After the approval of institutional review board (H-1202-065-398), we extracted the data of patients who were diagnosed of BPH in the year of 2008 and their follow-up visits for 3 years from HIRA database. The diagnosis of BPH was defined as 2 or more records of reimbursement with the International Classification of Diseases, 10th revision (ICD-10) diagnosis code of N40.0 used as a primary or secondary diagnosis. The first diagnosis was defined as an index diagnosis and patients with an index diagnosis of 2008 were traced for 3 years. Patients diagnosed with BPH within the previous 12 months before the index diagnosis were excluded. Patients diagnosed with prostate cancer (ICD-10, C61) within the 12 months after the index diagnosis or having claims suggesting a prior prostatic surgery, prostate cancers, inflammatory diseases of the prostate, neurological diseases or conditions that could affect LUTS were also excluded (Table 1).

Incidence was calculated as the number of new cases identified in 2008 divided by the number of at-risk individuals. Age, mean and median numbers of office visits and the tier of hospital where the index diagnosis was made were identified.

To determine treatment patterns, patient status was determined every 3 months as medication, no medication, or surgery status. Medication status was defined as having a prescription record for ≥25% of the 3 months. Surgery was assigned as having a record of the following procedures: transurethral resection of the prostate (TURP, R3975), photoselective vaporization of the prostate (PVP, R3976), holmium laser enucleation of prostate (HoLEP, R3977), open prostatectomy (R3950), and thermal therapy (R3516). All the other situations were assigned as no medication status. Medications were categorized as alpha-blockers, 5-alpha reductase inhibitors, or anticholinergics. Alfuzosin, doxazosin, tamsulosin, terazosin, and silodosin were categorized as alpha-blockers. Dutasteride and finasteride were included as 5-alpha reductase inhibitors. Oxybutynin, propiverine, fesoterodine, solifenacin, and tolterodine were categorized as anticholinergics.

For those who underwent surgery once during the 3 years of follow-up, preoperative medication status was assessed 1 month prior to the surgery. For those who were taking medication 1 month prior to the surgery, medication-free rate was calculated 6–36 months postoperatively in 3-month intervals.

The patient status such as medication, no medication and surgery were assessed as categorical variables. Age in a decade was treated as a categorical variable. Logistic regression analysis was used to calculate odds ratios (ORs). IBM SPSS Statistics ver. 21.0 (IBM Co., Armonk, NY, USA) was used for all analyses, and a p-value <0.05 was considered statistically significant.

A total of 386,873 men were identified and included in the BPH cohort of 2008. The incidence of BPH was 2,105 per 100,000 men, and increased with age (Fig. 1A). Mean age was 59.7 years. For index patients who were diagnosed with BPH in 2008, median number of office visits was 7. Median follow-up period was 428 days. Most patients were diagnosed in cities (Fig. 1B) or at clinics (60%) (Fig. 1C).

During the 3 years of follow-up, 27.1% of patients received medication for more than 9 months, 17.6% of patients took medication for more than a year. For the surgical treatment, 7,955 patients (2.1%) underwent surgery, including 252 patients who underwent surgery twice or more. The treatment rate increased with age until the age of 70 years (Fig. 2).

The timing of surgical treatment from the index diagnosis is shown in Fig. 3. During the first 3 months, 36.6% of patients underwent surgery. The cumulative incidence of surgical cases increased with time (Fig. 3A). The gradients of cumulative incidence were different between surgical methods (Fig. 3B).

TURP was the most commonly performed surgical procedure, followed by PVP and thermal therapy (Fig. 4). Open prostatectomy accounted for 1% of surgical treatments. As our cohort comprised patients diagnosed with BPH in 2008 and followed up until 2011, HoLEP constituted only 2% of BPH surgeries, as it has only been performed after 2010. Most surgically treated patients were in their 60s and 70s. Although the symptom severity was not assessed, older patients were more likely to receive treatment, with an OR of 1.56 for each additional decade (95% confidence interval, 1.55–1.58; p<0.001).

To assess medication free-rate, patients taking medication at 1 month preoperatively were assessed and traced until 3 years after the surgery (Fig. 5). At 1 month preoperatively, 50.5% of patients were taking medication. Alpha-blocker-containing regimens were the most common type of medication used, followed by 5-alpha reductase inhibitors. Medication-free rate at 1 year after the surgery was 82% on average. To calculate medication-free rate, we excluded HoLEP, because HoLEP had only been used in this cohort after 2010. Medication-free rate was highest for open prostatectomy (90.0%) and lowest for thermal therapy (78.1%). After 1 year, medication-free rate plateaued (average 82.9%), and at 36-month postsurgery, 83.9% of patients were not taking any medication. Among the patients who were taking medications at 36 months after the surgery, alpha-blockers were the most common type of medication used (alpha-blockers, A; 5-alpha reductase inhibitors, B; anticholinergics, C; A 5.2%, AB 3.9%, ABC 0.7%, AC 1.1%), and anticholinergic-containing regimens were maintained in 4.0% of patients, including anticholinergic monotherapy in 2.1% of patients. For those who were taking anticholinergics preoperatively, medication-free rate after 1 year was 73.3%, which was lower than the average, and 11.8% of them had maintained anticholinergics-containing regimens.