These 3 presentations address operational issues of paramount importance within the healthcare industry with a view to using statistics for the benefit of patients.
Topic: Model-informed decision support for planning and execution of clinical trials.
Suresh Ankolekar, Cytel
The planning and execution of clinical trials involve several decisions. What should be the target duration for the trial? What should be the enrolment levels and limits per regions/countries? How many enrolment sites per regions/countries? When do we open/close the sites? When do we plan the interim analyses? How do we optimise the enrolment plan? How do we optimise the drug supply? How do we re-optimise at interim stages? These are hard questions that are often addressed rather arbitrarily using ordinary spreadsheets. Decision-makers need to look beyond the averages and quantify uncertainty to be able to manage it. In this talk we will focus on statistical models to support enrolment planning. We will also talk about core simulation models used to optimise the drug supply. The emphasis will be on the simplicity of the models that can be easily implemented, even in spreadsheets.

Topic: Recruitment modelling on in flight studies.
Kirstie McKay, Covance
How modelling is used in Covance to support study teams make patient and study delivery decision on in flight studies with case study examples.  How do we address recruitment questions such as:
·      Assessing and comparing site recruitment performance?
·      If recruitment is not on track, what can we do?
·      What is the best way to identify sites to focus our attention on for recruitment? 
·      Once a study is underway and real recruitment data is available how do we assess if timelines are on track or that   specific study requirements may be met (target patient subgroup requirement)?  
Alongside patient recruitment we need to monitor and assess when an endpoint study may complete:
·       What if a patient is only assessed for endpoint once every 6 months – how do we assess if we’re tracking as expected? 
·       Or what if the endpoints are not tracking as expected how do we use the observed endpoint data to re-forecast the study end date?
David Worthington, Lancaster University
The NHS has a maximum waiting time target from referral to treatment (RTT) of 18 weeks, and hospital trusts that do not meet this target can face financial penalties. Hospital trusts that fear that they are in danger of exceeding this target for some of their patients typically review their outpatient and inpatient capacities for the coming months in order to decide whether extra capacity is needed, and if so where it is needed.
Important factors to consider in undertaking this exercise are the numbers and types of appointments that patients need between referral and treatment, the numbers of patients currently partway through their pathway, the numbers of new referrals expected to arrive and the capacities of the various clinics to see patients. And if current capacities do not seem sufficient to meet the target, what revised capacities are required?
This talk will describe work undertaken to develop a capacity planning tool for a Hospital Trust. It is a spreadsheet-based ‘what if …. ?’ model which first applies simple stocks and flows principles to calculate clinic throughputs, from which future patient waiting times can then be estimated.