Abstract
Introduction:
To investigate which patient, stone, infective, and surgical factors were most likely to increase the risk of postoperative urosepsis within 28 days of ureteroscopy (URS) and laser stone fragmentation for ureteral or renal stones.
Methods:
Data were collected prospectively in a single National Health Service institution. A logistic regression model was used to assess the association of factors with postoperative urosepsis. Two matched-pair analyses were used to assess the risk of postoperative urosepsis in patients with (a) an emergency presentation to hospital with urosepsis in the 90 days preceding URS and (b) a positive midstream sample of urine (MSSU) identified, but who were asymptomatic at preoperative assessment, who then received an appropriate course of antibiotics.
Results:
Four hundred sixty-two consecutive patients were included in the study. Thirty-four patients (7.4%) had an episode of urosepsis within 28 days of their operation. A positive preoperative MSSU was significantly associated with postoperative urosepsis on multivariable analysis, despite appropriate treatment with a preoperative course of antibiotics: odds ratio (OR) 4.88, 95% confidence interval (CI) 2.11, 11.31, p < 0.001. The presence of diabetes mellitus, presence of ischemic heart disease, patient American Society of Anesthesiologists score, same-session bilateral URS, and stone volume were the other variables significantly associated with postoperative infection on univariable analysis, but these ceased to be significantly associated on multivariable analysis. Subgroup analysis found that a positive MSSU in both patients with a preoperative ureteral stent and those without was significantly associated with postoperative urosepsis, however, the OR was much lower for the stented group (OR 3.23 vs OR 16.67). On matched-pair analysis, patients with a positive preoperative MSSU were significantly more likely to have postoperative urosepsis compared to controls (OR 17.46, 95% CI 2.18, 139.80, p = 0.007). There was no significant difference in the OR of postoperative urosepsis in patients who had a preceding urine infection requiring hospital admission in the 90 days preceding URS (OR 0.60, 95% CI 0.19, 1.92, p = 0.39).
Conclusions:
Positive preoperative MSSU was significantly associated with postoperative urosepsis by logistic regression and matched-pair analysis. These higher risk patients should be counseled appropriately before surgery, and should be the focus of vigilant postoperative monitoring. The study suggests particular caution in patients with a positive preoperative MSSU without a preoperative ureteral stent.
Introduction
U
We wanted to investigate which patient, stone, infective, and surgical factors were most likely to increase the risk of postoperative urosepsis within 28 days of surgery using univariable and multivariable logistic regression analysis. In particular, we hypothesized that having a preoperative stent may predispose to infection, especially when there was a positive midstream sample of urine (MSSU). In addition, given the interest in the role of preoperative urinary tract infection (UTI) from previous studies, 4,5 we also wanted to assess the impact of this in matched-pair analysis. For our matched-pair analysis, we identified two groups, which were then matched to their respective controls: (a) patients presenting emergency admission to hospital with symptomatic UTI or urosepsis in the 90 days preceding URS and (b) those who had a positive MSSU identified at preoperative assessment, but who were asymptomatic.
Methods
Data were collected from patients undergoing URS and stone fragmentation in a single National Health Service institution. Data were collected prospectively from November 2010 to April 2015. Patients undergoing elective procedures underwent preoperative assessment within 3 weeks of their operation and had a routine MSSU sent for culture. Patients with a positive MSSU were treated with a full treatment course (7 days) of an appropriate antibiotic (as per sensitivities). Repeat samples were then obtained to ensure clearance. All patients received a dose of prophylactic antibiotic at induction of anesthesia. Where the preoperative MSSU had been positive, antibiotic choice was as per sensitivities. Where the MSSU was sterile, patients received gentamicin at 5 mg/kg, except in patients with renal impairment (estimated glomerular filtration rate <30 mL/min/1.73 m2) where co-amoxiclav was used. All fragmentation was carried out by Holmium:YAG laser.
Postoperative urosepsis was defined as suspected or confirmed infection of urinary tract origin along with two or more of the following four factors: heart rate >90 beats/minute, temperature >38°C or <36°C, respiratory rate >20 breaths/minute or PaCO2 <4.3 kPa, and white blood cell count over 12 × 109/L or below 4 × 109/L. 6 Confirmed infection of urinary tract origin was by positive urine culture. Suspected infection was defined as sterile pyuria (>400 white cells/μL) with inhibitory substances present (in-keeping with antibiotic use) and a C-reactive protein over 10 mg/L, or the above plus a positive blood culture.
All patients were offered outpatient follow up at 3–4 months, with kidney, ureter, and bladder radiograph X-ray to determine if they were stone free. Stone-free status was classified as no fragments over 1 mm. If there was doubt regarding their stone status, patients underwent Computed Tomography (CT) scan.
Regression analysis
A logistic regression model was used to assess the association of factors with postoperative urosepsis. Patient factors included in the model were as follows: age, body mass index, presence of diabetes mellitus, presence of ischemic heart disease, and American Society of Anesthesiologists (ASA) score (as judged by anesthetist at the time of URS). Infection factors included were as follows: symptomatic UTI or urosepsis in the 90 days preceding URS requiring hospital attendance and positive MSSU at preoperative assessment. Stone factors were as follows: stone location, stone volume (measured in three perpendicular planes from CT on bone windowing), presence of multiple stones, and undergoing same-session bilateral URS. Surgical factors were as follows: whether the procedure was performed by consultant or specialty registrar, elective or emergency procedure, presence of a preoperative ureteral JJ stent or nephrostomy, presence of a postoperative JJ ureteral stent, and operation time.
Alezra et al. have previously found stone size over 1.5 cm and operation time over 70 minutes were significantly associated with postoperative pyelonephritis. 4 We studied this in our cohort by univariable regression analysis.
Matched-pair analysis
Strict criteria were used to match patients for each of our groups. First, patients with a history of an emergency admission to hospital with symptomatic UTI or urosepsis in the 90 days preceding URS were matched to patients without a history of UTI. In the second group, those patients found to have a positive MSSU identified at preoperative assessment were matched to patients with a negative MSSU preoperatively. Six parameters were employed in sequence: (a) gender; (b) age (divided into three cohorts: 20–40, 41–60, >60 years); (c) stone location (lower ureter, midureter, upper ureter, and renal pelvis/calices); (d) stone volume (<125 mm3, 125–1000 mm3 and >1000 mm3; these size cutoffs corresponding to 5 mm and 1 cm3); (e) presence of diabetes mellitus; and (f) ASA Score. When more than one match was identified, random numbers were assigned to rank potential matches.
All analyses were undertaken using IBM SPSS version 19 (SPSS). Significance was assumed at the 0.05 level.
Results
Four hundred sixty-two consecutive patients were included in the study. Thirty-four patients (7.4%) had an episode of urosepsis within 28 days of their operation. Positive preoperative and postoperative urine culture results are shown in Tables 1 and 2.
Sterile pyuria defined as >400 White cells/μL.
ESBL = extended spectrum beta-lactamase producing; CRP = C reactive protein.
Odds ratios (ORs) of the factors' association with postoperative urosepsis are shown in Tables 3 and 4.
p values in bold < 0.1 included in multivariable analysis.
ASA = American Society of Anesthesiologists; BMI = body mass index; CI = confidence interval; MSSU = midstream sample of urine; OR = odds ratio; UTI = urinary tract infection.
p value in bold < 0.05 significance level.
Those factors with an OR significantly impacting on urosepsis (p < 0.1) by univariable analysis (Table 1) were included in multivariable analysis.
One patient died of urosepsis despite antibiotics and postoperative care in the Intensive Treatment Unit. This patient had a past history of severe multiple sclerosis and had a large stone burden (multiple stones with total volume 1800 cm3). It was recognized as a high-risk procedure preoperatively and the patient counseled as such (further information regarding this patient is included in Supplementary Data; Supplementary Data are available online at
A positive preoperative MSSU was the most powerful predictor of postoperative urosepsis, despite adequate prior antibiotic treatment. Fourteen patients with positive preoperative MSSU developed postoperative urosepsis. Correlation between preoperative and postoperative cultures in patients with postoperative urosepsis is shown in Supplementary Table S1.
Four hundred twenty-eight patients (92.6%) attended follow-up clinic review, of which 357 (83.4%) were deemed stone free. Of the 71 patients who were not stone free at follow-up, 50 had small fragments and were treated conservatively. Twenty-one went on to have further URS. Being deemed to have residual fragments at follow-up was not associated with increased risk of postoperative urosepsis after the initial URS (OR 0.93, 95% CI 0.34, 2.49, p = 0.88).
Three hundred twenty-seven patients underwent treatment for stones in the upper ureter or kidney. An 11/13F access sheath (Navigator Sheath; Boston Scientific) was used in 240 of those patients, with 87 simply using a guidewire. Use of an access sheath was not associated with a significant reduction in postoperative urosepsis (OR 1.99, 95% CI 0.66, 5.97, p = 0.22).
In our study, there were 29 patients with stones measuring over 1.5 cm in one dimension. Stone size over 1.5 cm was not significantly associated with postoperative urosepsis (OR 2.06, 95% CI 0.68, 6.29, p = 0.20). We chose to measure stone volume in three dimensions to more accurately reflect stone burden. When patients were stratified according to stone volume below 1000 mm3 or over 1000 mm3 (i.e., equivalent of a 1 cm3), stone volume over 1000 mm3 was significantly associated with postoperative urosepsis (n = 81, OR 3.28, 95% CI 1.57, 6.86, p = 0.002). Sixty-eight patients had an operation longer than 70 minutes. There was no significant association between operative time over 70 minutes and postoperative infection: OR 1.89, 95% CI 0.82, 4.36, p = 0.14.
Same-session bilateral procedure (n = 20) was not associated with increased risk of postoperative infection compared with a second contralateral URS (n = 9) (bilateral URS OR 0.82, 95% CI 0.12, 5.57, p = 0.84).
Of the 52 patients with a positive preoperative MSSU, 32 had previously had a ureteral stent or nephrostomy inserted (22 for UTI in the preceding 90 days, and the remainder for pain or failed previous ureteroscopic access). In total, 182 patients in our study had a preoperative ureteral stent or nephrostomy. Subgroup analysis of those patients with a preoperative stent or nephrostomy found a significant association between positive preoperative MSSU and postoperative urosepsis: OR 3.23, 95% CI 1.08, 9.66, p = 0.04. We also reviewed the opposite patient group (i.e., those without a preoperative stent or nephrostomy). In that group, positive preoperative MSSU was strongly associated with the risk of postoperative infection: n = 20, OR 16.67, 95% CI 5.57, 49.84, p < 0.001.
Information on patients included in the matched-pair analyses is shown in Tables 5 and 6. Seventy-one patients with recent urine infection requiring hospital treatment in the 90 days preceding URS were matched to their respective controls. There was no significant difference in the OR of postoperative urosepsis in patients who had a preceding urine infection (OR 0.60, 95% CI 0.19, 1.92, p = 0.39).
SD = standard deviation.
Separately, 48 patients with a positive preoperative MSSU at preoperative assessment and treated with a course of appropriate antibiotics were matched to their respective controls. A preoperative positive MSSU was significantly associated with postoperative urosepsis (OR 17.46, 95% CI 2.18, 139.80, p = 0.007).
Discussion
A positive preoperative MSSU was the only factor significantly associated with postoperative urosepsis on multivariable analysis. Patients with a positive preoperative MSSU, despite being asymptomatic and receiving treatment with an appropriate preoperative course of antibiotics, were 4.88 × more likely to have postoperative urosepsis on multivariable analysis than those patients with a negative preoperative MSSU. The presence of diabetes mellitus, presence of ischemic heart disease, patient ASA score, undertaking same-session bilateral procedure, and stone volume were the other variables significantly associated with postoperative infection on univariable analysis, but these ceased to be significant on multivariable analysis.
Matched-pair analysis also showed that preoperative positive MSSU was significantly associated with postoperative urosepsis; in that analysis, preoperative positive MSSU increased risk 17.46 × compared to those with a negative MSSU. These results are striking and demonstrate both the importance of preoperative MSSU testing and close monitoring of patients post-URS with a positive preoperative MSSU.
Ureteral stents have a significant risk of colonization, and the sensitivity of urine culture to stent colonization is low. 7 Biofilm colonization has been shown to regularly occur on ureteral stents, with colonization usually indicated by asymptomatic bacteriuria in follow-up urine samples. 8 Bacteria cultured from biofilm or urine after stent insertion have been shown to be more resistant to antibiotics than bacteria cultured from urine before stent insertion. 7,8 We initially hypothesized that having a preoperative stent may predispose to infection, especially when there was a positive MSSU. Our subgroup analysis would seem to refute that concept: although a positive MSSU in patients with a preoperative stent and a positive MSSU in patients without a preoperative stent were both significantly associated with postoperative urosepsis, the OR was much lower for the patient group with stents (OR 3.23 vs OR 16.67). This would suggest that particular care should be taken in patients with a positive preoperative MSSU and no stent in situ. We speculate that patients without a stent were more likely to be obstructed and have stagnant urine, which could become infected. The drainage afforded by the stent may therefore outweigh the risk of the foreign body and biofilm formation.
There is a lack of consensus across the published literature regarding factors associated with post-URS infection. The Clinical Research Office of Endourological Society (CROES) URS Global study group reported that in 1325 patients with a negative baseline urine culture, female gender and ASA score of 2 or more were most strongly associated with postoperative infection. 9 One French group recently reported on 266 patients undergoing flexible URS, with postoperative acute pyelonephritis in 24 patients (7.4%). In their study, antibiotic treatment the week before URS was significantly associated with postoperative pyelonephritis on multivariable analysis (relative risk 5.8, p < 0.002). Preceding univariable analysis that found positive urine culture the day before surgery was also significantly associated. 4 These results would seem to corroborate our study findings. Similarly, Mitsuzuka et al. found that in their cohort of 153 patients, preoperative pyuria (>10 white blood cells per high power field: OR 3.62, p = 0.02) was significantly associated with postoperative febrile UTI. However, unlike our cohort, preoperative acute pyelonephritis (OR 4.43, p = 0.04) was also significantly associated with postoperative febrile UTI. 10
Previous studies have shown same-session bilateral URS and stone fragmentation to be safe. 4,11 Hollenbeck et al. 12 reported that same-session bilateral URS was associated with increased morbidity (OR 4.0, p = 0.02), but that the cumulative risk of performing staged bilateral procedures (14% per procedure) approximated that of bilateral URS in one sitting (29%). This study found that same-session bilateral URS was significantly associated with postoperative urosepsis on univariable analysis, but ceased to be significant on multivariable analysis. Same-session bilateral procedure was not associated with increased risk of postoperative infection compared with patients who underwent a second contralateral URS in the study period. This supports offering same-session bilateral URS.
Interestingly in Moses et al.'s recent study, compliance with the American Urological Association best practice statement on antibiotic prophylaxis was significantly associated with unplanned hospital return with UTI post-URS. 5 The reasons for this are unclear. In Chew et al.'s review of 81 patients undergoing URS for renal calculi, a course of postoperative antibiotics was not associated with lower risk of postoperative infection compared to a single prophylactic dose of antibiotic at anesthetic induction. 13 The numbers of patients in that study were small, with only eight patients actually having a postoperative UTI. Our review of a larger number of patients suggests that special attention should be paid to those with a positive preoperative MSSU. Further work is required to determine if these higher-risk patients would benefit from more intensive perioperative antibiotic prophylaxis.
The strengths of this study are the number of patients included, the broad range of factors included, and the confirmation of study findings by two different statistical methodologies, though this was a single-centre experience. The CROES URS Global Study group previously reported that in patients with a solitary kidney stone treated by flexible ureterorenoscopy, 52.2% were under 10 mm, 42.3% 10–20 mm, and 4.6% >20 mm. 14 Unlike that and other studies, we measured stone size in three dimensions on CT as opposed to a single dimension. We believe that this more accurately reflects stone burden, can be used to provide a total burden when there are multiple stones, and more accurately reflects the complexity of the stone both to the patient's symptoms and operatively; a 5 × 5 × 5 mm stone is quite different from a 5 × 1 × 1 mm one. As detailed above, many studies so far have involved single institutions and different outcome definitions and have not come to a consensus. A prospective, multi-institutional study of risk factors in post-URS infection would be helpful.
Conclusion
In conclusion, a positive preoperative MSSU was significantly associated with postoperative urosepsis. Further work is required to determine if these higher-risk patients would benefit from more intensive perioperative antibiotic prophylaxis, possibly by 24–48 hours of antibiotic therapy before and after surgery.
Footnotes
Acknowledgments
Early results from the above article were, in part, presented in poster form at the World Congress of Endourology, September 2015, and the Scottish Urological Society Meeting, April 2016.
Source of Study
Urology Department, Forth Valley Royal Hospital, Larbert, Scotland.
Author Disclosure Statement
No competing financial interests exist.
Abbreviations Used
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
