STEP Research

STEP is a large scale program of research around the use of diagnostic and screening tests in clinical and patient-based decision making and health policy.

It includes the projects listed below.

Fever Study (Febrile EValuation of children in the Emergency Room)

Fever is a very common symptom in childhood illnesses. Discriminating which febrile illnesses are serious and which are self limiting is difficult. For most serious bacterial diseases in children the definitive test results are not available immediately leading to the dilemma of failure or delay in treating serious illness or treating unnecessarily.

Traditional diagnostic decision making does not compile clinical information optimally. Ideally, the clinician should know how likely different diseases are in febrile children, how accurate near patient tests are and how influential each component of history and examination are.

This information has been collected and analysed through follow-up of approximately 16,000 febrile children, less than 5 years old, who attended the Emergency Department of The Children’s Hospital, Westmead, between July 2004 and June 2006 (Phase 1 of the study).

These data have informed a diagnostic algorithm that calculates the likelihood of serious bacterial infection in individual patients and displays it to the treating physician using interactive software. The new method was compared with the usual diagnostic processes using a cluster randomised controlled trial. The primary outcome of the trial was appropriate antibiotic treatment with secondary outcomes of patient care, such as testing, admission and duration of fever. During the trial emergency department doctors were asked to complete a survey on their attitudes and confidence in diagnosing a febrile child as well as their opinions on the intervention. Analysis of the trial is nearing completion.

Contact: Dr Gabrielle Williams, Senior Research Officer (STEP)

Colorectal Research in Symptom Prediction (CRISP)

Over 8000 people have recorded their bowel symptoms, using standardised questionnnaires. These data are being linked to colonoscopic finidngs to determine which bowel symptoms predict colorectal cancer. Publication is expected in 2008.

Contact: Les Irwig

Systematic Reviews of Diagnostic Test Accuracy

STEP has had a long term commitment to improving the methodology used for the systematic review of studies of diagnostic test accuracy, and STEP members have co-authored several chapters in the Cochrane handbook for diagnostic reviews.

We apply these methods in the systematic reviews conducted by STEP, and also provide leadership in the area through our involvement in Cochrane pilot reviews and the Cochrane Diagnostic and Screening Tests Methods Group.

Recent STEP systematic reviews include: liquid based versus conventional cytology, comparative accuracy of renal duplex ultrasound parameters for the diagnosis of renal artery stenosis, and the accuracy of MRI in breast cancer staging.

Contact: Petra Macaskill (

HPV testing Decision Aid for women with Minor Atypia on Pap smear (The IMAP Study)

This study will compare the psychosocial outcomes of different management strategies for women with a mildly abnormal Pap smear: (a) usual care (repeat Pap smear), (b) HPV testing, a new management proposed for this group, and (c) the choice of either management using a decision aid to support women to make their preferred choice. The study will assess psychosocial and quality of life outcomes and will model the effectiveness of each management strategy.

Contact: Kirsten McCaffery


Several projects are being done to conceptualise the issues in clinical monitoring and apply them to several areas, including cardiovascular disease, renal disease and osteoporosis. A book on the topic has just been published, with major contributions from STEP: Evidence-based Medical Monitoring. From principles to practice. Glasziou P, Irwig L, Aronson JK, (eds). Blackwell Publishing Inc, Massachusetts, January 2008. ISBN: 978-1-4051-5399-7 link to this book

Decision making in prostate cancer screening

This suite of studies examines a novel decision support system, AnnaLisa, and its acceptability and applicability to consumer and GP decision making concerning prostate cancer screening.

Contact: Glenn Salkeld (

Systematic review of criteria for assessing screening

A systematic review has been conducted of publications and web-based material concerning criteria for assessment of screening. This is part of a wider project that is analyzing the development of evidence-based policy for screening, using colorectal cancer screening in Australia as the lead case study.

Contact: Kathy Flitcroft (

Community preferences for prostate cancer screening

These studies will use discrete choice experiments to investigate men’s preferences for prostate cancer screening, the relative importance of various aspects of screening, and how men are willing to trade-off these aspects against each other.

Contact: Kirsten Howard

Screening for renal risk factors

A series of population based models has been developed to examine the cost effectiveness of screening for risk factors for chronic kidney disease including diabetes, hypertension and proteinuria.

Contact: Kirsten Howard

Cancer Screening in patients with renal compromise

Clinical effectiveness, costs and cost effectiveness of cancer screening (breast cancer, colorectal cancer and cervical cancer) in dialysis and transplant recipients will be assessed.

Contact: Germaine Wong