This is a guest post by @theMSguide who has multiple sclerosis. References supporting the view were supplied.
Disclaimer: As ever for all posts. These are to be considered the views of the author only. Constructive comments, as ever, are welcomed
The missing comparator: why MS drug risks are clear, but MS disease risks are not
People with multiple sclerosis (MS) are often handed a leaflet that can feel alarming. That reaction is understandable. The risks are real, and patients deserve clear, honest information.
But there is a structural imbalance in how we communicate risk in MS.
In the United Kingdom, patient information leaflets are regulated documents. They are required to include side effects and to present them in a structured way (including seriousness and frequency). Regulators also explicitly acknowledge that long lists of side effects can be frightening, even as they remain necessary for safety. I often suggest people read the datasheets for ibuprofen or paracetamol. The ones thrown away in our haste to get to the contents.
By contrast, the risks of poorly controlled MS (or delayed effective control) are often communicated in broad, variable terms, and without the kind of ‘decision-grade’ numbers that a person can actually use in clinic.
The result is not neutral. It disadvantages patients, and those the patients often turn to for advice and guidance.
The comparison patients are trying to make (and rarely get help with)
Many people are not really asking, ‘Do I want a disease-modifying therapy (DMT), yes or no?’ They are asking a more practical question: ‘Which option feels least harmful/lowest risk?’
That is a rational question in the face of a frightening leaflet. But it is also incomplete, because it ignores a second risk that is harder to see:
What is the trade-off between treatment risk and the risk of irreversible neurological loss if effective control is delayed?
Today, patients are often asked to compare two different kinds of information:
- Treatment risks: listed, categorised, and presented formally (because they must be).
- Disease risks: described variably, often without numbers, and rarely with concrete examples tied to outcomes that matter day-to-day.
This is not a criticism of clinicians. It is a systems problem.
Why the ‘disease risk sheet’ is hard (and why we still need it)
There are legitimate reasons disease-risk communication lags behind leaflet-style clarity:
- MS is heterogeneous: averages can mislead individuals.
- Modern untreated datasets are scarce: withholding effective therapy is ethically complex, so ‘clean’ contemporary untreated cohorts are limited.
- The disease itself has changed: population-level disability outcomes appear to have improved over time in some settings, plausibly reflecting earlier diagnosis, changing thresholds for treatment, and treatment effects.
- We over-rely on single scales: EDSS (Expanded Disability Status Scale) is useful, the timed 25ft walking test, the nine-hole peg test, but they compresses complex outcomes into one number and underweights cognition, fatigue, hand function, and other domains people actually care about. Quality of life measures.
None of these arguments justify the current imbalance. They simply define what a credible ‘disease risk sheet’ would need to do: stratify, quantify uncertainty, and talk about outcomes that matter to patients.
What we do know
If we want to communicate disease risk with the same seriousness as medicine risk, we should start with the strongest, most reproducible elements of evidence.
1) Untreated (or historically untreated) MS often accumulates disability over time — but with wide variability
Older population-based natural history studies (pre–high-efficacy era) show substantial disability accumulation at the group level, with median times to key disability landmarks measured in years to decades, depending on cohort characteristics and definitions.
Those studies are not a perfect counterfactual for a person diagnosed today. But they are still relevant in one crucial way: they make clear that MS can produce irreversible disability on clinically meaningful timescales, even when early disease looks ‘manageable’.
2) Relapses are not always ‘events you recover from’
A persistent misconception in MS, and something that reinforces the sense of clinical detachment patients feel is that relapses are temporary storms that pass without consequence. Recovery can be incomplete, and relapse-associated worsening contributes to long-term disability for many people. The abiding feeling that people think you ought to just get over it.
Evidence from multiple cohorts suggests that a substantial fraction of relapses leave residual deficit.
That matters because ‘I’ll wait and see’ is rarely a neutral stance from a patient point of view. If new and irreversible disability is the price of waiting then it feels cruel and dismissive. Damage is rarely reversible.
3) Disability progression can occur without obvious relapses
Another under-communicated risk is progression that is not temporally linked to relapses. This is now widely discussed under frameworks such as PIRA (progression independent of relapse activity) and related concepts describing disability accumulation outside classic relapse events.
This directly undermines a common patient strategy: ‘If I’m not relapsing, I’m safe to delay escalation.’ For a subset of people, that inference is wrong. Presently, we have no way of telling if we (the patients, for I am one too!) are in that subset. Any sensible person (In my Honest Opinion) goes in as heavy as they can even though it may transpire that they didn’t need to. It is a risk I am not willing to take. And one that is rarely explained effectively as – as a nurse once told me – ‘they didn’t want to scare the patient.’ If that isn’t patronising…
4) Delayed effective control is associated with worse long-term outcomes
We cannot randomise large groups of people to ‘treat late’ for a decade, for obvious ethical reasons. So much of the evidence on treatment delay comes from registries and comparative observational analyses.
Even with the limits of observational data, a consistent signal has emerged across multiple settings: earlier use of higher-efficacy strategies is associated with better disability outcomes than delayed escalation.
It does mean that ‘low perceived drug harm’ should not be the only anchor of decision-making. My opinion is that we can’t afford not to go in with the most efficacious therapies as soon as possible. Until we have a concrete way of determining who does not require a high-efficacy medicine/treatment then the profession is taking a gamble with the person. Presently, it is prudent to over-insure until there is a reliable way to stratify the way the disease will affect an individual.
Why this gap matters in real life
When information is asymmetrical, fear predictably fills the gap.
Across chronic diseases, anxiety about adverse effects and insufficient patient–provider communication are recognised contributors to poor adherence. In MS specifically, side effects are a commonly reported reason for discontinuation.
I am not claiming leaflets alone drive decisions. The causal chain is likely multi-factorial. But the pattern is familiar:
- leaflet read → risk salience spikes
- disease risk remains vague
- delay, de-escalation, or premature discontinuation becomes more likely
Sometimes that is the right decision for that individual. Often it is not. The point is that the decision environment is biased towards over-weighting medicine harms because those harms are presented with greater clarity.
Doctors and nurses, armed with the anecdotal/observational knowledge of many patients, who have seen them deteriorate, as well as the academic facts often seem content to watch/allow this to happen.
Existing decision tools can help but they do not fully solve the missing comparator
The UK has genuinely useful DMT decision resources (for example MS Society tools and MS Trust resources) and research-built decision aids such as CRIMSON (Considering Risk and Benefits in Multiple Sclerosis Treatment Selection). These support shared decision-making about treatments. They don’t ever discuss the risks of undertreated or untreated MS and they are not serving the patients as well as they could.
What remains much rarer is a patient-ready, clinic-usable way to quantify the counterfactual: the likely consequences of delayed effective control, expressed in outcomes patients care about, with uncertainty stated plainly.
We should build: a patient-facing ‘disease risk sheet’
A credible disease-risk summary should sit alongside the medicine leaflet and make the other side of the trade-off visible. It should not be a marketing tool. It should not be coercive. It should be explicit about uncertainty and heterogeneity.
At minimum, it should include:
- Expected disability trajectories, expressed as ranges and stratified by baseline risk features (not just a single average).
- Relapse recovery probabilities, including the reality of residual deficit.
- Progression without relapses (PIRA/related concepts), described in patient-facing terms.
- Outcomes people actually experience: walking, hand function, vision, bladder/bowel function, cognition, fatigue, work participation, and independence (not EDSS alone).
- Descriptions of how these outcomes manifest themselves: real world descriptions, not designed to frighten or terrify but actually what life is life when disability accrual becomes evident.
- The effect of timing: what is known (and unknown) about delay and long-term outcomes.
- Limitations: confounding, missing data, changing natural history, and why individual prediction remains imperfect.
Who could assemble it in practice?
I am not accusing any organisation of withholding anything. I am asking, pragmatically, who is best placed to build something usable at scale.
MS registries
Registries are an obvious candidate. Large registries exist with explicit ‘real-world evidence’ objectives, such as the Swedish MS Registry, and international platforms such as MSBase.
But registries also face real constraints: missingness, uneven capture across subgroups, and the need for consistent definitions and sustained funding.
Industry (collectively)
Companies run trials, extension studies, and pharmacovigilance systems. They have analytic resources and access to long-term follow-up in various forms. The problem is not capability. The problem is incentives and fragmentation.
Speculation (labelled as speculation): if the best comparative datasets exist, they may sit in silos. If they sit in silos, patients will not see them without a strong incentive — scientific, regulatory, reputational, or commercial — to pool and publish.
There may be benign reasons: analysis is hard; results may be too confounded for public use; or it is simply not anyone’s funded priority. But the practical outcome is the same: patients get asymmetrical information.
A challenge to the MS ecosystem
If you work in MS research, registries, charities, regulation, or industry:
- Do you believe a patient-ready dataset on delayed or undertreated MS exists in a usable form?
- If it does not, what is stopping us creating it from registry data, trial follow-up, and real-world evidence?
- Who should fund it, and who should be accountable for publishing it?
If strong work already exists that I have missed, I would rather be corrected than keep describing a gap that is already being solved.
Right now, too many people with MS are forced to choose between a frightening leaflet, misguided professional counsel, social media (yes, many people rely on it), and a vague or variable warning. That is not a fair nor an accurate decision environment.
The amount of people, me included, who have never been given a full-picture of the downsides is frightening. For the record: whatever upsides there may be are not worth having. Putting a brave face on decline is very dull and tiring.
COI: I have a consultancy called PatientSignal. I run @theMSguide on TikTok, Instagram, YouTube, Facebook. I am an Honorary Research Fellow at QMUL – if that even constitutes a COI? I advise – loosely – on ATTACKMS. I lead the patient engagement function for ASSISTMS and RESTORE – both paid roles. I head the Patient Advisory Board for Celestra Health (a medical device manufacturer) – unpaid
Some links
UK Medicines and Healthcare products Regulatory Agency (MHRA). Medicines: packaging, labelling and patient information leaflets. https://products.mhra.gov.uk/
MHRA. Best practice guidance on patient information leaflets (PILs). (Side effects listed by seriousness/frequency; acknowledgement that long lists can be frightening.) https://www.gov.uk/government/publications/best-practice-guidance-on-patient-information-leaflets
MS Society (UK). Disease modifying treatment (DMT) decision tool. https://www.mssociety.org.uk/living-with-ms/treatments-and-therapies/disease-modifying-therapies/disease-modifying-treatment-dmt-decision-tool
MS Trust. MS Decisions aid. https://mstrust.org.uk/information-support/ms-drugs-treatments/ms-decisions-aid
University of Leeds (CRIMSON project). Disease Modifying Treatments Decision Aid Booklet. https://crimson.leeds.ac.uk/
Source: multiple-sclerosis-research.org