european urology 51 (2007) 1394–1401

available at www.sciencedirect.com
journal homepage: www.europeanurology.com

Infections

A Single Procalcitonin Level Does Not Predict Adverse

Outcomes of Women with Pyelonephritis

Virginie Lemiale a, Bertrand Renaud a,*, Stéphane Moutereau b, Alfred N’Gako a,

Mirna Salloum a, Marie Jeanne Calmettes a, Jérôme Hervé a, Cyril Boraud a,
Aline Santin a, Jean-Claude Grégo a, François Braconnier b, Eric Roupie a
a
Emergency Department of Henri Mondor Hospital (AP-HP), Créteil, France
b
Biochimie Henri Mondor Hospital, Créteil, France

Article info Abstract

Article history: Objectives: Predicting medical outcomes for pyelonephritis in women is

Accepted December 11, 2006 difficult, leading to unnecessary hospitalization. Unlike other serious
Published online ahead of infectious diseases, high procalcitonin (PCT) level has never been associated
print on December 18, 2006 with 28-d adverse medical outcomes in women with pyelonephritis. There-
fore, we sought to determine the accuracy of PCT in discriminating between
Keywords: pyelonephritis with adverse medical outcome (PAMO) and pyelonephritis
Pyelonephritis without adverse medical outcome (PWAMO).
Procalcitonin Patients and methods: Adult women with pyelonephritis presenting to the
Prediction emergency department of a French tertiary care hospital were consecutively
Outcome included. Those patients who developed adverse medical outcomes during a
Emergency medicine 28-d follow-up period were identified as having PAMO. Baseline character-
Hospitalization istics and PCT level were compared between patients with PAMO and
PWAMO.
Results: Eleven women (19.0%) had PAMO and 47 (81%) had PWAMO. The
median PCT level was higher in PAMO compared with PWAMO 0.51 ng/ml
(IQR: 0.04–3.8) and 0.08 ng/ml (IQR: 0.01–1.0), but this difference was not
statistically significant ( p = 0.07). We failed to find a threshold value for PCT
that discriminated between PAMO and PWAMO (ROC, AUC = 0.67 [95%CI,
0.51–0.86]). All but one subject with PAMO had either a PCT level >0.1 ng/ml
or an underlying genitourinary abnormality by radiographic testing.
Conclusions: A single PCT level was a poor predictor of 28-d adverse
medical outcomes in women with pyelonephritis treated in the emergency
department. Prediction based on underlying genitourinary abnormality by
radiographic testing in addition to the PCT level should be investigated in
future studies.
# 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved.

* Corresponding author. Service d’Urgences, Centre Hospitalier Universitaire Henri Mondor

(AP-HP), 51, avenue du Maréchal Delattre de Tassigny, 94010 Créteil Cedex, France.
Tel. +33 1 49 81 24 82; Fax: +33 1 49 81 29 87.
E-mail address: bertrand.renaud@hmn.ap-hop-paris.fr (B. Renaud).
0302-2838/$ – see back matter # 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2006.12.018
 european urology 51 (2007) 1394–1401
1. Introduction
Patients with acute pyelonephritis are often treated
in emergency departments (EDs) [1]. At least 10% of
patients treated in EDs experience a complicated
course that requires hospitalization [1,2]. Despite
the establishment of objective criteria for admission
based on comorbid conditions and illness severity,
low-risk patients who could be safely managed as
outpatients are often admitted [1,3,4]. Furthermore,
implementation of an algorithm applying these
guidelines failed to increase the proportion of low-
risk patients treated in the outpatient setting [3].
Therefore, incorporating a biomarker that aids in
determining severity of illness might assist emer-
gency physicians in decisions regarding hospital
admission.
One such biomarker is procalcitonin (PCT), which
is a 116-amino acid prohormone of calcitonin
comprised of three constituent peptides: a 57-amino
acid sequence at the amino terminus (NProCT); the
centrally positioned immature CT that contains a
terminal glycine; and a 21-amino acid CT carboxy-
terminus peptide I (CCP-I). Procalcitonin is a pro-
hormone of calcitonin produced by the thyroid
gland. Extrathyroidal production of procalcitonin
occurs during bacterial infections and increases the
blood procalcitonin level [4]. In healthy volunteers
exposed to an intravenous administration of Escher-
ichia coli endotoxin, the blood procalcitonin level
increased during the first 3 h after injection and
returned to normal within 48 h [5]. As a result,
procalcitonin is frequently used as a marker of
bacterial infection in pediatric and intensive care
settings [6–12]. Studies of its behaviour in patients
with bacterial sepsis have led to the proposal that it
may be a useful marker of systemic bacterial
infection, with greater specificity and sensitivity
than acute phase proteins such as C-reactive protein
[13–15].
Adult patients hospitalized for acute pyelone-
phritis have higher levels of procalcitonin compared
with control patients without bacterial infection
[16]. Moreover, procalcitonin has been suggested as
a useful marker of severe bacterial infections among
adult patients with other infectious diseases diag-
nosed in the ED [10,17]. For children with pyelone-
phritis, PCT level is a marker of kidney damage [7].
Therefore, there is indirect evidence that procalci-
tonin might be a useful indicator of disease severity
in adults who are diagnosed with acute pyelone-
phritis in the ED.
We sought to determine the discriminatory
power and predictive accuracy of the procalci-
tonin level for adverse outcomes in patients with
1395
pyelonephritis managed in the ED. On the basis
of results of previous studies, we hypothesized
that low procalcitonin levels would identify a low-
risk group that could be safely managed as out-
patients.
2. Methods
2.1. Design
From May 2003 through April 2004, we conducted a prospec-
tive study of adults diagnosed with acute pyelonephritis in the
ED of a French tertiary care hospital, with an annual volume of
44,000 visits per year. The institutional review board of our
hospital approved the study protocol and the procedures for
patient informed consent.
2.2. Patients
2.2.1. Eligibility
Emergency department physicians screened the women with
a suspected diagnosis of pyelonephritis for enrollment. We
enrolled patients on weekdays, from 8:00 AM to 6:00 PM during
the 12-mo study period. On a daily basis, the main
investigator (V.L.) reviewed the medical records of the
patients treated in our ED during the previous 24 h to ensure
that all patients with suspected pyelonephritis were pro-
posed for study enrollment.
2.2.2. Inclusion criteria
Inclusion criteria were age 18 yr; temperature 38.5 8C;
positive dipstick test for nitrite reaction and urine leucocytes
activity, subsequently confirmed by urinalysis with >105 white
blood cells (WBCs) per milliliter and >104 colony-forming units
of bacteria per milliliter; acute onset of at least one of the
following signs or symptoms suggestive of pyelonephritis:
dysuria, nausea, flank pain, costovertebral angle tenderness
on physical examination; and provision of informed consent
to participate in the study [18].
2.2.3. Exclusion criteria
Exclusion criteria were pregnancy or lactation, evidence of
additional source of infection, history of lung or thyroid
neoplasia, presence of an indwelling urinary catheter or
nephrostomy, antibiotic administration within 7 d prior to
emergency department presentation, or <105 WBCs per
milliliter or <104 colony-forming units of bacteria per milliliter
at urine culture.
2.3. Baseline data collection and procedures
Baseline characteristics were recorded at presentation by ED
physicians through patient interviews. Clinical variables
included comorbid illnesses, symptoms, physical examina-
tion findings, laboratory tests, radiographic findings, micro-
biologic analyses, antibiotic regimen, and the initial site of
treatment. Emergency physicians were blinded to procalcito-
nin levels when they ordered laboratory, imaging studies,
treatment, and site of care. Inpatients were defined as female
1396 european urology 51 (2007) 1394–1401
subject who were admitted at the time of their initial
presentation to the ED, unlike outpatients who were not.
The term subsequent hospitalization was restricted to out-
patients who were hospitalized after an initial management in
the outpatient setting.
2.4. Follow-up data collection and outcomes
Investigative team members were blinded to procalcitonin
levels when they collected study data. One of us (V.L.)
collected the follow-up data 28 d after initial presentation
by a standardized review of medical records and tele-
phone interviews with the patient, a relative, or a medical
provider. The data collected consisted of the treatment
regime, symptom resolution, urine culture results, antibiotic
sensitivity, changes in antibiotic therapy, and results of
laboratory, microbiologic, radiographic data, and medical
outcomes.
We defined pyelonephritis with adverse medical outcome
(PAMO) as pyelonephritis with at least one adverse outcome
occurring within the 28 d following patient presentation and
pyelonephritis without adverse medical outcome (PWAMO) as
those without adverse outcomes. Adverse outcomes were
defined as follows: (1) a perceived need for hospitalization:
presence of severe sepsis defined by the presence of
concomitant systemic inflammatory response and organ
dysfunction [19]; urgent urologic surgical procedures related
to pyelonephritis; evidence of renal abscess; admission to
intensive care; (2) subsequent hospitalization; and (3) pyelo-
nephritis-related death.
2.5. Management of cases
In agreement with French guidelines [20], several laboratory
tests (WBC count, creatinine, blood urea nitrogen [BUN]
glucose) and microbiologic analyses (urine examination,
urine culture, and blood culture with antibiotic sensitivity)
were systematically performed. For patients with severity
criteria at presentation (septic shock, severe sepsis, and
suspicion of urinary tract obstruction), a renal ultrasound
was performed at patient presentation and within the first
24 h of presentation for patients without severity criteria. A
renal computed tomography (CT) was performed if the
abdominal ultrasound revealed abnormal findings or did
not reach quality standards because of patient character-
istics. Additional laboratory and radiographic tests, along
with procedures and hospital admission decisions, were
performed at the discretion of the managing physicians.
Patients with risk factors for pyelonephritis with an adverse
medical outcome such as diabetes mellitus, history of urinary
tract disease (ureteral calculus, anatomic urinary tract
anomaly, or kidney graft), or immunosuppression were
treated as inpatients [3]. First-line antimicrobial treatment
for outpatients and inpatients without severity criteria
consisted of either a fluoroquinolone (ofloxacin 200 mg bid
or ciprofloxacin 500 mg bid) or, in case of fluoroquinolone
intolerance, a third-generation cephalosporin (cefotaxim 1 g
tid or ceftriaxone 2 g a day). Inpatients with severity criteria
were treated by adding an aminoglycoside (gentamicin) to the
first-line treatment [20].
2.6. Measurement of procalcitonin level
Blood samples obtained for procalcitonin levels at presenta-
tion were initially centrifuged, and stored at 80 8C within
4 h. After the completion of all study enrollment and 28-d
follow-up for all enrolled patients, procalcitonin levels were
measured for all the blood samples stored during a single
session with the use of an immunoluminometric assay
(Brahms Diagnostica, Berlin, Germany).
2.7. Statistical analysis
We based the estimate of the sample size on the following
assumptions: (1) mean ( standard deviation) procalcitonin
levels of 9.1 ng/ml (18.2) for PAMO and 1.3 ng/ml (2.7) for
PWAMO and (2) a PAMO:PWAMO patient ratio of 1:5 [10,21].
Eight patients with PAMO and 40 patients with PWAMO were
required for enrollment to achieve a power of 80% with a type I
error of 5%. We monitored the enrollment rate (on-time
discounting post hoc exclusions) and increased enrollment by
15% to ensure adequate power in the event of protocol
violation or loss to follow-up.
The characteristics of the patients were compared with
the use of chi-square statistics or Fisher exact tests for
categoric variables, and t test or Wilcoxon rank sum tests for
continuous variables. We assessed the diagnostic perfor-
mance of the assay by using the receiver operating
characteristic (ROC) curve, formed by plotting sensitivity
on the y-axis and 1 – specificity on the x-axis for all possible
cutoff values of procalcitonin. The area under the ROC curve
shows the overall discriminatory ability of the test to predict
adverse outcome. In addition to the area under the curve, we
identified the cutoff value that maximized sensitivity with-
out much loss of specificity, with the aim to maximize the
negative predictive value of the test, which is appropriate to
identify low-risk patients [22]. We also calculated likelihood
ratios (LRs) as a measure of the extent to which the pretest
odds are altered by the test results; low LR (<0.1) and high LR
(>10) are considered useful in ruling out and ruling in
disease, respectivaly [23]. Concentrations of procalcitonin
were not normally distributed and were therefore log
transformed before analysis. All analyses were managed
with the use of STATA version 8.0 (Stata, College Station, TX,
USA).
3. Results
Of the 60 patients initially enrolled in our study, 2
had an indwelling urinary catheter and were
excluded because of the post hoc discovery of an
exclusion criterion. Therefore, a total of 58 patients
were included in the analysis, 11 (19%) with PAMO,
and 47 (81.0%) with PWAMO.
In this population of women, the mean age was
38.4 (17.0) yr, 18 had a prior history of urinary tract
disease, 4 had diabetes mellitus, and 4 were
immunosuppressed (2 patients with a kidney graft,
2 patients were on oral corticoisteroid therapy). The
 european urology 51 (2007) 1394–1401 1397

Table 1 – Baseline characteristics of women with pyelonephritis without (PWAMO) and with adverse medical outcome
(PAMO)

Patient characteristics PWAMO (N = 47)* PAMO (N = 11) p valuey

Age 36.4 (15.0) 47.5 (21.5) 0.11

Comorbidities
History of urinary tract diseasez 16 (34.4%) 2 (18.2%) 0.31
Chronic renal failure 2 (4.3%) 0 0.48
Immunosuppressionz 2 (4.2%) 2 (18.2%) 0.10
Diabetes mellitusz 3 (6.4%) 1 (9.1%) 0.75

Clinical signs and symptoms

Length of symptoms (d) 3.1 (4.1) 4.2 (2.9) 0.11
Temperature (8C) 39.4 (0.7) 39.5 (0.7) 0.55
Pulse (per minute) 94 (18) 98 (16) 0.49
Systolic blood pressure (mm Hg) 124 (16) 128 (24) 0.49
Dysuria 31 (66.0%) 7 (63.6%) 0.88
Flank pain 41 (87.2%) 9 (81.8%) 0.60
Inability to tolerate oral intake 18 (38.3%) 2 (18.2%) 0.20

Laboratory values
White blood cell count (103/mm3) 12.4 (4.7) 17.3 (7.2) 0.03
Creatinine (mmol/L) 91 (51) 129 (65) 0.04
Blood urea nitrogen (mmol/L) 5.0 (3.3) 9.9 (8.6) 0.11
Procalcitonin (ng/ml) 1.4 (3.7) 35.9 (113.8) 0.07

Radiographic findings
Congenital urinary tract defectz 3 (6.4%) 0 1.00
Urinary tract dilation 2 (4.3%) 3 (27.3%) 0.03

Microbiologic characteristics
Urine culture
Escherichia coli 44 (93.6%) 11 (100%) 0.70
Proteus mirabilis 2 (4.1%) 0 —
Other bacteria 1 (2.0%) 0 —

Positive blood culture 5 (10.6%) 5 (45.5%) <0.01

*
Results are presented as number of cases (%) for qualitative characteristics, and as mean ( standard deviation) for quantitative values.
y
p value of statistics comparing PWAMO to PAMO.
z
Risk factors for PAMO.

median duration of symptoms before presentation
was 2 d (interquartile range [IQR]: 1–4); mean tem-
perature was 39.4 8C (0.7), 51 (87.9%) had flank pain,
and 38 (65.5%) patients had dysuria (Table 1).
Comorbid conditions, clinical signs, and symptoms
did not differ significantly between the two groups
with the exception of WBC count and creatinine
level, which were higher for PAMO. Procalcitonin
level varied widely within both outcome groups and
its distribution was skewed. The median value of
PCT was higher for PAMO patients compared with
PWAMO patients: 0.08 ng/ml (IQR: 0.01–1.0) and
0.51 ng/ml (IQR: 0.04–3.8) for PWAMO and PAMO,
respectively. However this difference was not
statistically significant ( p = 0.07; Fig. 1).
All but three cases of pyelonephritis were due to
Escherichia coli (two Proteus mirabilis and one Enter-
obacter sp); all complicated courses were caused by
Escherichia coli strains. We diagnosed 10 cases of
bacteremia due to Escherichia coli, 5 in the PAMO
group and 5 in the PWAMO group. The median
procalcitonin level was 3.22 ng/ml (IQR: 1.70–6.06)
for bacteremic patients and 0.06 (IQR: 0.01–0.45) for
nonbacteremic patients ( p < 0.01). Two patients
with PWAMO were infected with Escherichia coli
strains that were resistant to empirical antimicro-
bial treatment; none was bacteremic.
Adverse outcomes are shown in Tables 2 and 3.
Eleven patients had PAMO: 2 inpatients with
Escherichia coli pyelonephritis had septic shock (a
27-year-old nonimmunosuppressed woman and an
84 year-old woman with diabetes mellitus; without
prior history of urinary tract disease), 4 inpatients
had acute organ failure, 2 inpatients required
urologic surgery within the first 48 h for the
treatment of a urinary tract obstruction, 1 inpatient
was diagnosed with a renal abscess on day 4, and 2
outpatients were subsequently hospitalized on day 3
and 4 after presentation for intravenous analgesia.
Twenty-eight patients (48.3%) were initially
hospitalized: 9/11 (81.8%) patients with PAMO and
19/47 (40.4%) patients with PWAMO. Four women of
1398 european urology 51 (2007) 1394–1401

Table 2 – Medical outcomes during the 28-d follow-up for

women with pyelonephritis with (PAMO) and without
adverse medical outcome (PWAMO)

Patient characteristics PWAMO PAMO p

(N = 47)* (N = 11) valuey

Treatment
Inpatients 19 (40.4%) 9 (81.8%) <0.01
Length of stay (d) 2.0 (4.0) 6.3 (4.5) <0.01

Outcomes
Septic shock — (—) 2 (18.2%) —
Severe sepsis — (—) 4 (36.4%) —
Subsequent hospitalization — (—) 2 (18.2%) —
Urologic procedure — (—) 2 (18.2%) —
Kidney abscess — (—) 1 (9.1%) —
Fig. 1 – Procalcitonin (PCT) for women with pyelonephritis
without adverse medical outcome (PWAMO) and women Death — (—) 0 —
with pyelonephritis with adverse medical outcome *
Results are presented as number and percentage for qualitative
(PAMO).
characteristics, and as mean and standard deviation for
The median line within boxes figures the median value of quantitative values.
PCT level, while the upper hinge and the lower hinge y
p value of statistics comparing PWAMO with PAMO.
figure the 75th and the 25th percentile of the PCT level
values, respectively.

presentation (Table 3). Five of those 11 women with

the PWAMO group stayed more than 5 d in the
hospital: 3 patients had poor psychosocial condi-
tions, and 1 patient had a kidney graft. All but one
patient with pyelonephritis with bacteremia were
hospitalized. The overall mean length of stay was
5.4 d (4.9) and was longer for PAMO compared with
PWAMO: 6.3 (1.4) d versus 2.0 (0.6) d, respectively
( p < 0.01; Table 2).
Among the 11 PAMO female subjects with a PCT
level >1 ng/ml, 1 was immunosuppressed, and the
other 10 did not report any risk factor for pyelone-
phritis with adverse medical outcome. However, 1
subject had symptoms lasting for 10 d prior to ED
PWAMO were hospitalized at first visit. The overall
ability of procalcitonin to discriminate between
PAMO and PWAMO, as given by the area under
the ROC curve, was 0.67 (95%CI, 0.51–0.86; Fig. 2).
Because 3 patients with a PAMO had low levels of
procalcitonin (<0.1) we could not determine a useful
procalcitonin threshold (Table 4). Those patients
had no underlying disease, but 2 patients had
ureteral obstruction requiring intervention and 1
patient developed a renal abscess. As a result, all but
1 of the patients with a PAMO had a procalcitonin
level 0.1 ng/ml or abnormal findings on renal
ultrasound or CT (Table 3).

Table 3 – Adverse outcomes, duration of symptoms, and procalcitonin levels during the 28-d follow-up for women with
pyelonephritis with adverse medical outcome (PAMO) (N = 11)

Outcomes Duration of symptoms Delay prior to onset of Procalcitonin

prior to presentation* (d) adverse outcomey (d) (ng/ml)

Septic shock 3 1 379.0

Acute organ failure (renal) 4 1 6.1
Acute organ failure (renal) 1 1 3.8
Acute organ failure (disseminated intravascular 5 1 2.6
coagulation)
Acute organ failure (renal) 6 1 2.2
Septic shock 2 1 0.5
Hospitalization for persistent fever and pain 8 4 0.2
Hospitalization for persistent fever and pain 2 3 0.1
Urologic surgery 10 1 <0.1
Kidney abscess 1 4z <0.1
Urologic surgery 4 2 <0.1

*
Duration of symptom refers to the time lag between the onset of symptoms and presentation to the emergency department.
y
Delay from presentation to onset of adverse outcome.
z
Delay from presentation to diagnosis.
 european urology 51 (2007) 1394–1401 1399

Table 4 – Characteristics of procalcitonin thresholds to discriminate between pyelonephritis with adverse medical
outcome (PAMO) and pyelonephritis without adverse medical outcome (PWAMO)

Procalcitonin threshold Sensitivity Specificity Likelihood ratio Likelihood ratio

(ng/ml) (%) (%) positive negative

0.1 72.7 (61.3–84.2) 53.2 (40.3–66.0) 1.55 (0.97–2.49) 0.51 (0.19–1.40)

0.2 63.6 (51.3–76.0) 61.7 (49.2–74.2) 1.66 (0.93–2.95) 0.59 (0.26–1.33)
0.4 54.5 (41.7–67.4) 68.1 (56.1–80.1) 1.71 (0.86–3.38) 0.67 (0.34–1.31)
0.5 54.5 (41.7–67.4) 70.2 (58.4–82.0) 1.83 (0.91–3.67) 0.65 (0.33–1.27)
1.0 45.4 (32.6–58.3) 76.6 (65.7–87.5) 1.94 (0.85–4.45) 0.71 (0.41–1.23)

Consistent with previous studies regarding other

Fig. 2 – Receiver operator characteristic (ROC) curve of
procalcitonin in predicting pyelonephritis with adverse
medical outcome.
In comparison, the areas under the ROC curve for
WBC count and creatinine level were 0.71 and 0.75,
respectively. The differences between the areas
under the ROC curves for procalcitonin level and
WBC count, and for procalcitonin level and creati-
nine level were not statistically significant ( p = 0.73
and p = 0.42, respectively).
4. Discussion
In this study of women with pyelonephritis pre-
senting to a French tertiary hospital ED, we enrolled
58 patients, 11 of whom (19.0%) had an adverse
medical outcome. The PCT levels were not signifi-
cantly higher in patients with PAMO compared with
those with PWAMO, and the discriminatory power
of the PCT level to predict adverse outcomes was low
(area under the curve [AUC] = 0.67 [95%CI, 0.51–
0.86]). Then, we could not determine a useful
threshold that accurately discriminated between
the two groups. Our findings suggested that a single
PCT measurement is not useful for predicting
adverse medical outcomes among women present-
ing to the ED with acute pyelonephritis.
infectious diseases, the PCT values of patients with
adverse outcomes overlapped the PCT values of
those without adverse outcomes [10,24]. Two
patients with PAMO had an undetectable level of
PCT (<0.1 ng/ml) and ureteral obstruction that
required an urgent urologic procedure. Moreover,
two additional patients who experienced adverse
outcomes on day 4 (hospitalization for renal abscess
and for increasing pain) had a low PCT level (<0.1
and 0.2 ng/ml). Patients with PWAMO had a PCT
level as high as 21.2 ng/ml. This value was measured
in a renal transplant patient, a finding consistent
with prior studies demonstrating high PCT levels for
immunosuppressed patients with an infection [4].
Contrary to other studies that have found the PCT
useful in discriminating between severe and non-
severe infections in the ED, ours did not [24,25]. In
the PWAMO group, we found 17 patients with a PCT
level >0.2 ng/ml and 13 with a PCT level >0.6 ng/ml
(Table 4).
In contrast to prior studies that based severity
definition on risk factors for pyelonephritis with
adverse medical outcome, we defined pyelonephritis
with adverse medical outcome on the basis of the
presence of adverse medical outcomes (septic shock,
acute renal failure, subsequent hospitalization due to
pyelonephritis, urologic procedure, and intensive
care admission) [3,11]. We believe that by using these
outcome measures to define PAMO, we are able to
develop a better algorithm for predicting adverse
outcomes from pyelonephritis. In our study, five
immunosuppressed patients and three patients with
diabetes mellitus were hospitalized at first visit but
had favorable outcomes, while seven women without
risk factors experienced an adverse outcome. This
difference in defining pyelonephritis with adverse
medical outcome might account for the higher
proportion of pyelonephritis with adverse medical
outcome identified in our study (19.0% vs. 9–10%).
Unlike prior studies of other infectious diseases
encountered in the ED, our findings do not support
the use of the PCT level to predict the outcome of
women with acute pyelonephritis [24,25].
1400 european urology 51 (2007) 1394–1401
Our study had several limitations that must be
acknowledged. First, there is important uncertainty
related to the AUC value because of the small
number of patients enrolled in our study. Specifi-
cally, the upper boundary of the AUC reached 0.86, a
potentially interesting value for predicting out-
comes (Fig. 1). Therefore, additional studies are
needed to confirm our preliminary findings. Second,
duration of symptoms prior to presentation ranged
from 1 to 10 d. Such variation might have influenced
our results since the PCT level varies with time [6].
Indeed, some authors have suggested that the peak
of PCT within the first 3 d of sepsis predicts severity
of sepsis in patients admitted to the intensive care
unit [26,27]. However, this approach is not relevant
to guide the site of treatment decision in the ED
setting. Third, two patients with PAMO and a PCT
level <0.1 ng/mg had a ureteral obstruction and
required urologic surgery within the first 3 d.
Urinary obstruction is a mechanical complication
in nature and a risk factor for adverse outcomes;
however, unlike PCT it is not directly related to
sepsis severity. As a matter of fact, a test based on
PCT level and underlying genitourinary abnormal-
ities found on screening on abdominal ultrasound or
CT might be more useful to discriminate between
PAMO and PWAMO.
5. Conclusions
In summary, in this preliminary study of women
with acute bacterial pyelonephritis seen in the ED,
unlike other infectious diseases, we found that
the discriminatory power of the PCT level was low
with regard to predicting outcome. We were there-
fore unable to determine a PCT level that would be
useful in guiding ED decision making. Further
studies are needed to confirm these findings in a
larger population.
Conflicts of interest
The corresponding author certifies, on behalf of all
authors, that no manufacturer of a product discussed
in the manuscript had a role, either directly or
through a third party, in the gathering or preparation
of data or in the writing of the manuscript.
References
[1] Elkharrat D, et al. Prévalence et prise en charge des infec-
tions urinaires communautaires. Etude prospective dans
78 SAU français. JEUR Suppl 2004;17:45.
[2] Foxman B, Klemstine KL, Brown PD. Acute pyelonephritis
in US hospitals in 1997: hospitalization and in-hospital
mortality. Ann Epidemiol 2003;13:144–50.
[3] Elkharrat D, et al. Relevance in the emergency department
of a decisional algorithm for outpatient care of women
with acute pyelonephritis. Eur J Emerg Med 1999;6:15–20.
[4] Ferriere F. Procalcitonin, a new marker for bacterial infec-
tions. Ann Biol Clin (Paris) 2000;58:49–59.
[5] Dandona P, et al. Procalcitonin increase after endotoxin
injection in normal subjects. J Clin Endocrinol Metab
1994;79:1605–8.
[6] Meisner M, et al. Comparison of procalcitonin (PCT) and C-
reactive protein (CRP) plasma concentrations at different
SOFA scores during the course of sepsis and MODS. Crit
Care (Lond) 1999;3:45–50.
[7] Benador N, et al. Procalcitonin is a marker of severity
of renal lesions in pyelonephritis. Pediatrics 1998;102:
1422–5.
[8] Gendrel D, et al. Measurement of procalcitonin levels in
children with bacterial or viral meningitis. Clin Infect Dis
1997;24:1240–2.
[9] Gendrel D, et al. Comparison of procalcitonin with C-
reactive protein, interleukin 6 and interferon-alpha for
differentiation of bacterial vs. viral infections. Pediatr
Infect Dis J 1999;18:875–81.
[10] Suprin E, et al. Procalcitonin: a valuable indicator of
infection in a medical ICU? Intensive Care Med 2000;26:
1232–8.
[11] Pecile P, et al. Procalcitonin: a marker of severity of acute
pyelonephritis among children. Pediatrics 2004;114:
e249–54.
[12] Reith HB, et al. Procalcitonin in early detection of post-
operative complications. Dig Surg 1998;15:260–5.
[13] Whicher J, Bienvenu J, Monneret G. Procalcitonin as an
acute phase marker. Ann Clin Biochem 2001;38:483–93.
[14] Meisner M. Biomarkers of sepsis: clinically useful? Curr
Opin Crit Care 2005;11:473–80.
[15] Castelli GP, et al. Procalcitonin and C-reactive protein
during systemic inflammatory response syndrome, sep-
sis and organ dysfunction. Crit Care 2004;8:R234–42.
[16] Martinot M, et al. Procalcitonin in pyelonephritis and
acute community-acquired pneumonia in adults. Presse
Med 2001;30:1091–6.
[17] Chirouze C, et al. Low serum procalcitonin level accu-
rately predicts the absence of bacteremia in adult patients
with acute fever. Clin Infect Dis 2002;35:156–61.
[18] Miller 2nd O, Hemphill RR. Urinary tract infection and
pyelonephritis. Emerg Med Clin North Am 2001;19:655–74.
[19] Levy MM, et al. 2001 SCCM/ESICM/ACCP/ATS/SIS Interna-
tional Sepsis Definitions Conference. Intensive Care Med
2003;29:530–8.
[20] SPLF, antibiothérapie des infections urinaires; 2é confér-
ence de consensus en thérapeutique anti-infectieuse. 1990.
[21] Behr MA, et al. Fever duration in hospitalized acute pye-
lonephritis patients. Am J Med 1996;101:277–80.
[22] Irwig L, et al. Designing studies to ensure that estimates of
test accuracy are transferable. BMJ 2002;324:669–71.
[23] Knottnerus JA, van Weel C, Muris JW. Evaluation of diag-
nostic procedures. BMJ 2002;324:477–80.
 european urology 51 (2007) 1394–1401 1401

[24] Hausfater P, et al. Usefulness of procalcitonin as a
marker of systemic infection in emergency department
patients: a prospective study. Clin Infect Dis 2002;34:
895–901.
[25] Chan YL, et al. Procalcitonin as a marker of bacterial
infection in the emergency department: an observational
study. Crit Care 2004;8:R12–20.
[26] Wanner GA, et al. Relationship between procalcitonin
plasma levels and severity of injury, sepsis, organ failure,
and mortality in injured patients. Crit Care Med
2000;28:950–7.
[27] Luyt CE, et al. Procalcitonin kinetics as a prognostic mar-
ker of ventilator-associated pneumonia. Am J Respir Crit
Care Med 2005;171:48–53.