- Email: [email protected]
Prognostic Factors and Percutaneous Nephrolithotomy Morbidity: A Multivariate Analysis of a Contemporary Series Using the Clavien Classification
Urolithiasis/Endourology Prognostic Factors and Percutaneous Nephrolithotomy Morbidity: A Multivariate Analysis of a Contemporary Series Using the Clavien Classification J. J. M. C. H. de la Rosette,*,† J. Rioja Zuazu, P. Tsakiris, A. M. Elsakka, J. J. Zudaire, M. P. Laguna and Th. M. de Reijke From the Department of Urology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands, and Department of Urology, Clinica Universitaria, Universidad de Navarra, Navarra, Spain
Purpose: We stratified factors affecting treatment morbidity, compared the outcomes of percutaneous nephrolithotomy procedures from a single department and provided evidence of treatment benefits when percutaneous nephrolithotomy is performed in an expert setting. Materials and Methods: Since the department became a dedicated endourological center in 2002 we grouped all percutaneous nephrolithotomy procedures into those performed before 2002 (group 1) and after 2002 (group 2). The modified Clavien classification was used to score morbidity. Independent variables with an influence on complications were studied including stone size, operating time, operative complications, dilation device, urine culture, group allocation and lithotripsy device. Contingency and logistic regression were used for univariate and multivariate analysis. Results: Of the 244 percutaneous nephrolithotomy procedures 68 comprised group 1 and 176 formed group 2. Statistical preoperative differences were patient age, the use of anticoagulants and positive urinary cultures. Group 1 had a complication rate of 56.8% and group 2 had a complication rate of 37.2%. There were significant differences between the groups (p ⫽ 0.007). Almost all complications were grade 1 to 2. On univariate analysis the influence variables were urine culture (OR 1.69), group allocation (OR 2.20), stone size (OR 2.28), dilation device (OR 4.8), lithotripsy device (OR 1.22), perioperative complications (OR 2.83) and surgical time (OR 1.87). On multivariate analysis the independent factors in the complicated outcome were stone size (OR 1.25), type of lithotripsy device (OR 1.35) and incidence of perioperative complications (OR 3.71). Conclusions: The dedicated setting for percutaneous nephrolithotomy at our center resulted in decreased operative time, more uneventful procedures and decreased hospitalization time. The modified Clavien morbidity score is a reliable tool for more objective outcome comparisons after renal stone treatment. Key Words: nephrostomy, percutaneous; classification; morbidity
lthough percutaneous nephrolithotomy for the treatment of urolithiasis is considered more invasive than other currently available minimally invasive techniques, trends in renal stone surgery show that PNL has been increasingly used in recent years.1 Among the contributing factors for this trend are the increasing incidence of stone disease, and the proven safety and efficacy of PNL for renal stones.2 This applies especially to stones larger than 2 cm or staghorn calculi that are resistant to fragmentation, stones occurring in kidneys with an abnormal anatomy and stones in more complicated patient groups.3,4 In addition, the decrease in efficacy of the latest generation shock wave lithotripsy devices5,6 and the limitations in their use7 have further accentuated the advantages and benefits of PNL.
A
Submitted for publication March 14, 2008. * Correspondence: Department of Urology G4-105, Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands (telephone: ⫹31 20 5666377; FAX: ⫹31 20 6669585; email: [email protected]). † Financial interest and/or other relationship with Galil Medical, BSC and AMS.
0022-5347/08/1806-2489/0 THE JOURNAL OF UROLOGY® Copyright © 2008 by AMERICAN UROLOGICAL ASSOCIATION
Another trend that seems to accompany the increasing use of PNL for renal stones is the greater percentage of PNL procedures shifting to specialized tertiary centers.8 At these centers a dedicated endourology team (including an endourologist, specialized nurses, an anesthesiologist and radiological support) is in place, and has access to a full range of state-of-the-art endourology equipment and instruments (including laser, mechanical lithotripters and SWL). This shift may be explained by the significantly improved outcomes and decreased expenditures resulting from the standardization of practice patterns. Supportive of this is the suggestion that regionalization of certain procedures that are technically more complex and demanding specialized medical centers may improve the overall quality of health care.9 Moreover, this may also result favorably in a reduction in medical related morbidity and complications.10 To define the treatment related benefits obtained when PNL is performed at a specialized center it is necessary to perform an outcome comparison with those procedures performed at a nonspecialized setting which is optimally done when there are minimum regional and institutional differences. In this retrospective study we present the results of PNL for renal stones performed at a single
2489
Vol. 180, 2489-2493, December 2008 Printed in U.S.A. DOI:10.1016/j.juro.2008.08.025
2490
PROGNOSTIC FACTORS AND PERCUTANEOUS NEPHROLITHOTOMY
TABLE 1. Preoperative group characteristics and clinical data Group 1 No. pts No. gender: Male Female Mean pt age (⫾SD) No. Pretreatment (%): SWL PNL No. diabetes mellitus (%) No. anticoagulants (%) No. pos urine culture (%) Mean cm stone burden (⫾SD) No. renal anomalies (%)
68
Group 2
p Value
176
38 30 46.2 (⫾16.3)
97 79 51.4 (⫾15.2)
0.914 0.020*
34 (50) 9 (13.2) 5 (7.4) 2 (2.9) 24 (35.3) 3.1 (⫾2.4) 14 (20.6)
48 (27) 19 (10.8) 13 (7.3) 27 (15.3) 77 (43.8) 2.4 (⫾1.9) 16 (10.8)
0.001* 0.609 0.402 0.007* 0.007* 0.083 0.209
* Statistically significant difference (p ⬍0.05).
department, and analyze and compare the outcomes of the PNL procedures, as well as provide evidence of treatment benefits when PNL is performed in an expert setting. MATERIALS AND METHODS The medical files of all patients who underwent PNL for renal stones at our department from January 1994 until June 2007 were collected and evaluated. PNL procedures were divided into 2 groups according to the date of operation. Group 1 included procedures performed before the year 2002 (alternatingly done by 5 urologists) while those procedures performed after 2002 formed group 2 (done by 1 urologist in the context of an endourology team). The year 2002 was used as a grouping parameter because since then the department has become an endourological center with expertise and dedication to PNL in terms of personnel and equipment. The modified Clavien grading system was used to evaluate the perioperative morbidity of PNL treatment (see Appendix).11 Stone burden was defined as the largest diameter of the stone in cm or the sum of the diameters of multiple stones, while anomalous kidneys are those with ureteropelvic junction stenosis, horseshoe kidneys or kidneys with previous partial nephrectomy. PNL was defined as uneventful when there was no reported problem in gaining access, no bleeding or perforation, and with uncomplicated stone fragmentation and extraction. The procedure was considered to have failed when it was not possible to gain access or when it had to be discontinued due to instrumentation problems. Preoperative clinical data and treatment outcomes were compared using Student’s t test. Odds ratios were calculated and statistical determinations were within the 95% CI. All p values were 2-tailed and p ⬍0.05 was considered statistically significant. For correlation 2 ⫻ 2 crosstabs (2) were used while logistic regression was used for the univariate and multivariate analysis. Analysis was performed with the SPSS® (version 12.0) statistical software package. RESULTS A total of 244 PNL procedures were performed. Of these procedures 68 were performed before 2002 (group 1) and 176 were performed after 2002 (group 2). The demographic characteristics of the patients and preoperative
clinical data are presented in table 1. There were no statistical differences between the groups with regard to gender, incidence of diabetes mellitus, mean stone burden and American Society of Anesthesiologists score, while the percentage of anomalous kidneys was equally distributed between the groups. However, there was a significant difference in mean patient age, type of preoperative treatment when PNL was performed as a second line therapy, use of anticoagulants and incidence of positive urinary culture. Table 2 shows operative and postoperative data. Balloon dilation was used more frequently to obtain the access tract in group 2 and more uneventful procedures were seen in this group. Mean operative time was much lower in group 2 (97.2 vs 121.2 minutes in group 1, p ⫽ 0.003) as was mean hospitalization time. However, group 2 had a significantly higher incidence of postoperative fever (18.2% vs 5.9% in group 1, p ⫽ 0.041). In terms of the procedural events in both groups there were significantly more uneventful cases in group 2 with less bleeding and fewer failed attempts observed. However, this group had a higher incidence of cases with perforations of the renal pelvis (7.9% vs 2.9%, respectively). Stone-free rates were similar for both groups. Sufficient data to grade complications with the modified Clavien scoring system were possible in 238 PNL procedures (66 group 1, 172 group 2), identifying complications in a total of 107 PNL cases (44.9%). In group 1 of 66 procedures 39 (59.1%) had complications of Clavien grade 1 or greater. On the other hand, only 68 of 172 (39.5%) patients had some complication in group 2, resulting in a statistically significant difference between the 2 groups (p ⫽ 0.007). Almost all complications were grade 1 to 2, while group 1 had a Clavien grade 3 (subtype a) case in which surgical intervention without general anesthesia was required due to pneumothorax. Group 2 had 1 grade 3 (subtype b) and 1 grade 4 case (subtype a) (table 3). Looking at the mean Clavien score over time we found that there is a negative correlation of these variables, indicating smaller scores in group 2 (see figure). A further analysis of the relationships of mean Clavien score with different parameters is presented in table 4. A univariate analysis was performed using as influence
TABLE 2. Operative and postoperative data Group 1 Mean operative mins (⫾SD) No. dilatations (%): Balloon Other No. procedures (%): Uneventful Failed Days hospitalization: Mean ⫾ SD (range) % Less than 5 % Greater than 10 No. postop fever (%) No. perforation (%) No. stone-free postop (%)†
121.1
(⫾58.2)
Group 2 97.2
p Value*
(⫾37.2)
0.003
24 35
(39.3) 145 (60.7) 21
(82.3) (17.7)
0.001
50 8
(73.5) (11.8)
(84.7) (2.8)
0.006
8.2 ⫾ 3.7 (4–16) 6.8 ⫾ 3.4 (3–14) 25 38 40 12 4 (5.9) 32 (18.2) 3 (4.4) 14 (7.9) 48 (71) 146 (83)
0.002
49 5
0.041 0.061 0.24
* Statistically significant difference (p ⬍0.05). † Nonclinically significant residual fragments (smaller than 2 mm).
PROGNOSTIC FACTORS AND PERCUTANEOUS NEPHROLITHOTOMY TABLE 3. Distribution of complications with the Clavien grading system No. (%)
TABLE 4. Correlation between Clavien grading system and other factors Variable
Clavien
Global
Group 1
Group 2
0 1 2 3a 3b 4a Missing
131 (53.7) 63 (25.8) 41 (16.8) 1 (0.4) 1 (0.4) 1 (0.4) 6 (2.5)
27 (39.7) 28 (41.2) 10 (14.7) 1 (1.4) — — 2 (2.9)
104 (59.1) 35 (19.9) 31 (17.6) — 1 (0.6) 1 (0.6) 4 (2.2)
Totals
244 (100)
68 (100)
176 (100)
2491
Group Urine culture Stone size (cm) Stone location Dilation Lithotripsy device Procedure Operative mins
Meaning
p Value
Before vs after 2002* Neg vs pos Less than 1 vs 1–2 vs greater than 2 Pyelum ⫹ calix vs lower vs upper vs pyelum vs staghorn Telescopic vs balloon Ultrasonic vs mechanical vs laser vs LithoClast® Uneventful vs eventful Less than 90 vs greater than 90
0.007 0.048 ⬍0.0001 0.046 0.006 ⬍0.0001 0.004 0.019
* Before 2002 surgery was performed by 5 urologists while after 2002 it was performed by 1 urologist within the context of an endourology team.
variables group allocation, urine culture, stone size, dilation type, lithotripsy device and operating time (table 5). On multivariate analysis stone size (OR 1.39), lithotripsy device (OR 1.25) and how complicated the procedure was (OR 3.71) remained independent variables for complications of PNL (table 5). DISCUSSION Standardization of the classification of complications for a certain procedure is necessary to allow comparison among different centers, within a center over time, or among instruments used or operating techniques. The modified Clavien system was proposed a few years ago11 and since then it has been reported mainly in laparoscopic procedures in urology.12,13 Recently Tefekli et al applied this system in their series of patients treated with PNL and presented results using each grade separately.14 An example of the type of comparison allowed by the Clavien system is that between theirs and our series. They reported any grade of complications in 29.2% (204 patients) of their 811 PNL procedures, with only 33 patients (4.06%) with a grade 1 complication and 171 (25.1%) with a complication higher than grade 1. However, in our series we had an overall higher complication rate of 43.8% (107 patients) but a lower Clavien score than Tefekli et al (63 patients [25.8%] had a Clavien grade 1 complication). Therefore, one may presume that a tertiary center focused
on endourology has lower complication rates. Moreover, complications are closely related to training, experience and caseload.15 Our study goal was an outcome comparison between the period when our department was not considered a dedicated endourological center and the period since it has been established as such. Therefore, comparing groups 1 and 2 should explain the observed trend of a greater percentage of PNL procedures shifting to specialized tertiary centers.8 Our results demonstrate for the first time to our knowledge that a dedicated center for PNL provides significant benefits in terms of a lower complication rate, decreased operative time, more uneventful procedures and decreased hospitalization. Three findings that require further explanation are the importance of operating time, the increased incidence of perforations and postoperative fever in group 2. This study is one of the first to identify operating time as an influencing factor related to morbidity and PNL complications. Regarding the incidence of perforation and fever no differences were found between the perforated cases and the rest of the group with regard to stone burden, access technique and type of lithotripter used, so the explanation of these differences is unclear and might just be coincidental. However, the consequence of the perforation is of minimal importance to the clinical course. As far as the increased incidence of postoperative fever is concerned there was no difference in the incidence of diabetes mellitus between the 2 groups but there were significantly more positive urine cultures in group 2 (p ⫽ 0.007) which could explain this fever related finding. The negative correlation of the mean Clavien score with each year over time, in addition to smaller scores in
TABLE 5. Univariate and multivariate analysis
Relationship of mean Clavien score with year of PNL with regard to number of patients (broken line).
Univariate: Lithotripsy device Urine culture Time Group Stone size Procedure Dilation Multivariate: Lithotripsy device Stone size Procedure
p Value
OR
95% CI
⬍0.0001 0.048 0.02 0.007 ⬍0.0001 0.005 0.012
1.23 1.69 1.87 2.21 2.28 2.83 4.8
1.127–1.411 1.004–2.844 1.106–3.194 1.239–3.938 1.155–4.533 1.37–5.83 1.14–20.1
⬍0.0001 0.003 0.006
1.25 1.39 3.71
1.13–1.48 1.17–1.92 1.46–9.42
2492
PROGNOSTIC FACTORS AND PERCUTANEOUS NEPHROLITHOTOMY
group 2, also shows that the Clavien score tends to decrease as familiarity with the technique increases (see figure). This finding supports studies showing that the learning curve for PNL represents a hyperbolic graph reaching a plateau when surgical competence and expertise are obtained.15–17 One of the important contributors to morbidity is operative time. The case may well be that as experience with PNL increases, operative time decreases and, thus, better PNL results are achieved. Another explanation may involve the balloon dilator used to obtain access. This type of access is obviously less timeconsuming than dilation with telescopic dilators. Moreover, this also emphasizes the significance of the specially trained team, including anesthesiologists and endourology related assistant nurses familiar with the procedure. Thus, time-consuming, unnecessarily delays are avoided. This is the first study to our knowledge to analyze the influence of prognostic factors on complications related to stone size, the course of the procedure and the type of lithotripter used. Other studies have been performed analyzing the influence of factors on the length of hospital stay including patient morbidity and complications of the procedure.18 Vascular complications after PNL have also been studied with the conclusion that the only influencing factor was stone size.19 Yamasaki reported that the size of the stone, the location and the presence of multiple stones were prognostic factors for stone recurrence and regrowth.20 In this study it was demonstrated that prognostic factors for an uncomplicated procedure are stone size, lithotripter used (much better for the ultrasonic) and if the course of the procedure was complicated (eventful vs uneventful). However, some factors that constrain the implications of our findings should be emphasized as well. The retrospective nature of the study is a degree of limitation. On the other hand these limitations account for both periods whereas the differences in outcome are significant. In addition, the extended study period causes a time effect with differences and heterogeneity in the study material due to the constant developments in instruments and stone disintegration methods. Data collection was often dependent on procedural reports by the surgeon especially in the earlier years. The modified Clavien system was proposed as a grading system for perioperative complications in general surgery and there are some limitations in classifying PNL complications, especially as auxiliary treatments (re-PNL, extracorporeal SWL, ureteroscopy) are part of the stone management as such and, therefore, should not really be considered complications. A specialized procedure or disease treatment modified Clavien system should solve this problem but this system has to be evaluated in trials addressing these questions. CONCLUSIONS The present study findings imply that when PNL procedures are performed at dedicated centers for stone treatment, significant benefits are provided in terms of decreased operative time, more uneventful procedures and decreased hospitalization. A significant parameter that influences the complication rate is the reduced operative time that is achieved with the experience and specialization of the whole surgical team. In addition, standardizing the classification of PNL morbidity with the modified Cla-
vien system constitutes a reliable tool for more objective outcome comparisons among centers, surgical techniques or tools used.
ACKNOWLEDGMENTS Sander Mettes reviewed patient files. APPENDIX Clavien System Grade 0 1 2 3a 3b 4a 4b 5
Meaning No complications Deviation from normal postoperative course without the need for intervention Minor complications requiring intervention Complications requiring intervention without general anesthesia Complications requiring intervention with general anesthesia Life threatening complications requiring intensive care management (single organ dysfunction) Life threatening complications requiring intensive care management (multiple organ dysfunction) Death
Abbreviations and Acronyms PNL ⫽ percutaneous nephrolithotomy SWL ⫽ shock wave lithotripsy
REFERENCES 1.
2.
3. 4.
5.
6.
7. 8.
9.
10. 11.
12.
Morris DS, Wei JT, Taub DA, Dunn RL, Wolf JS Jr and Hollenbeck BK: Temporal trends in the use of percutaneous nephrolithotomy. J Urol 2006; 175: 1731. Stamatelou KK, Francis ME, Jones CA, Nyberg LM and Curhan GC: Time trends in reported prevalence of kidney stones in the United States: 1976-1994. Kidney Int 2003; 63: 1817. Skolarikos A, Alivizatos G and de la Rosette JJ: Percutaneous nephrolithotomy and its legacy. Eur Urol 2005; 47: 22. Skolarikos A and de la Rosette JJ: Prevention and treatment of complications following percutaneous nephrolithotomy. Curr Opin Urol 2008; 18: 229. Miller NL and Lingeman JE: Treatment of kidney stones: current lithotripsy devices are proving less effective in some cases. Nat Clin Pract Urol 2006; 3: 236. Weizer AZ, Zhong P and Preminger GM: Twenty-five years of shockwave lithotripsy: back to the future? J Endourol 2005; 19: 929. de la Rosette JJ: Extracorporeal shock wave lithotripsy and the “end of the stone age”. Eur Urol 2006; 50: 400. Morris DS, Taub DA, Wei JT, Dunn RL, Wolf JS Jr and Hollenbeck BK: Regionalization of percutaneous nephrolithotomy: evidence for the increasing burden of care on tertiary centers. J Urol 2006; 176: 242. Birkmeyer JD: Should we regionalize major surgery? Potential benefits and policy considerations. J Am Coll Surg 2000; 190: 341. Michel MS, Trojan L and Rassweiler JJ: Complications in percutaneous nephrolithotomy. Eur Urol 2007; 51: 899. Dindo D, Demartines N and Clavien PA: Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004; 240: 205. Gonzalgo ML, Pavlovich CP, Trock BJ, Link RE, Sullivan W and Su LM: Classification and trends of perioperative mor-
PROGNOSTIC FACTORS AND PERCUTANEOUS NEPHROLITHOTOMY bidities following laparoscopic radical prostatectomy. J Urol 2005; 174: 135. 13. Permpongkosol S, Link RE, Su LM, Romero FR, Bagga HS, Pavlovich CP et al: Complications of 2,775 urological laparoscopic procedures: 1993 to 2005. J Urol 2007; 177: 580. 14. Tefekli A, Ali Karadag M, Tepeler K, Sari E, Berberoglu Y, Baykal M et al: Classification of percutaneous nephrolithotomy complications using the modified Clavien grading system: looking for a standard. Eur Urol 2008; 53: 184. 15. de la Rosette JJ, Laguna MP, Rassweiler JJ and Conort P: Training in percutaneous nephrolithotomy–a critical review. Eur Urol 2008; Epub ahead of print. 16. Allen D, O’Brien T, Tiptaft R and Glass J: Defining the learning curve for percutaneous nephrolithotomy. J Endourol 2005; 19: 279.
17.
2493
Tanriverdi O, Boylu U, Kendirci M, Kadihasanoglu M, Horasanli K and Miroglu C: The learning curve in the training of percutaneous nephrolithotomy. Eur Urol 2007; 52: 206. 18. Matlaga BR, Hodges SJ, Shah OD, Passmore L, Hart LJ and Assimos DG: Percutaneous nephrostolithotomy: predictors of length of stay. J Urol 2004; 172: 1351. 19. Srivastava A, Singh KJ, Suri A, Dubey D, Kumar A, Kapoor R et al: Vascular complications after percutaneous nephrolithotomy: are there any predictive factors? Urology 2005; 66: 38. 20. Yamasaki A: Long-term results of endourological treatment of urinary calculi: multivariate analyses of the risk factors related to recurrence or re-growth. Nippon Hinyokika Gakkai Zasshi 1994; 85: 1601.
Purpose: We stratified factors affecting treatment morbidity, compared the outcomes of percutaneous nephrolithotomy procedures from a single department and provided evidence of treatment benefits when percutaneous nephrolithotomy is performed in an expert setting. Materials and Methods: Since the department became a dedicated endourological center in 2002 we grouped all percutaneous nephrolithotomy procedures into those performed before 2002 (group 1) and after 2002 (group 2). The modified Clavien classification was used to score morbidity. Independent variables with an influence on complications were studied including stone size, operating time, operative complications, dilation device, urine culture, group allocation and lithotripsy device. Contingency and logistic regression were used for univariate and multivariate analysis. Results: Of the 244 percutaneous nephrolithotomy procedures 68 comprised group 1 and 176 formed group 2. Statistical preoperative differences were patient age, the use of anticoagulants and positive urinary cultures. Group 1 had a complication rate of 56.8% and group 2 had a complication rate of 37.2%. There were significant differences between the groups (p ⫽ 0.007). Almost all complications were grade 1 to 2. On univariate analysis the influence variables were urine culture (OR 1.69), group allocation (OR 2.20), stone size (OR 2.28), dilation device (OR 4.8), lithotripsy device (OR 1.22), perioperative complications (OR 2.83) and surgical time (OR 1.87). On multivariate analysis the independent factors in the complicated outcome were stone size (OR 1.25), type of lithotripsy device (OR 1.35) and incidence of perioperative complications (OR 3.71). Conclusions: The dedicated setting for percutaneous nephrolithotomy at our center resulted in decreased operative time, more uneventful procedures and decreased hospitalization time. The modified Clavien morbidity score is a reliable tool for more objective outcome comparisons after renal stone treatment. Key Words: nephrostomy, percutaneous; classification; morbidity
lthough percutaneous nephrolithotomy for the treatment of urolithiasis is considered more invasive than other currently available minimally invasive techniques, trends in renal stone surgery show that PNL has been increasingly used in recent years.1 Among the contributing factors for this trend are the increasing incidence of stone disease, and the proven safety and efficacy of PNL for renal stones.2 This applies especially to stones larger than 2 cm or staghorn calculi that are resistant to fragmentation, stones occurring in kidneys with an abnormal anatomy and stones in more complicated patient groups.3,4 In addition, the decrease in efficacy of the latest generation shock wave lithotripsy devices5,6 and the limitations in their use7 have further accentuated the advantages and benefits of PNL.
A
Submitted for publication March 14, 2008. * Correspondence: Department of Urology G4-105, Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands (telephone: ⫹31 20 5666377; FAX: ⫹31 20 6669585; email: [email protected]). † Financial interest and/or other relationship with Galil Medical, BSC and AMS.
0022-5347/08/1806-2489/0 THE JOURNAL OF UROLOGY® Copyright © 2008 by AMERICAN UROLOGICAL ASSOCIATION
Another trend that seems to accompany the increasing use of PNL for renal stones is the greater percentage of PNL procedures shifting to specialized tertiary centers.8 At these centers a dedicated endourology team (including an endourologist, specialized nurses, an anesthesiologist and radiological support) is in place, and has access to a full range of state-of-the-art endourology equipment and instruments (including laser, mechanical lithotripters and SWL). This shift may be explained by the significantly improved outcomes and decreased expenditures resulting from the standardization of practice patterns. Supportive of this is the suggestion that regionalization of certain procedures that are technically more complex and demanding specialized medical centers may improve the overall quality of health care.9 Moreover, this may also result favorably in a reduction in medical related morbidity and complications.10 To define the treatment related benefits obtained when PNL is performed at a specialized center it is necessary to perform an outcome comparison with those procedures performed at a nonspecialized setting which is optimally done when there are minimum regional and institutional differences. In this retrospective study we present the results of PNL for renal stones performed at a single
2489
Vol. 180, 2489-2493, December 2008 Printed in U.S.A. DOI:10.1016/j.juro.2008.08.025
2490
PROGNOSTIC FACTORS AND PERCUTANEOUS NEPHROLITHOTOMY
TABLE 1. Preoperative group characteristics and clinical data Group 1 No. pts No. gender: Male Female Mean pt age (⫾SD) No. Pretreatment (%): SWL PNL No. diabetes mellitus (%) No. anticoagulants (%) No. pos urine culture (%) Mean cm stone burden (⫾SD) No. renal anomalies (%)
68
Group 2
p Value
176
38 30 46.2 (⫾16.3)
97 79 51.4 (⫾15.2)
0.914 0.020*
34 (50) 9 (13.2) 5 (7.4) 2 (2.9) 24 (35.3) 3.1 (⫾2.4) 14 (20.6)
48 (27) 19 (10.8) 13 (7.3) 27 (15.3) 77 (43.8) 2.4 (⫾1.9) 16 (10.8)
0.001* 0.609 0.402 0.007* 0.007* 0.083 0.209
* Statistically significant difference (p ⬍0.05).
department, and analyze and compare the outcomes of the PNL procedures, as well as provide evidence of treatment benefits when PNL is performed in an expert setting. MATERIALS AND METHODS The medical files of all patients who underwent PNL for renal stones at our department from January 1994 until June 2007 were collected and evaluated. PNL procedures were divided into 2 groups according to the date of operation. Group 1 included procedures performed before the year 2002 (alternatingly done by 5 urologists) while those procedures performed after 2002 formed group 2 (done by 1 urologist in the context of an endourology team). The year 2002 was used as a grouping parameter because since then the department has become an endourological center with expertise and dedication to PNL in terms of personnel and equipment. The modified Clavien grading system was used to evaluate the perioperative morbidity of PNL treatment (see Appendix).11 Stone burden was defined as the largest diameter of the stone in cm or the sum of the diameters of multiple stones, while anomalous kidneys are those with ureteropelvic junction stenosis, horseshoe kidneys or kidneys with previous partial nephrectomy. PNL was defined as uneventful when there was no reported problem in gaining access, no bleeding or perforation, and with uncomplicated stone fragmentation and extraction. The procedure was considered to have failed when it was not possible to gain access or when it had to be discontinued due to instrumentation problems. Preoperative clinical data and treatment outcomes were compared using Student’s t test. Odds ratios were calculated and statistical determinations were within the 95% CI. All p values were 2-tailed and p ⬍0.05 was considered statistically significant. For correlation 2 ⫻ 2 crosstabs (2) were used while logistic regression was used for the univariate and multivariate analysis. Analysis was performed with the SPSS® (version 12.0) statistical software package. RESULTS A total of 244 PNL procedures were performed. Of these procedures 68 were performed before 2002 (group 1) and 176 were performed after 2002 (group 2). The demographic characteristics of the patients and preoperative
clinical data are presented in table 1. There were no statistical differences between the groups with regard to gender, incidence of diabetes mellitus, mean stone burden and American Society of Anesthesiologists score, while the percentage of anomalous kidneys was equally distributed between the groups. However, there was a significant difference in mean patient age, type of preoperative treatment when PNL was performed as a second line therapy, use of anticoagulants and incidence of positive urinary culture. Table 2 shows operative and postoperative data. Balloon dilation was used more frequently to obtain the access tract in group 2 and more uneventful procedures were seen in this group. Mean operative time was much lower in group 2 (97.2 vs 121.2 minutes in group 1, p ⫽ 0.003) as was mean hospitalization time. However, group 2 had a significantly higher incidence of postoperative fever (18.2% vs 5.9% in group 1, p ⫽ 0.041). In terms of the procedural events in both groups there were significantly more uneventful cases in group 2 with less bleeding and fewer failed attempts observed. However, this group had a higher incidence of cases with perforations of the renal pelvis (7.9% vs 2.9%, respectively). Stone-free rates were similar for both groups. Sufficient data to grade complications with the modified Clavien scoring system were possible in 238 PNL procedures (66 group 1, 172 group 2), identifying complications in a total of 107 PNL cases (44.9%). In group 1 of 66 procedures 39 (59.1%) had complications of Clavien grade 1 or greater. On the other hand, only 68 of 172 (39.5%) patients had some complication in group 2, resulting in a statistically significant difference between the 2 groups (p ⫽ 0.007). Almost all complications were grade 1 to 2, while group 1 had a Clavien grade 3 (subtype a) case in which surgical intervention without general anesthesia was required due to pneumothorax. Group 2 had 1 grade 3 (subtype b) and 1 grade 4 case (subtype a) (table 3). Looking at the mean Clavien score over time we found that there is a negative correlation of these variables, indicating smaller scores in group 2 (see figure). A further analysis of the relationships of mean Clavien score with different parameters is presented in table 4. A univariate analysis was performed using as influence
TABLE 2. Operative and postoperative data Group 1 Mean operative mins (⫾SD) No. dilatations (%): Balloon Other No. procedures (%): Uneventful Failed Days hospitalization: Mean ⫾ SD (range) % Less than 5 % Greater than 10 No. postop fever (%) No. perforation (%) No. stone-free postop (%)†
121.1
(⫾58.2)
Group 2 97.2
p Value*
(⫾37.2)
0.003
24 35
(39.3) 145 (60.7) 21
(82.3) (17.7)
0.001
50 8
(73.5) (11.8)
(84.7) (2.8)
0.006
8.2 ⫾ 3.7 (4–16) 6.8 ⫾ 3.4 (3–14) 25 38 40 12 4 (5.9) 32 (18.2) 3 (4.4) 14 (7.9) 48 (71) 146 (83)
0.002
49 5
0.041 0.061 0.24
* Statistically significant difference (p ⬍0.05). † Nonclinically significant residual fragments (smaller than 2 mm).
PROGNOSTIC FACTORS AND PERCUTANEOUS NEPHROLITHOTOMY TABLE 3. Distribution of complications with the Clavien grading system No. (%)
TABLE 4. Correlation between Clavien grading system and other factors Variable
Clavien
Global
Group 1
Group 2
0 1 2 3a 3b 4a Missing
131 (53.7) 63 (25.8) 41 (16.8) 1 (0.4) 1 (0.4) 1 (0.4) 6 (2.5)
27 (39.7) 28 (41.2) 10 (14.7) 1 (1.4) — — 2 (2.9)
104 (59.1) 35 (19.9) 31 (17.6) — 1 (0.6) 1 (0.6) 4 (2.2)
Totals
244 (100)
68 (100)
176 (100)
2491
Group Urine culture Stone size (cm) Stone location Dilation Lithotripsy device Procedure Operative mins
Meaning
p Value
Before vs after 2002* Neg vs pos Less than 1 vs 1–2 vs greater than 2 Pyelum ⫹ calix vs lower vs upper vs pyelum vs staghorn Telescopic vs balloon Ultrasonic vs mechanical vs laser vs LithoClast® Uneventful vs eventful Less than 90 vs greater than 90
0.007 0.048 ⬍0.0001 0.046 0.006 ⬍0.0001 0.004 0.019
* Before 2002 surgery was performed by 5 urologists while after 2002 it was performed by 1 urologist within the context of an endourology team.
variables group allocation, urine culture, stone size, dilation type, lithotripsy device and operating time (table 5). On multivariate analysis stone size (OR 1.39), lithotripsy device (OR 1.25) and how complicated the procedure was (OR 3.71) remained independent variables for complications of PNL (table 5). DISCUSSION Standardization of the classification of complications for a certain procedure is necessary to allow comparison among different centers, within a center over time, or among instruments used or operating techniques. The modified Clavien system was proposed a few years ago11 and since then it has been reported mainly in laparoscopic procedures in urology.12,13 Recently Tefekli et al applied this system in their series of patients treated with PNL and presented results using each grade separately.14 An example of the type of comparison allowed by the Clavien system is that between theirs and our series. They reported any grade of complications in 29.2% (204 patients) of their 811 PNL procedures, with only 33 patients (4.06%) with a grade 1 complication and 171 (25.1%) with a complication higher than grade 1. However, in our series we had an overall higher complication rate of 43.8% (107 patients) but a lower Clavien score than Tefekli et al (63 patients [25.8%] had a Clavien grade 1 complication). Therefore, one may presume that a tertiary center focused
on endourology has lower complication rates. Moreover, complications are closely related to training, experience and caseload.15 Our study goal was an outcome comparison between the period when our department was not considered a dedicated endourological center and the period since it has been established as such. Therefore, comparing groups 1 and 2 should explain the observed trend of a greater percentage of PNL procedures shifting to specialized tertiary centers.8 Our results demonstrate for the first time to our knowledge that a dedicated center for PNL provides significant benefits in terms of a lower complication rate, decreased operative time, more uneventful procedures and decreased hospitalization. Three findings that require further explanation are the importance of operating time, the increased incidence of perforations and postoperative fever in group 2. This study is one of the first to identify operating time as an influencing factor related to morbidity and PNL complications. Regarding the incidence of perforation and fever no differences were found between the perforated cases and the rest of the group with regard to stone burden, access technique and type of lithotripter used, so the explanation of these differences is unclear and might just be coincidental. However, the consequence of the perforation is of minimal importance to the clinical course. As far as the increased incidence of postoperative fever is concerned there was no difference in the incidence of diabetes mellitus between the 2 groups but there were significantly more positive urine cultures in group 2 (p ⫽ 0.007) which could explain this fever related finding. The negative correlation of the mean Clavien score with each year over time, in addition to smaller scores in
TABLE 5. Univariate and multivariate analysis
Relationship of mean Clavien score with year of PNL with regard to number of patients (broken line).
Univariate: Lithotripsy device Urine culture Time Group Stone size Procedure Dilation Multivariate: Lithotripsy device Stone size Procedure
p Value
OR
95% CI
⬍0.0001 0.048 0.02 0.007 ⬍0.0001 0.005 0.012
1.23 1.69 1.87 2.21 2.28 2.83 4.8
1.127–1.411 1.004–2.844 1.106–3.194 1.239–3.938 1.155–4.533 1.37–5.83 1.14–20.1
⬍0.0001 0.003 0.006
1.25 1.39 3.71
1.13–1.48 1.17–1.92 1.46–9.42
2492
PROGNOSTIC FACTORS AND PERCUTANEOUS NEPHROLITHOTOMY
group 2, also shows that the Clavien score tends to decrease as familiarity with the technique increases (see figure). This finding supports studies showing that the learning curve for PNL represents a hyperbolic graph reaching a plateau when surgical competence and expertise are obtained.15–17 One of the important contributors to morbidity is operative time. The case may well be that as experience with PNL increases, operative time decreases and, thus, better PNL results are achieved. Another explanation may involve the balloon dilator used to obtain access. This type of access is obviously less timeconsuming than dilation with telescopic dilators. Moreover, this also emphasizes the significance of the specially trained team, including anesthesiologists and endourology related assistant nurses familiar with the procedure. Thus, time-consuming, unnecessarily delays are avoided. This is the first study to our knowledge to analyze the influence of prognostic factors on complications related to stone size, the course of the procedure and the type of lithotripter used. Other studies have been performed analyzing the influence of factors on the length of hospital stay including patient morbidity and complications of the procedure.18 Vascular complications after PNL have also been studied with the conclusion that the only influencing factor was stone size.19 Yamasaki reported that the size of the stone, the location and the presence of multiple stones were prognostic factors for stone recurrence and regrowth.20 In this study it was demonstrated that prognostic factors for an uncomplicated procedure are stone size, lithotripter used (much better for the ultrasonic) and if the course of the procedure was complicated (eventful vs uneventful). However, some factors that constrain the implications of our findings should be emphasized as well. The retrospective nature of the study is a degree of limitation. On the other hand these limitations account for both periods whereas the differences in outcome are significant. In addition, the extended study period causes a time effect with differences and heterogeneity in the study material due to the constant developments in instruments and stone disintegration methods. Data collection was often dependent on procedural reports by the surgeon especially in the earlier years. The modified Clavien system was proposed as a grading system for perioperative complications in general surgery and there are some limitations in classifying PNL complications, especially as auxiliary treatments (re-PNL, extracorporeal SWL, ureteroscopy) are part of the stone management as such and, therefore, should not really be considered complications. A specialized procedure or disease treatment modified Clavien system should solve this problem but this system has to be evaluated in trials addressing these questions. CONCLUSIONS The present study findings imply that when PNL procedures are performed at dedicated centers for stone treatment, significant benefits are provided in terms of decreased operative time, more uneventful procedures and decreased hospitalization. A significant parameter that influences the complication rate is the reduced operative time that is achieved with the experience and specialization of the whole surgical team. In addition, standardizing the classification of PNL morbidity with the modified Cla-
vien system constitutes a reliable tool for more objective outcome comparisons among centers, surgical techniques or tools used.
ACKNOWLEDGMENTS Sander Mettes reviewed patient files. APPENDIX Clavien System Grade 0 1 2 3a 3b 4a 4b 5
Meaning No complications Deviation from normal postoperative course without the need for intervention Minor complications requiring intervention Complications requiring intervention without general anesthesia Complications requiring intervention with general anesthesia Life threatening complications requiring intensive care management (single organ dysfunction) Life threatening complications requiring intensive care management (multiple organ dysfunction) Death
Abbreviations and Acronyms PNL ⫽ percutaneous nephrolithotomy SWL ⫽ shock wave lithotripsy
REFERENCES 1.
2.
3. 4.
5.
6.
7. 8.
9.
10. 11.
12.
Morris DS, Wei JT, Taub DA, Dunn RL, Wolf JS Jr and Hollenbeck BK: Temporal trends in the use of percutaneous nephrolithotomy. J Urol 2006; 175: 1731. Stamatelou KK, Francis ME, Jones CA, Nyberg LM and Curhan GC: Time trends in reported prevalence of kidney stones in the United States: 1976-1994. Kidney Int 2003; 63: 1817. Skolarikos A, Alivizatos G and de la Rosette JJ: Percutaneous nephrolithotomy and its legacy. Eur Urol 2005; 47: 22. Skolarikos A and de la Rosette JJ: Prevention and treatment of complications following percutaneous nephrolithotomy. Curr Opin Urol 2008; 18: 229. Miller NL and Lingeman JE: Treatment of kidney stones: current lithotripsy devices are proving less effective in some cases. Nat Clin Pract Urol 2006; 3: 236. Weizer AZ, Zhong P and Preminger GM: Twenty-five years of shockwave lithotripsy: back to the future? J Endourol 2005; 19: 929. de la Rosette JJ: Extracorporeal shock wave lithotripsy and the “end of the stone age”. Eur Urol 2006; 50: 400. Morris DS, Taub DA, Wei JT, Dunn RL, Wolf JS Jr and Hollenbeck BK: Regionalization of percutaneous nephrolithotomy: evidence for the increasing burden of care on tertiary centers. J Urol 2006; 176: 242. Birkmeyer JD: Should we regionalize major surgery? Potential benefits and policy considerations. J Am Coll Surg 2000; 190: 341. Michel MS, Trojan L and Rassweiler JJ: Complications in percutaneous nephrolithotomy. Eur Urol 2007; 51: 899. Dindo D, Demartines N and Clavien PA: Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004; 240: 205. Gonzalgo ML, Pavlovich CP, Trock BJ, Link RE, Sullivan W and Su LM: Classification and trends of perioperative mor-
PROGNOSTIC FACTORS AND PERCUTANEOUS NEPHROLITHOTOMY bidities following laparoscopic radical prostatectomy. J Urol 2005; 174: 135. 13. Permpongkosol S, Link RE, Su LM, Romero FR, Bagga HS, Pavlovich CP et al: Complications of 2,775 urological laparoscopic procedures: 1993 to 2005. J Urol 2007; 177: 580. 14. Tefekli A, Ali Karadag M, Tepeler K, Sari E, Berberoglu Y, Baykal M et al: Classification of percutaneous nephrolithotomy complications using the modified Clavien grading system: looking for a standard. Eur Urol 2008; 53: 184. 15. de la Rosette JJ, Laguna MP, Rassweiler JJ and Conort P: Training in percutaneous nephrolithotomy–a critical review. Eur Urol 2008; Epub ahead of print. 16. Allen D, O’Brien T, Tiptaft R and Glass J: Defining the learning curve for percutaneous nephrolithotomy. J Endourol 2005; 19: 279.
17.
2493
Tanriverdi O, Boylu U, Kendirci M, Kadihasanoglu M, Horasanli K and Miroglu C: The learning curve in the training of percutaneous nephrolithotomy. Eur Urol 2007; 52: 206. 18. Matlaga BR, Hodges SJ, Shah OD, Passmore L, Hart LJ and Assimos DG: Percutaneous nephrostolithotomy: predictors of length of stay. J Urol 2004; 172: 1351. 19. Srivastava A, Singh KJ, Suri A, Dubey D, Kumar A, Kapoor R et al: Vascular complications after percutaneous nephrolithotomy: are there any predictive factors? Urology 2005; 66: 38. 20. Yamasaki A: Long-term results of endourological treatment of urinary calculi: multivariate analyses of the risk factors related to recurrence or re-growth. Nippon Hinyokika Gakkai Zasshi 1994; 85: 1601.