Matching the Intervention to the Need: How Programmes Have Approached Targeting

A befriending scheme paired a housebound 78-year-old with a weekly phone call. Her isolation score improved. Her loneliness score didn’t move. This is not a hypothetical failure — it’s the predictable result of applying a contact-based intervention to a problem that wasn’t primarily about contact frequency. She had one visitor a week either way. What she lacked was a relationship that felt reciprocal.

This is the targeting problem, and it is more consequential than most programme design decisions get treated as being. Isolation, loneliness, and weak social skills are three different deficits. They call for three different tools. Programmes that pick the wrong one don’t just underperform — they can produce a clean, measurable result on the wrong outcome and call it success.

Three deficits, not one

The National Academies’ 2020 consensus report on older adults draws a hard line between social isolation — an objective shortage of contacts and relationships — and loneliness, the subjective distress of feeling disconnected regardless of contact count. A 2024 study in Scientific Reports pushes further, showing the relationship between the two shifts with age: the mechanisms that produce isolation in a 40-year-old are not the mechanisms that produce it in an 80-year-old, and isolation and loneliness track together less tightly than programme designers tend to assume.

There’s a third deficit that gets less attention: a lack of the social skills or scripts needed to convert opportunity into relationship. Someone can be neither isolated (they have contact) nor lonely in the clinical sense, and still struggle to move an acquaintance toward something that functions as support. John Cacioppo’s work on loneliness frames the underlying signal as evolutionarily tuned to provoke reconnection — but the behavioral response to that signal has to be learned, and not everyone has had the chance to learn it.

Three deficits. Three different interventions:

Deficit	What’s actually missing	Matching intervention type
Isolation	Volume and diversity of social contact	Structured activity, volunteering, group participation
Loneliness	Perceived quality/reciprocity of relationship	Peer support, mutual-aid structures, therapeutic approaches
Skills deficit	Ability to initiate, sustain, or deepen contact	Coaching, rehearsal, social skills training

Social prescribing literature has effectively arrived at this distinction from a different direction. A 2022 qualitative meta-synthesis in BMC Health Services Research found that participants describe benefit extending beyond social contact itself, toward restored meaningful participation and purpose — and that structured, purposeful group activity appears to outperform contact alone. That’s a targeting finding, not just a satisfaction finding. It tells you what the mechanism actually is, and it isn’t “more people in the room.”

How social prescribing has approached this

Link workers operating within social prescribing schemes are, in effect, running a triage function — matching a referred person to one of several available activities based on a needs conversation, not a diagnosis. Systematic reviews of social prescribing report positive outcomes across the studies included, but the reviews themselves flag the limitation: heterogeneity across programmes is high, and the mechanism by which any given activity works for any given referral is rarely specified in advance. A 2021 systematic review in Perspectives in Public Health found all nine included studies reported positive individual impacts, with three showing reduced service use — but the review can’t tell you which referral matched which need, because the underlying studies mostly didn’t track that either.

This is the structural weakness of social prescribing as currently practiced: the targeting logic lives in the link worker’s judgment, not in a documented decision rule. That’s not a criticism of link workers — a skilled conversation can pick up cues that a checklist misses. But it means the approach doesn’t scale its targeting accuracy the way it scales its referral volume. Two link workers with different instincts will route the same person to different interventions, and neither outcome gets checked against the other.

Volunteering as a targeted case

The HEAL-HOA trial, published in The Lancet Healthy Longevity in late 2024, is worth attention precisely because it’s a randomized comparison rather than an uncontrolled programme evaluation — rare in this field. It tested prosocial engagement and volunteering, specifically, against a control group among lonely older adults in Hong Kong. Volunteering is a plausible intervention for isolation (it reliably produces contact) and a more uncertain one for loneliness (contact with strangers in a volunteering context doesn’t automatically produce felt reciprocity). The trial existing at all is notable: most of the intervention literature in this space is small and uncontrolled, which is part of why the American Heart Association’s 2022 scientific statement on social isolation explicitly names the absence of intervention evidence as the field’s central research gap. Programmes citing volunteering as a loneliness fix should be honest that the evidence for that specific claim is thinner than the evidence for volunteering as a contact-generating activity.

The failure mode: the contact substitution error

Call this the contact substitution error — treating any increase in social contact as evidence a loneliness problem has been addressed. It happens because contact is easy to count and loneliness is not. A programme logs attendance, sees attendance rise, and reports success. Meanwhile the participant’s subjective loneliness score, if anyone bothered to measure it, hasn’t moved, because the contact was low-quality, one-directional, or with people who don’t reciprocate.

The 2020 commentary on the National Academies report, written for clinicians, pushes for routine assessment of isolation in clinical settings specifically because contact and connection are conflated so easily in practice. A prevention-framing review by Julianne Holt-Lunstad makes the same point from the public health side: treating social connection as a modifiable risk factor requires measuring the right construct, not a proxy for it.

What this means in practice: before assigning anyone to an intervention, ask which of the three deficits is actually present — do a five-minute needs conversation that distinguishes “I don’t see many people” from “I don’t feel understood by the people I see” from “I don’t know how to keep a conversation going long enough to build something.” Route accordingly. Track the outcome that matches the deficit you targeted, not the outcome that was easiest to log.

Evidence status on targeting claims

Claim	Evidence status
Isolation and loneliness are distinct constructs requiring different measurement	Strong — consistent across consensus reports and recent empirical work
Structured group activity outperforms unstructured contact for restoring purpose	Moderate — qualitative synthesis, consistent finding but not RCT-based
Volunteering reduces loneliness specifically (not just isolation)	Weak-to-moderate — one randomized trial, single population, needs replication
Link worker judgment reliably matches referral type to underlying deficit	Untested — no systematic evidence this matching is done consistently or accurately
Skills-deficit is a distinct third category requiring coaching-style intervention	Plausible from theory, not directly tested in the cited intervention literature

What this does not solve

None of this addresses reach. A targeting framework is only useful for people who have already been referred, walked through a door, or picked up a phone — and the populations least likely to do any of those things are disproportionately the ones carrying the heaviest isolation burden. Getting the match right for someone already in the system says nothing about the person who never got assessed at all. Sharper targeting improves what happens to people already inside a programme; it does nothing to widen who gets there.