What your heart failure register is not telling you

Heart failure is recorded in two parts. An umbrella diagnosis code places a patient on the register, and a separate code says which kind of heart failure it is.¹ The second part is very often missing. A patient can carry the umbrella code alone, appear correctly on the register, and remain invisible to any search asking who has reduced ejection fraction. Work developing phenotype algorithms from UK records found a specific heart failure phenotype code in only around 18% of patients.²

That distinction now has consequences. From April 2026, HF009 replaces the separate ACE inhibitor and beta blocker indicators and asks instead what proportion of patients with a current diagnosis of heart failure with reduced ejection fraction are treated with all four pillars.³ It follows NICE NG106, updated in September 2025, which now recommends offering all four drug classes rather than following a fixed sequence in which one is optimised before the next is introduced.⁴ Much of the practice-facing commentary published earlier this year advised waiting for the v51 business rules before running searches. Those rules and the expanded cluster list were published on 17 July 2026, so the clusters driving the denominator can now be checked directly rather than inferred.⁵

The reason to act is not the indicator. In a linked cohort of 95,262 people with newly diagnosed heart failure in England, Nakao and colleagues found that those carrying an unspecified heart failure code, rather than a coded phenotype, were less likely to receive echocardiography, natriuretic peptide testing or specialist assessment, and less likely to be prescribed an ACE inhibitor, a beta blocker or a mineralocorticoid receptor antagonist. Hospitalisation and mortality were both higher in that group.⁶ The study is observational and cannot separate the code from everything the code stands in for, since vague coding often reflects a diagnosis that was never fully worked up. But at the point of care the effect is the same either way. If the phenotype is not in the record, no register, prompt or search can act on it.

The other half of the problem is the patients who are not there at all. A national English cohort covering 2000 to 2021 found that two thirds of people had symptoms suggestive of heart failure, or were taking a loop diuretic, on average two and a half years before diagnosis, and that almost half were first diagnosed as an inpatient. One year mortality reached 33% among patients diagnosed in hospital after long-term loop diuretic use without previous investigation in primary care.⁷ Earlier work using linked English data found that around four in five diagnoses were first recorded in hospital rather than in general practice.⁸

Audit recovers a meaningful share of these people. One English practice audit raised recorded prevalence from 0.63% to 1.12%, with inaccurate coding accounting for most of the patients who had been missing.⁹ A primary care heart failure service reviewing its caseload identified 2,916 additional patients requiring a left ventricular systolic dysfunction code, raising recorded prevalence from 0.7% to 1.05%.¹⁰ A systematic review of routinely collected data puts the same point more bluntly: coded records are usually right about the patients they contain, and miss roughly a third of true cases.¹¹

What those patients stand to gain is not marginal. In a comparative analysis of three randomised trials, Vaduganathan and colleagues estimated that comprehensive disease-modifying therapy could give a 55-year-old around 8.3 additional years free from cardiovascular death or first heart failure hospitalisation, compared with conventional ACE inhibitor or angiotensin receptor blocker plus beta blocker therapy.¹²That is a cross-trial estimate rather than a single trial result, though each component rests on its own randomised evidence.

The reverse error carries weight too. Preserved ejection fraction was once a distinction with little therapeutic consequence. Since EMPEROR-Preserved and DELIVER both met their primary composite endpoints in patients with ejection fraction above 40%, SGLT2 inhibition now has a role across the ejection fraction spectrum.^13,14 But phenotype still matters, because the September 2025 NG106 update recommends different treatment combinations according to ejection fraction: an ACE inhibitor, beta blocker, mineralocorticoid receptor antagonist and SGLT2 inhibitor in mildly reduced ejection fraction, and a mineralocorticoid receptor antagonist and SGLT2 inhibitor in preserved.⁴ Coding a patient as reduced when they are preserved, or the reverse, changes what they are offered.

The work is a search exercise before it is a prescribing one, and it starts with patients on the register carrying no ejection fraction code at all, then patients holding an ejection fraction code but no umbrella code, who are missing from the register entirely.¹ Most of the answers are already in the record. Echocardiogram reports sitting in filed correspondence rather than in structured fields are the richest single source, and repeat loop diuretic prescribing is the most reliable flag for someone who was never coded.

Two traps are worth naming. Patients whose ejection fraction has recovered stay on the register and stay on treatment. TRED-HF, a small open label randomised trial in recovered dilated cardiomyopathy, reported relapse in 44% of the withdrawal group within six months, against none of those continuing treatment.¹⁵ And where a patient is unsuitable for one pillar or declines it, that needs recording as a decision rather than appearing as an absence.

None of this reflects a failure of clinical care. It is the residue of coding conventions built over fifteen years to answer questions nobody is asking any more. HF009 is England only. The register problem is not.

References

1. Crundall A, Crawshaw-Ralli M, Fuat A, et al. Lessons learnt from HF coding in primary care. What might best practice look like? Br J Cardiol 2024;31:106-10. PMID 39917566

2. Sundaram V, Zakeri R, Witte K, Quint JK. Development of algorithms for determining heart failure with reduced and preserved ejection fraction using nationwide electronic healthcare records in the UK. Open Heart 2022;9:e002142. doi:10.1136/openhrt-2022-002142

3. NHS England. Quality and Outcomes Framework guidance for 2026/27. PRN02356. London: NHS England, 2026. england.nhs.uk

4. National Institute for Health and Care Excellence. Chronic heart failure in adults: diagnosis and management. NG106. Updated September 2025. nice.org.uk/guidance/ng106

5. NHS England Digital. Quality and Outcomes Framework business rules v51 2026. Published 17 July 2026. digital.nhs.uk

6. Nakao Y, Nakao K, Nadarajah R, et al. Prognosis, characteristics, and provision of care for patients with the unspecified heart failure electronic health record phenotype: a population-based linked cohort study of 95262 individuals. eClinicalMedicine 2023;63:102164. doi:10.1016/j.eclinm.2023.102164

7. Lawson CA, Ali M, McCann G, et al. Health inequalities and trends in heart failure diagnosis in primary care in England, 2000-21: a national retrospective cohort data-linkage study. Lancet Prim Care 2025;1:100060. doi:10.1016/j.lanprc.2025.100060

8. Bottle A, Kim D, Aylin P, et al. Routes to diagnosis of heart failure: observational study using linked data in England. Heart 2017;104:600-5. doi:10.1136/heartjnl-2017-312183

9. Cuthbert JJ, Gopal J, Crundall-Goode A, Clark AL. Are there patients missing from community heart failure registers? An audit of clinical practice. Eur J Prev Cardiol 2019;26:291-8. doi:10.1177/2047487318810839

10. Kahn M, Grayson AD, Chaggar PS, et al. Primary care heart failure service identifies a missed cohort of heart failure patients with reduced ejection fraction. Eur Heart J 2022;43:405-12. doi:10.1093/eurheartj/ehab629

11. Goonasekera M, Offer A, Karsan W, et al. Accuracy of heart failure ascertainment using routinely collected healthcare data: a systematic review and meta-analysis. Syst Rev 2024;13:79. doi:10.1186/s13643-024-02477-5

12. Vaduganathan M, Claggett BL, Jhund PS, et al. Estimating lifetime benefits of comprehensive disease-modifying pharmacological therapies in patients with heart failure with reduced ejection fraction: a comparative analysis of three randomised controlled trials. Lancet 2020;396:121-8. doi:10.1016/S0140-6736(20)30748-0

13. Anker SD, Butler J, Filippatos G, et al. Empagliflozin in heart failure with a preserved ejection fraction. N Engl J Med 2021;385:1451-61. doi:10.1056/NEJMoa2107038

14. Solomon SD, McMurray JJV, Claggett B, et al. Dapagliflozin in heart failure with mildly reduced or preserved ejection fraction. N Engl J Med 2022;387:1089-98. doi:10.1056/NEJMoa2206286

15. Halliday BP, Wassall R, Lota AS, et al. Withdrawal of pharmacological treatment for heart failure in patients with recovered dilated cardiomyopathy (TRED-HF): an open-label, pilot, randomised trial. Lancet 2019;393:61-73. doi:10.1016/S0140-6736(18)32484-X

Intended for UK healthcare professionals only. This article reports published guidance and study data and does not constitute prescribing advice.