Heart Failure: The Evidence Behind the Pathway

A 54-year-old autorickshaw driver is admitted at night, breathless lying flat, with swollen ankles and crackles at both bases. He has had diabetes for ten years. His echocardiogram the next morning shows an ejection fraction of 30 per cent. He is warm to touch, and wet. Three days of intravenous furosemide later he is comfortable, and the question on the ward round is what he goes home on. He supports a wife and two children in school. This is an illustrative case, not a real patient.

The Heart Failure Pathway will tell you what to do for him: which drugs are missing, the starting and target dose of each, the numbers that stop or slow a drug, and how to dose the diuretic. What it no longer carries is most of the why. The Indian registry figures, two trials, a little history, and the source of every row of its dose tables have moved here, so that the tool stays quick at the bedside. By the end of this article you will know why the tool asks you to start four drugs before he leaves, and which document each of its numbers comes from.

1. One ejection fraction, three diagnoses

His ejection fraction of 30 per cent is the easy case: below 35 per cent, all three current documents call it reduced. The trouble starts between 35 and 49 per cent. Here is the table the tool printed in its Doses tab; on a phone it is now two narrow tables.

“Why not just follow the newest guideline?” The instinct is to pick the newest. The reason not to is that two of the three are Indian national documents, and one of them is the workflow a district hospital is expected to follow. A tool that quietly chose one would hide from you that the disagreement exists. So the tool prints all three and settles none of it, and you should know that a patient with an ejection fraction of 45 per cent will be called something different depending on which document is open.

2. Our patients are younger, and they die sooner

Why should the age of Indian patients change how fast you start treatment? Look at what the registries recorded.

A mean age of about 60 means that many patients are in their forties and fifties, like our driver. One patient in seven in the national registry was dead within three months of going home (Harikrishnan 2022). In Trivandrum a third had died at one year, most of them under 70, and at ten years 82.7 per cent were dead (Harikrishnan 2017; Harikrishnan 2026). These are working-age patients with families who depend on them, and that should shape how urgently the treatment is started, not only whether it is.

3. The gap is not knowledge, it is prescription

“Surely everyone knows these drugs by now?” Everyone knows them. Not everyone writes them. In Trivandrum, only 25.4 per cent of patients with a reduced ejection fraction left hospital on the three disease-modifying drugs then available. Ten years later, mortality among those who did was 75.9 per cent, against 88.6 per cent among those who did not, a hazard ratio of 0.71. A hazard ratio of 0.71 means that, at any point in those ten years, the risk of dying was about 29 per cent lower in those who went home on the drugs. The National Heart Failure Registry did better, at 47.5 per cent, and still half the eligible patients went home without treatment that changes survival.

Nothing in this pathway is new science. The drugs are known, available and cheap, and are not written on the chart. That is why the tool reports what is missing before it reports anything else.

4. The four pillars, and where every dose comes from

The four pillars are a renin-angiotensin blocker (an ACE inhibitor, an angiotensin receptor blocker or an ARNI, the angiotensin receptor-neprilysin inhibitor), a beta blocker, a mineralocorticoid receptor antagonist, and an SGLT2 inhibitor. Every dose in the table is printed as it appears in HFAI 2025, the guideline of the Heart Failure Association of India, except the two rows that come from the ICMR workflow. The tool now shows the drug, the starting dose and the target dose; here is the table with its source column.

Notice the sacubitril-valsartan row. HFAI 2025 prints no ARNI dose at all, so the dose comes from the ICMR workflow of January 2026. When the tool recommends an ARNI in a result, it says so beside the dose.

“Should I not get one drug to its target before adding the next?” That was the old way. The old sequence titrated an ACE inhibitor to target over months before the beta blocker was added, and by the time the third drug was due the patient had often been readmitted or lost to follow-up. HFAI 2025 now asks for all four to be started at whatever doses are possible, because even small doses have shown benefit. A quarter of the target dose of all four beats a full dose of one.

5. The discharge is the intervention

What happens after discharge decides whether the doses ever reach target. STRONG-HF randomised 1078 patients to usual care or to up-titration to full doses within two weeks of discharge, with four outpatient visits over two months. Readmission for heart failure or death by 180 days fell from 23.3 to 15.2 per cent, a risk ratio of 0.66, and the trial was stopped early (Mebazaa 2022).

The visits are as much the active ingredient as the doses: somebody measured the blood pressure, the potassium and the creatinine often enough to keep pushing. Where follow-up is uncertain, as it often is for a man who loses a day’s earnings to come to the clinic, do as much of the titration as possible before discharge, and write the first review date on the discharge summary instead of “come if unwell”.

6. The numbers that stop or slow a pillar

Every titration meets a potassium, a creatinine, a blood pressure or a heart rate. The tool keeps every threshold and what it means; here is the table with the source of each row. Read the source column with care: two of the starting conditions come only from the ICMR workflow, and HFAI 2025 prints no blood pressure threshold at all.

This is where the common mistake lives. A creatinine rises by 15 per cent after the ACE inhibitor is started, and the drug is stopped. HFAI 2025 says a rise of up to 20 per cent is expected and is not a reason to stop. The harm of stopping cannot be seen on the day, and it is large: years without a drug that keeps patients alive.

7. The diuretic, and the kidney

Our driver came in on no diuretic, so his first intravenous dose was taken from the tool’s range for a patient on no loop diuretic. Had he been on furosemide at home, the rule would have been twice his regular total daily oral dose, given intravenously.

“Bolus or infusion? And how much?” Both are acceptable. DOSE found no difference between them, and no significant symptom difference between high and low dose strategies, although the high dose gave more diuresis with a transient rise in creatinine that recovered (Felker 2011).

That rise in creatinine is worth understanding. A rising creatinine during successful decongestion is not automatically kidney injury. The congested kidney is a venous problem as much as an arterial one: raised pressure in the renal veins reduces filtration, so relieving the congestion can help the kidney.

Back to the ward

Our driver is warm, now dry, with a blood pressure of 112 systolic, a heart rate of 84, a potassium of 4.6 mEq/L and a creatinine of 1.1 mg/dL. Before he goes home he is started on dapagliflozin 10 mg and spironolactone 25 mg, then bisoprolol 1.25 mg, and then an ACE inhibitor at its starting dose, or sacubitril-valsartan 50 mg twice daily if it can be afforded. His potassium and creatinine are checked before each step. He leaves with a written review date two weeks away, and a note that each dose goes up every 2 weeks if the numbers allow. Each of those steps is a line in the tool. Now you know what stands behind each line.

Try this

A 62-year-old woman on enalapril 5 mg twice daily, bisoprolol 5 mg and spironolactone 25 mg comes for review. Her potassium is 5.3 mEq/L and her creatinine has gone from 1.2 to 1.4 mg/dL. Her resident suggests stopping the spironolactone. What do the tool’s numbers say? Illustrative case.

Answer. A potassium above 5.0 mEq/L calls for caution and a recheck before any increment; stopping is for a potassium above 5.5 mEq/L. A creatinine rise of about 17 per cent is within the expected 20 per cent. So hold the next increment, recheck, and look for the reversible cause: a potassium supplement, an over-the-counter painkiller, over-diuresis, or a salt substitute, which in India is very often potassium chloride. Do not stop the spironolactone on these numbers.

Take-home points

• Three current documents divide the ejection fraction differently. Below 35 per cent they agree; between 35 and 49 per cent, know which one you are using.
• Indian patients with heart failure are often of working age, and one in seven is dead within three months of discharge.
• The gap is prescription. Start all four pillars, at whatever dose is possible, before the patient goes home.
• Fix the first review date in writing. The visits are what get the doses to target.
• A creatinine rise of up to 20 per cent is expected. A potassium above 5.0 mEq/L is a caution; above 5.5 mEq/L, stop.

For patients and families, the author’s own explainer Heart Attack, Cardiac Arrest, Heart Failure: Are They All the Same? is written for them. For the pharmacology of one of the four pillars, see Beta-Blockers: Clinical Pharmacology & Practice Pearls.

Declaration: the author built and maintains the tool. It is free to use. The figures in this article are those the tool carried in its Doses and Evidence tabs on 4 November 2026, each from the source named beside it. The Heart Failure Pathway is still in administrator preview, pending clinical review.

References

1. Chopra VK, Sengupta S, Sastry BKS, et al. 2025 HFAI guidelines for diagnosis and management of heart failure. Heart Failure Journal of India. 2025;3(Suppl 1):S1-S48.
2. Indian Council of Medical Research, Department of Health Research. Standard Treatment Workflow: Heart Failure, a Breathless Patient (ICD-11 BD10-BD1Z). January 2026.
3. Kober L, Adamo M, et al. 2026 ESC guidelines for the management of heart failure. Eur Heart J. 2026. doi:10.1093/eurheartj/ehag100. Published 28 August 2026.
4. Guha S, Harikrishnan S, Ray S, et al. CSI position statement on management of heart failure in India. Indian Heart J. 2018;70(Suppl 1):S1-S72.
5. Harikrishnan S, Sanjay G, Agarwal A, et al. One-year mortality outcomes and hospital readmissions of patients admitted with acute heart failure: data from the Trivandrum Heart Failure Registry in Kerala, India. Am Heart J. 2017;189:193-199.
6. Harikrishnan S, Ganapathi S, Huffman MD, et al. Long-term prognosis in heart failure: findings from the ten-year follow-up of the Trivandrum registry. J Card Fail. 2026. doi:10.1016/j.cardfail.2026.07.028.
7. Harikrishnan S, Bahl A, Roy A, et al. Clinical profile and 90 day outcomes of 10 851 heart failure patients across India: National Heart Failure Registry. ESC Heart Fail. 2022;9(6):3898-3908.
8. Mebazaa A, Davison B, Chioncel O, et al. Safety, tolerability and efficacy of up-titration of guideline-directed medical therapies for acute heart failure (STRONG-HF): a multinational, open-label, randomised trial. Lancet. 2022;400(10367):1938-1952.
9. McMurray JJV, Packer M, Desai AS, et al. Angiotensin-neprilysin inhibition versus enalapril in heart failure. N Engl J Med. 2014;371(11):993-1004.
10. McMurray JJV, Solomon SD, Inzucchi SE, et al. Dapagliflozin in patients with heart failure and reduced ejection fraction. N Engl J Med. 2019;381(21):1995-2008.
11. Felker GM, Lee KL, Bull DA, et al. Diuretic strategies in patients with acute decompensated heart failure. N Engl J Med. 2011;364(9):797-805.
12. Mullens W, Dauw J, Martens P, et al. Acetazolamide in acute decompensated heart failure with volume overload. N Engl J Med. 2022;387(13):1185-1195.
13. Bauersachs J, Konig T, van der Meer P, et al. Pathophysiology, diagnosis and management of peripartum cardiomyopathy: a position statement from the Heart Failure Association of the European Society of Cardiology Study Group on peripartum cardiomyopathy. Eur J Heart Fail. 2019;21(7):827-843.