I have written about uncertainty and how it affects the minds of people with chronic diseases such as MS (please see ‘Uncertainty associated with MS: are you comfortable with it?’, 20-Apr-2024).

Uncertainty has a dark underbelly, and unless you can come to terms with it, it will paralyse you with rumination and intrusive thoughts about what life would be like if only this or that happened. I have seen too many of my patients succumb to this problem whilst others come to terms with what has happened to them and get on with their lives. I urge you to try and embrace the latter. For example, the following are some questions you may have about your MS:

* Why did it take so long for my general practitioner or neurologist to take my symptoms seriously?

* Why did it take so long to be diagnosed with MS?

* Would an earlier diagnosis of MS have made a difference to my outcome?

* Why are there no reliable prognostic calculators for people with MS to know what is going to happen in the future?

* Will I have benign MS and not have any disabilities when I am older?

* Will I become disabled and need a wheelchair in the next 15 years?

* Will MS affect my cognition and prevent me from being able to work?

* If I start on a low-efficacy DMT, will it mean I won’t do as well if I switch to a high-efficacy treatment in 5-10 years?

* If I have AHSCT, will I be cured?

* If I start on an anti-CD20 therapy, will I recover function?

* I have smouldering MS; if I start tolebrutinib, will my physical functioning improve?

* …….

Uncertainty underlies the potential answers to all of these questions. Although uncertainty surrounds us, the human brain deals poorly with it. A field of science deals with uncertainty, which some call the ‘science of uncertainty’ and deals with different concepts depending on the context. The following are some of the issues around uncertainty and the practice of medicine. I suspect many of you have been ‘victims’ of uncertainty without necessarily knowing about it.

Uncertainty as a core component of the scientific and medical process

1. Measurement Uncertainty:

All measurements in neurology and medicine have inherent uncertainty due to the limitations of the neurological examination and our diagnostic tests. In addition, there is human error and variability in the measured neurological phenomena. Clinicians rarely quantify this uncertainty and report it alongside their findings. How confident are they that the neurological examination is normal, and is that white matter abnormality on MRI non-specific, or could it be the first sign of multiple sclerosis? If your initial MRI was passed off as normal despite having non-specific white matter lesions, you may not realise that you fell into this grey zone of diagnostic uncertainty.

2. Statistical Uncertainty:

When analysing data, we use statistical methods to estimate the range of possible values for a quantity and the confidence level associated with that estimate. There is always a chance that a so-called positive finding is a false positive, and sometimes a normal or negative finding is a false negative. This is why the sensitivity and specificity of tests are so important and give us confidence that what we are measuring is likely to be correct. No neurologist can be 100% sure about a diagnosis of MS or saying you don’t have MS, i.e. not diagnosing MS. They rarely tell you about their uncertainty, and it is even rarer for them to provide figures of how likely they are to make a diagnostic error.

3. Natural Variability:

Many natural phenomena are inherently variable, making it difficult to predict their exact behaviour. This can apply to neuroradiological findings. Many aspects of neurological function, its measurement and the impact of disease on these attributes are fundamentally uncertain, meaning their outcomes cannot be predicted with certainty. This explains why decisions in medicine can seem so variable and depend on who you see and may even depend on their cultural background. For example, if you have MS in Sweden, you are likely to get a different decision about the treatment of your MS compared to if you live in Denmark. This variability even happens at a local level. For example, one MS centre in London may have a very different way of treating MS compared to another. Even within centres, one neurologist may give you different advice than their colleague working in the same centre. The general public rarely gets exposed to this side of medical practice. This variability also applies to diagnosis. One neurologist may be prepared to diagnose MS when the diagnostic certainty is 70%, and another would prefer to wait until something happens to improve their diagnostic certainty to above 90%. I know many of you find this variability in medicine unacceptable.

Decision theory

Decision theory is the field that studies how to make optimal decisions in the face of uncertainty. Sadly, decision theory is not taught in medical school—at least not in my medical school, and I am not aware of it being taught in my current medical school where I work. I think this is a mistake.

Clinicians need the skills to do clinical risk assessments in near real-time. If they make a premature diagnosis of MS based on insufficient data, what are the consequences for the patient and their reputation? The general population think the diagnosis of MS is black and white. It is not. It is rather grey and may get greyer when the new MS diagnostic criteria are implemented. In general, neurologists prefer to err on not making a diagnosis of MS when they are not sure. If you are diagnosed with MS in the future, it is because, with time, the diagnosis of MS declares itself. This does not mean the neurologist was negligent; he or she was likely to be uncertain when you first presented.

Could diagnostic uncertainty or diagnostic dithering be considered medical gaslighting? Yes, this would have been gaslighting if the correct interpretation of symptoms or ignoring a patient’s questions had led to further investigations and an earlier diagnosis. And possibly not if interpreting and investigating those symptoms would not have changed the diagnosis in terms of fulfilling the MS criteria for dissemination in time and space. Gaslighting can be subtle; if, for example, the neurologist didn’t acknowledge and validate a patient's symptoms, then this is a form of gaslighting. The ‘science of uncertainty’ encompasses understanding and managing uncertainty in clinical practice, medical research, and medical decision-making. It involves quantifying uncertainty, analysing its sources, and developing strategies to cope with it and communicate it to people with diseases. Much needs to be done concerning the latter. Many of you email me with issues that could be avoided with a better understanding of uncertainty.

Some think AI and large-language models (LLMs) will resolve diagnostic uncertainty. I doubt it will. While AI may make diagnostic algorithms more efficient and include more data, uncertainty will always exist. An example that is relevant to MS is MS prognosis.

Prognostic profiling

How many of you want to know your prognosis, e.g. what are my chances of needing a walking stick in 10 years, given my current state? For the pre-DMT era, I could give you an answer based on natural history data. The average time needed for a walking stick from symptom onset was about 17 years, which means 50% of pwMS needed a walking stick in less time than this, and in the other 50%, it took longer than 17 years or not at all. Don’t forget that about 15% of pwMS in this pre-DMT era never needed a walking stick.

A question worth asking is, “How many people who don’t have MS end up needing a walking stick due to ageing?” Data from the 2011-2012 National Health and Aging Trends Study estimated that 15.9% of people aged 70-74 used a cane. This number increased with age, reaching 50.1% for those aged 85-89 and 70.6% for those 90 and older. Therefore, ageing and its effects on physical functioning must be considered when predicting MS outcomes. I have given many talks about MS prognosis and include the following poor prognostic factors on the list. You can add up how many you have on the list. The question is whether this will help you manage your MS.

* Age of onset >40 years

* Male sex

* Ancestry - Afro-American, Afro-Caribbean, African, South Asian, Asian

* Multifocal disease onset

* Motor involvement

* Cerebellar involement

* Bladder dysfunction

* Bowel dysfunction

* Cognitive involvement

* Residual disability after your first attack

* Progressive worsening

* Diagnosis of primary or progressive MS

* Ability to improve cognition (learning on the PASAT test)

* Partial or no recovery from any relapses

* More than two relapses in the first 2 years

* An EDSS ≥ 3.0 within 5 years of diagnosis

* High baseline MRI ≥ 9 T2 lesions

* Gd-enhancing lesions on baseline MRI scan

* Posterior fossa lesions (brain stem or cerebellum)

* Spinal cord lesions

* Two or more paramagnetic rim lesions or PRLs

* Brain atrophy on MRI

* Advanced brain age

* Retinal thinning on OCT

* Locally synthesised OCBs (oligoclonal IgG bands) in your spinal fluid

* Raised neurofilament levels in your spinal fluid or blood

* Low vitamin D levels

* Smoker

* Diabetes or prediabetes

* Hypertension

* Hypercholesterolaemia

* Obesity

* Failed a platform or low-efficacy DMT...