Is Outpatient Totally Tubeless Standard Percutaneous Nephrolithotomy Safe and Efficacious?

Introduction

Kidney stones are common with a prevalence of 11% in the United States according to a recent NHANES analysis. ¹ Treatment depends on the size and location. ^2,3 The American Urological Association (AUA) guidelines recommend percutaneous nephrolithotomy (PCNL) as first-line therapy for stones >2 cm or >1 cm and in the lower pole. ^2,3 When PCNL was described in 1976, the standard of care was to leave a nephrostomy tube postoperatively (post-op). ^4,5 However, because of nephrostomy tubes being bothersome and leading to reno-cutaneous fistula and urinomas, tubeless PCNL (t-PCNL), in which only a Double-J stent is left for post-op drainage, was introduced in the late 1990s. ⁶ Internalized stents were associated with lower pain scores and shorter length of stay (LOS), but caused renal colic, bladder spasms, urinary frequency, and hematuria, leading to pain, longer hospitalizations, ED visits, and readmissions. ^6,7 In fact, one study showed that a stent following PCNL had more morbidity and worse effect on quality of life (QOL) measures than a nephrostomy tube. ⁸ Subsequently, totally tubeless PCNL (tt-PCNL) where no internal or external tubes are left at the time of discharge after a short hospital stay was reported. ^4,9 However, most studies on tt-PCNL have left at least a ureteral catheter or foley catheter for several hours or overnight, and to our knowledge, no series has described same-day discharge for tt-PCNL.

In this study, we describe our experience with tt-PCNL on a totally outpatient basis—with the patient going home within 2–6 hours postprocedure with absolutely no tube, catheter, or stent at all, evaluating its safety and efficacy vis-à-vis a comparison with outpatient t-PCNL patients discharged with an indwelling stent.

Methods

Institutional review board (IRB # 14-00879) approval for data collection and evaluation of patients scheduled for outpatient PCNL was obtained. Consecutive patients scheduled for outpatient PCNL were prospectively enrolled in a nonrandomized manner from August 2023 to January 2024. All procedures were performed by two experienced fellowship trained surgeons (M.G. and W.M.A.) in a regular hospital. Patients with a solitary kidney, Chronic kidney disease (CKD) 3 or higher, specific anatomical variations such as horseshoe kidney, pelvic kidney, ptotic kidney, or urinary diversion, or a preexisting nephrostomy were excluded, as well as patients undergoing a mini-PCNL, since they typically have lower stone burdens and are more likely to be left without a stent. Patients with preexisting stents or ureteral stones were not excluded. Intravenous antibiotics were administered as per AUA guidelines on antibiotic prophylaxis. Patients were positioned in a modified Barts flank-free position. Cystoscopy was performed, followed by retrograde placement of a 0.035-inch guidewire in the renal pelvis, followed by a ureteral occlusion balloon (Occluder™, Boston Scientific, Inc.). Ultrasound (US)-guided access was established, the tract balloon dilated to 24F with sheath placement (NephroMax™, 24F), and nephroscopy and lithotripsy were conducted with gravity irrigation at a standardized height of 80 cm. Lithotripsy was performed using a 3.9-mm dual energy probe (Swiss LithoClast Trilogy®, Boston Scientific, Inc.).

At case conclusion, decision was made for totally tubeless vs placing a stent. Usual criteria for tt-PCNL included no evidence of active infection at surgery, no lithotripsy of a ureteral stone, absence of severe or massive hydronephrosis, minimal ureteral/renal pelvic inflammation, no narrowing at the ureteropelvic junction (UPJ), minimal intraoperative bleeding, and no endoscopic, ultrasonic, and fluoroscopic evidence of residual fragments.

For tt-PCNL patients, an open-ended ureteral catheter was passed over a wire into the renal pelvis. For t-PCNL patients, a double pigtail silicone stent (Black Silicone stent, Cook Urological Inc.) was placed under vision with proximal coil in the renal pelvis and distal coil in the bladder. In both groups, a hemostatic agent (SURGIFLO®, Ethicon, Inc.) was injected into the tract during sheath removal, and a foley catheter was placed. External pressure was applied to the kidney for 2 minutes to help with hemostasis. For tt-PCNL patients, the external aspect of the ureteral catheter was tunneled into the foley. Foley catheters (and external ureteral catheter in the tt-PCNL patients) were removed exactly 1 hour following the procedure in the postanesthesia care unit (PACU). After voiding, all patients were discharged from the PACU, with an indwelling stent in the tubeless patients and without any stents, catheters, or tubes in tt-PCNL patients.

Demographic information and baseline characteristics, including age, sex, body mass index (BMI), diabetes, prior stone surgery, Guy’s stone score, stone burden, and preoperative hydronephrosis, were collected. Intraoperatively, access location, needlestick attempts, and total operative (OR) time were recorded. Post-op data included pain scores using visual analog scales (VASs) during recovery and discharge from PACU and time spent in PACU.

Primary outcomes included 30-day complications and ED visits or readmissions. Secondary outcome was stone-free rate (SFR) evaluated with kidney ureter bladder X-ray (KUB) and/or renal/bladder ultrasound (RBUS). Statistical analysis was performed using SPSS. Categorical variables were compared using Chi-square or Fisher’s exact tests, and continuous variables were analyzed via Mann–Whitney U tests.

Results

Consecutive patients (n = 130) scheduled for outpatient PCNL during the study period were prospectively enrolled and after exclusions 103 patients analyzed (Fig. 1). No patient was admitted immediately post-op. No patients required a nephrostomy. Fifty patients were discharged with an indwelling stent (t-PCNL), whereas 53 patients were discharged with no stent (tt-PCNL). There were no significant differences in mean age, BMI, gender, diabetes, prior PCNL, Guy’s stone scores, stone burden, and preoperative hydronephrosis between groups (Table 1).

OPEN IN VIEWER

FIG. 1.

Flow Diagram.

Location of access did not differ between groups and was most commonly lower pole (70.0% vs 77.4%, p-value = 0.547) and subcostal (84% vs 81%, p-value = 0.893). Needlestick attempts (62% one time, 64.2% one time, p = 0.663) did not differ between groups. Total OR time (80 minutes vs 57 minutes, p < 0.001) was longer in the t-PCNL (stented) group (Table 2).

Patients were stented for the following reasons: stones at UPJ (n = 15), tight ureterovesical junction (n = 4), concern for infection/inflammation (n = 26), poorly draining kidney (n = 1), cortical scarring/renal insufficiency (n = 1), difficult anatomy (n = 1), and tight ureter (n = 1).

Post-op data and complications are detailed in Table 3. No visceral or pleural injuries occurred in either group. There were no immediate post-op transfusions. There were also no immediate post-op admissions or prolonged observations, with all patients being discharged from PACU. Delayed transfusion during readmission was similar (1.9% vs 2.0%, p = 1). ED visits without admissions (2.0% vs 0%, p = 0.485), readmissions (16% vs 9.4%, p = 0.316), and SFR (78.8% vs 93.5%, p = 0.084) were also comparable (Table 3). Reasons for readmissions in the tt-PCNL cohort included two pseudoaneurysms, two episodes of hydronephrosis necessitating stent replacement, and fever. Reasons for readmission in the t-PCNL cohort included fever, gross hematuria, urinary retention, and severe AKI. Clavien–Dindo 2 and 3 complications were similar (6.0% vs 1.9%, p = 0.354, 2.0% vs 7.5%, p = 0.205), but Clavien–Dindo 1 complications were higher with t-PCNL compared with tt-PCNL (16% vs 3.8%, p = 0.048).

PACU time was comparable (4.35 vs 3.96 hours, p = 0.274). Average PACU pain (1.74 vs 1.89, p = 0.666), maximum PACU pain (4.32 vs 4.92, p = 0.356), and pain at discharge (0.88 vs 1.77, p = 0.078) were similar. The mean PACU morphine equivalents were higher for t-PCNL than for tt-PCNL groups (8.18 vs 6.75), but did not reach statistical significance (p = 0.7). US was the most common post-op imaging modality for both t-PCNL and tt-PCNL patients (US: t-PCNL = 67%; tt-PCNL = 78%, CT: t-PCNL = 27%, tt-PCNL = 11%, KUB: t-PCNL = 6%, tt-PCNL = 11%, p < 0.05).

Discussion

Leaving a patient totally tubeless on the day of PCNL has not been common practice because of a concern for potential complications such as post-op urinary leakage and urinary obstruction and to avoid phone calls and return visits to the office or emergency room. We conducted this study to assess the feasibility and efficacy of rendering a patient totally tubeless and to assess the feasibility of same day discharge without tubes. Although discharged totally tubeless, we prefer temporary internal drainage in the form of an open-ended catheter for 1 hour. The reason is to ensure that the path of least resistance is antegrade, preventing formation of a urinoma from leakage out of the tract and allowing the kidney parenchyma to coapt and seal the tract by decompression, which also helps with hemostasis.

To our knowledge, our study is the first to report same day discharge for tt-PCNL and to compare outcomes with outpatient t-PCNL. We found no differences in intraoperative complications, transfusion rates, Clavien 2 or 3 complications, and readmissions. Two tt-PCNL patients had reoperations, both for delayed stent placement for ureteral obstruction, one for clots and the other for hydroureteronephrosis caused by edema at the ureteral orifice. There were no cases of urine leak or urinoma. Two tt-PCNL patients had residual obstructing stones. One patient in the t-PCNL and one in the tubed PCNL group had AVMs necessitating embolization. Another study found that 2/43 (4.6%) tt-PCNL, 3/41 (7.3%) t-PCNL, and 7/92 (7.6%) tubed PCNL patients developed complications (p = 0.783). ⁷ We did find a higher rate of Clavien 1 complications with t-PCNL, in particular hematuria and bladder pain, likely because of the indwelling stent. A ureteroscopy study found a higher frequency of LUTS and hematuria in stented vs nonstented patients. ¹⁰ In a randomized controlled trial comparing nephrostomy drainage vs ureteral stent following PCNL, patients with a ureteral stent had significantly worse QOL. ⁸ Our data confirm that tt-PCNL is safe and reduces symptoms in select patients probably by virtue of not leaving an indwelling stent.

PACU and discharge VAS scores in our patients were comparable. Previous studies showed that tt-PCNL improved post-op pain scores. ^{7,8,11 –13} However, pain was assessed during admission in these studies. Yun and coworkers found that analgesic requirements (pain regimen not specified) during hospitalization were lower with tt-PCNL. ¹⁴ Istanbulluoglu and associates found no difference in NSAID but lower narcotic requirements with tt-PCNL vs both t-PCNL and tubed PCNL. ⁷ Bhat and colleagues found lower tramadol requirements and pain scores during hospitalization with tt-PCNL vs both t-PCNL and tubed PCNL (VAS 3.04 vs 4.76 vs 6.64, respectively, p < 0.0001). ¹¹ We only gathered pain scores at two points—during recovery and at discharge from PACU, which was 2–4 hours for our patients. Getting VAS scores 24 hours after operation may provide better evidence of reduced pain with tt-PCNL, without the influence of parenteral medications and general anesthesia. For this reason, we are performing a prospective randomized control trial comparing t-PCNL with tt-PCNL with VAS scores on days 1–3 and Patient-Reported Outcomes Measurement Information System (PROMIS) scores on days 10 and 30 post-op.

Another potential benefit of tt-PCNL is shorter post-op stay. ^6,7,11,12,14 The vast majority of PCNLs in our center are outpatient with most patients being discharged within 4 hours, making it difficult to discern a difference between LOS for tt-PCNL vs t-PCNL. PACU time was comparable between groups (4.35 vs 3.96 hours, p = 0.274). In other comparative studies, tt-PCNL patients stayed longer than t-PCNL or tubed PCNL patients and for a much longer time than in our study. In five studies, tt-PCNL had a shorter LOS in comparison with that of tubed PCNL. ^6,7,11,13,14 However, the two studies that compared tt-PCNL with t-PCNL had conflicting data regarding LOS. Although one study showed slightly shorter LOS when comparing tt-PCNL with t-PCNL (2.48 vs 3.20 days, respectively), another study showed a similar LOS for tt-PCNL (2.09 vs 1.74 days, respectively). ^7,11 Our tt-PCNL and t-PCNL patients had short post-op stays with minimal complications and ED visits, demonstrating the safety of outpatient tt-PCNL.

Another potential benefit of tt-PCNL includes shorter OR time. We found total OR time (80 minutes vs 57 minutes, p < 0.001) longer in the stented group. Often, we find that the procedural steps for stent placement, including the assessment of the position of the distal coil, can add to the OR time. For our tt-PCNL patients, since the distal end of the open-ended catheter does not need to be assessed, the OR time can be slightly shorter, but this does not account for the significant difference we saw. The more likely explanation is that our study was not a randomized trial, so there could be selection bias that would result in OR time differences. Patients with ureteral and renal stones, infected stones, and severe hydronephrosis are likely to have longer procedures and also more likely to have stent placement. Yun and associates in their nonrandomized retrospective study found no difference in total OR time between tt-PCNL and PCNL with nephrostomy tube (128.7 vs 148.5 minutes, respectively, p = 0.06). ¹⁴ Another study also found no difference in OR time for tt-PCNL, t-PCNL, and tubed PCNL (53, 51.3, and 60.1 minutes, respectively, p = 0.130). ⁷ Overall, these findings demonstrate that tt-PCNL approach has the potential to decrease OR time, but this would need to be confirmed in a prospective randomized trial of outpatient PCNL, which we are currently conducting.

One important concern when transitioning from tubed PCNL to t-PCNL is SFR. Proponents of a second look have been slow to adopt t-PCNL because an important avenue for re-entering the kidney to remove residual stones seen on a post-op CT is no longer available. A similar concern about going stentless and tubeless has been raised, but more because of the fear a residual fragment or missed stone would cause obstruction post-op, preventing the tract from closing and resulting in pain, obstruction, and/or urine leak/urinoma. We evaluated 30-day SFR using either KUB and/or RBUS in all patients. Although we noticed no statistical difference in SFR comparing outpatient t-PCNL with tt-PCNL (78.8% to 93.5%, p = 0.08), the SFR was higher with tt-PCNL, again possibly because of selection bias in that patients with more severe hydronephrosis or infection stones were not candidates for tt-PCNL. This difference in SFR was not apparently because of Guy’s stone score, total stone burden, number of stones, or number of calices with stones, which were all similar in the two groups. Our SFR is comparable with prior studies. Yun and coworkers evaluated SFR based on a post-op KUB and CT scan with unspecified timing, finding a SFR of 73% for tubed vs 78% for tt-PCNL patients (p = 0.725). ¹⁴ Crook and associates reported SFR based on an intraoperative assessment only, with rates of 84% in tubed vs 96% in tt-PCNL. ¹³ Our study is consistent with prior literature, where no difference in SFR was observed post-op. These findings suggest that even in the setting of more complex stones with large stone burdens, leaving patients totally tubeless made no difference in SFR.

Although our study highlights the safety and efficacy of performing outpatient PCNL totally tubeless, it is not without limitations. One of the shortcomings of our study includes the absence of post-op imaging in all patients. Being a tertiary referral center, over 50% of our patients are referred by urologists outside our system. The majority of these do come from the tristate area, but it is generally more convenient for them to follow-up with their referring urologist in their local town. This is true for totally tubeless patients who do not need a procedure for stent removal, but even more so for patients with a stent, who have stent removal by us but then are even more reluctant to come back for yet another visit into New York City. Although we do advise all patients to do a 24-hour urine and have imaging at 1-month postoperation or stent removal to evaluate for hydronephrosis and residual stones, we depend on the referring urologist to do this and to review the results with their patients, and do not always have access to these results. It is a standard of care in our practice to perform an US and/or KUB 1 month following an outpatient PCNL to assess for silent hydronephrosis and residual stone burden. Despite the limitations noted, the image follow-up rate in this series was 77%, but this does not reflect the true percentage of patients who had post-op imaging. The national average is reported to be 46% for KUB 3 months following PCNL. ¹⁵ Another limitation is that we did not record pain scores 24 hours and beyond the day of the procedure. In addition, since our patients were not randomized, the possibility of selection bias could affect OR and post-op factors. Prospective randomized controlled trials should be performed to compare tt-PCNL with t-PCNL to confirm our findings. We are currently carrying out a randomized controlled trial where we randomized patients to tt-PCNL and t-PCNL and assess pain from POD 1–3, perform a PROMIS at POD 10 and 30, and assess post-op complications.

Conclusion

Outpatient tt-PCNL is a safe and effective option in selected patients. When comparing stent omission with stent placement after t-PCNL, there is no difference in intraoperative complications, Clavien 2/3 complications, readmissions, ED visits, and transfusion rates. Furthermore, the patients who are left totally tubeless have a lower rate of Clavien 1 complications and shorter OR time. A randomized controlled trial should help confirm these findings.