Harnessing the power of intelligent machines to enhance primary care

Moore et al have recently published an editorial on ‘Harnessing the power of intelligent machines to enhance primary care’.¹ Whilst we appreciate the main messages of the editorial, we would like to correct the impression given by the authors in relation to one of our studies QRISK3.²  As stated in our QRISK3 paper,² we included seven additional parameters in QRISK3 in response to the 2014 update to the NICE guidelines and published literature.^3-8 These were chronic kidney disease (stages 3, 4, 5); a measure of systolic blood pressure variability; migraine; corticosteroids; SLE; atypical antipsychotics; severe mental illness.

The paper by Weng et al highlighted only identified two of these,⁹ and was published the year after our paper was submitted to the BMJ.¹⁰ It’s inaccurate therefore to attribute the development of QRISK3 to machine learning techniques. Also for clarification the improvement in accuracy of 3.6% referred to from the paper by Weng et al was not in comparison to QRISK2 but to the published equations in the 2013 American ACC/AHA guidelines. Their highest c statistic value of 0.764 from the machine-learning algorithms was lower than ours for either the current version of QRISK2 (0.879 in women and 0.858 in men) or the new QRISK3 (0.880 in women and 0.858 in men).

References

1. Moore SF, Hamilton W, Llewellyn DJ. Harnessing the power of intelligent machines to enhance primary care. Br J Gen Pract 2018;68(666):6-7. doi: 10.3399/bjgp17X693965.
2. Hippisley-Cox J, Coupland C, Brindle P. Development and validation of QRISK3 risk prediction algorithms to estimate future risk of cardiovascular disease: prospective cohort study. BMJ 2017;357 doi: 10.1136/bmj.j2099.
3. National Clinical Guideline Centre. Lipid modification: cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease. London, 2014:286.
4. Thompson IM, Tangen CM, Goodman PJ, et al. Erectile dysfunction and subsequent cardiovascular disease. JAMA 2005;294(23):2996-3002. doi: 10.1001/jama.294.23.2996.
5. Vlachopoulos CV, Terentes-Printzios DG, Ioakeimidis NK, et al. Prediction of Cardiovascular Events and All-Cause Mortality With Erectile Dysfunction: A Systematic Review and Meta-Analysis of Cohort Studies. Circ Cardiovasc Qual Outcomes 2013;6(1):99-109. doi: 10.1161/circoutcomes.112.966903.
6. Shamloul R, Ghanem H. Erectile dysfunction. Lancet; 381(9861):153-65. doi: 10.1016/S0140-6736(12)60520-0.
7. Kurth T, Winter AC, Eliassen AH, et al. Migraine and risk of cardiovascular disease in women: prospective cohort study. BMJ 2016;353 doi: 10.1136/bmj.i2610.
8. Rothwell PM, Howard SC, Dolan E, et al. Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet 2010; 375(9718):895-905. doi: 10.1016/S0140-6736(10)60308-X.
9. Weng SF, Reps J, Kai J, et al. Can machine-learning improve cardiovascular risk prediction using routine clinical data? PLOS ONE 2017;12(4):e0174944. doi: 10.1371/journal.pone.0174944.
10. QRISK3 Peer review. http://www.bmj.com/content/357/bmj.j2099/peer-review BMJ 2016.