The SREL pathway is one of the most consequential provisions in the UK benefits system. It allows people living with terminal illness to fast-track their claims for Personal Independence Payment (PIP), a benefit designed to help with the extra costs of disability. Under the special rules, claimants avoid the standard face-to-face assessments, receive the highest rate of support, and in many cases have their claims processed within days rather than months. For people who may have only weeks or months to live, that speed can make the difference between dignity and destitution. Before April 2022, eligibility hinged on a clinical judgment that death could reasonably be expected within six months. The 2022 change extended that window to twelve months, a move intended to spare more dying people the stress and paperwork of a standard claim.

To measure the effect, the research team turned to a powerful epidemiological tool: interrupted time series analysis. Rather than comparing simple before-and-after snapshots—which can be confounded by seasonal effects, administrative changes, or secular trends—the method models the underlying trend in a time series and asks whether the policy intervention produced a statistically detectable change in level or slope. The researchers obtained monthly caseload data for SREL-linked PIP claims in England and Wales from the Department for Work and Pensions via its public Stat-Xplore platform, covering January 2019 through July 2025. In total, 2,370,328 PIP cases were analyzed, a dataset large enough to detect shifts in trend with considerable statistical confidence. The study protocol was prospectively registered on the Open Science Framework, and the analysis was reported in line with RECORD guidelines for studies using routinely collected data.

The headline result is unambiguous: the change accelerated growth in the SREL caseload dramatically. Before the reform, the caseload was already expanding, at an annual growth rate of 4.3 percent as of 2019. After April 2023—the point at which the new policy’s effects became fully visible in the data—that annual growth more than doubled to 9.3 percent. In practical terms, the number of people receiving fast-tracked end-of-life benefits was rising more than twice as fast after the threshold change as it had been before. This is precisely the kind of step change in trajectory that interrupted time series methods are designed to isolate, and its magnitude suggests the policy did what it was designed to do: widen the door.

But the aggregate number conceals a more textured story about who crossed that threshold. Growth was strongest among people with non-cancer diagnoses, whose annual caseload growth jumped from 11.9 percent to 18.8 percent. This matters because the six-month rule had long been criticized as biased toward cancer patients, whose disease trajectories are often more predictable to clinicians. People with organ failure, neurological disease, or frailty tend to have illness courses marked by gradual decline punctuated by acute crises, making any six-month prognosis a practical guessing game. The twelve-month rule appears to have relaxed that constraint enough to bring substantially more non-cancer patients into the system.

The most striking shift, however, appeared among older adults. Among claimants at or above state pension age, the trend inverted entirely: from an annual decline of 2.6 percent before the change to an increase of 17.9 percent after it. That reversal—roughly a twenty-percentage-point swing in annual growth—suggests the previous threshold had effectively excluded large numbers of elderly people who were dying but whose prognoses could not be confidently compressed into six months. For this population, the reform was not a marginal adjustment but a wholesale change in access.

Geography added yet another layer of variation. Regional analysis revealed marked disparities in how the policy played out across England and Wales. The West Midlands shifted from an annual decline in SREL caseload to an 11.1 percent increase, while growth in the East of England accelerated more than fivefold. Such heterogeneity is difficult to explain by patient characteristics alone and points instead to differences in clinical practice: how willing or able clinicians in different regions are to certify a terminal prognosis, how aware local services are of the special rules, and how referral pathways are organized. A benefit that depends on a doctor’s signature will inevitably inherit the unevenness of the health system that produces those signatures.

Yet the study’s most sobering finding lies not in the data on awards but in what the researchers describe as a fundamental mismatch at the heart of prognosis-based eligibility. Approximately half of the people deemed eligible for SREL under the twelve-month criterion are not actually in the last year of life—their prognoses were overestimated, or their illness trajectories diverged from expectations. Meanwhile, more than 200,000 people die each year from a life-limiting illness without ever being correctly identified as suitable for referral. The authors conclude that prognosis may simply not be the best mechanism for determining eligibility: the inherent uncertainty of clinical forecasting produces a system that simultaneously gives fast-tracked support to many people who are not dying within the statutory window and withholds it from many who are.

This is a problem that palliative medicine has grappled with for decades. Clinician estimates of survival are known to be systematically optimistic, with accuracy degrading further out from the point of prediction. A six-month threshold demands a degree of precision that prognostication rarely delivers; a twelve-month threshold widens the target but does not eliminate the error. The UCL team’s data quantify, at national scale, what that uncertainty costs: a benefits system in which roughly half of awards go to people outside the intended window while hundreds of thousands of dying people each year pass through unrecognized. The money matters, but so does the timing—people who are identified late, or not at all, may never receive the financial relief, care coordination, and formal recognition of their situation that the pathway is meant to provide.

The policy implications are likely to be debated on both sides of the Atlantic. The United States faces analogous challenges with hospice eligibility, where a six-month prognosis requirement forces clinicians to make similarly uncertain predictions, and where fears of fraud can deter clinicians from referring patients who might outlive the window. The UK experience offers a rare natural experiment: when the threshold was doubled, access expanded substantially, particularly for the groups historically least well served—older adults and people with non-cancer illnesses. But expansion alone did not fix the identification problem, and regional and diagnostic disparities persisted even after the reform.

For now, the study provides the clearest empirical picture yet of what happens when a government relaxes the prognostic gatekeeping of end-of-life support. The caseload grew, faster and more equitably than before, but the system still misses a large share of the people it was built to catch. Whether the answer lies in abandoning prognosis altogether—as some palliative care researchers have argued, in favor of criteria based on need rather than expected survival—or simply in continuing to widen the window, the evidence from England and Wales suggests that the current architecture, however calibrated, will keep failing a significant fraction of dying people as long as it rests on the inherently unreliable art of predicting death.