1 INTRODUCTION
Urinary problems are widespread among older men. A Swedish study of men aged 40–80 years reported a 24% prevalence of lower urinary tract symptoms (LUTS), with only 4% seeking medical care[
1]. Similarly, a Danish survey found that 22% of men aged 20 and older experienced LUTS, yet 24% did not discuss their symptoms, and 59% avoided healthcare altogether[
2]. The prevalence of LUTS is anticipated to increase with an aging population.
Artificial intelligence (AI) is gaining prominence in healthcare, particularly in radiology and pathology. However, its application in primary care remains underutilized. In the United States, studies have shown that digital anamnesis for gastrointestinal issues is more complete and better organized than physician-documented histories[
3]. In Sweden, digital anamnesis in emergency care has demonstrated potential for safer and more efficient care[
4].
Our hypothesis is that an AI-driven app can assist men aged 50 and above with LUTS by:
• Helping them recognize that their symptoms may reflect natural aging rather than disease.
• Supporting self-management with lifestyle changes and tailored training programs.
• Providing guidance on when and where to seek professional care.
• Offering decision support to primary care providers.
This paper outlines our preliminary work and discusses future developments.
2 MATERIALS AND METHODS
The project has progressed through the following phases:
Phase 1: A retrospective virtual database (
n = 50) was constructed, including patient data age, body mass index (BMI), comorbidities, medications, International Prostate Symptom Scores (IPSS), urinary diaries, and timed micturition assessments. Urologists’ diagnoses based on these variables were used to train with AI in the form of machine learning and Bayesian models[
5]. The resulting app will provide feedback on symptom severity, categorizing potential etiologies into bladder, prostate, or nocturia-related causes. Each category includes information on the etiology, lifestyle factors, and tailored training programs (e.g., pelvic floor muscle exercises for overactive bladder).
A prospective clinical study began in 2021, collecting real-life data to validate the app's initial version.
Phase 2: The next phase involves a randomized study, recruiting participants via a website. One group will receive AI-based advice, while the control group will not. Both groups will complete follow-up questionnaires after three months. A feasibility study will precede this phase to ensure usability and enrich the database.
3 RESULTS
In its first test in a new set of men, the app demonstrated 90% concordance between algorithmic assessments and urologists’ evaluations. A prospective study, approved in 2021 (Dnr 2021-01178), has up to now enrolled 50 patients from three primary health centers. The average age was 70 years (range 44–86 years). The mean score for the IPSS questions 1–7 was 14, and for the quality of life (QoL) question three. The flow measurement was missing in many cases. One-third of the patients did not get a recommendation at the first visit. Of the remaining, less than half received the same treatment recommendation at this visit as the AI algorithm would have suggested.
A web-based study, also ethically approved (Dnr 2024-07439-01), is currently being initiated. Recruitment strategies include public campaigns and targeted invitations. Participants complete a digital questionnaire and are randomized if they meet the inclusion criteria.
4 DISCUSSION
This study represents a novel approach to managing LUTS in men by leveraging AI and digital health technologies (Figures 1 and 2). For women with urinary incontinence, an app has been tested with good results[
6]. In contrast to our approach, their app is introduced by the treating physician and only has a pelvic floor muscle training programme. We take advantage of a validated anamnesis protocol used in clinical routine that is easily digitalized and then analyzed by AI.
The app has demonstrated a high degree of concordance with urologist assessments, suggesting its potential as a reliable first-line tool for LUTS management. Such tools can empower patients to take an active role in their health, fostering better engagement and adherence to management strategies. This is particularly important in populations where reluctance to seek medical advice is common. Moreover, by enabling patients to self-manage or identify the appropriate level of care, the app could help reduce the burden on primary care providers. In overburdened healthcare systems, this shift may free up resources for more complex cases. Additionally, the app may contribute to the earlier identification of more serious conditions by encouraging patients to engage with their symptoms earlier.
While promising, this approach is not without challenges. For older adults, who constitute a significant portion of the target demographic, digital literacy and comfort with mobile technology may present barriers to adoption. A lack of familiarity with smartphones, poor eyesight, or cognitive impairments could hinder the app's effectiveness in this group. Another concern is the accuracy of self-reported data. While tools like digital questionnaires improve standardization, some users may misinterpret questions or provide incomplete information. To mitigate this, future versions could integrate dynamic question flows that adapt based on user responses, ensuring clarity and completeness.
Integrating such an app into existing healthcare pathways will require careful planning. For example, primary care providers may need training to interpret app-generated reports effectively and incorporate them into decision-making processes. Additionally, regulatory frameworks must ensure that these tools meet safety and efficacy standards while protecting user data. Ethical considerations also play a critical role. Although the app is not intended to replace medical consultations, it must minimize the risk of delayed diagnoses, especially for serious conditions. The worst case is if it delays a cancer diagnosis. LUTS is a late symptom for prostate cancer and in a large screening study, men with urinary symptoms had a decreased risk of cancer compared to an age-matched group without[
7]. Ensuring that the app clearly communicates when professional medical advice is necessary will be crucial.
The next steps for this project include expanding the app's features to improve usability and diagnostic accuracy. For example, incorporating smartphone-based uroflowmetry, which analyses the sound of urine flow, could provide more objective data[
8]. Similarly, visual tools like the Visual Prostate Symptom Score (VPSS) could simplify symptom reporting for users with limited literacy or cognitive abilities[
9]. Another promising avenue is the integration of wearable technologies like Vesica, iUFLOW, and Bladderlys Smart Bladder Diary. These devices are capable of tracking bladder activity and fluid intake could enhance the app's ability to provide personalized advice. Combining these data streams with machine learning algorithms may further refine the app's predictive capabilities.
Finally, long-term studies are needed to evaluate the app's impact on clinical outcomes, patient satisfaction, and healthcare costs. Randomized controlled trials, as we plan, comparing the app to standard care will be essential in establishing its efficacy. Additionally, examining how the app performs across different demographics and healthcare systems will provide insights into its scalability and adaptability.
This study aligns with the broader trend of digital transformation in healthcare. AI-driven tools have the potential to democratize access to quality care, particularly in underserved areas or among populations with limited access to specialists. However, achieving this vision will require addressing disparities in technology access and digital literacy. The success of this app could serve as a model for similar initiatives targeting other common conditions. For instance, combining digital anamnesis with AI-based decision support could be applied to areas like cardiovascular disease, diabetes, or mental health. These efforts could collectively shift healthcare toward a more patient-centered and preventive model.
2025 The Author(s). UroPrecision published by John Wiley & Sons Australia, Ltd on behalf of Higher Education Press.
PDF
(321KB)
