Comparison of 90-day re-admission rates between open retropubic radical prostatectomy (RRP), laparoscopic RP (LRP) and robot-assisted laparoscopic prostatectomy (RALP)

Authors


  • S.-D.C. and J.J.K have equal contribution to this manuscript.

Herng-Ching Lin, School of Health Care Administration, Taipei Medical University, 250 Wu-Hsing St., Taipei 110, Taiwan. e-mail: henry11111@tmu.edu.tw

Abstract

Study Type – Therapy (case series)

Level of Evidence 4

What's known on the subject? and What does the study add?

With the increased use of laparoscopic radical prostatectomy (LRP) and robot-assisted laparoscopic prostatectomy (RALP), a growing number of publications have sought to compare these more advanced techniques to retropubic RP (RRP). Many studies have found RALP and LRP to be associated with lower blood loss, postoperative pain, and hospital stay when compared with RRP.

The present study showed that, after adjusting for potential confounders, patients undergoing RALP had a lower risk of 90-day re-admission than patients undergoing RRP. However, there was no significant difference in the odds of being re-admitted ≤90 days after RP between patients undergoing a LRP and RRP.

OBJECTIVE

  • • To examine the risk of 90-day re-admission among patients undergoing retropubic radical prostatectomy (RRP), laparoscopic RP (LRP), and robot-assisted laparoscopic prostatectomy (RALP) in Taiwan.

PATIENTS AND METHODS

  • • We identified 2741 hospitalised patients who underwent a RP. Of these 2741 cases, 1773 patients underwent RRP, 694 LRP, and 274 RALP.
  • • We performed a conditional (fixed-effect) logistic regression model to explore the odds of 90-day re-admission from RP among patients undergoing RRP, LRP, and RALP.

RESULTS

  • • In all, 257 of the 2741 (9.4%) sampled subjects were re-admitted ≤90 days of the index RP.
  • • Patients undergoing a RALP had a significantly lower incidence rate of 90-day re-admission than patients undergoing a RRP or LRP (3.6% vs 10.7% vs 8.2%, P < 0.001).
  • • Compared with patients undergoing a RRP, the odds ratio (OR) of 90-day re-admission for patients undergoing a RALP was only 0.35 (95% confidence interval [CI] 0.19–0.68) after adjusting for patient age, geographic region, year of surgery, Charlson Co-morbidity Index score, and surgeon age and the number of RP cases/year.
  • • However, there was no significant difference in the odds of being re-admitted ≤90 days of RP between patients undergoing a LRP and RRP.
  • • The adjusted odds of 90-day re-admission for patients undergoing a RALP were 0.46 (95% CI 0.23–0.94) those of patients undergoing a LRP.

CONCLUSIONS

  • • Our study shows that patients undergoing a RALP had a lower adjusted risk of 90-day re-admission than patients undergoing RRP. However, no significant differences were identified between LRP and RRP.
Abbreviations
(R)(L)RP

(retropubic) (laparoscopic) radical prostatectomy

RALP

robot-assisted laparoscopic prostatectomy

NHI(RD)

National Health Insurance (Research Database)

IRB

Institutional Review Board

CCI

Charlson Co-morbidity Index.

INTRODUCTION

Prostate cancer is the leading cancer diagnosis and second leading cause of cancer-related mortality for men in the USA [1]. Surgery is the most commonly used treatment for clinically localised prostate cancer and the only treatment shown to provide a survival advantage [2,3]. Retropubic radical prostatectomy (RRP) has long been the standard surgical treatment for clinically localised prostate cancer [4,5]. However, the procedure is associated with significant blood loss, incontinence, erectile dysfunction and prolonged convalescence. Recent surgical advances have led to the development of minimally invasive RP techniques, e.g. laparoscopic RP (LRP) and robot-assisted laparoscopic prostatectomy (RALP). Over the last decade, minimally invasive RP techniques have come to be considered standard methods of treatment in urology departments across North America [6–10].

Following the increased use of LRP and RALP, a growing number of publications have sought to compare these more advanced techniques to RRP. Although many studies have found that RALP and LRP are associated with lower blood loss, postoperative pain, and hospital stay when compared with RRP [11–19], few studies have investigated the rates of re-admission among patients undergoing these surgical techniques. As far as we know, only three studies comparing the rates of re-admission among patients undergoing RRP, LRP and RALP have been reported. Touijer et al. [19] reported the rate of re-admissions after LRP to be relatively higher than after RRP. Similarly, Rabbani et al. [20] also found a higher rate of re-admission in their LRP group. Furthermore, another study by Nelson et al. [21] suggested that re-admission rates were similar in RALP and RRP. However, to date, there is a lack of data directly comparing the risk of re-admission of all three surgical techniques (RRP vs LRP vs RALP).

To fill this gap in the literature, using a nationwide population-based dataset in Taiwan, the present study aimed to examine the risk of 90-day re-admission among patients undergoing RRP, LRP and RALP during a 5-year period. To the best of our knowledge, this is the largest and most complete nationwide population-based study to compare the prevalence and risk of re-admission among patients undergoing RRP, LRP, and RALP.

PATIENTS AND METHODS

This nationwide population-based study used data sourced from Taiwan's National Health Insurance Research Database (NHIRD). The NHIRD is published by the National Health Research Institutes and includes all the medical claims for 22.60 million of Taiwan's 22.96 million population under the Taiwan NHI. Data in the NHIRD also include a registry of contracted medical facilities, a registry of board-certified physicians, a registry of catastrophic illness patients, and expenditures for prescriptions dispensed at contracted pharmacies. The inpatient medical claims provide one principal diagnosis coded according to the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) and up to four secondary diagnoses for each patient. The NHIRD offers an exclusive opportunity for the researchers to trace all the medical utilisations of all the enrolees under the Taiwan NHI.

This study was exempted from full review by an Institutional Review Board (IRB) after consulting with the director of the Taipei Medical University IRB, as the NHIRD consists of de-identified secondary data released to the public for research purposes.

We identified 2741 hospitalised patients who underwent a RP between January 2006 and December 2009 based on the ICD-9-CM procedure code 605. Of these 2741 cases, we further identified those cases who underwent a LRP or RALP by the additional ICD-9-CM procedure code 542.1. While there is no ICD-9-CM procedure code to differentiate LRP and RALP, in Taiwan there were only three urologists capable of performing RALP during the period 2006 to 2009. These three urologists only perform RALP. Therefore, 1773 patients underwent RRP, 694 LRP, and 274 RALP during the period 2006 to 2009.

The primary study outcome was the binary variable: ‘90-day re-admission’, which was defined as whether or not a patient was re-admitted to a hospital ≤90 days of their RP. The diagnoses for re-admission included intestinal infections, UTIs, haematuria, pneumonia, prostatic hypertrophy, retention of urine, cardiac complications during or resulting from the procedure, respiratory complications resulting from the procedure, and postoperative infection. In addition, the reasons for being re-admitted could include any of the following procedures: operations on the ureter, operations on the urinary bladder, operations on the urethra, operations on the urinary tract, or operations on the prostate and seminal vesicles.

In this study, we also took patient and surgeon characteristics into consideration in the regression modelling. Patient characteristics consisted of age, geographic region, and severity of illness. As no illness severity index is currently available in the NHIRD, we used the Charlson Co-morbidity Index (CCI) to quantify pre-existing co-morbidities as a means of adjusting for the higher potential mortality risks associated with 19 conditions (congestive heart failure, myocardial infarction, liver disease, cancer, dementia, etc.) with the score set at zero in the absence of co-morbid conditions. In this study, a patient was stratified into three levels based on CCI score: ≤1, 2, and ≥3.

Surgeon characteristics included the surgeon's age (≤40, 41–50 and >50 years) and their RP case volume during the study period. Surgeons were sorted in ascending order of their total volume of RPs, with the threshold points (high, medium and low) being determined by the volume that most closely sorted the sample patients into three roughly equally sized groups. This method is consistent with the methodologies adopted in many previous studies [22,23]. The sample of 2741 patients was subsequently divided into three surgeon volume groups: <7 cases/year (hereafter referred to as low volume), 7–18 cases/year (medium volume), and >18 cases/year (high volume).

Pearson chi-squared tests were used to examine the differences among patients undergoing RRP, LRP and RALP for patient and surgeon characteristics. We then performed a conditional (fixed-effect) logistic regression model (conditioned on hospitals in order to partition out systematic hospital-specific variation) to explore the odds of 90-day re-admission after an index RP among patients undergoing RRP, LRP, and RALP. A two-sided P < 0.05 was considered to indicate statistical significance.

RESULTS

Of the 2741 sampled subjects, the mean (sd) age was 66.1 (7.1) years; 66.2, 66.0 and 65.9 years for patients undergoing RRP, LRP, and RALP, respectively (P= 0.653). Table 1 presents the distribution of patient and surgeon characteristics. It shows that patients undergoing RALP were more likely to reside in the northern part of Taiwan (P < 0.001) and to have a CCI score of ≤1 (P < 0.001) than their counterparts undergoing RRP or LRP. Patients receiving a RALP were also more like to have surgeons aged ≤40 years than other age groups (P < 0.001). In addition, patients receiving a RALP had a greater tendency than their counterparts to have a surgeon who had a high number of RP cases/year (P < 0.001).

Table 1. Comparisons of patients undergoing open retropubic radical prostatectomy, laparoscopic radical prostatectomy and robotic-assisted laparoscopic radical prostatectomy in relation to patients and physician characteristics in Taiwan, 2006 ∼ 2009 (n= 2741)
VariableOpen RRPLRPRALPChi-square value
N (%) N (%) N (%)(P)
Patients characteristics    
 N1773694274 
 Age, years:   10.42 (0.108)
  <60324 (18.3)116 (16.7)59 (21.5) 
  60–69784 (44.2)337 (48.6)111 (40.5) 
  70–79598 (33.7)208 (30.0)89 (32.5) 
  >7067 (3.8)33 (4.7)15 (5.5) 
 Geographic region:   141.75 (<0.001)
  Northern710 (40.1)176 (25.4)182 (66.4) 
  Others1063 (59.9)518 (74.6)92 (33.6) 
 CCI score:   20.69 (<0.001)
  ≤11334 (75.2)514 (74.1)226 (82.5) 
  2294 (16.6)107 (15.4)43 (15.7) 
  ≥3145 (8.2)73 (10.5)5 (1.8) 
Physicians characteristics    
 Age, years:   130.89 (<0.001)
  ≤40239 (13.5)130 (18.7)59 (21.5) 
  41–50942 (53.1)444 (64.0)197 (71.9) 
  ≥51592 (33.4)120 (17.3)18 (6.6) 
 RP cases/year:   99.31 (<0.001)
  <7623 (35.1)244 (35.2)50 (18.2) 
  7–18554 (31.3)264 (38.0)61 (22.3) 
  >18596 (33.6)186 (26.8)163 (59.5) 

Table 2 shows the incidence of 90-day re-admission among patients undergoing RRP, LRP, and RALP. In all, 257 of the 2741 (9.4%) sampled subjects were re-admitted ≤90 days of the index RP. Patients undergoing a RALP had a significantly lower 90-day re-admission incidence rate than patients undergoing an RRP or LRP (3.6% vs 10.7% vs 8.2%, P < 0.001, chi-squared value 15.43). The conditional logistic regression model showed that patients undergoing a RALP were at 0.32 (95% CI 0.16–0.60) times the risk of their counterparts who underwent an RRP to be re-admitted ≤90 days after their index RPs. However, there was no significant difference in the odds of being re-admitted ≤90 days between patients undergoing an LRP and RRP. We further analysed the relationship between LRP and RALP for 90-day re-admission. We found that patients undergoing RALP only had 0.42 (95% CI 0.21–0.84) times the odds of those undergoing LRP for 90-day re-admission.

Table 2. The distribution of 90-day re-admission of patients undergoing open RRP, LRP and RALP (n= 2741)
VariableTotal sampleOpen RRPLRPRALP
  1. *P < 0.05; †P < 0.001.

N 27411773694274
90-day re-admission, n (%):    
Yes257 (9.4)190 (10.7)57 (8.2)10 (3.6)
No2484 (90.6)1583 (89.3)637 (91.8)264 (96.4)
Crude OR (95% CI)1.000.75 (0.55–1.02)0.32 (0.16–0.60)
Crude OR (95% CI)1.34 (0.98–1.83)1.000.42* (0.21–0.84)

Table 3 presents the covariate-adjusted odds ratios (ORs) for 90-day re-admission. We found that compared with patients undergoing an RRP, the OR of 90-day re-admission for patients undergoing an RALP was only 0.35 (95% CI 0.18–0.68) after adjusting for patient age, geographic region, CCI score, year of surgery and surgeon age and the number of RP cases/year. In addition, as expected, patients who had CCI score of ≥3 were more likely to be re-admitted ≤90 days of the index RP than patients who had CCI score of ≤1 (OR 1.94; 95% CI 1.31–2.89).

Table 3. Adjusted ORs of 90-day re-admission of the sampled patients (n= 2741)
Variable90-day re-admission
Adjusted OR (95% CI) P
Type of prostatectomy surgical technique
 RALP0.35 (0.18–0.68)0.002
 LRP0.76 (0.55–1.05)0.096
 Open RRP1.00 
Patients characteristics
 Age, years:  
  <601.00 
  60–690.86 (0.59–1.25)0.426
  70–791.25 (0.85–1.82)0.253
  >700.91 (0.44–1.89)0.798
 Geographic region:
  Northern1.00 
  Others0.75 (0.57–0.98)0.037
 CCI score:  
  ≤11.00 
  21.50 (1.08–2.09)0.017
  ≥31.94 (1.31–2.89)0.001
Physicians characteristics
 Age, years  
  ≤401.00 
  41–500.97 (0.67–1.42)0.893
  ≥510.92 (0.61–1.39)0.696
 RP cases/year:
  <71.00 
  7–180.82 (0.60–1.12)0.207
  >180.59 (0.41–0.85)0.004
 Year of surgery:
  20061.00 
  20071.29 (0.92–1.81)0.146
  20081.06 (0.73–1.53)0.767
  20090.89 (0.61–1.30)0.552

We further analysed the adjusted ORs of 90-day re-admission between LRP and RALP. The odds of 90-day re-admission for patients undergoing a RALP were 0.46 (95% CI 0.23–0.94) those of patients undergoing a LRP after adjusting for patient age, geographic region, year of surgery, CCI score, and surgeon age, and the number of RP cases/year.

DISCUSSION

RRP represents one of the major treatment options for patients with clinically localised prostate cancer [4,5]. Over the past decade, minimally invasive RP approaches for surgical management of prostate cancer have become increasingly utilised [24]. A good number of studies have recently purported to address the question of whether the open, pure laparoscopic or robotic approach to RP leads to better results [11–19]. The re-admission rate after surgery is a tool used to evaluate the total quality of clinical care. However, no published data about the re-admission rate after RP, LRP and RALP in Asia has been made available to date, with the risk of re-admission among patients undergoing these procedures remaining unclear. As far as we know, this is the first nationwide population-based study to compare the risk of re-admission among patients undergoing RRP, LRP, and RALP in Asia. After adjusting for patient age, geographic region, year of surgery, CCI score, and physician age and the number of RP cases/year, we found that patients undergoing RALP were at 0.35 the risk of 90-day re-admission of patients undergoing RRP. Furthermore, we found no significant differences in the 90-day risk of re-admission between patients undergoing LRP and RRP.

Nevertheless, previous studies by Touijer et al. [19] and Rabbani et al. [20] found the rate of re-admissions after LRP to be relatively higher than after RRP. Nelson et al. [21] suggested that re-admission rates were similar in patients undergoing RALP and RRP. Such inconsistent results may be due to differences in patient selection, number of patients included in the series, and methodological differences in study design. In addition, previous studies suffered from several methodological limitations, which could have undermined the strength of their findings. For example, unlike the present study, which has taken patient co-morbidities into consideration, previous investigations only performed univariate analyses that cannot be used to adequately control for potential confounders. Furthermore, previous studies have tended to use hospital-based data and/or used questionnaire-based collection methods. On account of being questionnaire-based studies, their investigations were vulnerable to recall biases, and on account of being hospital-based were vulnerable to selection biases. Additionally, their small case numbers may have precluded their ability to garnish the necessary statistical power to detect differences. Our population-based study therefore extends the published findings as it avoided the biases of previous studies and can be generalised to the overall population.

Recently, Fridriksson et al. [25] suggested that the 90-day re-admission rate after RP decreased over time. Using data from the Prostate Cancer Data Bases in Sweden (PCBaSe), they found that only three LRPs were registered in the period between 1996 and 1999. The re-admission rate after RRP was highest in that period (437/1638, 27%). In the period between 2000 and 2003 the re-admission rate was slightly lower for RRP (612/4539, 14%) compared with LRP (100/502, 20%). In the period between 2004 and 2006 the re-admission rate was only 11% for both RRP (600/5480) and LRP (166/1499). Similarly, in the present study, we found the re-admission rate to be slightly lower for LRP (8.2%) compared with RRP (10.7%) in the period between 2006 and 2009. In addition, we found that the re-admission rate after RALP (3.6%) to be significantly lower than that of RRP or LRP in Taiwan. As the most common surgery-related causes for re-admission were postoperative infections and bleeding, we suggest that the quality of clinical care for patients with prostate cancer who underwent RALP was better than RRP and LRP.

RALP has rapidly overtaken both LRP and RRP in the surgical management of prostate cancer in the USA, accounting for >75% of all the RPs currently being performed [26]. RALP is more attractive than LRP to many surgeons because of the more rapid transfer of open-surgery skills to a minimally invasive setting, three-dimensional vision, six degrees of freedom, downscaling of movements and elimination of physiological tremor [27–29]. Despite the rampant adoption of RALP in the USA, the comparison of RALP, LRP, and RRP remains relevant to current urological practice, as RRP and LRP continue to thrive outside of USA, e.g. in Europe and Asia. In the present study, we clearly showed that patients undergoing RALP had a lower risk of 90-day re-admission than patients undergoing RRP. We think that the present study is an important step in helping to define the relative efficacy of RALP, LRP, and RRP as treatment options for clinically localised prostate cancer. Moreover, the present findings may prove valuable to patients trying to make an objective decision about the various treatment options.

Because the present study used nationwide population-based datasets, its robust findings can be generalised to the population as a whole. Furthermore, the present data include the availability of multiple patient and hospital level variables that allow for analyses of specific treatment methods and their clinical outcomes. However, several limitations of the present study merit note. First, because we identified patients diagnosed with prostate cancer by ICD-9-CM codes from administrative claims data, the validity of diagnoses could be compromised. Furthermore, because patients with true prostate cancer may be misidentified as non- prostate cancer cases if they do not seek appropriate medical care, misclassification would probably bias our results towards the null. Second, this dataset did not allow us to account for differences in prostate cancer severity among patients. Moreover, we could not adjust for tumour characteristics, e.g. tumour grade or stage. Third, a selection bias may be present for the type of surgical technique offered to each patient, as well as for the different patient referral patterns that may exist for each surgeon. Only a randomised study incorporating all three surgical approaches for RP would be likely to produce better matching of preoperative variables. Furthermore, not all re-admissions are the same; but unfortunately on account of the few cases characterising each re-admission diagnostic category, we were unable to perform an analysis after stratification for re-admission diagnosis. Finally, it is possible that the residual confounding of some unmeasured socioeconomic characteristics contributed to the results detected in the present study. In effort to better adjust for these factors we divided the present analysis into two regional categories: ‘Northern’ and ‘Others’. We chose to adjust for ‘Northern’ on account of Taiwan's capital, Taipei, being located in the North as well as for the North's relatively higher level of development and infrastructure.

Ultimately, as the technology of robotic surgery becomes more affordable, the decisive factor affecting the worldwide adoption of RALP will be the relative benefits to the patient. The present study clearly showed that, after adjusting for potential confounders, patients undergoing RALP had a lower risk of 90-day re-admission than patients undergoing RRP. However, there was no significant difference in the odds of being re-admitted ≤90 days of RP between patients undergoing an LRP and RRP. The present results provide the best available summary of recent population-based outcomes data, and thus are a potentially important source of information for treatment planning and patient counselling. We recommend large-scale studies in other regions or countries to be carried out to confirm the findings of the present one.

CONFLICT OF INTEREST

All authors report no disclosures.

Ancillary