Read past the headline.
A promising finding can be real and still leave important questions unanswered. Start with three questions: What happened, to whom, and for how long? Then look at the comparison, harms and practical burden.
What does “cure” mean?
Here, Cure research names an area of research, not a claim that every entry cures type 1 diabetes. Look for the exact outcome: less injected insulin, measurable insulin production, freedom from severe lows, or insulin independence.
Insulin independence means a participant did not need external insulin under the study’s definition and during its observation period. It does not by itself show that the autoimmune process is gone, that the effect will last for life, or that no other treatment is needed.
For example, the FDA approved donor-islet therapy Lantidra for a narrowly defined group of adults with repeated severe hypoglycaemia despite intensive management. It requires immunosuppressive medicines to maintain the transplanted cells; the procedure and those medicines can cause serious harm. Always ask what treatment burden has been exchanged for less insulin. FDA approval explanation, 28 June 2023.
A phase is a question, not a promise.
Drug trials usually move from early safety and dose questions in phase 1, through initial tests of benefit and safety in phase 2, to larger confirmatory studies in phase 3. Monitoring continues after approval. Phases can overlap, and device studies do not necessarily use this drug-phase framework.
A later phase means a programme is being tested at a different stage. It does not tell you that it will succeed, when it will be available, or what share of a “cure” has been achieved. FDA guide to clinical research.
Plans are not results.
A registry describes what researchers intend to test, who can take part and when they expect to finish. A planned completion date is not a promised publication date. “Completed” describes study activity; it does not mean the treatment worked.
Check whether a number is planned enrolment, actual enrolment, or the participants included in a particular analysis. Look separately for posted results or a published paper. “No results posted” is not evidence that the result was negative; “results submitted” is also different from publicly posted results. Read a registry record · Read study results.
Our trial index links to registries and reported evidence. A registry listing is not an invitation to enrol or a recommendation to participate.
How big is the benefit?
Statistical significance asks whether a finding crosses a statistical threshold under the analysis assumptions. Clinical importance asks whether the size of the benefit matters in life: fewer severe lows, more time in range, less burden or better health.
A small p-value does not measure the size of a benefit or prove that a treatment is safe. Read the effect size, its confidence interval, follow-up and harms together. Conversely, a small study that does not find a statistically significant difference may be too uncertain to rule out an important effect. NIH guide to interpreting results.
Illustrative example, not a study result: a change from 60% to 65% time in range is 5 percentage points, equivalent to 72 minutes a day. Whether that tradeoff is worthwhile also depends on lows, effort, side effects and cost.
Can I compare those two studies?
Check whether participants had similar starting glucose levels, ages, disease duration and treatment experience. Also compare the control treatment, targets, follow-up and definition of the outcome. A device tested in people starting far above target is not directly comparable with one tested in people already doing well.
A randomised head-to-head trial is designed to compare treatments within one study. Putting results from unrelated trials next to each other does not recreate that comparison. Small uncontrolled studies and manufacturer announcements can be useful signals, but leave more room for alternative explanations. NIH guide to reading a scientific article.
Use Compare to inspect differences and evidence, not to infer a proven clinical winner.
What do the site’s scores mean?
Our 0–100 scores combine editorial ratings using published weights. They are tools for exploring priorities, not measured effectiveness, success probabilities or personalised medical advice. A two-point gap has no established clinical meaning.
Compare scores within a category and under the same weights. Evidence labels are separate from the arithmetic; there is no hidden evidence-quality multiplier. Missing ratings are omitted and remaining weights adjusted, so missing information must not be mistaken for good performance.
Named products and research programmes can have provisional ratings based on limited evidence or intended features. Design concepts have no score or rank. Trial records are not scored. The methodology explains the formula and its limits; review status explains what has actually been checked.
Can I actually get it?
Approval, commercial launch, funding and individual eligibility answer different questions. A country label cannot establish all four. Check the exact indication, compatible devices, local supply and coverage. A recruiting study may have no open site near you, or may not fit your circumstances.
Use Availability to understand these distinctions, and take the relevant label or study record to your care team. The reader digest separates original source dates from the date we reviewed a highlight.