Start Date Jul 2017
UK primary-care referral guidance published by The National Institute for Health and Clinical Excellence (NICE) describes the signs, symptoms, and test results (“features”) of undiagnosed cancer. In 2015 (2011 for ovarian cancer), the NICE guidance was revised to liberalise investigation, by introducing more features representing a low-risk of undiagnosed cancer. To some extent, the revised NICE guidelines codified trends in clinical practice.
Our study used observational primary-care data from the Clinical Practice Research Datalink (CPRD) to explore trends over the period 2006–2017 in: (a) the time to diagnosis (clinical arm) and (b) diagnostic activity (health economics arm) for adults with cancer.
We hypothesised that times to diagnosis would reduce over time for participants presenting with features newly introduced during guidance revision, as evidence on these low-risk cancer features emerged and was translated and codified into practice by the revised guidance.
We explored trends in diagnostic activity over the period 2006–2017.
Aims & objectives
The aims of this research are to:
Explore changes over the period 2006-2017 in: (a) time to cancer diagnosis and (b) cancer diagnostic activity for 11 common cancer sites for which new features of undiagnosed cancer were introduced during guidance revision, to:
- Estimate the early impact of guidance revision on the time to diagnosis of these 10 cancer sites
- Explore the impact of guidance revision on diagnostic resource use within primary care and the costs associated with these
The CPRD provided electronic medical records for 83,935 adults with an incident diagnosis of myeloma or bladder, breast, colorectal, lung, oesophageal, ovarian, pancreatic, prostate, stomach or uterine cancer between 1 January 2006 and 31 December 2017. Participants were grouped by their first presenting (index) feature of their undiagnosed cancer in the year before diagnosis:
- “New-NICE” participants had index features that were introduced during guidance revision.
- “Old-NICE” participants had index features listed in the original 2005 NICE guidance.The CPRD provided electronic medical records for 83,935 adults with an incident diagnosis of myeloma or bladder, breast, colorectal, lung, oesophageal, ovarian, pancreatic, prostate, stomach or uterine cancer between 1 January 2006 and 31 December 2017. Participants were grouped by their first presenting (index) feature of their undiagnosed cancer in the year before diagnosis:
For analysis of time to diagnosis, the outcome variable was diagnostic interval (number of days from index feature to diagnosis). Semiparametric varying-coefficient analyses compared diagnostic intervals between New-NICE and Old-NICE groups over the period 1 January 2006 to 31 December 2017.
For resource use, diagnostic tests ordered in primary care will be identified, and trends in the quantity and costs will be explored over the period 1 January 2006 to 31 December 2017. Potential methods for identifying the impact of guidance revision include difference-in-difference analyses.
Outputs & impact
To date the outputs include:
- One methodological paper on application of the semiparametric varying-coefficient model to uncover trends in healthcare data. This work was presented to the CPRD journal club on 11 August 2020.
- Two clinical papers on trends in diagnostic interval between 2006 and 2017, and trends in lung cancer symptoms over time.
A further health economics publication on the impact of revised NICE guidance on diagnostic activity is planned.
Now that the clinical arm of this study is complete, the focus has moved to the health economics analyses, which are planned for completion in 2020.
- Price SJ, Zhang X, Spencer AE. Measuring the impact of national guidelines: What methods can be used to uncover time-varying effects for healthcare evaluations? Social Science and Medicine 2020; 258:113021 https://doi.org/10.1016/j.socscimed.2020.113021
- Price SJ, Spencer AE, Zhang X, Ball S, Lyratzopoulos G, Mujica-Mota R, Stapley S, Ukoumunne OC, Hamilton W. Trends in time to cancer diagnosis around the period of changing national guidance on referral of symptomatic patients: a serial cross-sectional study using UK electronic healthcare records from 2006–17. Cancer Epidemiology (in press)
- Chowienczyk S, Price S, Hamilton W. Changes in the presenting symptoms of lung cancer from 2000-2017: a serial cross-sectional study of observational records in UK primary care. British Journal of General Practice 2020;70(692):e193-e199 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6988682/