We did not ask our participants whether they used an online aid to support their decision‐making. However, the wider variability in responses among EAU users indicates that simple, intuitive guidelines that clinicians can readily internalise may be more effective, particularly as decision‐support tools do not address discrepancies between competing guideline frameworks [7].

Previous studies have suggested that clinician experience may influence adherence to guideline‐based care, with more senior clinicians relying more on individual clinical judgement [14]. However, in our study, adherence rates were similar between trainees and consultants. This was further supported by multivariable analysis, which demonstrated that training level was not independently associated with guideline adherence after adjusting for guideline type. These findings suggest that experience may not be a major determinant of compliance in this setting and that variation in adherence is more likely driven by differences in guideline structure and applicability.

Limitations of the current study include that we were unable to calculate a reliable response rate because of the limited dissemination, which led to potential selection bias. Reasons for misclassification were not fully explored, so whether errors reflected ambiguity within guidelines, recall, or individual clinical judgement remains uncertain. This study provides an evidence base to advocate for clearer, unified recommendations that safeguard patient equality and rational use of NHS resources. The heterogeneity in participant experience and subspecialty interest may represent a limitation; however, this variability reflects real‐world clinical practice and enhances the generalisability of the findings.

Conclusion

Substantial variation exists in the adoption of NMIBC guidelines and their real‐world application in the UK. NICE guidance was associated with greater consistency and usability than EAU guidance for low‐ and intermediate‐risk disease. These findings highlight the need for clearer, unified recommendations to promote equitable care and the rational use of surveillance resources.

Funding

No funding received.

Disclosure of Interests

No conflicts of interest.

Supporting information

Appendix S1. Online Survey.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available because of privacy or ethical restrictions.