Implementation of a Hospital-Wide Protocol Reduces Time to Decompression and Length of Stay in Patients with Stone-Related Obstructive Pyelonephritis with Sepsis

Abstract

Introduction and Objectives:

Patients with obstructive pyelonephritis (OPN) require urgent decompression through retrograde ureteral stent (RUS) or percutaneous nephrostomy (PCN). In 2016, the urology and interventional radiology (IR) departments at our institution established a protocol for patients with OPN with sepsis. The primary objectives were to assess this protocol's impact on improving time to decompression and whether more expedient decompression decreased length of stay (LOS). Secondarily, we assessed the impact of the protocol and clinical factors on receipt of PCN over RUS.

Materials and Methods:

One hundred forty-seven patients at our institution who underwent PCN from 2012 to 2017 or stent from 2014 to 2017 for stone-related OPN meeting sepsis criteria were retrospectively reviewed. Univariate descriptive statistics compared patient characteristics and outcomes between RUS and PCN pre- and postprotocol implementation. Multivariable logistic regression assessed predictors of decompression with PCN (vs RUS) and of prolonged LOS (pLOS; >5 days).

Results:

Utilization of PCN increased after implementation of the protocol from 4 to 14 PCN/year with a decrease in the median time from urologic consultation to PCN from 9.2 to 4.3 hours (p = 0.001) with overall median time to decompression decreasing from 5.4 to 4.5 hours (p = 0.017). Predictors of undergoing PCN (vs RUS) included increasing comorbidity and ≥1 cm obstructing stone. On multivariable analysis controlling for comorbidity, leukocytosis, and septic shock, increasing hours to decompression increased odds of pLOS (1.08, 95% confidence interval 1.02–1.15, p = 0.014).

Conclusions:

After implementing our OPN with sepsis protocol, time to decompression decreased with dramatic improvement in time to PCN. Quicker decompression was independently associated with reduced odds of prolonged hospital stay. A well-designed protocol engages both urology and IR in the management of these acutely ill patients and improves outcomes.

Introduction and Objectives

Kidney stones are a common problem affecting roughly 9% of all Americans.¹ Although many stones can be monitored or nonurgently managed, a minority of nephrolithiasis cases require urgent management. Most commonly this intervention is in the form of decompression, either with a retrograde ureteral stent (RUS) or a percutaneous nephrostomy (PCN) tube. Among the indications for urgent decompression, obstructive pyelonephritis (OPN) may have the direst consequences if not intervened upon. Large administrative data have demonstrated more than a twofold mortality risk increase (19.2% vs 8.8%) for those patients having sepsis from OPN who were not treated with decompression as opposed to those who were decompressed.²

Research has now shown that cases of OPN secondary to calculi have risen by as much as 95.6% in the 10-year period between 1999 and 2009.³ OPN results when a calculus blocks the passage of urine thereby allowing bacteria to ultimately enter the renal parenchyma, with the potential for escalation into severe sepsis.⁴ Severe sepsis and septic shock result in a mortality rate between 20% and 40%.⁵ In the specific case of OPN with sepsis, this rate is lower but still severe.² Haas and colleagues recently demonstrated that in patients presenting with OPN, a delay in decompression of >2 days has been shown to worsen mortality outcomes as demonstrated within the National Inpatient Sample.⁶ Upon recognition of possible OPN, the patient should receive antibiotics and imaging should be performed as soon as possible.⁷ Once a complicated urinary infection with obstruction is recognized (i.e., OPN), the next step is to immediately relieve the obstruction with a RUS or PCN.⁸

In our hospital system, there previously was no defined protocol in place for determining the modality for relief of obstruction in OPN with sepsis. Anecdotally, this can lead to potential disagreements at many hospitals in the method of decompression based on many factors, for example, availability of operating room vs interventional radiology (IR) suite, time of presentation, and institutional resources. In an effort to optimize time to upper tract decompression for patients presenting with OPN secondary to urolithiasis, we established a protocol between the urology and IR departments (Fig. 1). The primary purpose of this study was to determine the protocol's impact on time to decompression. Secondarily, we sought to determine predictors of depression modality (PCN vs RUS) while accounting for protocol implementation. Finally, we investigated whether time to decompression was associated with prolonged length of stay (pLOS) after decompression to validate time to decompression as a meaningful quality measure in patients with OPN secondary to urolithiasis.

FIG. 1.

Obstructive pyelonephritis with sepsis protocol.

Materials and Methods

The OPN with sepsis protocol was designed to give objective criteria whereby the treating urologist could recommend decompression with PCN through the IR team, especially when the consulting urologist felt that a percutaneous intervention may be safer and/or more expeditious (Fig. 1). In brief, the IR team could be consulted when patients with OPN met sepsis criteria and had hydronephrosis or an anatomic variation that would make RUS difficult. Implementation of the OPN algorithm went into effect in 2016 after approval by both departments as to the acuity of these patients and need for expeditious decompression.

An IRB-approved retrospective chart review was done on all patients who underwent PCN from 2012 to 2018 and all patients who underwent RUS from 2014 to 2018 as inpatients. As there were fewer PCNs performed preprotocol, it was necessary to screen an additional 2 years before obtain additional numbers for more accurate group comparisons. Patients were screened through current procedural terminology codes for PCN insertion and RUS placement. Keyword searches from operative notes such as “pyelonephritis,” “sepsis,” “calculus,” “stone,” and “lithiasis,” were done to initially evaluate which of the >4000 RUS and PCN procedures should be more carefully chart reviewed.

All patients included in the study were required to have obstruction from a kidney/ureteral stone with indications for urgent decompression in which the urologist could have activated the OPN with sepsis protocol for IR decompression with PCN if the treating urologist thought that PCN was the preferable modality for decompression. Patient demographic information, comorbidities, illness severity measures, CT findings, decompression modality, intensive care unit (ICU) admission, and time points of initial emergency room (ER) presentation, urology consult time, decompression time, and discharge time were all recorded. The presence of septic shock before decompression was defined as persistent hypotension requiring vasopressors or ICU level care. Hydronephrosis severity was coded on a scale from 0 to 3 with 0 representing no hydronephrosis and 3 representing severe hydronephrosis. In case of either modality's failure, the patient and time points were categorized as the first chosen decompression modality in an intention to treat manner.

Descriptive statistics were generated to compare characteristics and outcomes of the cohorts of RUS vs PCN both before and after the protocol as well as comparisons before vs after the protocol for the RUS, PCN, and combined cohorts. The chi-square test was used for categorical variables, whereas the t-test was employed for normally distributed continuous variables and the Mann–Whitney U test for non-normally distributed variables. Two separate multivariable logistic regression models were utilized to assess the odds of decompression with PCN (vs RUS) and a pLOS. A pLOS was defined as ≥5 days as 5 days was the upper quartile of length of stay (LOS) after decompression. Covariates were chosen based on clinical significance and statistical significance in the univariate comparisons with total number of covariates limited to not more than 1 per 10 pLOS outcome. A sensitivity analysis subdividing the RUS and PCN cohorts was also performed to assess if the impact of timeliness of decompression on LOS was similar regardless of decompression modality. In analyzing the relationship between time to decompression and pLOS, time to decompression outliers were excluded through Tukey's rule of either 1.5 × interquartile range (IQR) below the first quartile or 1.5 × IQR above the third quartile. All statistics were performed using SPSS version 25.

Results

There was a total of 140 patients decompressed at our institution for OPN with sepsis from 2014 to 2018 and an additional 7 PCNs obtained from 2012 to 2013 that met study entry criteria. Table 1 breaks down the patient characteristics and outcomes by RUS vs PCN and by pre- vs postprotocol implementation with descriptive statistics comparisons. After the protocol's implementation, PCNs were done far more frequently at a rate of 14/year compared with only 4/year before the protocol, whereas the rate of RUS remained roughly stable (22/year to 25/year). Both RUS failure rates (3/94, 3%) and PCN failure rates (2/53, 4%) were low with RUS failures occurring because of an impacted stone and PCN failures occurring because of difficult access at the time of procedure.

Table 1.

Cohort Characteristics and Outcomes Stratified by retrograde ureteral stent vs Percutaneous Nephrostomy and by Pre- vs Postobstructive Pyelonephritis Protocol

Parameter	Preprotocol				Pre- vs postprotocol p-values			Postprotocol
Parameter	RUS	PCN	Total	p	RUS	PCN	Total	RUS	PCN	Total	p
Count	44 (72%)	17 (28%)	61 (100%)	—	—	—	—	50 (58%)	36 (42%)	86 (100%)	—
Female gender	33 (75%)	11 (65%)	44 (72%)	0.421	0.743	0.658	0.451	36 (72%)	21 (58%)	57 (66%)	0.186
Age, mean ± SD	52.6 ± 18	64.4 ± 14	55.9 ± 18	0.019	0.53	0.325	0.56	50.2 ± 18	59.5 ± 18	54.1 ± 18	0.022
CCI, mean ± SD	2.1 ± 2.6	4.7 ± 2.6	2.8 ± 2.8	0.002	0.473	0.032	0.175	1.8 ± 2.1	2.9 ± 2.8	2.2 ± 2.5	0.046
Tmax, mean ± SD	38.3 ± 1	37.9 ± 1.7	38.2 ± 1.2	0.232	0.554	0.328	0.363	38.4 ± 0.9	38.3 ± 1	38.4 ± 0.9	0.434
Min SBP, mean ± SD	108 ± 22	90 ± 16	103 ± 22	0.004	0.81	0.07	0.449	111 ± 26	102 ± 26	108 ± 26	0.144
WBC, mean ± SD	14.6 ± 6	14.9 ± 6	14.7 ± 6	0.864	0.407	0.469	0.233	15.6 ± 6	16.3 ± 7	16 ± 6	0.590
Septic shock (before decompression)	4 (9%)	8 (47%)	12 (20%)	0.001	0.138	0.037	0.989	10 (20%)	7 (19%)	17 (20%)	0.949
≥1 cm stone	7 (16%)	10 (59%)	17 (28%)	0.001	0.454	0.243	0.756	11 (22%)	15 (42%)	26 (30%)	0.050
Hydro severity, mean ± SD	1.4 ± 0.73	1.8 ± 0.81	1.5 ± 0.76	0.057	0.396	0.964	0.275	1.5 ± 0.76	1.8 ± 0.70	1.7 ± 0.75	0.072
Outcomes
Hours ER arrival to GU consult, median (IQR)	7.7 (4.6–9)	6.1 (3.5–12)	7.6 (5–10)	0.841	0.342	0.334	0.627	6.2 (4–10)	9.5 (5–16)	7.4 (4–15)	0.009
Hours GU consult to decompression, median (IQR)	4.5 (3.4–7.6)	9.2 (6.6–14)	5.4 (3.5–9)	0.001	0.609	0.001	0.017	4.5 (3.2–5.8)	4.3 (3.2–6.6)	4.5 (3.2–6.1)	0.786
Hours ER to decompression, median (IQR)	13 (9–20)	18 (11–28)	14 (10–21)	0.074	0.369	0.561	0.457	11 (8–18)	16 (11–26)	13 (8–21)	0.041
LOS s/p decompression, median (IQR)	2.7 (1.7–4.6)	7.8 (4.3–15)	3.2 (2–7)	0.001	0.458	0.045	0.55	2.9 (1.9–4.4)	3.9 (2.7–8.0)	3.1 (2.2–5.4)	0.017
pLOS >5 days s/p decompression	6 (14%)	10 (63%)	16 (28%)	<0.001	0.511	0.047	0.702	9 (20%)	10 (32%)	19 (25%)	0.205
ICU admission	5 (11%)	11 (65%)	16 (26%)	<0.001	0.072	0.01	0.944	13 (26%)	10 (28%)	23 (27%)	0.854
Death	0	1 (6%)	1 (2%)	0.105	0.18	0.58	0.497	2 (4%)	1 (3%)	3 (4%)	0.761

Boldface indicates statistical significance.

CCI = Charlson comorbidity index; ER = emergency room; GU = urology; ICU = intensive care unit; IQR = interquartile range; LOS = length of stay; PCN = percutaneous nephrostomy; pLOS = prolonged length of stay; RUS = retrograde ureteral stent; s/p = status post; SBP = systolic blood pressure; SD = standard deviation; Tmax = maximum recorded temperature; WBC = white blood cell.

Preprotocol, the population of patients who underwent PCN vs RUS were older (64.4 years vs 52.6 years, p = 0.019), had a higher Charlson Comorbidity Index (CCI; 4.7 vs 2.1, p = 0.002), and were more acutely ill with lower minimum systolic blood pressures (90 vs 108, p = 0.004) and a higher percentage of septic shock (47% vs 9%, p = 0.001). On univariate comparison, PCN was favored over RUS in patients with ≥1 cm stones both before and after the protocol, whereas more severe hydronephrosis trended toward decompression with PCN. Multivariable logistic regression (Table 2) of the entire cohort confirmed predictors of undergoing PCN as CCI (odds ratio [OR] 1.25, 95% confidence interval [CI] 1.07–1.46, p = 0.005), ≥1 cm stone (OR 3.18, 95% CI 1.42–7.15, p = 0.005), whereas hydronephrosis severity trended toward favoring PCN (OR 1.62, 95% CI 0.97–2.72, p = 0.066). Although patients were more likely to undergo PCN after the protocol (OR 2.25, 95% CI 1.00–5.05, p = 0.049), septic shock was not a significant driver for decompression with PCN.

Table 2.

Logistic Regression Predictors for Undergoing Decompression with Percutaneous Nephrostomy (Percutaneous Nephrostomy n = 53)

Parameter	Univariate OR (95% CI)	p	Multivariate OR (95% CI)	p
Age	1.03 (1.01–1.05)	0.002	—	—
Female gender	0.55 (0.27–1.13)	0.104	—	—
CCI	1.24 (1.09–1.42)	0.002	1.25 (1.07–1.46)	0.005
WBC	1.02 (0.97–1.08)	0.480	—	—
≥1 cm stone	3.77 (1.79–7.94)	<0.001	3.18 (1.42–7.15)	0.005
Hydro scale (0–3)	1.92 (1.19–3.09)	0.008	1.62 (0.97–2.72)	0.066
Septic shock (before decompression)	2.26 (0.99–5.14)	0.053	1.35 (0.52–3.51)	0.538
Postprotocol (vs preprotocol)	1.86 (0.92–3.77)	0.083	2.25 (1.00–5.05)	0.049

Boldface indicates statistical significance.

CI = confidence interval; OR = odds ratio.

The primary objective of lowering time from urology consult to decompression with implementation of the OPN with sepsis protocol was met (Fig. 2). Median time from urology consultation to receipt of PCN decreased from a median 9.2 hours to a median of 4.3 hours (p = 0.001) after protocol implementation, whereas median time of the entire cohort (PCN+RUS) decreased from 5.4 to 4.5 hours (p = 0.017). Although time to decompression with RUS did not significantly decrease with a stable median of 4.5 hours, the mean time to RUS dropped from 6.0 to 4.9 hours with fewer patients at the upper extremes.

FIG. 2.

Box and whisker plot of time from urology consult to decompression with pre- and post-OPN protocol comparisons. Median = middle horizontal line; X = mean. OPN = obstructive pyelonephritis.

Although time from urology consult to decompression was the primary time measure that was intervened upon, time from ER presentation to decompression is more clinically relevant and thus was chosen for use in the analysis on impact on pLOS >5 days after decompression. Results of the univariate and multivariable logistic regression analysis are shown in Table 3. All variables (age, female gender, CCI, white blood cells (WBC), PCN (vs RUS), septic shock before decompression, and time from ER presentation to decompression) were all significant on univariable analysis. On multivariable analysis, hours from ER presentation to decompression remained significant with a similar OR on univariable analysis (OR 1.08, 95% CI 1.02–1.15, p = 0.014). This represents an additional 8% increased odds of pLOS after decompression per hour of delay in decompression. Not surprisingly, presence of septic shock before decompression had the greatest impact on prolonging LOS (OR 8.28, 95% CI 2.53–27.1, p = 0.001). We also found the WBC to be an independent predictor of pLOS with each unit increase resulting in a 10% increased odds in pLOS (OR 1.10, 95% CI 1.03–1.18, p = 0.008). On sensitivity analysis subdividing the cohort into RUS and PCN groups, hours from ER presentation to decompression was found to be a significant predictor within the RUS cohort (OR 1.11, 95% CI 1.01–1.21, p = 0.028) on multivariable analysis, but only trended toward significance within the PCN group (OR 1.06, 95% CI 0.98–1.15, p = 0.098), likely because of the diminished power of the PCN group (n = 53) compared with the RUS group (n = 94).

Table 3.

Logistic Regression on Prolonged Length of Stay Status Post Decompression >5 Days (Prolonged Length of Stay n = 35)

Parameter	Univariate OR (95% CI)	p	Multivariable OR (95% CI)	p
Age	1.03 (1.01–1.05	0.009	—	—
Female gender	0.30 (0.13–0.67)	0.003	—	—
CCI	1.41 (1.20–1.65)	<0.001	1.27 (1.05–1.54)	0.014
WBC	1.10 (1.03–1.18)	0.008	1.11 (1.03–1.21)	0.010
PCN	3.6 (1.62–8.04)	0.002	—	—
Septic shock (before decompression)	11.2 (4.05–30.9)	<0.001	8.28 (2.53–27.1)	<0.001
Hours from ER to decompression	1.09 (1.04–1.15)	<0.001	1.08 (1.02–1.15)	0.014

Boldface indicates statistical significance.

Discussion

OPN is a urologic emergency warranting prompt decompression. In addition to good clinical judgment dictating the importance of prompt decompression, there is now evidence that delaying decompression either by missed diagnosis or delay in intervention can negatively impact mortality.⁶ This study reinforces that a prompt decompression impacts patient outcomes—for each additional hour of delay we observed 8% increased odds of the patient having a pLOS on multivariable analysis. A previous study by Srougi and colleagues analyzing 40 patients with obstructive ureteral stones and systemic inflammatory response syndrome for risk factors for septic shock did not find a correlation between time to decompression and septic shock.⁹ However, this study was limited in sample size and confounding whereby patients exhibiting signs of septic shock were likely decompressed more urgently than less acute patients. A strength of this study was its determination of septic shock before decompression and inclusion of this variable within multivariable analysis to control for confounding between time to decompression, septic shock, and LOS.

As time to decompression was the only modifiable parameter that affected outcomes, it should be considered as an important quality measure in the care of patients with OPN. This study has shown that reductions in time to decompression can be attained through collaboration with IR and the institution of an OPN with sepsis protocol. After instituting the protocol, the time from urologic consultation to decompression improved from a median of 5.4 to 4.5 hours with most of the improvement attributable to quicker receipt of PCN. Although the majority of reduction in time to decompression likely resulted from lessening decisional conflict with increased willingness by IR to assist in the management of stone-related OPN with sepsis patients, additional improvements may have resulted from the increased awareness of situational urgency by the attending and resident physicians of all services involved. The protocol did not address the importance of prompt diagnosis of the condition, which could also reduce time from presentation to decompression and improve outcomes. This would involve efforts at education of ER staff that was beyond the scope of the study.

From a practical standpoint, adapting such a protocol that gives the urologist choice in decompression modality requires a cooperative IR department willing to follow a treatment algorithm and a urology staff that justly chooses PCN in appropriate scenarios to serve the patient's interests—namely selecting the modality that will be the most likely to succeed while also being the most expeditious. Although the protocol strives to streamline the decision-making process, there are of course gray areas where discussion of risk/benefits of each approach is warranted between attending urologist and interventional radiologist. Although all patients in this study met criteria for OPN with sepsis and either anatomic variant or hydronephrosis could have been chosen for PCN, the majority of patients (58%) still received RUS after implementation of the protocol. Historically and somewhat dogmatically, PCN is generally preferred for more acutely ill patients^10,11; however, beyond relative contraindications to general anesthesia with RUS, there is no evidence to suggest that either modality is superior to the other with two randomized trials showing no difference in outcomes.^12
–14 Multivariable analysis showed the primary drivers for choice of PCN over RUS in our study to be anatomically related with larger stones and increased hydronephrosis tending to receive PCN. Beyond the expediency of care, should the patient not be in florid sepsis, consideration must be given to the patient's preference for quality of life considerations as RUS and PCN have different side effect profiles.¹⁵

Limitations of our study arise from its retrospective single-institution design. First, the intervention (initiation of the OPN with sepsis protocol) was instituted across all patients at a set point in time and, therefore, the control group was a historical cohort in which evolving practices may have changed over time regardless of protocol initiation. Randomization of the intervention (utilizing the protocol vs usual care) could not practically be achieved because of the acute setting of OPN and the necessary shift in mind-set that the OPN protocol required. Although the study was done in a retrospective manner, retrospective screening techniques utilizing keywords from operative notes followed by a detailed chart review limited the impact of case omission. Furthermore, we recognize that this specific protocol may not be necessary at institutions without interdepartmental conflict; however, in those institutions where such disagreements may be a source of delay, these results should serve to spark improved cooperation and adoption of a similar-minded protocol. Such a protocol may also not be feasible or practical at institutions that possess more limited IR or urology resources. Nevertheless, we strongly believe that a mutually agreed upon OPN with sepsis management protocol primarily based on objective criteria will improve decisiveness in modality choice and expediency of decompression, which is integral to improving patient outcomes.

Conclusions

After implementing our OPN with sepsis protocol, time to decompression decreased with dramatic improvement seen in time to PCN. Furthermore, a timely decompression was shown to reduce odds of prolonged hospital stay after controlling for patient comorbidity and illness measures. A well-designed protocol that streamlines the decision of decompression modality while engaging both urology and IR in the management of these acutely ill patients expedites decompression and improves outcomes.

Footnotes

Acknowledgment

We would like to acknowledge Dr. Christopher Sayegh for his assistance in data acquisition, creating the abstract and poster presentation for the 2019 World Congress of Endourology that was further adapted for the manuscript.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

Abbreviations Used

References

Scales

, Smith

, Hanley

, et al. Prevalence of kidney stones in the United States. Eur Urol, 2012; 62:160–165.

Borofsky

, Walter

, Shah

, et al. Surgical decompression is associated with decreased mortality in patients with sepsis and ureteral calculi. J Urol, 2013; 189:946–951.

Sammon

, Ghani

, Karakiewicz

, et al. Temporal trends, practice patterns, and treatment outcomes for infected upper urinary tract stones in the United States. Eur Urol, 2013; 64:85–92.

Talner

, Davidson

, Lebowitz

, et al. Acute pyelonephritis: Can we agree on terminology?. Radiology, 1994; 192:297–305.

Martin

, Mannino

, Eaton

, et al. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med, 2003; 348:1546–1554.

Haas

, Li

, Hyams

, Shah

Delayed decompression of obstructing stones with urinary tract infection is associated with increased odds of death. J Urol, 2020. DOI: 10.1097/JU.0000000000001182.

Wagenlehner

FME

, Pilatz

, Naber

, et al. Therapeutic challenges of urosepsis. Eur J Clin Invest, 2008; 38 Suppl 2:45–49.

Assimos

, Krambeck

, Miller

, et al. Surgical management of stones: American urological association/endourological society guideline, PART I. J Urol, 2016; 196:1153–1160.

Srougi

, Moscardi

, Marchini

, et al. Septic shock following surgical decompression of obstructing ureteral stones: A prospective analysis. J Endourol, 2018; 32:446–450.

10.

Goldsmith

, Oredein-McCoy

, Gerber

, et al. Emergent ureteric stent vs percutaneous nephrostomy for obstructive urolithiasis with sepsis: Patterns of use and outcomes from a 15-year experience. BJU Int, 2013; 112:E122–E128.

11.

Lynch

, Anson

, Patel

. Current opinion amongst radiologists and urologists in the UK on percutaneous nephrostomy and ureteric stent insertion for acute renal unobstruction: Results of a postal survey. BJU Int, 2006; 98:1143–1144.

12.

Pearle

, Pierce

, Miller

, et al. Optimal method of urgent decompression of the collecting system for obstruction and infection due to ureteral calculi. J Urol, 1998; 160:1260–1264.

13.

Ramsey

, Robertson

, Ablett

, et al. Evidence-based drainage of infected hydronephrosis secondary to ureteric calculi. J Endourol, 2010; 24:185–189.

14.

Wang

, Hsu

, Chen

, et al. Percutaneous nephrostomy versus ureteroscopic management of sepsis associated with ureteral stone impaction: A randomized controlled trial. Urolithiasis, 2016; 44:415–419.

15.

Shoshany

, Erlich

, Golan

, et al. Ureteric stent versus percutaneous nephrostomy for acute ureteral obstruction—clinical outcome and quality of life: A bi-center prospective study. BMC Urol, 2019; 19:79.