Comparison of Emergency Room Visits and Rehospitalization for Bleeding Complications following Transurethral Procedures for the Treatment of Benign Prostatic Hyperplasia: A Population-Based Retrospective Cohort Study

Background: The postoperative bleeding complications associated with laser surgery of the prostate and transurethral resection of the prostate (TURP) were compared. Methods: We used the Taiwan National Health Insurance Research Database to conduct an observational population-based cohort study. All eligible patients who received transurethral procedures between January 2015 and September 2018 were enrolled. Patients who received laser surgery or TURP were matched at a ratio of 1:1 by using propensity score matching, and the association of these procedures with bleeding events was evaluated. Results: A total of 3302 patients who underwent elective transurethral procedures were included. The multivariable Cox regression analysis revealed that diode laser enucleation of the prostate (DiLEP) resulted in significantly higher emergency room risks within 90 days after surgery due to clot retention than the Monopolar transurethral resection of the prostate (M-TURP) (Hazard Ratio: 1.52; 95% Confidence Interval [CI], 1.06–2.16, p = 0.022). Moreover, GreenLight photovaporization of the prostate (PVP) (0.61; 95% CI, 0.38–1.00 p = 0.050) and thulium laser vaporesection of the prostate (ThuVARP) (0.67; 95% CI, 0.47–0.95, p = 0.024) resulted in significantly fewer rehospitalization due to clot retention than did M-TURP. No significant increase in blood clots were observed in patients using comedications and those with different demographic characteristics and comorbidities. Conclusions: Among the investigated six transurethral procedures for Benign prostatic hyperplasia, PVP and ThuVARP were safer than M-TURP because bleeding events and clot retention were less likely to occur, even in patients receiving anticoagulant or antiplatelet therapy. However, DiLEP and holmium laser enucleation of the prostate (HoLEP) did not result in fewer bleeding events than M-TURP.


Introduction
Benign prostatic hyperplasia (BPH) is the most prevalent chronic disease among middle-aged and older men worldwide [1]. BPH surgery is covered by the Taiwan National Health Insurance (NHI) program. Monopolar transurethral resection of the prostate (M-TURP) is the standard treatment for BPH with small-sized prostate glands. However, 2 of 16 M-TURP is associated with short-term postoperative bleeding-related complications such as hematuria and clot retention [2]. Bipolar TURP (B-TURP) is a safe and effective procedure with a significantly shorter operating time and efficacy similar to that of conventional M-TURP [3].
The introduction of TURP was followed by the advent of laser technology, which led to the development of some therapies with greater efficacy and fewer complications than TURP. GreenLight photovaporization of the prostate (PVP), thulium laser vaporesection of the prostate (ThuVARP), holmium laser enucleation of the prostate (HoLEP), and diode laser (980 nm) enucleation of the prostate (DiLEP) are more efficient laser techniques compared to TURP, with reported reproducible clinical results and few bleeding-related complications [4][5][6].
Because of the increasing prevalence of comorbidities and indications, the number of patients with BPH undergoing transurethral laser procedures increased. Transurethral laser procedures are associated with lower intraoperative blood loss and fewer discharge days [7][8][9]. However, no definite conclusions were drawn because of insufficient evidence. Moreover, few studies compared severe bleeding tendencies associated with all transurethral laser procedures and TURP in the same baseline population. This study evaluated the postoperative bleeding complications among common laser techniques and TURP for BPH.

Data Sources
The National Health Insurance Research Database (NHIRD) is a population-level dataset derived from the claims data of more than 99% of people in Taiwan enrolled in the NHI program [10]. The research database includes data on patients' medical history, medication use, surgical intervention history, personal data, and diagnosis identified on the basis of the International Classification of Diseases, Ninth and Ten Revision, Clinical Modification (ICD-9-CM and ICD-10-CM, respectively) diagnosis codes [11,12].
The study protocol was approved by the Research Ethics Committee of Show Chwan Memorial Hospital (IRB-No: 1091213), and the requirement for informed consent was waived because the NHIRD datasets contain no identifiable personal information.

Study Design
We conducted a nationwide cohort study by retrieving NHIRD data on hospitalized patients who underwent their first TURP or laser surgery between 2015 and 2018. Followup data before enrollment were used to evaluate comorbidities. The discharge date of patients after surgery was considered their index date.
In order to evaluate the postoperative status of patients, we performed a follow-up for 15, 30, 60, and 90 days after surgery and calculated the occurrence of emergency room (ER) visits and rehospitalizations during the postsurgical period. Furthermore, we determined the statistical differences between these rates according to patient characteristics and surgery type.

Surgery Types
The M-TURP uses a single active electrode at the surgery site with a non-conductive hypo-osmolar irrigation medium. The B-TURP incorporates monopolar technology and is performed in a normal saline environment, addressing the dilutional hyponatremia of conventional monopolar TURP [13,14].
Type of laser techniques includes PVP, ThuVARP, HoLEP, and DiLEP. Based on the different wavelength-dependent laser-prostatic tissue interactions, the main techniques are coagulation, vaporization, resection, and enucleation. Although each laser type is different in design, their principle is not distinctive [15,16].

Covariate Assessment
We identified the following covariates that are potential confounders: Charlson comorbidity index (CCI), age, and medication history. The medication history of the following drugs was assessed: statins, angiotensin-converting enzyme inhibitors (ACEIs), β-

Postoperative Bleeding Complications Leading to ER Visits
We evaluated the bleeding events leading to ER visits in patients with urine retention, acute urine retention, hematuria, using tranexamic acid, and who underwent diagnostic or treatment procedures, including bladder sonography, bladder instillation, bladder irrigation with a foley catheter, or bladder blood clot evacuation with a Toomey bladder evacuator.

Postoperative Bleeding Complications Leading to Rehospitalization
We evaluated the bleeding events leading to rehospitalization in patients using tranexamic acid and who underwent diagnostic or treatment procedures, including bladder sonography or cystoscopy (the inspection of the bladder and urethra with cystoscopy and removal of clots with suction), bladder blood clot evacuation with a Toomey bladder evacuator, bladder instillation, and intermittent or continuous bladder irrigation with a foley catheter.

Statistical Methods
We performed between-group comparisons by using the paired t-test [17] for continuous variables and McNemar's [18] test for categorical variables. Cox regression [19] analysis with covariates was used to estimate the relationship and differences in the risk of bleeding between the TURP and laser surgery groups. The hazard ratios (HRs) and 95% confidence intervals (CIs) for the outcomes were measured for all groups. The Kaplan-Meier method was used to estimate the outcomes of the study cohorts. The differences between the curves were examined using the log-rank test [20].

Propensity Score Matching
Propensity score matching (PSM) is a popular approach for estimating treatment effects by using observational data [21]. In order to reduce selection bias and the effects of confounders, we used robust PSM to create matched sets of patients who underwent TURP and those who underwent laser surgery at a ratio of 1:1 with full matching without replacement. Logistic regression was used for propensity score calculation [22]; the covariates used in the logistic regression model were age, DM, HCD, CKD, hyperlipidemia, CCI, and the index year at the start of the follow-up. The flowchart of surgery type and matching is presented in Figure 1.
Baseline characteristics were matched using PSM to reduce potential selection bias. PSM was performed using multivariate logistic regression, and matching was performed using the package of R Statistical Software "MatchIt" (version 4.4.0; R Core Team 2021, Vienna, Austria). Statistical analysis was performed using SPSS 21.0 (SPSS Inc., Chicago, IL, USA), and statistical significance was set at p < 0.05.

Patient Characteristics
Patients with noncancerous BPH who underwent surgery between 2015 and 2018 were included in this study ( Figure 1). We performed PSM and included 3302 patients; of them, 1651 underwent laser surgery, and 1651 underwent B-TURP or M-TURP. The mean age and CCI of patients were 70.9 ± 8.4 years and 3.2 ± 2.6. The demographic characteristics of the TURP group and laser surgery groups are listed in Table 1. Patients in both groups were men with similar age and comorbidity distributions. No differences in common event-related comorbidities and medication history were observed between the groups.  Table 2 presents the results of further subgroup analysis to investigate the baseline differences among the M-TURP, B-TURP, PVP, ThuVARP, HoLEP, and DiLEP groups ( Table 2). No differences in common comorbidities and medication history were observed among the six groups.

Comparison of ER Visits with Rehospitalization Due to Postoperative Bleeding
We estimated the proportion of patients who returned to our ER due to clot retent within 15, 30, 60, and 90 days after surgery (Table 3).

Comparison of ER Visits with Rehospitalization Due to Postoperative Bleeding
We estimated the proportion of patients who returned to our ER due to clot retention within 15, 30, 60, and 90 days after surgery ( Table 3).

Effect of Surgery Type on Bleeding Events
The effect of six surgeries on the risk of bleeding events was assessed using the multivariate Cox regression analysis with adjustment for age, sex, CCI, comedications (statins, ACEIs, β-blockers, anticoagulants, antiplatelets, spironolactone, and 5-α-reductase inhibitors), comorbidities (HCD, hyperlipidemia, DM, and CKD), and the index year at the start of follow-up.
We evaluated the bleeding events leading to ER visits within 90 days after surgery. The  Table 4).
The analysis of the bleeding visits leading to rehospitalization within 90 days after surgery revealed that the adjusted HRs of the PVP (0.61; 95% CI, 0.38-1.00 p = 0.050) and ThuVARP (0.67; 95% CI, 0.47-0.95, p = 0.024) groups were significantly lower than that of the M-TURP group. Moreover, the HR of the DiLEP (1.12; 95% CI, 0.71-1.75, p = 0.634) group was higher than that of the M-TURP group (Table 5).
Regarding bleeding events leading to ER visits, the 90-day cumulative incidence following laser surgery was not significantly different from that of TURP (p = 0.796; Figure 3A), and the same was verified between PVP, ThuVARP, and B-TURP and M-TURP. Moreover, the bleeding risk following HoLEP and DiLEP was higher than after M-TURP (p = 0.050, p = 0.708; Figure 3B).
Regarding the bleeding events leading to rehospitalization, the 90-day cumulative incidence post-laser surgery was not significantly different after TURP (p = 0.145; Figure 4A). The cumulative incidence of bleeding events leading to rehospitalization was lower in the PVP and ThuVARP groups than that of the M-TURP cohort (p = 0.026, p = 0.069; Figure 4B).

Comparison of Postoperative Bleeding between the Subgroups of Demographics, Comorbidities, and Comedications
Significantly higher rates of clot retention leading to rehospitalization were noted in patients aged >80 years (2.93; 95% CI, 1.52-5.65, p = 0.001). A high CCI score (>4) was observed in patients visiting the ER due to bleeding events after 90 days of surgery (2.36; 95% CI, 1.43-3.92, p = 0.001). No significant differences in bleeding events were observed between patients visiting the ER and those who were rehospitalized in the subgroups of comorbidities (DM, HCD, CKD, hyperlipidemia, and heart failure) and comedications (statins, ACEIs, β-blockers, anticoagulants, antiplatelets, spironolactone, and 5-α-reductase inhibitors).  Regarding bleeding events leading to ER visits, the 90-day cumulative incidence following laser surgery was not significantly different from that of TURP (p = 0.796; Figure  3A), and the same was verified between PVP, ThuVARP, and B-TURP and M-TURP. Moreover, the bleeding risk following HoLEP and DiLEP was higher than after M-TURP (p = 0.050, p = 0.708; Figure 3B). Regarding the bleeding events leading to rehospitalization, the 90-day cumulative incidence post-laser surgery was not significantly different after TURP (p = 0.145; Figure  4A). The cumulative incidence of bleeding events leading to rehospitalization was lower in the PVP and ThuVARP groups than that of the M-TURP cohort (p = 0.026, p = 0.069; Figure 4B).   Regarding the bleeding events leading to rehospitalization, the 90-day cumulative incidence post-laser surgery was not significantly different after TURP (p = 0.145; Figure  4A). The cumulative incidence of bleeding events leading to rehospitalization was lower in the PVP and ThuVARP groups than that of the M-TURP cohort (p = 0.026, p = 0.069; Figure 4B).

Discussion
To the best of our knowledge, our study is the first to compare six transurethral procedures for the treatment of BPH or lower urinary tract infection and to investigate differences in postoperative bleeding. No study has compared six transurethral procedures by using balanced baseline characteristics.

Participants in Balanced Groups
We presented the results of laser surgery and TURP with adjustments for demographics, comorbidities, and comedications by using PSM. Most studies did not perform a multivariate analysis with adjustments for confounders [5]. We adjusted the outcomes be-tween the age, comorbidities, and comedication cohorts. In order to avoid the inconsistency of comorbidities, we used CCI [23].
Shamout et al. reported that 28.6% and 11.1% of the patients returned to ER or were rehospitalized post-M-TURP, respectively. This was higher than in our study [29]. In another study, the reported the percentage of patients who returned to ER or were rehospitalized post-M-TURP (8.0%, 2.8%), HoLEP (9.8%, 0.9%), and PVP (7.5%, 1.7%) was lower than in our study [30].
The evidence involving ThuVARP and DiLEP and their bleeding risk and other complications is still scarce [31]. Therefore, we believe that a complete follow-up study of all six transurethral procedures is needed to be able to fully compare and balance their risk/benefit.

Differences in Postoperative Bleeding between Different Surgery Types
Early postoperative bleeding is a frequent complication occurring within 1 to 3 months after BPH surgery [32,33] and is treated with bladder irrigation and clot removal, if necessary [34].
PVP and ThuVARP are safer than M-TURP because blood transfusion, clot retention, hemoglobin decline, or transurethral resection syndrome is less likely to occur [35]. Our findings support changes in the surgical treatment of BPH from M-TURP to new laser methods [24,36,37]. However, DiLEP and HoLEP resulted in more bleeding events than M-TURP. Early DiLEP and HoLEP therapies are associated with increased postoperative complications, although some later studies reported lower morbidity in patients undergoing these therapies [38,39].

Comparison between the Subgroups of Demographics, Comorbidities, and Comedications
The high CCI score subgroup demonstrated increased HRs for ER visits within 90 days after surgery, whereas patients aged >80 years exhibited a significant risk of bleeding leading to rehospitalization. Our study suggested that laser surgery is safer than M-TURP for BPH because of reduced catheter time and risk of bleeding, even in patients receiving anticoagulant or antiplatelet therapy [23,35,40].
We investigated four types of laser surgery and two types of TURP and demonstrated that laser therapies are associated with a shorter hospital stay, less bleeding, and lower transfusion requirements than TURP, even in patients receiving anticoagulant or antiplatelet therapy [41,42]. However, the rates of ER visits and rehospitalization revealed that DiLEP and HoLEP were not associated with reduced bleeding complications [43,44]. In the recent decade, favorable outcomes were obtained after laser surgery, but a high percentage of urologists still prefer TURP. This may be because of the perceived precipitous learning curve of laser surgery [45][46][47].

Limitations
First, we used a retrospective cohort design. We matched all potential confounders between the surgery cohorts, but the selection and observational bias may still exist. Meanwhile, the sample size differs between all sub-groups studied, and this may result in bias. However, the use of data from the NHIRD, a large and well-validated database, may control the bias.
The second limitation is the accuracy of diagnosis. Potential misdiagnosis exists in the NHIRD because of the possible misclassification of ICD-9-CM and ICD-10-CM codes.
Third, the NHIRD lacks specific clinical (e.g., the weight of the resected prostate, catheterization time, and hemoglobin decline) and lifestyle behavior (e.g., Body Mass Index, smoking behavior, and alcohol use) data.
Finally, the study lacks information regarding the surgeons' experience and training. These limitations do not compromise the conclusions of this study. However, largescale prospective studies should be conducted to further validate our results.

Conclusions
Among the six investigated, transurethral procedures for BPH, PVP, and ThuVARP were safer than M-TURP because bleeding events and clot retention were less likely to occur, even in patients receiving anticoagulant or antiplatelet therapy. However, DiLEP and HoLEP did not result in fewer bleeding events than M-TURP. Our findings suggest that PVP and ThuVARP are effective alternatives to M-TURP for the surgical treatment of BPH. These results would provide useful information for urologists and patients with BPH.  Informed Consent Statement: According to the THIRD dataset incorporating encrypted secondary data, all participants are unfeasible to be identified; hence, the informed consent requirement was waived.

Data Availability Statement:
The personal electronic data of NHIRD protects by the Computer-Processed Personal Data Protection Law. The results of the academic study are available for researchers from the NHIRD of Taiwan. All researchers accord with the criteria for access to confidential data, which cannot be shared publicly because of legal guidelines imposed by the government of Taiwan under the "Personal Information Protection Act". Requests for data can be applied as an official proposal to the NHIRD (https://dep.mohw.gov.tw/dos/np-2497-113.html (accessed on 7 August 2022)). The contact information for needed data is: 886-2-85906828; Email: sthuiying@mohw.gov.tw.

Conflicts of Interest:
The authors declare no conflict of interest.