Implementation of a clinical decision rule to optimize hospital admission policy of patients with febrile urinary tract infection

In daily clinical practice the risk of a complicated course and thereby need
for clinical observation and hospital-based treatment for febrile urinary tract
infection (FUTI) is assessed on basis of history, underlying disease and on
severity of local and vital signs. Fever is a sign with little specificity, and
may reflect the mere presence of local kidney infection or of impending
urosepsis. Prompt recognition and start of preferably intravenous antibiotic
treatment of the latter condition is of utmost importance, to prevent
progression to septic shock and death.
Currently, in the Netherlands about 90% of patients presenting at emergency
departments with febrile UTI will be admitted, because of a low chance of
life-threatening complications that cannot be reliably estimated. Clinical
tools helping to classify risks in patients with febrile UTI are urgently
needed and should be helpful to better identify those who benefit from hospital
admission, and those who can be safely treated at home.
We recently derived and validated a clinical prediction rule that adequately
predicts a complicated outcome or death in patients with FUTI. This prediction
rule, called the Prediction Rule for Admission policy in Complicated urinary
Tract InfeCtion LEiden (PRACTICE), is a bedside clinical score which allocates
points to readily available variables including age, sex, co-morbidity and
vital signs.
There is sufficient evidence that young non-pregnant women without co-morbidity
are good candidates for outpatient treatment. In previous studies, we
demonstrated that selected elderly, men and patients with co-morbidity might
also be eligible for outpatient home-based treatment. Previous studies indicate
that the costs of FUTI are most sensitive to hospitalization rates.
We assume that the PRACTICE is a good bedside clinical tool to identify
low-risk patients with FUTI who can be managed as outpatients. Introduction of
the PRACTICE into clinical practice may thus lead to lower hospitalization
rates without compromising clinical outcome, and lower medical costs. This
study, which is unique with regard to its subject and population, will evaluate
this hypothesis. Known and recently identified biomarkers will be assessed to
correlate to the success of application of the PRACTICE in the future.

To validate a clinical decision rule to guide need for hospitalization in
patients presenting with febrile urinary tract infection (FUTI) with the aim to
reduce the hospitalization rate without compromising clinical outcome.

A stepped wedge cluster-randomized trial. All the participating centers will
start with a 3-month control period. Thereafter, every 3 months one of the
centers will start with the intervention period until the end of the trial. In
the end, all study sites have received the intervention.

Use of the Prediction Rule for Admission policy in Complicated urinary Tract
InfeCtion LEiden (PRACTICE) to help guide physicians. The PRACTICE-score can be
divided into several risk classes.

According to risk class the following recommendations will apply:
< 75 points: strongly recommendation towards home based management
75-100 points: consider home based management
>100 points: strongly recommendation towards hospital admission

For more details about the PRACTICE, see paragraph 5.1

The burden of study procedures is limited. Besides the clinical decision rule
to determine which patient to admit and which patient to refer safely back to
primary care, the diagnostic and therapeutic approach to the patient will be
the standard clinical practice in all study centers. Subjects are contacted
twice by phone and twice in person by a home visit to evaluate their medical
condition. One extra urine culture will be performed and two additional blood
and urine samples will be taken. Subjects will be asked to fill a short
standardized satisfaction questionnaire.
Implementation of the clinical decision rule by definition holds a risk that
certain patients will initially be managed as outpatients, while at a later
stage, hospitalization is deemed mandatory. Based on own experiences with
home-based treatment, it is not expected that this may alter patients* clinical
outcome. Furthermore, the risk of unnecessary hospitalization (e.g.
hospital-acquired infections and * antimicrobial resistance) should not be
underestimated.