International Urology and Nephrology

, Volume 46, Issue 3, pp 493–497

Risk factors for septic shock in acute obstructive pyelonephritis requiring emergency drainage of the upper urinary tract

Authors

  • Jun Kamei
    • Department of UrologyThe University of Tokyo Hospital
  • Hiroaki Nishimatsu
    • Department of UrologyThe University of Tokyo Hospital
  • Tohru Nakagawa
    • Department of UrologyThe University of Tokyo Hospital
  • Motofumi Suzuki
    • Department of UrologyThe University of Tokyo Hospital
  • Tetsuya Fujimura
    • Department of UrologyThe University of Tokyo Hospital
  • Hiroshi Fukuhara
    • Department of UrologyThe University of Tokyo Hospital
  • Yasuhiko Igawa
    • Department of Continence Medicine, Graduate School of MedicineThe University of Tokyo
    • Department of UrologyThe University of Tokyo Hospital
  • Yukio Homma
    • Department of UrologyThe University of Tokyo Hospital
Urology - Original Paper

DOI: 10.1007/s11255-013-0545-5

Cite this article as:
Kamei, J., Nishimatsu, H., Nakagawa, T. et al. Int Urol Nephrol (2014) 46: 493. doi:10.1007/s11255-013-0545-5

Abstract

Purpose

To assess the risk factors for septic shock in patients with acute obstructive pyelonephritis requiring emergency drainage of the upper urinary tract.

Methods

We retrospectively reviewed the records of 48 patients who underwent emergency drainage of the upper urinary tract for sepsis associated with acute obstructive pyelonephritis at our institute. Univariate and multivariate analyses were performed to identify the risk factors.

Results

Among 54 events of sepsis, we identified 20 events of septic shock requiring vasopressor therapy. Cases with shock were more likely than those without shock to have ureteral stone (70 vs 38 %, p = 0.024) and positive blood culture results (81 vs 28 %, p = 0.006). They received drainage significantly earlier than those without shock (1.0 vs 3.5 days, p < 0.001). Univariate analysis demonstrated that acute obstructive pyelonephritis by ureteral stone, rapid progression (the occurrence of symptoms to drainage ≤1 day), positive blood culture, leukocytopenia (<4,000/mm3), thrombocytopenia (<120,000/mm3), and prothrombin time international normalized ratio ≥1.20 were correlated with septic shock. Multivariate logistic regression analysis identified thrombocytopenia (p = 0.005) and positive blood culture (p = 0.040) as independent risk factors for septic shock.

Conclusions

Thrombocytopenia and positive blood culture were independent risk factors for septic shock in acute obstructive pyelonephritis requiring emergency drainage. Thrombocytopenia would be practically useful as a predictor of septic shock.

Keywords

Acute obstructive pyelonephritisSeptic shockThrombocytopeniaUreteral drainage

Introduction

Acute obstructive pyelonephritis sometimes requires emergency drainages of the upper urinary tract by percutaneous nephrostomy or retrograde ureteral stenting [1]; however, septic shock may develop despite appropriate emergency drainage.

The Surviving Sepsis Campaign Guidelines, first published in 2004 and updated in 2008 [2, 3], are now regarded as the international standard for treatment of severe sepsis including urosepsis. Although the guidelines recommend emergency drainage for acute obstructive pyelonephritis, little information is known for the risk for developing septic shock in spite of drainage. Recently, Yamamoto et al. reported that age and presence of paralysis were independent risk factors for septic shock in patients receiving emergency drainage for acute pyelonephritis with ureteral calculi. However, in this study, they analyzed limited cases only with ureteral calculi [4], and for cases with other causes, the risk factor remains uncertain. Thus, in this study, we have attempted to identify the risk factors for septic shock in cases with other causes as well as ureteral calculi.

Methods

We retrospectively reviewed the records of patients who underwent emergency drainage of the upper urinary tract for sepsis associated with acute obstructive pyelonephritis at our institute from April 2006 to September 2011. The diagnosis of sepsis was made by the criteria for systemic inflammatory response syndrome (SIRS), which included two or more of the following conditions: (1) body temperature greater than 38 °C or less than 36 °C, (2) heart rate greater than 90 beats per minute, (3) tachypnea, as manifested by a respiratory rate greater than 20 breaths per minute or hyperventilation, as indicated by a partial CO2 pressure less than 32 mmHg, and (4) white blood cell count greater than 12,000/mm3 or less than 4,000/mm3, or more than 10 % immature neutrophils [5, 6]. Patients’ age, sex, underlying comorbidities, performance status, the side of infected kidney, type of drainage, cause of obstruction, and the time interval between the occurrence of symptoms of pyelonephritis and drainage were recorded. We also evaluated SIRS score, the results of blood and urine culture before antibiotic treatment, and laboratory examinations including blood leukocyte count, blood thrombocyte count at nadir, C-reactive protein (CRP), serum creatinine and total bilirubin level, and prothrombin time international normalized ratio (PT-INR) prior to drainage.

Septic shock was defined as sepsis with a systolic blood pressure less than 90 mmHg despite adequate fluid replacement or using vasopressors for at least 1 h [7, 8]. By comparing patients who did or did not progress to septic shock, we identified risk factors for septic shock.

Chi-square test, Student’s t test, and Mann–Whitney U test were used for univariate analysis to compare variables between cases with septic shock and without. All variables that were significant on univariate analysis were entered into multivariate analysis, and logistic regression analysis was used for multivariate analysis. Analyses were performed by JMP 9 (SAS institute Inc., Tokyo, Japan). P values less than 0.05 were considered significant. This study was approved by the institutional ethical committee (3,124).

Results

We identified 48 patients who met SIRS criteria and required emergency ureteral drainage for acute obstructive pyelonephritis. A total of 54 emergency drainage procedures were undertaken, with once in 44, twice in two, and three times in two patients. The patients’ clinical characteristics are shown in Table 1. The cause of obstruction was ureteral stone in 27, malignant neoplasm in 14, obstruction of ureteral stent in six, anastomotic stenosis of ileal conduit in three, adhesion after lower abdominal operation in two, ureteropelvic obstruction in one, and unknown in one. The patients’ laboratory data prior to drainage is shown in Table 2. Culture samples prior to treatment were taken from blood and urine in 34 events (63 %) and 53 events (99 %), respectively. Organisms isolated from blood and/or urine culture were Escherichia coli (n = 22), Pseudomonas aeruginosa (n = 5), and other gram-negative bacilli (n = 12).
Table 1

Characteristics of patients

 

Septic shock (+) (n = 20)

Septic shock (−) (n = 34)

p value

Sex

 

 Male

5

11

0.568

 Female

15

23

 

Age

 

 Mean (range)

60 (40–81)a

64 (38–84)a

0.200

Underlying comorbidities

 

 Diabetes mellitus

5

10

0.972

 Malignancy

5

16

0.188

 Steroid use

3

7

0.883

Laterality

 

 Right

12

19

0.768

 Left

8

15

 

Performance status

 

 0

13

20

0.157

 1

4

8

 

 2

1

6

 

 3

2

0

 

Drainage

 

 Percutaneous nephrostomy

5

15

0.160

 Retrograde ureteral stent

15

19

 

Cause of obstruction

 

 Urinary stone

14

13

0.024

 Others

6

21

 

SIRS score

 

 2

4

11

0.388

 3

8

15

 

 4

8

8

 

Interval between onset and drainage (days)

 

 Median (range)

1.0 (0–3)b

3.5 (0–20)b

<0.001

 ≤1 day

13

7

0.001

 ≥2 days

7

27

 

aMean (range),  chi-square test,  Student’s t test, b Median (range),  Mann–Whitney U test

Table 2

Laboratory data of patients prior to drainage

 

Septic shock (+) (n = 20)

Septic shock (−) (n = 34)

p value

White blood cell count (/μL)

11,200 (1,200–47,900)a

11,650 (4,000–35,200)a

<0.001

Leukocytopenia (<4,000/μL %)

5

0

0.01

Thrombocyte count nadir (×10,000/μL)

8.15 (0.80–41.90)a

22.0 (6.6–54.9)a

<0.001

Thrombocytopenia (<1.2 × 105/μL %)

16

3

<0.001

PT-INR

1.31 (0.92–3.51)a

1.13 (0.90–1.49)a

0.005

Elongation of PT-INR (≥1.20 %)

16

12

0.004

CRP (mg/dL)

13.69 (4.33–29.74)a

12.84 (0.88–37.68)a

0.425

Serum creatinine level (mg/dL)

2.26 (0.93–6.50)a

2.13 (0.45–19.79)a

0.802

Serum total bilirubin level (mg/dL)

0.80 (0.20–1.70)a

0.60 (0.20–2.40)a

0.914

Blood culture

 

 Positive

13

5

0.006

 Negative

3

13

 

Urine culture

 

 Positive

19

23

0.064

 Negative

1

10

 

aMedian (range),  Mann–Whitney U test,  chi-square test

Septic shock occurred in 20 events of drainage. No patients without septic shock died, but one female 74-year-old patient died of septic shock in 8 days after drainage. She had end-stage bladder cancer and poor performance status but no other complications. The other 19 cases with septic shock finally recovered from septic shock with intensive care and the median length of hospital stay was 15 days (5–75). Patients with septic shock were treated with antibiotics, vasopressors in all cases, blood infusion in 10 cases, gamma globulin in 10 cases, antithrombin III agents in 6 cases, blood purification therapy in 4 cases, and recombinant human soluble thrombomodulin in 3 cases (Table 3).
Table 3

Risk factors for septic shock

Variables

Number

Multivariate analysis

OR (95 %CI)

p value

Cause of obstruction

 

 Ureteral stone

27

Reference

0.951

 Other

27

1.81 (0.13–23.75)

 

Blood culture

 

 Positive

16

Reference

0.040

 Negative

18

9.11 (1.11–74.79)

 

Leukocytopenia (<4,000/μL)

 

 No

49

Reference

0.999

 Yes

5

>50 (0.00 to >200.0)

 

Thrombocytopenia (<1.2 × 105/μL)

 

 No

35

Reference

0.005

 Yes

19

23.90 (2.64–216.18)

 

PT-INR

 

 <1.20

26

Reference

0.266

 ≥1.20

28

3.24 (0.41–25.75)

 

Onset to drainage

 

 ≥2 days

34

3.25 (0.37–28.65)

0.288

 ≤1 day

20

Reference

 

Cases with shock were more likely than those without shock to have ureteral stone (p = 0.024) and positive blood culture results (p = 0.006). They received emergency drainage significantly earlier than those without shock (p < 0.001).

Univariate analysis demonstrated that septic shock was significantly associated with ureteral stone (p = 0.024), rapid progression (the occurrence of symptoms to drainage ≤1 day, p = 0.01), positive blood culture (p = 0.006), leukocytopenia (<4,000/mm3) (p = 0.01), thrombocytopenia (<120,000/mm3) (p < 0.001), and coagulopathy (PT-INR ≥ 1.20) prior to drainage (p = 0.004). Multivariate logistic regression analysis identified thrombocytopenia [odds ratio (OR) 23.90; 95 % CI 2.64–216.18; p = 0.005] and positive blood culture [OR 9.11; 95 % CI 1.11–74.79; p = 0.040] as independent risk factors for septic shock.

Discussion

We have found that thrombocytopenia and positive blood culture were significantly associated with septic shock independently in acute obstructive pyelonephritis requiring emergency ureteral drainage.

Previous studies documented coagulopathy and thrombocytopenia as the predictors of progression of sepsis [9, 10]. This would be natural in considering the process of multiple organ dysfunction and septic shock [911]. Septic conditions stimulate the release of local and systemic proinflammatory mediators, which would result in low systemic vascular resistance and hypotension [12, 13]. These mediators also activate coagulation cascade and promote fibrin clot formation and platelet activation, inducing coagulopathy and thrombocytopenia [1214]. Our univariate analysis also demonstrated that thrombocytopenia and prolonged PT-INR were risk factors, although the latter did not remain as a significant factor in multivariate analysis. Thrombocytopenia may be clinically useful as a predictor of septic shock, since blood platelet count is a quick test.

Positive blood culture was marginally associated with septic shock in our study. Hsu et al. [15] reported that patients with complicated acute pyelonephritis with positive blood culture were likely to present severe sepsis or shock. However, the results of blood culture need at least 8 h, which may weaken the utility of blood culture in emergency situation [16, 17]. Moreover, in our study, blood culture was not taken correctly in 20 patients; in 14 cases, antibiotic therapy had been already begun before the referral to our institute, which might lead potential selection bias because these cases might have been under serious condition.

Yamamoto et al. reported that the interval from the occurrence of symptoms to drainage was significantly shorter in patients with septic shock in univariate analysis [14], which is similar to our results. Thus, the rapid progression may be an important risk factor but multivariate analysis in our study failed to show the significance.

Serum creatinine and bilirubin were not significant factors in our study, although they are included in the sequential organ failure assessment score [18], which was commonly used to evaluate the organ damage or predict outcomes in septic patients [18, 19]. This may be because serum creatinine level in patients with acute obstructive pyelonephritis, sometimes elevated by obstructive uropathy rather than SIRS, may not reflect the severity of sepsis. Serum bilirubin level was elevated in only 5 patients with shock in our study.

Comorbidities such as malignancy, diabetes mellitus, and steroid use were known to be associated with infectious diseases [20], although they were not necessarily risk factors for septic shock in acute obstructive pyelonephritis. Yoshimura et al. [21] reported that diabetes mellitus and immune suppression status were not associated with septic shock by examining 473 patients with urosepsis associated with upper urinary tract calculi. Similarly, Lee et al. [22] found that diabetes mellitus and malignancy were not related to septic shock in 208 bacteremic acute pyelonephritis patients. However, the reason for the lack of association was not discussed in these papers.

The weakness of our study is a retrospective analysis of a single institution with a limited number of cases. Confirmatory studies with a larger population may be required.

Conclusion

Our study revealed thrombocytopenia would be clinically more useful as a predictor of septic shock in the emergency room.

Conflict of interest

None declared.

Copyright information

© Springer Science+Business Media Dordrecht 2013