Rehospitalization after Radical Prostatectomy in a Nationwide, Population Based Study € Friðriksson,* Erik Holmberg, Jan Adolfsson, Mats Lambe, n Om Jo €r Stattin Anna Bill-Axelson, Stefan Carlsson, Jonas Hugosson† and Pa ¨ F, PS), Umea˚, Department Department of Surgical and Perioperative Sciences, Urology and Andrology, Umea˚ University (JO of Oncology, Institute of Clinical Sciences, Sahlgrenska Academy at University of Go¨teborg (EH) and Department of Urology, Sahlgrenska University Hospital (JH), Go¨teborg, Departments of Clinical Science, Intervention and Technology (JA) and Medical Epidemiology and Biostatistics (ML, AB-A) and Section of Urology, Department of Molecular Medicine and Surgery (SC), Karolinska Institutet, Stockholm and Regional Cancer Centre, Uppsala University Hospital (ML) and Department of Surgical Sciences, Uppsala University (AB-A), Uppsala, Sweden
Abbreviations and Acronyms CCI ¼ Charlson comorbidity index LRP ¼ laparoscopic RP NOMESCO ¼ Nordic Medico-Statistical Committee NPCR ¼ National Prostate Cancer Register PCBaSe ¼ Prostate Cancer data Base Sweden PSA ¼ prostate specific antigen RALP ¼ robot-assisted RP RP ¼ radical prostatectomy RRP ¼ retropubic RP Accepted for publication January 30, 2014. Study received Ume a University Hospital research ethical review board approval. Supported by the Swedish Research Council 825-2012-5047, Swedish Cancer Foundation 11 0471, V€asterbotten County Council and Lion’s Cancer Research Foundation at Ume a University. * Correspondence: Department of Surgery and perioperative Sciences, Urology and Andrology, Ume a University, Ume a University Hospital, 90187 Ume a, Sweden (telephone: þ46(0) 90 785 0000; FAX: þ46(0)90 143024; e-mail: jon. [email protected]). † Financial interest and/or other relationship with Pfizer.
Purpose: We investigated hospital readmission frequency during the 90 days after radical prostatectomy and assessed the readmission risk associated with potentially related variables. Materials and Methods: Using the population based, nationwide PCBaSe (Prostate Cancer data Base Sweden) we identified men diagnosed with incident prostate cancer between 2000 and 2011 who underwent radical prostatectomy as primary treatment. We used logistic regression analysis to examine the association of the risk of 90-day postoperative readmission with surgical method, calendar period, tumor risk category, hospital case load and patient characteristics. Results: During 90 postoperative days 2,317 of the 24,122 men (10%) identified were nonelectively readmitted, specifically 10% after retropubic, 9% after robotassisted and 11% after laparoscopic radical prostatectomy. The range in readmission frequency among hospitals was 0% to 35%. Higher readmission risk was associated with the early calendar period (2009 to 2011 vs 2000 to 2002 OR 0.71, 95% CI 0.61e0.83), greater age (70 or greater vs less than 60 years OR 1.17, 95% CI 1.00e1.36), higher risk category (high vs low OR 1.78, 95% CI 1.57e2.03), high comorbidity (Charlson comorbidity index 3 or greater vs 0 OR 1.77, 95% CI 1.29e2.44) and low hospital surgical volume (150 or greater vs fewer than 30 radical prostatectomies per year OR 0.70, 95% CI 0.60e0.81). Conclusions: Readmission rates after different radical prostatectomy methods were similar, ranging from 9% to 11%, with wide variation among hospitals. Readmission rates can be used as an indicator of perioperative care quality but potential confounders must be adjusted to avoid bias. Key Words: prostate, prostatic neoplasms, prostatectomy, patient readmission, perioperative care
See Editorial on page 10.
HOSPITAL readmission rates are considered a broad indicator of quality of care in medicine.1 In several fields of surgery they are used as a quality indicator of perioperative and postoperative clinical care.2 The
112
j
www.jurology.com
readmission rate after RP is approximately 5% at some centers with a high case load, ie high volume hospitals.3,4 However, this is likely not representative of all health care providers since lower perioperative
0022-5347/14/1921-0112/0 THE JOURNAL OF UROLOGY® © 2014 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
http://dx.doi.org/10.1016/j.juro.2014.01.109 Vol. 192, 112-119, July 2014 Printed in U.S.A.
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
mortality, fewer perioperative complications and shorter hospitalization have been noted for procedures performed by high volume surgeons at high volume hospitals.5e7 In the United Kingdom a registry based study showed a fourfold higher readmission rate of approximately 20% within a year of RP.8 In the United States 30-day postoperative readmission rates after 101,604 RPs covered by Medicare were 30% higher at low than at high volume hospitals.9 In a study from Taiwan 90-day readmissions were less frequent after RALP than after RRP.10 Because there are few studies with data on factors other than hospital volume and surgical method that can also influence the readmission risk after RP, we analyzed readmission risk in relation to surgical method, calendar period, tumor risk category, patient age, comorbidity and hospital surgical volume.
113
postoperative complications, group 4dserious medical conditions (ischemic heart disease, heart failure, heart arrhythmias, thromboembolic disease, pulmonary embolism, liver or kidney failure and cerebral vascular disease) and group 5dother diagnoses. We regarded groups 1 to 3 as surgical complications, and groups 4 and 5 as medical complications. We classified interventions within 90 days after RP as related or not related to postoperative complications. We further classified interventions related to postoperative complications as performed with local anesthesia, transurethral intervention or general anesthesia.
Hospital Surgical Volume During the study period RP was done at 52 hospitals with wide variation in the number at each hospital (see figure). Based on this distribution we categorized hospitals into 3 groups by the mean number of RPs performed per year, including lowdfewer than 30, intermediated30 to 149 and highd150 or greater.
MATERIALS AND METHODS
Risk Category
Study Population and Data Collection
Based on NPCR data we defined 3 risk categories, including lowdlocal clinical stage T1-2, Gleason score 6 or less and serum PSA less than 10 ng/ml, intermediated clinical stage T1-2, Gleason score 7 and/or PSA 10 to less than 20 ng/ml and highdclinical stage T3, Gleason score 8-10 and/or PSA 20 to less than 50 ng/ml according to a modification to NCCNÒ risk categorization.16
The NPCR of Sweden has captured approximately 98% of all newly diagnosed prostate cancer cases since 1998, in contrast to the Swedish Cancer Registry, for which registration is compulsory and mandated by law.11 NPCR data include time of diagnosis, patient age, tumor stage and differentiation, and serum PSA as well as primary treatment given or planned up to 6 months after diagnosis. The unique Swedish personal identity number in PCBaSe links NPCR to a number of other national health care registries and demographic databases.12 We analyzed data on men who underwent RP as primary treatment for prostate cancer between 2000 and 2011. We used PCBaSe to determine date of surgery, surgical method according to the NOMESCO Classification of Surgical Procedures (endoscopic KEC01 or retropubic RP KEC00), duration of perioperative hospitalization, comorbidity, readmission within 90 days of surgery, diagnoses at discharge after readmission according to ICD-9 or 10 and interventions at readmission according to the NOMESCO Classification of Surgical Procedures. We calculated CCI by adding the scores of each comorbid condition based on discharge diagnoses up to 10 years before the date of prostate cancer diagnosis and cancer diagnoses at other sites in the cancer registry.13,14 As a measure of socioeconomic status, we used education level based on information in LISA (Longitudinal Integration Database for Health Insurance and Labour Market Studies), which also provided information on marital status.15 Education level categories were lowdgrade school (9 years), mediumdhigh school (10 to 12 years) and highdcollege (greater than 12 years). Data on education level were available until 2009, covering 19,426 cases. We gathered information from each hospital in Sweden that had documented endoscopic RP to learn whether and when RALP was introduced and we classified the procedure accordingly as LRP or RALP. We classified readmissions by discharge diagnoses into 5 groups, including group 1dpostoperative infection, group 2dpostoperative bleeding, group 3dother
Statistical Methods We used multivariable logistic regression to calculate the OR as a measure of the risk of 90-day readmission, considering RP type, calendar period, patient age, tumor risk category, perioperative hospital stay, CCI, education level and hospital surgical volume. To analyze the risk of surgical methods and calendar periods we performed multivariable logistic regression analysis for each surgical method and for the 3 latest calendar periods17 using STATAÒ, release 12.1. The Umea˚ University Hospital research ethical review board approved the study.
Readmission rate vs surgical volume at individual hospitals for surgeries performed in 2000 to 2011.
114
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
2011 (7%) while the frequency after RRP and RALP changed only marginally. The range of readmission frequency varied particularly widely among hospitals with a small surgical volume (see figure). Of the 2,895 readmissions 2,530 showed a discharge diagnosis or a first contributing diagnosis other than prostate cancer and could be further analyzed. Of those readmissions 1,726 (68%) were related to surgical complications and 804 (32%) were related to medical conditions (Appendix 1). The most common causes of readmission related to surgery were postoperative infection (3%) and bleeding (1%). Of the 2,317 men who were readmitted a total of 1,192 interventions were performed, of which 815 (68%) were related to postoperative complications. The most common postoperative complications were cystoscopy, bladder catheterization and transurethral bladder neck incision (Appendix 2). Table 3 shows the results of multivariable logistic regression analysis of readmission after RP. The risk was lower in later calendar periods. It increased with patient age, tumor risk category and CCI, and decreased as hospital surgical volume
RESULTS Table 1 shows the number of men who underwent RP between 2000 and 2011, and the type of surgery. The number of RPs increased substantially during the study period with the proportion of RALPs increasing dramatically. Compared with men in the RRP group, those treated with RALP and LRP were more often younger than 60 years and in a lower tumor risk category, and had a higher education level and a shorter perioperative hospital stay. The proportion of married men and the comorbidity index were comparable regardless of surgical method (table 1). Table 2 shows the number of RPs and readmissions within 90 days by calendar period, excluding 298 readmissions that were planned, usually for catheter removal. Since 445 men were readmitted more than once, the total number of readmissions was 2,895. Readmission frequency for all RPs was slightly higher between 2000 and 2002 (12%) than between 2003 and 2011 (9% to 10%) but the frequency after LRP decreased dramatically from 2000 to 2002 (22%) to 2009 to
Table 1. Baseline characteristics of men who underwent prostatectomy from 2000 to 2011 in PCBaSe RRP No. pts Age: Median (IQR) No. less than 60 No. 60e64 No. 65e69 No. 70 or greater No. calendar period (%): 2000e2002 2003e2005 2006e2008 2009e2011 No. marital status (%): Married Divorced Widower Never married Missing data No. risk category (%):* Low Intermediate High Periop stay (days): Median (IQR) No. 1e2 (%) No. 3e4 (%) No. 5e6 (%) No. 7 or greater (%) No. CCI (%): 0 1 2 3 or Greater No. education level (%):† Low Medium High
LRP
16,375
RALP
1,354
6,393
Totals 24,122
63 (59e67) 4,128 (25.2) 5,290 (32.3) 5,073 (31.0) 1,884 (11.5)
63 (59e67) 404 (29.8) 424 (31.3) 407 (30.1) 119 (8.8)
63 (59e67) 1,870 (29.3) 1,945 (30.4) 1,865 (29.2) 713 (11.2)
63 (59e67) 6,402 (26.5) 7,659 (31.8) 7,345 (30.4) 2,716 (11.3)
2,780 5,396 4,511 3,688
(17.0) (33.0) (27.5) (22.5)
118 390 505 341
(8.7) (28.8) (37.3) (25.2)
0 580 1,813 4,000
(9.1) (28.4) (62.6)
2,898 6,366 6,829 8,029
(12.0) (26.4) (28.3) (33.3)
10,650 1,919 2,175 1,421 210
(65.0) (11.7) (13.3) (8.7) (1.3)
869 153 184 126 22
(64.2) (11.3) (13.6) (9.3) (1.6)
4,066 714 965 551 97
(63.6) (11.2) (15.1) (8.6) (1.5)
15,585 2,786 3,324 2,098 329
(64.6) (11.6) (13.8) (8.7) (1.4)
7,274 6,346 2,212
(45.9) (40.1) (14.0)
736 463 128
(55.5) (34.9) (9.6)
3,073 2,625 519
(49.4) (42.2) (8.3)
11,083 9,434 2,859
(47.4) (40.4) (12.2)
5 1,080 6,104 5,530 3,659
(4e6) (6.6) (37.3) (33.8) (22.3)
4 291 684 250 129
(3e5) (21.5) (50.5) (18.5) (9.5)
2 3,576 2,393 228 196
(2e3) (55.9) (37.4) (3.6) (3.1)
4 4,947 9,181 6,008 3,984
(3e6) (20.5) (38.1) (24.9) (16.5)
13,423 1,969 748 235
(82.0) (12.0) (4.6) (1.4)
1,138 163 45 8
(84.0) (12.0) (3.3) (0.6)
5,373 672 272 76
(84.0) (10.5) (4.3) (1.2)
19,934 2,804 1,065 319
(82.6) (11.6) (4.4) (1.3)
4,476 5,833 3,731
(31.9) (41.6) (26.6)
309 509 340
(26.7) (44.0) (29.4)
820 1,688 1,720
(19.4) (39.9) (40.7)
5,605 8,030 5,791
(28.9) (41.3) (29.8)
* Excluding 746 men with distant or regional metastasis, or missing data. † Information only available until 2009.
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
Table 2. RPs and readmissions within 90 days by calendar period Period 2000e2002: All RPs RRP LRP RALP 2003e2005: All RPs RRP LRP RALP 2006e2008: All RPs RRP LRP RALP 2009e2011: All RPs RRP LRP RALP Full study period: All RPs RRP LRP RALP
Table 3. Multivariable regression analysis of readmissions after RP by preoperative factors
No. RPs/Readmissions (%) 2,898/336 (11.6) 2,780/310 (11.2) 118/26 (22.0) 0 6,366/573 (9.0) 5,396/467 (8.7) 390/56 (14.4) 580/50 (8.6) 6,829/660 (9.7) 4,511/452 (10.0) 505/40 (7.9) 1,813/168 (9.3) 8,029/748 (9.3) 3,688/380 (10.3) 341/23 (6.7) 4,000/345 (8.6) 24,122/2,317 (9.6) 16,375/1,609 (9.8) 1,354/145 (10.7) 6,393/563 (8.8)
increased. A higher readmission risk was associated with longer perioperative hospital stay (7 days or greater vs 1 to 2 OR 1.90, 95% CI 1.60e2.27). Readmission risk was not associated with education level (table 3). Multivariable regression analysis of operative methods revealed that differences associated with calendar period, patient age, tumor risk category, comorbidity and surgical volume were stronger for RRP than for RALP (table 4). Table 5 shows the results of multiple regression analysis for the 3 latest calendar periods. In 2003 to 2005 the risk was higher after LRP than after RRP but in 2009 to 2011 the risk was quite similar for the surgical methods. In 2000 to 2011 the incidence of 90-day postoperative mortality was 33 of 16,375 men (0.20%) for RRP, 10 of 6,393 (0.16%) for RALP and 0 of 1,354 for LRP. To assess readmissions after RP performed by private health care providers we used NPCR data on 2007 to 2011. Of 1,040 men treated with RP (RRP in 601 and RALP in 439) at a private hospital 296 (28%) were also documented in the national patient registry and 39 (4%) were readmitted to a public hospital within 90 days of RP.
DISCUSSION In this nationwide, population based study of more than 24,000 RPs performed in Sweden during 2000 to 2011 the overall frequency of readmissions within 90 days postoperatively was approximately 10%. Except for LRP performed during 2000 to 2002,
115
Surgical method: RRP LRP RALP Calendar period: 2000e2002 2003e2005 2006e2008 2009e2011 Age: Less than 60 60e64 65e69 70 or Greater Risk category:* Low Intermediate High CCI: 0 1 2 3 or Greater No RPs/unit/yr: Less than 30 30e149 150 or Greater Education level:† Low Medium High
No. Readmissions (%)
OR (95% CI)
1,609 (9.8) 145 (10.7) 563 (8.8)
1.00 (referent) 1.27 (1.06e1.53) 1.13 (0.99e1.29)
336 573 660 748
(11.6) (9.0) (9.7) (9.3)
1.00 (referent) 0.74 (0.64e0.86) 0.76 (0.66e0.88) 0.71 (0.61e0.83)
563 718 734 302
(8.8) (9.4) (10.0) (11.1)
1.00 (referent) 1.04 (0.92e1.17) 1.10 (0.98e1.24) 1.17 (1.00e1.36)
881 (8.0) 955 (10.1) 391 (13.7)
1.00 (referent) 1.29 (1.17e1.42) 1.78 (1.57e2.03)
1,844 291 132 50
(9.3) (10.4) (12.1) (15.7)
1.00 (referent) 1.12 (0.98e1.28) 1.31 (1.08e1.60) 1.77 (1.29e2.44)
404 (11.3) 1,250 (9.8) 663 (8.6)
1.00 (referent) 0.80 (0.71e0.91) 0.70 (0.60e0.81)
581 (10.7) 810 (10.4) 585 (10.5)
1.00 (referent) 1.01 (0.90e1.14) 1.03 (0.91e1.17)
* Excluding 746 men with distant or regional metastasis, or missing data. † Information only available until 2009.
when it was first introduced, we found relatively small differences in readmission frequency among the 3 surgical methods. The high initial readmission rates for LRP were likely due to the learning curve. In contrast, readmission rates 2 years after RALP was introduced were similar to rates after RRP. The number of RALPs increased rapidly during the study period from no procedures in 2000 to 2002 to 50% of all RPs in 2009 to 2011. Clinical characteristics related to an increased risk of readmission were age, tumor risk category and CCI, particularly during 2000 to 2002. Of men who underwent RP 89% were younger than 70 years. Age distribution differed slightly among the surgical groups. The association of increased age with a somewhat increased risk of readmission was no longer evident during 2009 to 2011. The association between high risk category and higher readmission frequency weakened somewhat during the later calendar periods and 88% of men had low or intermediate risk cancer. The association between high CCI and increased readmission risk was particularly evident for CCI 3 or greater. However, men with CCI 3 or greater represented only 1% of the study group and in 94% CCI was 1 or less. Readmission frequency differences among hospitals were large, especially at those with low surgical
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
116
Table 4. ORs of readmission after RP by preoperative factors and surgical method
Table 5. ORs of readmission after RP by preoperative factors and period
OR (95% CI)
Calendar period: 2000e2002 2003e2005 2006e2008 2009e2011 Age: Less than 60 60e64 65e69 70 or Greater Risk category:* Low Intermediate High CCI: 0 1 2 3 or Greater No RPs/unit/yr: Less than 30 30e149 150 or greater
OR (95% CI)
RRP
LRP
RALP
1.00 (referent) 0.73 (0.63e0.86) 0.81 (0.69e0.94) 0.79 (0.66e0.93)
1.00 (referent) 0.79 (0.45e1.41) 0.49 (0.27e0.90) 0.21 (0.11e0.40)
Not available 1.00 (referent) 1.08 (0.77e1.51) 0.96 (0.70e1.32)
1.00 (referent) 1.11 (0.96e1.29) 1.22 (1.06e1.42) 1.34 (1.11e1.61)
1.00 (referent) 0.76 (0.48e1.21) 0.75 (0.46e1.21) 0.76 (0.36e1.60)
1.00 (referent) 0.99 (0.79e1.25) 0.95 (0.76e1.20) 0.90 (0.66e1.25)
1.00 (referent) 1.38 (1.22e1.56) 1.94 (1.67e2.25)
1.00 (referent) 1.48 (0.99e2.21) 1.02 (0.51e2.03)
1.00 (referent) 1.09 (0.91e1.32) 1.52 (1.13e2.05)
1.00 (referent) 1.10 (0.93e1.29) 1.33 (1.05e1.67) 2.04 (1.43e2.90)
1.00 (referent) 1.26 (0.73e2.17) 1.92 (0.80e4.54) Not available
1.00 (referent) 1.13 (0.85e1.48) 1.10 (0.72e1.68) 1.14 (0.52e2.50)
1.00 (referent) 0.91 (0.80e1.04) 0.75 (0.62e0.90)
1.00 (referent) 0.16 (0.10e0.25) 0.11 (0.06e0.20)
1.00 (referent) 1.31 (0.78e2.18) 1.27 (0.77e2.08)
Surgical method: RRP LRP RALP Age: Less than 60 60e64 65e69 70 or Greater Risk category:* Low Intermediate High CCI: 0 1 2 3 or Greater No RPs/unit/yr: Less than 30 30e149 150 or Greater
2003e2005
2006e2008
2009e2011
1.00 (referent) 2.24 (1.64e3.05) 1.37 (0.96e1.95)
1.00 (referent) 0.86 (0.61e1.22) 1.08 (0.87e1.34)
1.00 (referent) 0.66 (0.43e1.02) 0.82 (0.66e1.03)
1.00 (referent) 1.07 (0.84e1.35) 1.24 (0.98e1.58) 1.56 (1.15e2.11)
1.00 (referent) 0.99 (0.79e1.22) 0.84 (0.67e1.06) 1.24 (0.94e1.64)
1.00 (referent) 0.95 (0.76e1.19) 1.17 (0.95e1.44) 1.03 (0.79e1.35)
1.00 (referent) 1.46 (1.20e1.78) 2.13 (1.65e2.75)
1.00 (referent) 1.36 (1.13e1.63) 1.76 (1.37e2.25)
1.00 (referent) 1.10 (0.93e1.30) 1.76 (1.41e2.20)
1.00 (referent) 1.10 (0.84e1.44) 1.46 (0.99e2.15) 2.87 (1.66e4.99)
1.00 (referent) 1.30 (1.01e1.66) 1.52 (1.06e2.16) 0.96 (0.41e2.23)
1.00 (referent) 1.03 (0.82e1.30) 1.20 (0.86e1.68) 1.43 (0.85e2.40)
1.00 (referent) 0.70 (0.55e0.89) 0.61 (0.46e0.82)
1.00 (referent) 1.05 (0.83e1.34) 0.92 (0.70e1.22)
1.00 (referent) 1.05 (0.84e1.31) 1.11 (0.82e1.50)
* Excluding 746 men with distant or regional metastasis, or missing data.
* Excluding 746 men with distant or regional metastasis, or missing data.
volume. The latter hospitals showed the highest and lowest readmission rates, likely a result of the larger effect of random variation in hospitals with smaller numbers of events. Nevertheless, it might be prudent to assess perioperative care at hospitals with high readmission rates. Our finding of no association between readmission risk and education level is in contrast to a study from the United States, which showed a particularly high risk of readmission for men of low socioeconomic status.18 The equal access, tax financed health care and narrow distribution of socioeconomic status in Sweden likely explain the difference. As previously reported,19e21 median perioperative stay was shorter after RALP than after RRP. Of men who underwent RALP 56% had a hospital stay of 48 hours or less compared with 7% treated with RRP. The risk of readmission was not independent of the duration of perioperative stay because shortterm postoperative complications led to longer perioperative stay as well as higher readmission frequency. Consequently, men with the shortest perioperative stay had the lowest readmission frequency, suggesting that protocols aiming to shorten the postoperative hospital stay are not likely to lead to an increased readmission rate. Mortality was low regardless of surgical method. We found that 90-day postoperative all-cause mortality was less than 0.20%, in line with previous NPCR reports.22 The strengths of our study include the population based, nationwide cohort including data on virtually all RPs performed in Sweden during the study
period. Our study also has some limitations. We did not have prospective per protocol data on readmissions but used administrative data. However, discharge diagnosis codes are 88% to 90% accurate23 and there is little reason to believe that these variations were related to differences in surgical method. Furthermore, the total number of readmissions is less likely to be misclassified than the discharge diagnosis regardless of what those diagnoses are. Approximately 1,040 of 9,843 RPs (11%) documented in NPCR between 2007 and 2011 were performed at private hospitals.11 Swedish private hospitals where RP is performed generally do not provide emergency readmissions. However, we cannot exclude the possibility of some patients being admitted without being documented in the national patient registry and, thus, not captured as a readmission in this study. In line with this we found little documentation in the national patient registry of RPs performed at private hospitals. Despite these limitations we interpret our results to mean that the proportion of readmissions after RP in the subgroup of men who underwent RP at private hospitals was substantially lower than the proportion treated with RP at public hospitals (4% vs 10%). Because the former subgroup was quite small, the frequency of readmission in our main analysis should be representative of the Swedish population. Our readmission rates are comparable to those in other countries. In a 2006 to 2009 study in Taiwan 9% of patients were readmitted within 90 days.10 Readmission rates were approximately 5% in 2 single center studies from the United States3,4
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
and approximately 20% within a year in a registry based study from England.8
CONCLUSIONS Readmission rates after different RP methods during 2000 to 2011 in Sweden were similar, ranging from 9% to 11%, with wide variation among hospitals. Overall administrative data on readmission rates after RP are readily available. Our data on readmissions after prostatectomy are in accordance with previous observations that such data can be used to assess the quality of perioperative care. However, to avoid bias adjustment for possible confounders must be performed.
117
ACKNOWLEDGMENTS NPCR Steering Group: P€ ar Stattin (Chair), Anders Widmark, Camilla Thellenberg, Ove Andren, Anna Bill-Axelson, Ann-Sofi Fransson, €lmMagnus To¨rnblom, Stefan Karlsson, Marie Hja Eriksson, Bodil Westman, Bill Pettersson, David Robinson, Mats Anden, Jan-Erik Damber, Jonas Hugosson, Ingela Frank-Lissbrant, Maria Nyberg, Go¨ran Ahlgren, Ola Bratt, Rene Blom, Rolf Lundgren, Lars Egevad, Calle Walller, Jan-Erik Johansson, Olof Akre, Per Fransson, Eva Johansson, Fredrik Sandin, Hans Garmo, Mats Lambe, Karin Hellstro¨m, Annette Wigertz and Erik Holmberg.
APPENDIX 1 Main indications for readmission by surgical complications and medical conditions Main Indications (ICD-9 or 10/No.) Surgical: Bleeding Hemorrhage þ hematoma complicating procedure (T81.0/189) Hematuria (R31.9/74) Anemia, unspecified (D64.9/5) Acute posthemorrhagic anaemia (D62.9/3) Blood transfusion (without reported diagnosis) (Z51.3/2) Infection: Infection following procedure (T81.4/300) Urinary tract infection (N39.0/101) Sepsis (A40 þ A41/77) Orchitis þ epididymitis (N45/38) Infection þ inflammatory reaction due to prosthetic device (T83.5/16) Pneumonia (J15 þ J18/12) Cystitis (N30/10) Other postop complications: Other complications of procedures (T81.8/186) Urinary retention (R33.9/80) Abdominal þ pelvic pain (R10/57) Bladder neck obstruction (N32.0/54) Urethral stricture (N35/41) Post-procedural urethral stricture (N99.1/38) Complication of surgical þ medical care (T88/38) Disruption of operation wound (T81.3/34) Mechanical complication of urinary (indwelling) catheter (T83.0/30) Noninfective disorders of lymphatic vessels and lymph nodes (I89/24) Post-procedural disorders of circulatory system (I97/20) Constipation (K59.0/20) Unspecified complication of procedure (T81.9/18) Obstructive þ reflux uropathy (N13/17) Fitting þ adjustment of urinary device (Z46.6/15) Fever (R50.9/13) Serious medical conditions:* Pulmonary embolism without mention of acute or pulmonary (I26.9/87) Atrial fibrillation þ flutter (I48.9/65) Acute myocardial infarction (I21/61) Cerebral infarction (I63/34) Angina pectoris (I20/32) Phlebitis þ thrombophlebitis (I80/22) Acute tubulo-interstitial nephritis (N10/18) Other medical diagnoses: Diseases of musculoskeletal system þ connective tissue (M00-M99/39) Mental þ behavioral disorders (F00-F99/35) Malignant neoplasms other than prostate Ca (C00-C97/34) Essential (primary) hypertension (I10/27) Calculus of kidney þ ureter (N20/16) Chest pain, unspecified (R07.4/13)
Diagnoses at Readmission (total No./% all RPs) All RPs (276/1.1) RRP (151/0.9) LRP (17/1.3) RALP (108/1.7)
All RPs (611/2.5) RRP (424/2.6) LRP (23/1.7) RALP (164/2.6)
All RPs (839/3.5) RRP (599/3.7) LRP (40/3.0) RALP (200/3.1)
All RPs (418/1.7) RRP (330/2.0) LRP (15/1.1) RALP (73/1.1)
All RPs (386/1.6) RRP (263/1.6) LRP (32/2.4) RALP (91/1.4)
* Ischemic heart disease, heart failure, heart arrhythmia, thromboembolic disease, pulmonary embolism, liver or kidney failure and cerebral vascular disease.
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
118
APPENDIX 2 Main interventions within 90 days after RP by postoperative complications and other interventions Main Interventions (NOMESCO/No.)
Interventions at Readmission (total No./% all RPs)
Related to postop complications
All RPs (815/3.4) RRP (592/3.6) LRP (39/2.9) RALP (184/2.9)
Transurethral: Cystoscopy (UKC02/142) Transurethral incision or resection of bladder neck (KCH42/71) Urethroscopic internal urethrotomy (KDV12/58) Transluminal endoscopic evacuation of blood clots from bladder (KCV22/21) Transurethral incision of prostate (KED32/15) Dilatation of urethra (TKD00/13) Urethroscopy (UKD02/11) With local anesthesia: Catheterization of bladder (TKC20/82) Percutaneous puncture of bladder (TKC10/39) Percutaneous local drainage of peritoneal cavity (TJA40/27) Percutaneous puncture of kidney or pelvis of kidney (TKA10/27) Puncture of retroperitoneal space (TKK10/24) Other minor surgical procedure in urology (TKW99/14) With general anesthesia: Laparotomy (JAH00/21) Vacuum treatment of wound (DQ023/21) Repair of wound dehiscence (KWA00/12) Reoperation for deep hemorrhage (KWE00/10) Laparotomy þ drainage of peritoneal cavity (JAK00/8) Reoperation for deep infection in urological surgery (KWC00/9) Repair of wound dehiscence in gastroenterological surgery (JWA00/8) Reoperation for insufficiency of anastomosis or suture (KWF00/8) Other: Expansion þ recanalization of coronary artery (FNG00, 02, 05/28) Gastroscopy (UJD02, 05/19) Total cardiopulmonary bypass in normothermia or moderate hypothermia at concurrent surgical procedure (FXA00/12) Aortocoronary venous bypass (FNC10, 20, 30/10) Anastomosis to coronary artery from internal mammary artery (FNA00/10) Cholecystectomy (JKA20, 21/9)
All RPs (375/1.6) RRP (324/2.0) LRP (17/1.3) RALP (34/0.5)
All RPs (289/1.2) RRP (184/1.1) LRP (18/1.3) RALP (87/1.4)
All RPs (151/0.6) RRP (84/0.5) LRP (4/0.3) RALP (63/1.0)
All RPs (377/1.6) RRP (255/1.6) LRP (12/0.9) RALP (110/1.7)
REFERENCES 1. Joynt KE and Jha AK: A path forward on Medicare readmissions. N Engl J Med 2013; 368: 1175. 2. Brown RE, Qadan M, Martin RC 2nd et al: The evolving importance of readmission data to the practicing surgeon. J Am Coll Surg 2010; 211: 558. 3. Nelson B, Kaufman M, Broughton G et al: Comparison of length of hospital stay between radical retropubic prostatectomy and robotic assisted laparoscopic prostatectomy. J Urol 2007; 177: 929. 4. Rabbani F, Yunis LH, Pinochet R et al: Comprehensive standardized report of complications of retropubic and laparoscopic radical prostatectomy. Eur Urol 2010; 57: 371. 5. Begg CB, Riedel ER, Bach PB et al: Variations in morbidity after radical prostatectomy. N Engl J Med 2002; 346: 1138. 6. Barocas DA, Mitchell R, Chang SS et al: Impact of surgeon and hospital volume on outcomes of radical prostatectomy. Urol Oncol 2010; 28: 243.
7. Wilt TJ, Shamliyan TA, Taylor BC et al: Association between hospital and surgeon radical prostatectomy volume and patient outcomes: a systematic review. J Urol 2008; 180: 820. 8. Judge A, Evans S, Gunnell DJ et al: Patient outcomes and length of hospital stay after radical prostatectomy for prostate cancer: analysis of hospital episodes statistics for England. BJU Int 2007; 100: 1040. 9. Yao SL and Lu-Yao G: Population-based study of relationships between hospital volume of prostatectomies, patient outcomes, and length of hospital stay. J Natl Cancer Inst 1999; 91: 1950. 10. Chung SD, Kelle JJ, Huang CY et al: Comparison of 90-day re-admission rates between open retropubic radical prostatectomy (RRP), laparoscopic RP (LRP) and robot-assisted laparoscopic prostatectomy (RALP). BJU Int 2012; 110: 966. 11. National Prostate Cancer Register: National Quality Report for the Year of Diagnosis 2012. Available at http://www.npcr.se. Accessed January 17, 2014.
12. Van Hemelrijck M, Wigertz A, Sandin F et al: Cohort profile: the National Prostate Cancer Register of Sweden and Prostate Cancer data Base Sweden 2.0. Int J Epidemiol 2013; 42: 956. 13. Charlson ME, Pompei P, Ales KL et al: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40: 373. 14. Berglund A, Garmo H, Tishelman C et al: Comorbidity, treatment and mortality: a population based cohort study of prostate cancer in PCBaSe Sweden. J Urol 2011; 185: 833. 15. Statistics Sweden: Longitudinal Integration Database for Health Insurance and Labour Market Studies. Available at http://www.scb.se. Accessed January 17, 2014. 16. National Comprehensive Cancer Network: Prostate Cancer. NCCN Clinical Practice Guidelines in Oncology (NCCN Guideline), Version 1.2014. Available at http://www.nccn.org. Accessed January 17, 2014.
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
17. Hosmer DWJ and Lemeshow S: Applied Logistic Regression, 2nd ed. New York: Wiley 2000. 18. Arbaje AI, Wolff JL, Yu Q et al: Postdischarge environmental and socioeconomic factors and the likelihood of early hospital readmission among community-dwelling Medicare beneficiaries. Gerontologist 2008; 48: 495. 19. Bolenz C, Gupta A, Hotze T et al: Cost comparison of robotic, laparoscopic, and open radical
119
prostatectomy for prostate cancer. Eur Urol 2010; 57: 453.
radical prostatectomy conducted in 2003 to 2005. J Urol 2008; 179: 1811.
20. Ficarra V, Novara G, Artibani W et al: Retropubic, laparoscopic, and robot-assisted radical prostatectomy: a systematic review and cumulative analysis of comparative studies. Eur Urol 2009; 55: 1037.
22. Carlsson S, Adolfsson J, Bratt O et al: Nationwide population-based study on 30-day mortality after radical prostatectomy in Sweden. Scand J Urol Nephrol 2009; 43: 350.
21. Touijer K, Eastham JA, Secin FP et al: Comprehensive prospective comparative analysis of outcomes between open and laparoscopic
23. Inpatient Diseases in Sweden 1987-2011. Available at http://www.socialstyrelsen.se. Accessed January 17, 2014.
Abbreviations and Acronyms CCI ¼ Charlson comorbidity index LRP ¼ laparoscopic RP NOMESCO ¼ Nordic Medico-Statistical Committee NPCR ¼ National Prostate Cancer Register PCBaSe ¼ Prostate Cancer data Base Sweden PSA ¼ prostate specific antigen RALP ¼ robot-assisted RP RP ¼ radical prostatectomy RRP ¼ retropubic RP Accepted for publication January 30, 2014. Study received Ume a University Hospital research ethical review board approval. Supported by the Swedish Research Council 825-2012-5047, Swedish Cancer Foundation 11 0471, V€asterbotten County Council and Lion’s Cancer Research Foundation at Ume a University. * Correspondence: Department of Surgery and perioperative Sciences, Urology and Andrology, Ume a University, Ume a University Hospital, 90187 Ume a, Sweden (telephone: þ46(0) 90 785 0000; FAX: þ46(0)90 143024; e-mail: jon. [email protected]). † Financial interest and/or other relationship with Pfizer.
Purpose: We investigated hospital readmission frequency during the 90 days after radical prostatectomy and assessed the readmission risk associated with potentially related variables. Materials and Methods: Using the population based, nationwide PCBaSe (Prostate Cancer data Base Sweden) we identified men diagnosed with incident prostate cancer between 2000 and 2011 who underwent radical prostatectomy as primary treatment. We used logistic regression analysis to examine the association of the risk of 90-day postoperative readmission with surgical method, calendar period, tumor risk category, hospital case load and patient characteristics. Results: During 90 postoperative days 2,317 of the 24,122 men (10%) identified were nonelectively readmitted, specifically 10% after retropubic, 9% after robotassisted and 11% after laparoscopic radical prostatectomy. The range in readmission frequency among hospitals was 0% to 35%. Higher readmission risk was associated with the early calendar period (2009 to 2011 vs 2000 to 2002 OR 0.71, 95% CI 0.61e0.83), greater age (70 or greater vs less than 60 years OR 1.17, 95% CI 1.00e1.36), higher risk category (high vs low OR 1.78, 95% CI 1.57e2.03), high comorbidity (Charlson comorbidity index 3 or greater vs 0 OR 1.77, 95% CI 1.29e2.44) and low hospital surgical volume (150 or greater vs fewer than 30 radical prostatectomies per year OR 0.70, 95% CI 0.60e0.81). Conclusions: Readmission rates after different radical prostatectomy methods were similar, ranging from 9% to 11%, with wide variation among hospitals. Readmission rates can be used as an indicator of perioperative care quality but potential confounders must be adjusted to avoid bias. Key Words: prostate, prostatic neoplasms, prostatectomy, patient readmission, perioperative care
See Editorial on page 10.
HOSPITAL readmission rates are considered a broad indicator of quality of care in medicine.1 In several fields of surgery they are used as a quality indicator of perioperative and postoperative clinical care.2 The
112
j
www.jurology.com
readmission rate after RP is approximately 5% at some centers with a high case load, ie high volume hospitals.3,4 However, this is likely not representative of all health care providers since lower perioperative
0022-5347/14/1921-0112/0 THE JOURNAL OF UROLOGY® © 2014 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
http://dx.doi.org/10.1016/j.juro.2014.01.109 Vol. 192, 112-119, July 2014 Printed in U.S.A.
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
mortality, fewer perioperative complications and shorter hospitalization have been noted for procedures performed by high volume surgeons at high volume hospitals.5e7 In the United Kingdom a registry based study showed a fourfold higher readmission rate of approximately 20% within a year of RP.8 In the United States 30-day postoperative readmission rates after 101,604 RPs covered by Medicare were 30% higher at low than at high volume hospitals.9 In a study from Taiwan 90-day readmissions were less frequent after RALP than after RRP.10 Because there are few studies with data on factors other than hospital volume and surgical method that can also influence the readmission risk after RP, we analyzed readmission risk in relation to surgical method, calendar period, tumor risk category, patient age, comorbidity and hospital surgical volume.
113
postoperative complications, group 4dserious medical conditions (ischemic heart disease, heart failure, heart arrhythmias, thromboembolic disease, pulmonary embolism, liver or kidney failure and cerebral vascular disease) and group 5dother diagnoses. We regarded groups 1 to 3 as surgical complications, and groups 4 and 5 as medical complications. We classified interventions within 90 days after RP as related or not related to postoperative complications. We further classified interventions related to postoperative complications as performed with local anesthesia, transurethral intervention or general anesthesia.
Hospital Surgical Volume During the study period RP was done at 52 hospitals with wide variation in the number at each hospital (see figure). Based on this distribution we categorized hospitals into 3 groups by the mean number of RPs performed per year, including lowdfewer than 30, intermediated30 to 149 and highd150 or greater.
MATERIALS AND METHODS
Risk Category
Study Population and Data Collection
Based on NPCR data we defined 3 risk categories, including lowdlocal clinical stage T1-2, Gleason score 6 or less and serum PSA less than 10 ng/ml, intermediated clinical stage T1-2, Gleason score 7 and/or PSA 10 to less than 20 ng/ml and highdclinical stage T3, Gleason score 8-10 and/or PSA 20 to less than 50 ng/ml according to a modification to NCCNÒ risk categorization.16
The NPCR of Sweden has captured approximately 98% of all newly diagnosed prostate cancer cases since 1998, in contrast to the Swedish Cancer Registry, for which registration is compulsory and mandated by law.11 NPCR data include time of diagnosis, patient age, tumor stage and differentiation, and serum PSA as well as primary treatment given or planned up to 6 months after diagnosis. The unique Swedish personal identity number in PCBaSe links NPCR to a number of other national health care registries and demographic databases.12 We analyzed data on men who underwent RP as primary treatment for prostate cancer between 2000 and 2011. We used PCBaSe to determine date of surgery, surgical method according to the NOMESCO Classification of Surgical Procedures (endoscopic KEC01 or retropubic RP KEC00), duration of perioperative hospitalization, comorbidity, readmission within 90 days of surgery, diagnoses at discharge after readmission according to ICD-9 or 10 and interventions at readmission according to the NOMESCO Classification of Surgical Procedures. We calculated CCI by adding the scores of each comorbid condition based on discharge diagnoses up to 10 years before the date of prostate cancer diagnosis and cancer diagnoses at other sites in the cancer registry.13,14 As a measure of socioeconomic status, we used education level based on information in LISA (Longitudinal Integration Database for Health Insurance and Labour Market Studies), which also provided information on marital status.15 Education level categories were lowdgrade school (9 years), mediumdhigh school (10 to 12 years) and highdcollege (greater than 12 years). Data on education level were available until 2009, covering 19,426 cases. We gathered information from each hospital in Sweden that had documented endoscopic RP to learn whether and when RALP was introduced and we classified the procedure accordingly as LRP or RALP. We classified readmissions by discharge diagnoses into 5 groups, including group 1dpostoperative infection, group 2dpostoperative bleeding, group 3dother
Statistical Methods We used multivariable logistic regression to calculate the OR as a measure of the risk of 90-day readmission, considering RP type, calendar period, patient age, tumor risk category, perioperative hospital stay, CCI, education level and hospital surgical volume. To analyze the risk of surgical methods and calendar periods we performed multivariable logistic regression analysis for each surgical method and for the 3 latest calendar periods17 using STATAÒ, release 12.1. The Umea˚ University Hospital research ethical review board approved the study.
Readmission rate vs surgical volume at individual hospitals for surgeries performed in 2000 to 2011.
114
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
2011 (7%) while the frequency after RRP and RALP changed only marginally. The range of readmission frequency varied particularly widely among hospitals with a small surgical volume (see figure). Of the 2,895 readmissions 2,530 showed a discharge diagnosis or a first contributing diagnosis other than prostate cancer and could be further analyzed. Of those readmissions 1,726 (68%) were related to surgical complications and 804 (32%) were related to medical conditions (Appendix 1). The most common causes of readmission related to surgery were postoperative infection (3%) and bleeding (1%). Of the 2,317 men who were readmitted a total of 1,192 interventions were performed, of which 815 (68%) were related to postoperative complications. The most common postoperative complications were cystoscopy, bladder catheterization and transurethral bladder neck incision (Appendix 2). Table 3 shows the results of multivariable logistic regression analysis of readmission after RP. The risk was lower in later calendar periods. It increased with patient age, tumor risk category and CCI, and decreased as hospital surgical volume
RESULTS Table 1 shows the number of men who underwent RP between 2000 and 2011, and the type of surgery. The number of RPs increased substantially during the study period with the proportion of RALPs increasing dramatically. Compared with men in the RRP group, those treated with RALP and LRP were more often younger than 60 years and in a lower tumor risk category, and had a higher education level and a shorter perioperative hospital stay. The proportion of married men and the comorbidity index were comparable regardless of surgical method (table 1). Table 2 shows the number of RPs and readmissions within 90 days by calendar period, excluding 298 readmissions that were planned, usually for catheter removal. Since 445 men were readmitted more than once, the total number of readmissions was 2,895. Readmission frequency for all RPs was slightly higher between 2000 and 2002 (12%) than between 2003 and 2011 (9% to 10%) but the frequency after LRP decreased dramatically from 2000 to 2002 (22%) to 2009 to
Table 1. Baseline characteristics of men who underwent prostatectomy from 2000 to 2011 in PCBaSe RRP No. pts Age: Median (IQR) No. less than 60 No. 60e64 No. 65e69 No. 70 or greater No. calendar period (%): 2000e2002 2003e2005 2006e2008 2009e2011 No. marital status (%): Married Divorced Widower Never married Missing data No. risk category (%):* Low Intermediate High Periop stay (days): Median (IQR) No. 1e2 (%) No. 3e4 (%) No. 5e6 (%) No. 7 or greater (%) No. CCI (%): 0 1 2 3 or Greater No. education level (%):† Low Medium High
LRP
16,375
RALP
1,354
6,393
Totals 24,122
63 (59e67) 4,128 (25.2) 5,290 (32.3) 5,073 (31.0) 1,884 (11.5)
63 (59e67) 404 (29.8) 424 (31.3) 407 (30.1) 119 (8.8)
63 (59e67) 1,870 (29.3) 1,945 (30.4) 1,865 (29.2) 713 (11.2)
63 (59e67) 6,402 (26.5) 7,659 (31.8) 7,345 (30.4) 2,716 (11.3)
2,780 5,396 4,511 3,688
(17.0) (33.0) (27.5) (22.5)
118 390 505 341
(8.7) (28.8) (37.3) (25.2)
0 580 1,813 4,000
(9.1) (28.4) (62.6)
2,898 6,366 6,829 8,029
(12.0) (26.4) (28.3) (33.3)
10,650 1,919 2,175 1,421 210
(65.0) (11.7) (13.3) (8.7) (1.3)
869 153 184 126 22
(64.2) (11.3) (13.6) (9.3) (1.6)
4,066 714 965 551 97
(63.6) (11.2) (15.1) (8.6) (1.5)
15,585 2,786 3,324 2,098 329
(64.6) (11.6) (13.8) (8.7) (1.4)
7,274 6,346 2,212
(45.9) (40.1) (14.0)
736 463 128
(55.5) (34.9) (9.6)
3,073 2,625 519
(49.4) (42.2) (8.3)
11,083 9,434 2,859
(47.4) (40.4) (12.2)
5 1,080 6,104 5,530 3,659
(4e6) (6.6) (37.3) (33.8) (22.3)
4 291 684 250 129
(3e5) (21.5) (50.5) (18.5) (9.5)
2 3,576 2,393 228 196
(2e3) (55.9) (37.4) (3.6) (3.1)
4 4,947 9,181 6,008 3,984
(3e6) (20.5) (38.1) (24.9) (16.5)
13,423 1,969 748 235
(82.0) (12.0) (4.6) (1.4)
1,138 163 45 8
(84.0) (12.0) (3.3) (0.6)
5,373 672 272 76
(84.0) (10.5) (4.3) (1.2)
19,934 2,804 1,065 319
(82.6) (11.6) (4.4) (1.3)
4,476 5,833 3,731
(31.9) (41.6) (26.6)
309 509 340
(26.7) (44.0) (29.4)
820 1,688 1,720
(19.4) (39.9) (40.7)
5,605 8,030 5,791
(28.9) (41.3) (29.8)
* Excluding 746 men with distant or regional metastasis, or missing data. † Information only available until 2009.
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
Table 2. RPs and readmissions within 90 days by calendar period Period 2000e2002: All RPs RRP LRP RALP 2003e2005: All RPs RRP LRP RALP 2006e2008: All RPs RRP LRP RALP 2009e2011: All RPs RRP LRP RALP Full study period: All RPs RRP LRP RALP
Table 3. Multivariable regression analysis of readmissions after RP by preoperative factors
No. RPs/Readmissions (%) 2,898/336 (11.6) 2,780/310 (11.2) 118/26 (22.0) 0 6,366/573 (9.0) 5,396/467 (8.7) 390/56 (14.4) 580/50 (8.6) 6,829/660 (9.7) 4,511/452 (10.0) 505/40 (7.9) 1,813/168 (9.3) 8,029/748 (9.3) 3,688/380 (10.3) 341/23 (6.7) 4,000/345 (8.6) 24,122/2,317 (9.6) 16,375/1,609 (9.8) 1,354/145 (10.7) 6,393/563 (8.8)
increased. A higher readmission risk was associated with longer perioperative hospital stay (7 days or greater vs 1 to 2 OR 1.90, 95% CI 1.60e2.27). Readmission risk was not associated with education level (table 3). Multivariable regression analysis of operative methods revealed that differences associated with calendar period, patient age, tumor risk category, comorbidity and surgical volume were stronger for RRP than for RALP (table 4). Table 5 shows the results of multiple regression analysis for the 3 latest calendar periods. In 2003 to 2005 the risk was higher after LRP than after RRP but in 2009 to 2011 the risk was quite similar for the surgical methods. In 2000 to 2011 the incidence of 90-day postoperative mortality was 33 of 16,375 men (0.20%) for RRP, 10 of 6,393 (0.16%) for RALP and 0 of 1,354 for LRP. To assess readmissions after RP performed by private health care providers we used NPCR data on 2007 to 2011. Of 1,040 men treated with RP (RRP in 601 and RALP in 439) at a private hospital 296 (28%) were also documented in the national patient registry and 39 (4%) were readmitted to a public hospital within 90 days of RP.
DISCUSSION In this nationwide, population based study of more than 24,000 RPs performed in Sweden during 2000 to 2011 the overall frequency of readmissions within 90 days postoperatively was approximately 10%. Except for LRP performed during 2000 to 2002,
115
Surgical method: RRP LRP RALP Calendar period: 2000e2002 2003e2005 2006e2008 2009e2011 Age: Less than 60 60e64 65e69 70 or Greater Risk category:* Low Intermediate High CCI: 0 1 2 3 or Greater No RPs/unit/yr: Less than 30 30e149 150 or Greater Education level:† Low Medium High
No. Readmissions (%)
OR (95% CI)
1,609 (9.8) 145 (10.7) 563 (8.8)
1.00 (referent) 1.27 (1.06e1.53) 1.13 (0.99e1.29)
336 573 660 748
(11.6) (9.0) (9.7) (9.3)
1.00 (referent) 0.74 (0.64e0.86) 0.76 (0.66e0.88) 0.71 (0.61e0.83)
563 718 734 302
(8.8) (9.4) (10.0) (11.1)
1.00 (referent) 1.04 (0.92e1.17) 1.10 (0.98e1.24) 1.17 (1.00e1.36)
881 (8.0) 955 (10.1) 391 (13.7)
1.00 (referent) 1.29 (1.17e1.42) 1.78 (1.57e2.03)
1,844 291 132 50
(9.3) (10.4) (12.1) (15.7)
1.00 (referent) 1.12 (0.98e1.28) 1.31 (1.08e1.60) 1.77 (1.29e2.44)
404 (11.3) 1,250 (9.8) 663 (8.6)
1.00 (referent) 0.80 (0.71e0.91) 0.70 (0.60e0.81)
581 (10.7) 810 (10.4) 585 (10.5)
1.00 (referent) 1.01 (0.90e1.14) 1.03 (0.91e1.17)
* Excluding 746 men with distant or regional metastasis, or missing data. † Information only available until 2009.
when it was first introduced, we found relatively small differences in readmission frequency among the 3 surgical methods. The high initial readmission rates for LRP were likely due to the learning curve. In contrast, readmission rates 2 years after RALP was introduced were similar to rates after RRP. The number of RALPs increased rapidly during the study period from no procedures in 2000 to 2002 to 50% of all RPs in 2009 to 2011. Clinical characteristics related to an increased risk of readmission were age, tumor risk category and CCI, particularly during 2000 to 2002. Of men who underwent RP 89% were younger than 70 years. Age distribution differed slightly among the surgical groups. The association of increased age with a somewhat increased risk of readmission was no longer evident during 2009 to 2011. The association between high risk category and higher readmission frequency weakened somewhat during the later calendar periods and 88% of men had low or intermediate risk cancer. The association between high CCI and increased readmission risk was particularly evident for CCI 3 or greater. However, men with CCI 3 or greater represented only 1% of the study group and in 94% CCI was 1 or less. Readmission frequency differences among hospitals were large, especially at those with low surgical
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
116
Table 4. ORs of readmission after RP by preoperative factors and surgical method
Table 5. ORs of readmission after RP by preoperative factors and period
OR (95% CI)
Calendar period: 2000e2002 2003e2005 2006e2008 2009e2011 Age: Less than 60 60e64 65e69 70 or Greater Risk category:* Low Intermediate High CCI: 0 1 2 3 or Greater No RPs/unit/yr: Less than 30 30e149 150 or greater
OR (95% CI)
RRP
LRP
RALP
1.00 (referent) 0.73 (0.63e0.86) 0.81 (0.69e0.94) 0.79 (0.66e0.93)
1.00 (referent) 0.79 (0.45e1.41) 0.49 (0.27e0.90) 0.21 (0.11e0.40)
Not available 1.00 (referent) 1.08 (0.77e1.51) 0.96 (0.70e1.32)
1.00 (referent) 1.11 (0.96e1.29) 1.22 (1.06e1.42) 1.34 (1.11e1.61)
1.00 (referent) 0.76 (0.48e1.21) 0.75 (0.46e1.21) 0.76 (0.36e1.60)
1.00 (referent) 0.99 (0.79e1.25) 0.95 (0.76e1.20) 0.90 (0.66e1.25)
1.00 (referent) 1.38 (1.22e1.56) 1.94 (1.67e2.25)
1.00 (referent) 1.48 (0.99e2.21) 1.02 (0.51e2.03)
1.00 (referent) 1.09 (0.91e1.32) 1.52 (1.13e2.05)
1.00 (referent) 1.10 (0.93e1.29) 1.33 (1.05e1.67) 2.04 (1.43e2.90)
1.00 (referent) 1.26 (0.73e2.17) 1.92 (0.80e4.54) Not available
1.00 (referent) 1.13 (0.85e1.48) 1.10 (0.72e1.68) 1.14 (0.52e2.50)
1.00 (referent) 0.91 (0.80e1.04) 0.75 (0.62e0.90)
1.00 (referent) 0.16 (0.10e0.25) 0.11 (0.06e0.20)
1.00 (referent) 1.31 (0.78e2.18) 1.27 (0.77e2.08)
Surgical method: RRP LRP RALP Age: Less than 60 60e64 65e69 70 or Greater Risk category:* Low Intermediate High CCI: 0 1 2 3 or Greater No RPs/unit/yr: Less than 30 30e149 150 or Greater
2003e2005
2006e2008
2009e2011
1.00 (referent) 2.24 (1.64e3.05) 1.37 (0.96e1.95)
1.00 (referent) 0.86 (0.61e1.22) 1.08 (0.87e1.34)
1.00 (referent) 0.66 (0.43e1.02) 0.82 (0.66e1.03)
1.00 (referent) 1.07 (0.84e1.35) 1.24 (0.98e1.58) 1.56 (1.15e2.11)
1.00 (referent) 0.99 (0.79e1.22) 0.84 (0.67e1.06) 1.24 (0.94e1.64)
1.00 (referent) 0.95 (0.76e1.19) 1.17 (0.95e1.44) 1.03 (0.79e1.35)
1.00 (referent) 1.46 (1.20e1.78) 2.13 (1.65e2.75)
1.00 (referent) 1.36 (1.13e1.63) 1.76 (1.37e2.25)
1.00 (referent) 1.10 (0.93e1.30) 1.76 (1.41e2.20)
1.00 (referent) 1.10 (0.84e1.44) 1.46 (0.99e2.15) 2.87 (1.66e4.99)
1.00 (referent) 1.30 (1.01e1.66) 1.52 (1.06e2.16) 0.96 (0.41e2.23)
1.00 (referent) 1.03 (0.82e1.30) 1.20 (0.86e1.68) 1.43 (0.85e2.40)
1.00 (referent) 0.70 (0.55e0.89) 0.61 (0.46e0.82)
1.00 (referent) 1.05 (0.83e1.34) 0.92 (0.70e1.22)
1.00 (referent) 1.05 (0.84e1.31) 1.11 (0.82e1.50)
* Excluding 746 men with distant or regional metastasis, or missing data.
* Excluding 746 men with distant or regional metastasis, or missing data.
volume. The latter hospitals showed the highest and lowest readmission rates, likely a result of the larger effect of random variation in hospitals with smaller numbers of events. Nevertheless, it might be prudent to assess perioperative care at hospitals with high readmission rates. Our finding of no association between readmission risk and education level is in contrast to a study from the United States, which showed a particularly high risk of readmission for men of low socioeconomic status.18 The equal access, tax financed health care and narrow distribution of socioeconomic status in Sweden likely explain the difference. As previously reported,19e21 median perioperative stay was shorter after RALP than after RRP. Of men who underwent RALP 56% had a hospital stay of 48 hours or less compared with 7% treated with RRP. The risk of readmission was not independent of the duration of perioperative stay because shortterm postoperative complications led to longer perioperative stay as well as higher readmission frequency. Consequently, men with the shortest perioperative stay had the lowest readmission frequency, suggesting that protocols aiming to shorten the postoperative hospital stay are not likely to lead to an increased readmission rate. Mortality was low regardless of surgical method. We found that 90-day postoperative all-cause mortality was less than 0.20%, in line with previous NPCR reports.22 The strengths of our study include the population based, nationwide cohort including data on virtually all RPs performed in Sweden during the study
period. Our study also has some limitations. We did not have prospective per protocol data on readmissions but used administrative data. However, discharge diagnosis codes are 88% to 90% accurate23 and there is little reason to believe that these variations were related to differences in surgical method. Furthermore, the total number of readmissions is less likely to be misclassified than the discharge diagnosis regardless of what those diagnoses are. Approximately 1,040 of 9,843 RPs (11%) documented in NPCR between 2007 and 2011 were performed at private hospitals.11 Swedish private hospitals where RP is performed generally do not provide emergency readmissions. However, we cannot exclude the possibility of some patients being admitted without being documented in the national patient registry and, thus, not captured as a readmission in this study. In line with this we found little documentation in the national patient registry of RPs performed at private hospitals. Despite these limitations we interpret our results to mean that the proportion of readmissions after RP in the subgroup of men who underwent RP at private hospitals was substantially lower than the proportion treated with RP at public hospitals (4% vs 10%). Because the former subgroup was quite small, the frequency of readmission in our main analysis should be representative of the Swedish population. Our readmission rates are comparable to those in other countries. In a 2006 to 2009 study in Taiwan 9% of patients were readmitted within 90 days.10 Readmission rates were approximately 5% in 2 single center studies from the United States3,4
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
and approximately 20% within a year in a registry based study from England.8
CONCLUSIONS Readmission rates after different RP methods during 2000 to 2011 in Sweden were similar, ranging from 9% to 11%, with wide variation among hospitals. Overall administrative data on readmission rates after RP are readily available. Our data on readmissions after prostatectomy are in accordance with previous observations that such data can be used to assess the quality of perioperative care. However, to avoid bias adjustment for possible confounders must be performed.
117
ACKNOWLEDGMENTS NPCR Steering Group: P€ ar Stattin (Chair), Anders Widmark, Camilla Thellenberg, Ove Andren, Anna Bill-Axelson, Ann-Sofi Fransson, €lmMagnus To¨rnblom, Stefan Karlsson, Marie Hja Eriksson, Bodil Westman, Bill Pettersson, David Robinson, Mats Anden, Jan-Erik Damber, Jonas Hugosson, Ingela Frank-Lissbrant, Maria Nyberg, Go¨ran Ahlgren, Ola Bratt, Rene Blom, Rolf Lundgren, Lars Egevad, Calle Walller, Jan-Erik Johansson, Olof Akre, Per Fransson, Eva Johansson, Fredrik Sandin, Hans Garmo, Mats Lambe, Karin Hellstro¨m, Annette Wigertz and Erik Holmberg.
APPENDIX 1 Main indications for readmission by surgical complications and medical conditions Main Indications (ICD-9 or 10/No.) Surgical: Bleeding Hemorrhage þ hematoma complicating procedure (T81.0/189) Hematuria (R31.9/74) Anemia, unspecified (D64.9/5) Acute posthemorrhagic anaemia (D62.9/3) Blood transfusion (without reported diagnosis) (Z51.3/2) Infection: Infection following procedure (T81.4/300) Urinary tract infection (N39.0/101) Sepsis (A40 þ A41/77) Orchitis þ epididymitis (N45/38) Infection þ inflammatory reaction due to prosthetic device (T83.5/16) Pneumonia (J15 þ J18/12) Cystitis (N30/10) Other postop complications: Other complications of procedures (T81.8/186) Urinary retention (R33.9/80) Abdominal þ pelvic pain (R10/57) Bladder neck obstruction (N32.0/54) Urethral stricture (N35/41) Post-procedural urethral stricture (N99.1/38) Complication of surgical þ medical care (T88/38) Disruption of operation wound (T81.3/34) Mechanical complication of urinary (indwelling) catheter (T83.0/30) Noninfective disorders of lymphatic vessels and lymph nodes (I89/24) Post-procedural disorders of circulatory system (I97/20) Constipation (K59.0/20) Unspecified complication of procedure (T81.9/18) Obstructive þ reflux uropathy (N13/17) Fitting þ adjustment of urinary device (Z46.6/15) Fever (R50.9/13) Serious medical conditions:* Pulmonary embolism without mention of acute or pulmonary (I26.9/87) Atrial fibrillation þ flutter (I48.9/65) Acute myocardial infarction (I21/61) Cerebral infarction (I63/34) Angina pectoris (I20/32) Phlebitis þ thrombophlebitis (I80/22) Acute tubulo-interstitial nephritis (N10/18) Other medical diagnoses: Diseases of musculoskeletal system þ connective tissue (M00-M99/39) Mental þ behavioral disorders (F00-F99/35) Malignant neoplasms other than prostate Ca (C00-C97/34) Essential (primary) hypertension (I10/27) Calculus of kidney þ ureter (N20/16) Chest pain, unspecified (R07.4/13)
Diagnoses at Readmission (total No./% all RPs) All RPs (276/1.1) RRP (151/0.9) LRP (17/1.3) RALP (108/1.7)
All RPs (611/2.5) RRP (424/2.6) LRP (23/1.7) RALP (164/2.6)
All RPs (839/3.5) RRP (599/3.7) LRP (40/3.0) RALP (200/3.1)
All RPs (418/1.7) RRP (330/2.0) LRP (15/1.1) RALP (73/1.1)
All RPs (386/1.6) RRP (263/1.6) LRP (32/2.4) RALP (91/1.4)
* Ischemic heart disease, heart failure, heart arrhythmia, thromboembolic disease, pulmonary embolism, liver or kidney failure and cerebral vascular disease.
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
118
APPENDIX 2 Main interventions within 90 days after RP by postoperative complications and other interventions Main Interventions (NOMESCO/No.)
Interventions at Readmission (total No./% all RPs)
Related to postop complications
All RPs (815/3.4) RRP (592/3.6) LRP (39/2.9) RALP (184/2.9)
Transurethral: Cystoscopy (UKC02/142) Transurethral incision or resection of bladder neck (KCH42/71) Urethroscopic internal urethrotomy (KDV12/58) Transluminal endoscopic evacuation of blood clots from bladder (KCV22/21) Transurethral incision of prostate (KED32/15) Dilatation of urethra (TKD00/13) Urethroscopy (UKD02/11) With local anesthesia: Catheterization of bladder (TKC20/82) Percutaneous puncture of bladder (TKC10/39) Percutaneous local drainage of peritoneal cavity (TJA40/27) Percutaneous puncture of kidney or pelvis of kidney (TKA10/27) Puncture of retroperitoneal space (TKK10/24) Other minor surgical procedure in urology (TKW99/14) With general anesthesia: Laparotomy (JAH00/21) Vacuum treatment of wound (DQ023/21) Repair of wound dehiscence (KWA00/12) Reoperation for deep hemorrhage (KWE00/10) Laparotomy þ drainage of peritoneal cavity (JAK00/8) Reoperation for deep infection in urological surgery (KWC00/9) Repair of wound dehiscence in gastroenterological surgery (JWA00/8) Reoperation for insufficiency of anastomosis or suture (KWF00/8) Other: Expansion þ recanalization of coronary artery (FNG00, 02, 05/28) Gastroscopy (UJD02, 05/19) Total cardiopulmonary bypass in normothermia or moderate hypothermia at concurrent surgical procedure (FXA00/12) Aortocoronary venous bypass (FNC10, 20, 30/10) Anastomosis to coronary artery from internal mammary artery (FNA00/10) Cholecystectomy (JKA20, 21/9)
All RPs (375/1.6) RRP (324/2.0) LRP (17/1.3) RALP (34/0.5)
All RPs (289/1.2) RRP (184/1.1) LRP (18/1.3) RALP (87/1.4)
All RPs (151/0.6) RRP (84/0.5) LRP (4/0.3) RALP (63/1.0)
All RPs (377/1.6) RRP (255/1.6) LRP (12/0.9) RALP (110/1.7)
REFERENCES 1. Joynt KE and Jha AK: A path forward on Medicare readmissions. N Engl J Med 2013; 368: 1175. 2. Brown RE, Qadan M, Martin RC 2nd et al: The evolving importance of readmission data to the practicing surgeon. J Am Coll Surg 2010; 211: 558. 3. Nelson B, Kaufman M, Broughton G et al: Comparison of length of hospital stay between radical retropubic prostatectomy and robotic assisted laparoscopic prostatectomy. J Urol 2007; 177: 929. 4. Rabbani F, Yunis LH, Pinochet R et al: Comprehensive standardized report of complications of retropubic and laparoscopic radical prostatectomy. Eur Urol 2010; 57: 371. 5. Begg CB, Riedel ER, Bach PB et al: Variations in morbidity after radical prostatectomy. N Engl J Med 2002; 346: 1138. 6. Barocas DA, Mitchell R, Chang SS et al: Impact of surgeon and hospital volume on outcomes of radical prostatectomy. Urol Oncol 2010; 28: 243.
7. Wilt TJ, Shamliyan TA, Taylor BC et al: Association between hospital and surgeon radical prostatectomy volume and patient outcomes: a systematic review. J Urol 2008; 180: 820. 8. Judge A, Evans S, Gunnell DJ et al: Patient outcomes and length of hospital stay after radical prostatectomy for prostate cancer: analysis of hospital episodes statistics for England. BJU Int 2007; 100: 1040. 9. Yao SL and Lu-Yao G: Population-based study of relationships between hospital volume of prostatectomies, patient outcomes, and length of hospital stay. J Natl Cancer Inst 1999; 91: 1950. 10. Chung SD, Kelle JJ, Huang CY et al: Comparison of 90-day re-admission rates between open retropubic radical prostatectomy (RRP), laparoscopic RP (LRP) and robot-assisted laparoscopic prostatectomy (RALP). BJU Int 2012; 110: 966. 11. National Prostate Cancer Register: National Quality Report for the Year of Diagnosis 2012. Available at http://www.npcr.se. Accessed January 17, 2014.
12. Van Hemelrijck M, Wigertz A, Sandin F et al: Cohort profile: the National Prostate Cancer Register of Sweden and Prostate Cancer data Base Sweden 2.0. Int J Epidemiol 2013; 42: 956. 13. Charlson ME, Pompei P, Ales KL et al: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40: 373. 14. Berglund A, Garmo H, Tishelman C et al: Comorbidity, treatment and mortality: a population based cohort study of prostate cancer in PCBaSe Sweden. J Urol 2011; 185: 833. 15. Statistics Sweden: Longitudinal Integration Database for Health Insurance and Labour Market Studies. Available at http://www.scb.se. Accessed January 17, 2014. 16. National Comprehensive Cancer Network: Prostate Cancer. NCCN Clinical Practice Guidelines in Oncology (NCCN Guideline), Version 1.2014. Available at http://www.nccn.org. Accessed January 17, 2014.
REHOSPITALIZATION AFTER RADICAL PROSTATECTOMY
17. Hosmer DWJ and Lemeshow S: Applied Logistic Regression, 2nd ed. New York: Wiley 2000. 18. Arbaje AI, Wolff JL, Yu Q et al: Postdischarge environmental and socioeconomic factors and the likelihood of early hospital readmission among community-dwelling Medicare beneficiaries. Gerontologist 2008; 48: 495. 19. Bolenz C, Gupta A, Hotze T et al: Cost comparison of robotic, laparoscopic, and open radical
119
prostatectomy for prostate cancer. Eur Urol 2010; 57: 453.
radical prostatectomy conducted in 2003 to 2005. J Urol 2008; 179: 1811.
20. Ficarra V, Novara G, Artibani W et al: Retropubic, laparoscopic, and robot-assisted radical prostatectomy: a systematic review and cumulative analysis of comparative studies. Eur Urol 2009; 55: 1037.
22. Carlsson S, Adolfsson J, Bratt O et al: Nationwide population-based study on 30-day mortality after radical prostatectomy in Sweden. Scand J Urol Nephrol 2009; 43: 350.
21. Touijer K, Eastham JA, Secin FP et al: Comprehensive prospective comparative analysis of outcomes between open and laparoscopic
23. Inpatient Diseases in Sweden 1987-2011. Available at http://www.socialstyrelsen.se. Accessed January 17, 2014.
