Value-added time

Every assessment on this site is judged against one standard. It is not a number and not a target. It is a way of reading evidence that can be applied the same way every time and checked by anyone.

Value-added time is the difference care makes to the time a person has left, as that person would judge it, net of the time the care itself consumes. Care adds value in two ways: by improving that time, or by accompanying the person through time that cannot be improved. It is judged for the person, and separately for those who care for them.

The conviction underneath it, that a person’s worth does not depend on their capacity or their capacity to improve, is set out on what this publication stands on.

Why time

The people this publication writes about have little time and know it: frail older people, people in the last year of life, people on long courses of intravenous treatment. Care is paid for in that time, in waits, appointments, lines and days in a bed. So the standard counts what care takes from a person’s time as seriously as what it gives. It also asks whether the care came at the moment it could help, in the measure that fitted.

Two forms of added value

Time improved. The care changes what the remaining time holds, such as function, symptoms, being able to do what matters, or where it is lived. Or it reduces how much of that time the care itself uses up.

Time accompanied. The person is not left alone in time that cannot be changed. Nothing is added to the time; someone is added to it. Watching with a dying person changes no outcome, and a standard that counted only outcomes would score it at zero. This one does not. It does count accompaniment only where it has been measured, never where it is assumed.

What it is not

It is not the lean meaning of the words. In the NHS’s Red and Green Bed Days system, “A RED day is a day of no value for a patient”, and a green day is one of care “that progresses their progress towards discharge” (Emergency Care Improvement Programme). That judges value by the process. A family’s last day at a dying person’s bedside would be a red day, and is not one here.

It is not a single number. Days alive and at home record where a person was, not what the day held. In survivors of critical illness they were “poorly associated” with health-related quality of life (Granholm et al., 2023).

It is not a funding rule. It says whether care achieved its end for the person who received it. It cannot say whether that care is worth more than the care it displaces.

How a claim is read against it

Every claim on this site about the effect of care gets a VAT reading. That covers the settled findings and open questions in each domain, and the claims of each pattern in the register. Each reading answers six questions about the evidence behind the claim.

1. How far did the evidence reach? This is the highest rung of the ladder below that the studies measured and found moved.
2. Was an outcome the person would judge measured, and what did it show? Shown, measured but not shown, or never measured.
3. What happened to the time and burden the care takes from the person? Reduced, unchanged, increased, or never measured.
4. Was accompaniment measured?
5. Was the effect on carers measured?
6. Were results reported separately for people with frailty, cognitive impairment, deprivation or distance from services?

Each reading also names the result that would show the claim wrong.

This is a different question from how sure we are. Every claim also carries a confidence grade, from high to very low, which says how firmly the evidence supports it. The two are shown side by side and never combined: a claim can be certain and reach no further than the system, or reach the person on thin evidence.

The ladder runs from the person’s own judgement outwards:

1. What the person judges. Goals they set, what matters to them, their capability, and their self-reported health or quality of life. Satisfaction alone does not count.
2. Harms the person experiences. Delirium, falls, drug reactions, line complications, pain.
3. Being alive, and where the person chooses. Survival, and living at home.
4. Function in daily life.
5. Outcomes that serve the system. Length of stay, readmission, documentation.

Survival is not lower on the ladder because it matters less. It sits there because it says nothing about what the time held.

A claim can reach far on one question and not at all on another. Hospital at home makes living in residential care six months later less likely. It “probably results in little to no difference” in self-reported health status (Edgar et al., 2024). Its reading records both, because the disagreement is the finding.

Who makes the judgement matters too. In one Dutch geriatric outpatient clinic, doctors’ agreement with older patients’ own top priority was about two in five, and no better than chance (Festen et al., 2021). So where the person cannot say, the reading uses their earlier wishes first, then what can be observed of their experience, and a proxy’s view only last, labelled as a proxy’s.

Every reading is made from the studies themselves, read at source, and checked by a second pass that cannot see the first. Where the two passes differ, the more cautious reading stands.

The profile: how the readings add up

The readings are added up into a profile for each domain and for the pattern register. It is a set of counts, never a score. The figures below are computed from the readings stored with each claim; nobody types them, and the site’s checks fail if they drift from their source.

• Acute frailty. Of 8 claims graded moderate or high, an outcome the person would judge is shown for 0 and was never measured for 6.
• Clinical psychology. Of 12 claims graded moderate or high, an outcome the person would judge is shown for 5 (for 1 of them, that outcome is itself graded moderate or high) and was never measured for 6.
• Palliative care. Of 2 claims graded moderate or high, an outcome the person would judge is shown for 0 and was never measured for 2.
• Therapies and AHPs. Of 13 claims graded moderate or high, an outcome the person would judge is shown for 2 (for 2 of them, that outcome is itself graded moderate or high) and was never measured for 6.
• OPAT. Of 5 claims graded moderate or high, an outcome the person would judge is shown for 0 and was never measured for 4.
• Pattern register. Of 4 claims graded moderate or high, an outcome the person would judge is shown for 0 and was never measured for 3.

How to read it. The headline for each domain counts its claims graded moderate or high: how many show an outcome the person would judge, and how many never measured one. Where an outcome is shown, it also says for how many that outcome is itself graded moderate or high, since a strong claim about survival can carry only weak evidence about quality of life. A domain can have strong evidence that care works, with every claim reaching survival or home at high certainty, and still show none. That means the evidence has not yet reached what people would judge for themselves. It does not mean the care has failed, and the profile is never a verdict on any service, hospital or system.

Tracking it over time

Three things are tracked separately:

• the profile of each domain’s evidence, which moves when a domain is revised;
• the profile of the pattern register, which moves as patterns are graded;
• the profile of what the publication publishes each month.

A snapshot is kept for each month, and a month’s snapshot is never rewritten after the month has closed. A correction appears in the next month, with a note of what changed.

What the trend can show. It shows whether the evidence this publication reads reaches the person more often over time: whether more claims measure what people value, and whether carers and the people furthest from good care are counted. It cannot show whether anyone’s care has improved. A rising line means the evidence is getting closer to the person, which is the gap this publication exists to close. It is not a claim about the quality of care.

What has been published. The weekly Signals will carry a reading too, once the routines that write them record one.

Of 7 months of published Signals, none yet carries a VAT reading. This series is not yet read, which is not the same as zero. It begins when the routines that write each Signal record a reading for it.

The rules that keep it honest

Ten rules bind every use of the standard. These five do most of the work:

• The outcomes that count are fixed before the evidence is read. If a trial’s main outcome did not move, a secondary outcome that did can never appear in a headline.
• An outcome nobody measured is reported as unmeasured. It is never described as valuable anyway.
• Time is never reported without the person’s judgement beside it. Where the two disagree, the disagreement is the finding.
• Improvability is never the measure of a person’s claim to care. Where evidence rations care by the chance of improvement, this publication asks what is then owed to the person judged unlikely to improve, and whether anyone measured it.
• The standard is a lens, never a target. It is not offered as any organisation’s performance indicator, because every measure it uses can be gamed once it becomes one.