Variation in Direct Access to Tests to Investigate Cancer: A Survey of English General Practitioners

Background: The 2015 NICE guidelines for suspected cancer recommend that English General Practitioners have direct access to diagnostic tests to investigate symptoms of cancer that do not meet the criteria for urgent referral. We aimed to identify the proportion of GPs in England with direct access to these tests. Methods: We recruited 533 English GPs through a national clinical research network to complete an online survey about direct access to laboratory, radiology, and endoscopy tests in the three months leading up to the release of the 2015 NICE guidance. If they had direct access to a diagnostic test, GPs were asked about the time necessary to arrange a test and receive a report. Results are reported by NHS sub-region and, adjusting for sampling, for England as a whole. Results: Almost all GPs reported direct access to x-ray and laboratory investigations except faecal occult blood testing (54%, 95% CI 49–59%) and urine protein electrophoresis (89%, 95% CI 84–92%). Fewer GPs had direct access to CT scans (54%, 95% CI 49–59%) or endoscopy (colonoscopy 32%, 95% CI 28–37%; gastroscopy 72%, 95% CI 67–77%). There was significant variation in direct access between NHS regions for the majority of imaging tests—for example, from 20 to 85% to MRI. Apart from x-ray, very few GPs (1–22%) could access radiology and endoscopy within the timescales recommended by NICE. The modal request to test time was 2–4 weeks for routine radiology and 4–6 weeks for routine endoscopy with results taking another 1–2 weeks. Conclusion: At the time that the 2015 NICE guideline was released, local investment was required to not only provide direct access but also reduce the interval between request and test and speed up reporting. Further research using our data as a benchmark is now required to identify whether local improvements in direct access have been achieved in response to the NICE targets. If alternative approaches to test access are to be proposed they must be piloted comprehensively and underpinned by robust effectiveness data.