Economic issues in the delivery and organisation of services for early detection of cancer

Abstract

Earlier detection and diagnosis of cancer is a key goal for health care systems internationally due to evidence of improved patient outcomes associated with cancers diagnosed at an earlier stage. Much of the current health economics literature focuses on evaluating the efficiency of early detection initiatives, such as screening, for asymptomatic patients. This thesis aims to provide evidence on the organisation and delivery of services for symptomatic patients with suspected cancer, particularly the system-level issues of waiting times and provider payment methods. I examine the relationship between waiting times and General Practitioner (GP) demand for an urgent specialist cancer visit using annual nationally representative data on practices in England referring to 185 hospital Trusts between April 2012 and March 2018. Waiting times at local hospitals were not associated with urgent suspected cancer GP referrals once practice fixed effects were included. The results from this chapter suggest that the demand for diagnostic cancer services is inelastic to waiting times. The validity of the approach in this chapter relies on the assumption that local GP practice referrals do not have a significant impact on the waiting times at local hospitals. I re-examine the relationship between waiting times and GP demand, using an exogenous change to cancer waiting times policy. I examine the impact of a pilot programme in five English hospitals to reduce waiting times for urgent diagnosis of suspected cancer. I applied difference-in-differences methods to two cancer types separately (bowel and lung cancer). The proportion of referred patients breaching two-week waiting times targets for suspected bowel cancer fell by 3.9 percentage points in pilot hospitals from a baseline of 4.8%, and GP referrals increased by 10.8%. The pilot was not successful for lung cancer; with some evidence that waiting times increased, and referrals reduced by -10.5%. This suggests family doctor referrals for suspected cancer are responsive to waiting times at the margin. I examine a change in methods of hospital payment for imaging scans. These services were "unbundled" from the national outpatient tariff in England in 2013 to encourage their use. First, I examine the effect of payment unbundling on utilisation using difference-in-differences methods applied to data from all hospitals in England. Hospitals recorded substantial increases in scan use of 134%. The unbundling increased annual national provider payments by £21.7 million, of which £15.8 million was due to the behavioural response to the reform. However, 15% of scans recorded after the unbundling were at providers that previously recorded no scans, suggesting some of the observed increase may reflect previous under-coding of activity. Finally, I examine the effect of this payment unbundling on patient outcomes. I apply interrupted time series and continuous difference-in-differences methods to data on 4.7 million suspected cancer patients. I found evidence of small reductions in patient mortality, emergency cancer admissions and cancer bed days (elective and emergency), and days until first admission. The improvements to patient outcomes suggest some of the increase in recorded utilisation was real and was beneficial. This thesis shows that GP demand responds to waiting times and hospitals respond to increased reimbursements in the diagnostic process for suspected cancer patients. It is important to consider demand and supply factors within healthcare systems, and how these interact will influence the effectiveness of new early detection of cancer initiatives.

Date of Award	1 Aug 2024
Original language	English
Awarding Institution	The University of Manchester
Supervisor	David Shackley (Supervisor), Matthew Sutton (Supervisor), Rachel Meacock (Supervisor) & Jonathan Stokes (Supervisor)