Building a Cost Case for a Connection Programme Without Overselling It

A finance director does not want to hear that loneliness is “as bad as smoking 15 cigarettes a day.” They want to know what the programme costs, what it is expected to reduce, and whether that reduction is big enough to matter against the budget line it sits in. Most cost cases for connection programmes fail at exactly this point: they cite population-level mortality statistics and then, three sentences later, ask for £40,000 to run a walking group. The numbers do not connect to each other. This is the gap a cost case has to close, and it has to be closed honestly or it will be closed for you, later, by someone auditing outcomes you never actually measured.

Step 1: separate the headline statistic from your programme’s mechanism

The Surgeon General’s 2023 advisory states that the mortality risk of social disconnection is comparable to smoking up to 15 cigarettes a day. Julianne Holt-Lunstad’s 2015 meta-analysis put the odds ratio for early mortality from social isolation at 1.29, from loneliness at 1.26, and from living alone at 1.32, drawing on data pooled across studies. The American Heart Association’s 2022 scientific statement found isolation and loneliness associated with roughly a 29% increased risk of heart attack or death from heart disease and a 32% increased risk of stroke.

None of these numbers describe what your programme does. They describe population associations between a state — being isolated — and an outcome measured years or decades later. Your twelve-week peer support group is not going to move all-cause mortality, and no funder should be told it will. What you can plausibly claim is that the programme moves people along the pathway these studies describe: reducing isolation, which is associated with these downstream risks. That is a weaker claim. It is also the true one, and it is the one that survives a second reading.

Step 2: find the nearest utilisation evidence, not the nearest mortality evidence

The link between connection and downstream cost is more defensible when you use evidence about service use rather than evidence about death. A systematic review of social prescribing and loneliness found that three of nine included studies reported reductions in GP, emergency, social worker, or inpatient service use. That is a smaller, more specific, more auditable claim than a mortality statistic, and it is closer to what a commissioner actually pays for.

Cigna’s 2020 workplace survey found lonely workers miss work roughly twice as often for illness and five times as often for stress. If your programme sits inside an employer, this is the number to build from — not because it is a robust causal estimate, but because it is at least measured in the same currency (absence days) that the employer already tracks.

Step 3: use ranges, and say why the range is wide

The National Academies’ 2020 report on isolation in older adults calls for routine assessment within the health care system but does not offer a cost-per-participant figure, because the underlying trial evidence is not there to support one. A qualitative meta-synthesis of social prescribing found that participants describe benefit extending beyond contact itself to restored purpose and participation — useful for programme design, useless for a spreadsheet.

Where the evidence permits a number, present it as a range with the underlying study named, not as a single figure presented as fact. “Reduces avoidable GP contact, though the trial base is small and heterogeneous” is a defensible sentence in front of a funder. “Saves the NHS £X per participant” is not, unless you can show the study that produced X and explain why your programme resembles it.

Step 4: build the case in this order

1. State the mechanism your programme actually targets (isolation, network size, or loneliness — these are not interchangeable).
2. Cite the utilisation or service-use evidence closest to that mechanism, not the mortality literature.
3. Name the local cost this maps to (GP appointments, sickness absence, emergency contacts, social care hours).
4. Apply a conservative estimate of how many participants you expect to shift, and show your working.
5. State the range of avoided cost this implies, with the caveat attached, not footnoted.
6. Name what you will measure to check the claim, and commit to reporting it even if it disappoints.

What this means in practice: a cost case that says “we expect to reduce isolation for 60% of participants and this is associated with reduced avoidable GP contact in some trials, though the evidence base is small” will survive scrutiny better than one that says “this programme will save £1,200 per person.” The first sentence is defensible for years. The second one is disprovable in eighteen months.

The avoided-cost mirage

The most common failure mode in these cost cases is what might be called the avoided-cost mirage: taking a population-level statistic (people who are isolated cost the system more) and applying it forward as if enrolling someone in your programme retroactively erases that cost. It does not work that way. A person who was isolated and is now less isolated does not instantly stop generating the healthcare contacts, absences, or care needs that accumulated under the old state. The saving, if it exists, shows up slowly and partially, and only for the portion of participants whose isolation was actually driving the cost rather than correlating with something else — poor health, low income, bereavement — that the programme does not touch. Cost cases that promise year-one savings are almost always running on this mirage.

Evidence status for common cost-case claims

Claim	Evidence status
Isolation is associated with higher mortality risk	Strong — large meta-analyses, consistent effect sizes
Isolation is associated with higher cardiovascular risk	Strong — AHA scientific statement, explicit about the intervention gap
Connection programmes reduce GP or emergency contact	Weak to moderate — some trials show it, evidence base small and heterogeneous
Connection programmes reduce workplace absence	Weak — plausible mechanism (Cigna’s absence data), no strong programme-level trial evidence
A specific £ figure can be attached per participant	Not supported by current evidence — treat any such figure with suspicion

The UK’s 2018 loneliness strategy embedded loneliness measurement into national statistics precisely because the evidence base was too thin to cost properly without consistent data. Five years on, that measurement infrastructure exists at national level; most local cost cases still do not have it, and are built instead on the mortality statistics that happen to be quotable rather than the utilisation statistics that would actually hold up.

What this does not solve

A well-built cost case will get a programme funded once. It does not solve the deeper measurement problem: most connection programmes cannot yet show, with their own data, that the utilisation reduction they are claiming actually occurred in their population, because most do not track healthcare contacts, absence days, or care hours before and after participation. Until that tracking exists as standard practice, the cost case remains an estimate borrowed from someone else’s trial, applied to your programme on the strength of resemblance rather than proof. Say that plainly to the funder. It is more persuasive than pretending otherwise, and it is the only honest way to ask for the money to build the tracking that would let the next cost case be a real one.