Summary and interpretation of main findings
Among a population of men with moderate-severity LUTS allocated to the TRIUMPH intervention (mean IPSS at baseline of 13.6) or usual care (mean IPSS at baseline of 14.6), the difference between the two arms in mean IPSS score at 12 months (primary outcome) was –1.81 in favour of the intervention arm (p < 0.001) after adjustment for baseline IPSS and minimisation variables (see Table 14). Secondary outcomes, notably the incontinence measure (ICIQ-UI-SF) (difference in means –0.74 at 1 year; see Table 17) and IPSS-QoL (–0.31; see Table 18), likewise showed evidence of improvement, compared with the change in symptoms seen with usual care. A greater decrease was also seen in the B-IPQ (see Table 19). The outcomes consistently showed advantage for the intervention. The results suggest strong evidence of a treatment effect.
The trial population was recruited by inviting men who were already known to have LUTS, that is men who were already recorded in GP EMRs as having previously presented with one or more of these symptoms (‘prevalent LUTS’). Hence, the changes achieved by the intervention are in addition to the prior care put in place by practices, potentially including medication. There was no selection based on the type(s) of LUTS present in men joining the trial. Hence this is a pragmatic trial, reflecting the real-world spectrum and the recognised co-existence of LUTS types in many affected people.
The assessment of participant LUTS was based on an understanding of the range of LUTS people can experience, categorised as storage, voiding and post voiding (also known as post micturition). The IPSS was used for the primary outcome, because of its extensive use in research and the detailed understanding of its performance. However, it was previously identified that the IPSS does not measure two symptoms that can be highly bothersome for men, specifically incontinence and post-micturition dribble.64 Hence, IPSS alone was not considered a sufficient diagnostic assessment in determining which symptoms to prioritise in selecting three sections of the intervention booklet to recommend for individual participants, hence the inclusion of ICIQ-UI-SF, which is an incontinence symptom patient-reported outcome measure and includes questions that can elicit the presence of post-micturition dribble.
The bladder diary used in LUTS assessment was the ICIQ bladder diary,65 which is the only validated diary. This is considered a vital tool for understanding the mechanism of increased voiding frequency, and nocturia in particular.10,66 However, bladder diaries do seem to represent a barrier for many patients, in terms of the inconvenience of measuring voided volumes over 3 full days, and also reliable provision of complete information. Varied willingness to complete a bladder diary at baseline may have contributed to the lower consent rate in the intervention arm. Only participants in the intervention arm were asked to complete one for symptom assessment to inform intervention delivery. Comparisons with the usual-care arm, therefore, cannot be made. Baseline measures of other symptom outcomes remained largely balanced between the two arms despite the disparity in consent rate. It is, however, noteworthy that the mean IPSS in the intervention arm was approximately 1 unit less than that of the usual-care arm.
The intervention was developed with extensive PPI contribution to achieve clear use of language and to convey the concepts effectively. The intention was to ensure that fundamental understanding would underpin effectiveness and long-term benefit. Overall, the sections did achieve clarity, as indicated in Chapter 5, Theme 4: experiences of the intervention. In the qualitative interviews, men spoke positively about ease of use, and the informative, explanatory and engaging content and tone. Two sections in particular took considerable development: ‘Getting rid of the last drops’ and ‘Reducing sleep disturbance caused by needing to pass urine’. These aim to treat highly bothersome symptoms (post-micturition dribble and nocturia, respectively), and, in both cases, materials previously available for patients failed to achieve insight and benefit. We note that the unadjusted mean nocturia scores were lower in the intervention arm than in the CG at 12 months.
Appendix 1, Table 41 shows that approximately two-thirds of the patients were referred to the sections on ‘Drinks and fluid intake’, ‘Controlling an urgent need to pass urine’ and ‘Reducing sleep disturbance caused by needing to pass urine’. For the other booklet sections, approximately one-third of participants were referred to each. These recommendations were made by the HCP, based on the evaluation of IPSS, ICIQ-UI-SF and bladder diary findings. Hence, choices were individualised according to the severity of individual symptoms and the bother (which does not always correspond with severity). Of course, the booklet was a standardised printing, so that participants were able to go through the entire booklet out of personal choice.
The type of HCP (nurse or HCA) undertaking the assessment and intervention did not appear to affect outcomes. Accordingly, the intervention appears well suited to delivery in clinical practice by either of these professionals. Site initiation training covered examples of bladder diaries and how primary care staff should interpret them for the purposes of the trial, including how this would direct the recommendation for relevant sections of the intervention booklet, using a checklist developed for the trial (see Report Supplementary Material 1). Equivalent training would be needed for introduction into routine clinical practice. A key part of the training relates to the interpretation of a bladder diary, as completion of a diary varies considerably according to willingness of patients to complete it; hence, extracting the meaningful information is not always straightforward.
The sustained response to the intervention over time is particularly noteworthy. The intervention was delivered with one individual session and three follow-up contacts using a contact modality based on the patient’s preference. The last follow-up was scheduled for 12 weeks, and hence the primary outcome capture at 12 months indicates that response is sustained for the medium term. Indeed, Tables 15 and 16 show there is minimal change in IPSS between 6 and 12 months.
The trial aimed to recruit a diverse population of participants, but the ethnicity data given in Table 7 show that white participants overwhelmingly outnumbered all other ethnic representation. The reasons for this are not clear, and the observation does have implications for generalisability. The deprivation index also showed a marked predominance of men from the two least deprived quintiles. Likewise, this will affect the generalisability of results, given the nature of the intervention. The COVID-19 pandemic took hold during the follow-up phase of the trial. A shift towards electronic questionnaire completion when possible, which was already an option for trial participants, enabled follow-up to continue. There did not appear to be any clear difference in the baseline characteristics of patients participating at that time.
There was no difference in secondary care referral rates between the two arms. We were unable to capture reasons for referral, although prostate cancer screening was commonly mentioned during qualitative interviews. Adverse event rates did not substantially differ between the two arms, except that life-threatening adverse events were more commonly seen in the intervention arm, as reported in Table 30. This did not appear to be a causative relationship.
The primary analysis of IPSS at 12 months was robust to ways of accounting for clustering and multiple imputation of missing data. High levels of adherence to the intervention delivery meant that per-protocol analyses largely reflected the primary analysis.
Economic evaluation
This is the first economic evaluation of a manualised and standardised non-pharmacological intervention for male LUTS. From an NHS perspective, mean costs and QALYs were similar across both trial arms, with a slightly lower mean cost (£30 difference) in the intervention arm. Hence, there was a small positive INMB at the NICE willingness-to-pay threshold of £20,000 per QALY, and a 63% probability of the intervention being the cost-effective option.
Electronic medical record data were available from one of England’s two main GP IT systems (EMIS Web or SystmOne) for each site; this minimised the number of missing data and burden on the participants. However, SystmOne was unable to extract data related to the reason any given consultation was undertaken, and aspects of resulting prescriptions. Accordingly, steps were taken to minimise the potential for bias from using the EMR systems, and adjustments were made in the regression analysis. This made it difficult to maintain masking to randomisation allocation for the health economic analysis. Completeness of data was affected by time constraints in sites, notably absence of EMR data from one site in the intervention arm (covering 5.5% of the intervention participants), and lack of secondary care letters for participants with missing data. Accordingly, it was assumed that no secondary care resources were consumed if the participant had no urology referral recorded in their EMR data set and had completed other questions in the questionnaire booklet. Sensitivity analyses tested these assumptions.
The analysis included 80.4% (n = 866) of the trial participants. For 16.1% of participants, questionnaire data were incomplete, and ‘missing not at random’ could not be ruled out. Some participants had their 12-month follow-up after 11 March 2020, and hence may have reduced healthcare use as a result of the impact of the COVID-19 pandemic in the UK. This may have affected the usual-care arm more than the intervention arm.
Overall, the results of the economic evaluation indicate that the costs should not be seen as a limiting factor for the roll-out of the TRIUMPH trial intervention.
Qualitative study
From qualitative interviews, it is apparent that the intervention was well received, and had a positive effect on men’s symptoms and attitude to their LUTS. Men in the intervention arm described relief and gratitude at being better informed, and reported symptom-related benefits. A minority of men were not receptive to the guidance. This was mainly linked to experiences of LUTS and of primary care.
Participants were men with a full range of LUTS, including long-term and severe symptoms, which caused distress, tiredness, anxiety, embarrassment, frustration and annoyance, impinging on and worsened by the pressures of occupational activities and, at worse, leading to complete withdrawal from valued social activities. For many men, justifying a visit to the GP reflected that symptoms had become very disruptive or distressing. Severely disruptive symptoms were particularly important, such as a severe UTI, incontinence, urinating very frequently in the day and/or night, acute retention, blood in urine and/or prostate cancer concerns. These were also factors influential to the willingness, engagement and uptake of health-related information and guidance.
The interviews also highlighted that structured, in-depth self-management is not sufficiently embedded within GP consultations. Consultation accounts ranged widely, covering simple reassurance, assessment, medication, explanation and self-management, PSA testing and referral inconsistency. Detailed explanations and self-management guidance were particularly lacking, and it is clear that men often remained ill-informed after their primary care contact. Men were broadly unfamiliar with self-help information to support LUTS, and were unlikely to seek it out. Hence, the TRIUMPH qualitative study supports dissemination of trial guidance, particularly when distressing symptoms are present and men are willing to make lifestyle changes, but also more broadly to support wider population-level awareness of the availability of tailored self-management options to support LUTS among men.
Non-participation was often due to the perception that the trial could be time-consuming and effortful, with some men saying that they felt too old and others saying that they were too busy for this additional burden. Of course, attitudes varied substantially, and have to be considered in the light of the demographic make-up of the trial population. Some ‘declining’ men said that the trial did not seem relevant to their current circumstances, in particular their symptom experiences; this strongly indicates that personal relevance is crucial for engagement with the intervention.
Barriers to the dissemination of LUTS self-management guidance in primary care are hugely important. GPs’ consultations have to cover multiple issues in a constrained time. Discussions around the implications of PSA testing can be time-consuming and take precedence, with implications for the inclusion of self-management information and guidance. Achieving delivery of this material without relying on GP availability is potentially very beneficial.
Men do not attend GPs seeking self-management remedies, and are much more likely to go the GP with concerns about prostate cancer or after lengthy periods enduring severe LUTS. The resulting consultation may well focus on diagnostic tests and medication, and overlook assessment to guide self-management. It is vital that GPs identify that a presentation with LUTS (even when symptoms are dismissively described) should not conclude solely with prostate cancer screening; once excluded, it is vital to return to the actual presenting complaint, and to structured self-management guidance. A further key point is to review professional treatment recommendations, notably the use of basic reassurance, because we identified that men have a strong tendency to tolerate and downplay their LUTS.
In this population with prevalent LUTS, the intervention was enthusiastically received because it brought information, explanation and self-management where they had not previously been provided. There was a willingness to engage with self-help and a sense that this was a good option, contrasting with the situation for men in the usual-care arm who continued to tolerate distressing symptoms without access to potentially helpful support and guidance. Indeed, men demonstrated an appetite for guidance and some intervention arm men spontaneously sometimes shared content with others. In some ways this could serve an advocacy role to help bring acceptability to discussion in this area.
Strengths and limitations
Strengths of the trial include recruitment to target and excellent follow-up rates, resulting in informative and robust conclusions. The primary outcome was also intentionally timed to allow the longer-term benefit of the TRIUMPH trial intervention to be captured, at 12 months post consent, following a 12-week intervention delivery. This is key, as continuing motivation is a concern in the longer-term effectiveness of lifestyle and conservative interventions.
Limitations of the trial include the need for trial participants to be unmasked, as a patient-reported primary outcome was used. In addition, the trial population was self-selected for trial inclusion, and included only those who had previously sought help from their GP for their LUTS. The small number of men of black, Asian or mixed-race ethnicity was also a limitation, particularly in view of potential ethnic differences in health beliefs regarding LUTS.67 A potential limitation is that sites recruiting to the usual-care arm would potentially have developed renewed awareness of the condition and the current approaches to management as a result of participating in the trial. Potentially this would decrease the observed differences between randomised groups, given that there was some improvement in the usual-care arm in the primary outcome.
Overall evidence and generalisability
The trial findings show a significant improvement in prevalent LUTS, based on an assessment by trained staff, individualised recommendations for priority sections in a standardised booklet and HCP follow-up. HCPs used their best estimates to make up for missing data in the assessments; some staff may not feel able to undertake this step.
A study of men undergoing drug therapy supported the practice of powering studies to detect group mean differences in IPSS of at least 3 points.22 This study had a recruitment pool that was rather different from that of the TRIUMPH trial: it enrolled participants from secondary care, and participants had to be aged ≥ 45 years, have a flow rate of at least 4 ml/second and have an IPSS of between 8 and 24 on two screening visits. The finding of a group mean difference of –1.8 is less than the accepted minimal clinically important difference of 3 points. In developing the trial, we considered that a lower difference in mean IPSS of 2 points could be meaningful if affecting particularly bothersome symptoms, such as nocturia or post-micturition dribble. Because the manualisation of the intervention meant that participants were specifically directed to advice relating to their more bothersome symptoms, the selective reduction in IPSS would be more likely in the clinically important areas. Considerations on the clinical importance of this difference need to factor in that the trial (1) used a non-drug intervention, (2) was based in primary care and (3) was unselective in terms of type or severity of LUTS. Indeed, many men were already established on medications used to treat LUTS, as indicated in Table 8.
It is important to recognise that the trial population was predominantly white. Furthermore, based on practice postcode, there was less representation from low-affluence areas. Hence, generalisability will be affected by the nature of the population served by the practice, willingness of staff to undertake training and facilities for follow-up.
Recommendations for research
In order of priority, the trial results suggest the following recommendations for further research:
- Establishing the role of manualised and standardised care earlier in the clinical pathway. The trial population was drawn from the prevalent LUTS patients identified through diagnosis codes at practices. It will be important to establish whether or not equivalent, or better, changes in IPSS are achieved at the time of first presentation, that is incident LUTS. Intuitively, manualised and standardised care based on self-help measures would precede prescription of LUTS medications, to enhance compliance and potentially reduce medication use. Looking at incident LUTS patients would determine whether or not manualised and standardised care reduces prescriptions issued to treat LUTS.
- Generalising the findings to a more diverse population. The non-white participants were a small part of the trial, so conclusions and recommendations from the TRIUMPH trial have to be considered accordingly. To explore the potential for this approach in a more diverse population, research is needed to understand the barriers to participation, to determine what adaptations to study design are needed to enable participation in a future study.
- Rationalising the assessment and combining it with the intervention. Development of a unified assessment and outcome tool is highly desirable; potentially, this could include a bladder diary adapted for use in primary care, with fewer data collection requirements (notably volume measurement throughout the entire 3 days, bladder sensation reporting and fluid intake) and indicators of key information (e.g. ‘sleep’ and ‘wake’ times) to maximise completion and facilitate interpretation. This would then require further evaluation when used in conjunction with the intervention booklet.
- Adapting the design of the health economic research. Future health economic studies should consider stratifying for practice EMR systems within the randomisation process and controlling for this variable.
- Evaluating the need for repeated follow-up. Three follows-up contacts were made. It is possible that fewer would be sufficient. The trial had very high follow-up rates, such that there are very few data relating to men who had fewer follow-ups.
Implications for health care
Nurses and HCAs are suitable for the assessment and delivery of conservative treatment of LUTS using the TRIUMPH trial booklet.
Training in the approach to assessment and delivery of the intervention can be given in approximately 1 hour, which is likely to be a crucial requirement for clinical practice, and could potentially be delivered online.
In a prevalent LUTS population, we found a mean difference in IPSS of –1.8 points between the intervention and usual-care arms, associated with congruent improvements in quality of life and other secondary outcomes. The improvement was not accompanied by an increase in adverse events, and was sustained for many months after conclusion of the intervention.
Qualitative findings indicate that rigorous self-management guidance is insufficiently embedded within primary care for men with LUTS. Men had little awareness that LUTS is a condition that can be alleviated through a range of clinically recommended self-management techniques. Men who were receptive were willing and often eager to trial lifestyle approaches, found these helpful, and enthusiastically welcomed the guidance and intervention, saying it should be available to all men.
Stigmatised representations of LUTS as an inevitable and irreparable aspect of age-related decline were prevalent. Men often neglected enduring and distressing symptoms as ‘just a part of ageing’ and visited their GP only when symptoms became severe or alarming, often with concern about the possibility of prostate cancer. It appears that GP consultations tend, as a consequence, to orient around PSA testing and other tests to rule out possible underlying conditions, and move quickly into medication discussions, meaning that, ultimately, in-depth self-management easily slips out of primary care for men with LUTS, despite cycles of PSA testing (which we have termed ‘the PSA test loop’).
Inadequate self-management guidance within primary care is potentially a missed opportunity. The qualitative findings support the development of practice-based nurse- or HCA-facilitated primary care LUTS services for men that can better support self-management guidance beyond the constraints of the GP consultation, and might also help to foster and embed a culture of supported self-management guidance for men with LUTS within primary care. GPs and men with LUTS were supportive of self-management as an appropriate treatment option (alone or alongside medication, dependent on symptoms and aetiology).
More broadly, there is potentially a need to reach men in other settings, to challenge misconceptions and address the paucity of knowledge, awareness and understanding of LUTS and LUTS treatment options, with a focus on self-management. A broader, targeted public health campaign for men with LUTS could be extremely helpful to increase men’s awareness of the availability of self-help (and other clinically recommended) guidance to support LUTS. Without this, the information men receive about LUTS is often related to campaigns focused on prostate cancer, which can be alarming and unhelpful with respect to LUTS care.
Finally, the reassurance offered to men with LUTS by clinicians needs to be adjusted in recognition of denigrating and stigmatised beliefs about LUTS. Telling men that LUTS are ‘nothing to worry about’ and ‘just a normal part of ageing’ is inadequate and insensitive to men’s tendency to downplay and disregard their LUTS; dismissive reassurance feeds despondency, inaction and a sense of helpless decline. Conversely, self-management guidance promoted increased self-efficacy and better symptom management and outcomes.
Conclusion
The trial evaluated an intervention using a standardised information booklet, with manualisation by a nurse or HCA guiding the participants to the most appropriate sections of the booklet based on key individual assessments, after which three follow-up contacts were undertaken. This intervention consistently showed benefit for men’s LUTS and quality of life across a range of outcome measures, when compared against usual practice in a UK primary care setting. Improvements were sustained for a full year. Effect sizes were relatively small, but had the potential to achieve valuable clinical impact, with minimal risk of adverse effects. Qualitative data showed that men valued the intervention highly, and intervention costs were marginally lower than usual care.