Prostatic urethral lift (UroLift): a real-world analysis of outcomes using hospital episodes statistics

Prostatic urethral lift (PUL) is a minimally invasive treatment for men with lower urinary tract symptoms (LUTS), which involves placing non-absorbable sutures with a nitinol prostate capsular anchor and a stainless steel urethral end piece to mechanically open the anterior prostatic fossa and disobstruct the urethra [1]. As the treatment does not use thermal energy to excise or ablate tissue it also reduces some adverse effects, such as erectile dysfunction and ejaculatory dysfunction, which can be associated with traditional treatments such as TURP [2]. Other minimally invasive treatments are available; prostate artery embolization (PAE) and iTIND also avoid the use of thermal energy, Rezum therapy uses steam heat energy to remove prostate tissue, whilst Aquablation uses high pressure water jet to remove adenoma. Currently in the UK there is no consensus as to which treatment should be offered to which men and at what point in the treatment pathway. Recent UK National audits suggest the majority of men are still treated with traditional TURP, and that less than 10% of men are being treated with PUL, even less with PAE and Rezum, and that Aquablation is only offered at one centre in the UK [3].

PUL marketed as UroLift (manufactured by NeoTract Inc.) was first performed in the United Kingdom (UK) as part of the commercially sponsored BPH-6 trial [4]. The UK National Institute for Health and Care Excellence (NICE) published interventional procedures guidance for urethral lift in January 2014, recommending its use for the treatment of men with lower urinary tract symptoms [5], and Medical Technology Guidance on the UroLift System in September 2015 [6]. This recommended UroLift as an alternative to surgical procedures, in a day-case surgery setting, in men over the age of 50, who have a prostate of less than 100 ml without an obstructing middle lobe. Further studies support the wider use of PUL in men with obstructive median lobes [7], large prostates [8] and in men with retention [9]. Yet despite the positive published outcomes of UroLift, adoption of the procedure in the NHS has been slow. To encourage widespread adoption, UroLift was added to the Innovation Technology Tariff (ITT) in April 2017, and subsequently selected as a Rapid Uptake Product by the National Health Service (NHS) Accelerated Access Collaborative in 2018 [10]. Both schemes aim to support adoption of innovative technologies within the NHS.

Hospital Episode Statistics (HES) is a data warehouse containing episodes of care under a single consultant for patients at NHS hospitals in England. HES datasets include Admitted Patient Care (all admissions including day-case procedures), outpatient appointments and attendance at accident and emergency departments [11]. Clinical coding of procedures uses the Classification of Intervention and Procedures, (OPCS-4) and coding of diagnoses uses the International Classification of Diseases (ICD-10). A specific procedure code was introduced into the UK National Clinical Coding Standards for UroLift in 2017 (“M68.3: Endoscopic insertion of prosthesis to compress lobe of prostate”) enabling robust identification of procedures in HES data from that point.

The aim of this study is to use national administrative data from HES to determine uptake as well as real-world in-hospital and longitudinal outcomes of prostatic urethral lift (UroLift) procedures conducted in an NHS hospital setting in England.

Episodes of UroLift implantation were identified from the presence of procedure code “M68.3: Endoscopic insertion of prosthesis to compress lobe of prostate” in the HES Admitted Patient Care (APC) dataset with a discharge date between 1st April 2017 and 31st January 2020. This dataset also includes day-case surgeries where the patient is admitted to hospital for a planned surgical procedure and returns home on the same day. Individual episodes of care from HES were aggregated into admissions (single periods of care within a treating hospital) [12]. Analysis was restricted to the earliest UroLift implantation admission for each patient (index admission) conducted within NHS hospitals, which also included a diagnosis code relating to benign prostate hyperplasia (Additional file 1). Those discharged after 1st January 2020 were excluded to ensure 30 day follow up.

Pseudonymised data from HES and the Civil Registration (formerly, the Office of National Statistics) Mortality datasets were supplied under Data Access Request Service (DARS) agreement DARS-NIC-170211-Z1B4J. No patient identifiable information was used and ethical approval was not sought or required. All scripts for applying eligibility criteria, data cleaning, processing and statistical analysis were written in the statistical programming language R [13].

Patient characteristics from the index UroLift procedural admission were summarised using descriptive statistics. Patients catheterized on admission were identified by the presence of ICD10 code Z96.0 “Presence of urogenital implants”. In-hospital outcomes included complications [14], catheterization due to retention (procedure code M47 “Urethral catheterization of bladder”), subsequent removal of the catheter (M47.3 “removal of catheter from bladder”), length of hospital stay and death.

All hospital activity (including APC episodes, day-case surgeries, outpatients, and accident and emergency attendances) and all-cause mortality occurring after discharge from the index UroLift implantation were extracted for the cohort. Outcomes at 30-days included catheter status by analyzing catheterization code in both APC and outpatient attendances. Longitudinal outcomes included retreatment (Additional file 2), other bladder/prostate intervention (Additional file 3) and all-cause mortality. Kaplan–Meier analysis was applied to the time from the discharge date of the index UroLift procedure to the date of retreatment (by UroLift of other endoscopic interventions) or date of death. The timing of events was estimated using the hazard function [15]. Patients with no events and known to be alive at the end of the study were considered censored.

This HES analysis represents the largest cohort study of UroLift in a NHS hospital setting, with robust identification of procedures, comprehensive coverage across England, median follow-up of 1.2 years [min 11 days; max 2.8 years]. Our study demonstrates an increase in uptake of the procedure across the NHS during the study period, with 80 NHS trusts providing the procedure in England. In our study 3.7% of patients were aged less than 50 years old, this is higher than the rate reported in a retrospective case review conducted in the US and Australia (17 in 1413, 1.2%) [16]. However, given the major benefit of PUL is to maintain sexual function there continues to be a strong argument for treating younger men who may wish to preserve normal ejaculation for conception. Our results suggest a potential selection bias towards older men as a new procedure is introduced into practice, which may represent a learning curve as clinicians become comfortable with the technique and its results. Additionally in the UK the majority of men opt for pharmaceutical treatment rather than procedure based treatment for mild to moderate symptoms due to national guidelines.

We confirmed that the majority of procedures were conducted as day-case surgeries (85.3%) and that there was a low rate of in-hospital complications (3.4%). This is in keeping with previous reports that UroLift has several advantages including the ability to be undertaken under local anesthesia as a day case, a short learning curve and shorter procedure time (compared with traditional bladder outlet procedures) [17]. The majority (> 95%) of complications reported in the literature have been Clavien grade 1 with the most common being pelvic pain and dysuria [17]. The complication rate we found was low but we are not able to directly compare ICD-10 classification with Clavien–Dindo without introducing reporter bias in the assessment of the severity of the ICD-10 code in the absence of detailed patient notes. The most common adverse event we noted was urinary retention followed by bleeding.

By day 30, 205 men (7.0%) remained catheter dependent, and an additional 243 men required catheterization after 30 days (variable follow-up). In the LIFT study, 32% of patients required catheterization for failed voiding trial with a mean catheter duration of 0.9 days average for the whole cohort of 206 patients across the trial [1]. In the BPH-6 study, 45% of UroLift patients had a postoperative catheter for more than 24 h [4]. The post-operative catheterization rate we report is lower than in the two trials and may represent more cautious patient selection for implementation of a new procedure in an NHS setting compared to a trial setting, however the clinical profile of patients cannot be determined from HES coding. The lower rate may also be due to increasing clinician confidence and experience in the post-operative period after treatment.

Although the overall reported complication rate from UroLift is low, there is a paucity of longitudinal data on emergency re-presentations in the real-life setting. We noted a large proportion of our cohort (12%) attended A&E within 30 days of their PUL procedure. This will be an overestimate of complications, as some attendances maybe unrelated to UroLift, nevertheless it represents an upper limit used to demonstrate hospital service usage following UroLift implantation. However, there is little published data for comparable bladder outlet procedures for A&E admissions in the UK.

Due to large cohort size, our study is able to provide a more robust estimate of retreatment rates (11.9% at 2 years) following UroLift implantation. Retreatment rates were higher than those found in patients randomized to UroLift (n = 140) in the LIFT study (7.5% [95% CI 3.8 to 13.6%] at 2 years [18], but similar to the outcomes at 3 years, 10.7% [95% CI 6.3 to 17.3%] [19]. This may represent a lower threshold to offer retreatment outside of clinical trials.

Whilst this is a retrospective cohort study, it reports real-world outcomes from all UroLift procedures conducted across NHS hospitals in England, with comprehensive follow-up, which reduces both selection and reporting bias [11]. This approach allows normalization of outcome as it reflects both the high volume experienced surgeon and those surgeons who perform fewer cases. Given the large number of cases we have captured and described, the effect of outliers, surgical selection and retreatment bias is minimized, and the overall results offer a valid real-world reflection of the outcomes and hospital resource usage associated with PUL when conducted in an English NHS hospital.

There are however some limitations of using the HES data as it can only be used to produce overall performance indicators (e.g. readmission, complications and length of stay) and it does not allow assessment of all individual patient characteristics as some are not coded (e.g. severity of symptoms, size of prostate, presence of median lobe, International Prostate Symptom Score (IPSS), uroflowmetry, medication etc.) which makes it difficult for our study to comment on efficacy of UroLift. Additionally, due to lack of coding (both procedural and diagnosis) administrative data cannot be used to make meaningful comparisons of outcomes between the different prostate procedures (e.g. TURP, GreenLight, HoLEP etc.). However due to the creation of a procedure code introduced specifically for UroLift, and implemented in 2017, we have been able to use routine administrative data to identify and follow a cohort of men having this intervention across all NHS hospitals in England. This makes HES a powerful tool in investigating patient pathways, hospital resource usage and safety outcomes following an intervention due its comprehensive coverage of hospital activity.

This HES data analysis of UroLift shows that it can be delivered safely in a day-case surgery setting with minimal morbidity. The hospital resource usage in terms of catheterization is lower than in published trial data but emergency hospital attendance in the first 30 days following UroLift implantation is higher than would have been expected for a minimally invasive treatment. The need for further surgical intervention seen in this study appears higher than reported in published trial data at 2 years which may have an impact on the health economic calculations used to assess the place of this treatment in NHS hospital pathways.