Reflections from the Frontline

How I got into this field — and what nobody tells you at the start 

By Dr. Ram Kumar

I am a consultant paediatric neurologist, 19 years in post, still doing clinical work in the NHS and private practice in the North West. My field is brain injury, cerebral palsy, and epilepsy in children and young people. I have been doing expert witness work for thirteen years, starting after I had been a consultant for six or seven years. 

I did not stumble into medico-legal work through ambition or design. A colleague pointed out, rather practically, that it made no sense to start from a standing start at retirement. That observation lodged somewhere and eventually got me thinking. But knowing you have relevant experience and knowing how to translate that into expert witness work are entirely different problems. 

I had no contacts, no idea who the instructing solicitors were, and no clear understanding of what the work actually involved. I attended an Expert Witness Institute meeting about fourteen years ago — the content was largely about how to get paid. Useful, eventually, but I still had no route in. What actually got me started was an accident of personnel: my secretary had previously worked for a senior colleague who was well known in the field and had been practising for decades. I got in contact, worked with him as an associate for a couple of years, and gained the equivalent of several years of learning through mentorship that would otherwise have taken much longer by trial and error. 

My clinical and research work is in children with cerebral palsy, epilepsy, and brain injury. In the clinical world, this is a niche. In medico-legal terms, brain injury in children — including birth injury and traumatic cases — is the subject of major claims resulting in multi-million pound settlements, driven by the cost of round-the-clock care over decades of remaining life. The opinion of a paediatric neurologist answers questions with that level of financial consequence. 

That gap between clinical expertise and medico-legal competence is worth naming directly. You can be highly experienced in your clinical field and still be entirely unprepared for the expert witness role: the CPR Part 35 obligations, the concept of balance of probabilities, the difference between what is medically common and what is legally probable. I completed the Cardiff University-Bond Solon Certificate in Civil Litigation relatively early in my journey. It gave me a wider context than I would have acquired just from writing reports. 

What has actually changed in thirteen years — and what has not 

The most visible change has been technological, though not uniformly for the better. 

Thirteen years ago, I was still receiving some approaches by hard copy letter, and records arrived in ring folders, sometimes in several large boxes. I was already pushing towards paperless working before it became standard. The inefficiency of those arrangements was obvious — boxes delivered while no one was in, stored in a corner, then difficult to navigate when eventually opened weeks later. 

The Covid-19 pandemic accelerated changes that were already underway: telemedicine as routine clinical practice, electronic records, and a shift in expert witness work to remote conferences via Teams and Zoom. Face-to-face conferences in London chambers — involving a half-day of travel for a two-hour meeting — have almost entirely disappeared. This has been a genuine improvement. 

The volume and velocity of information has increased, I think exponentially. The findability of information has improved. It is much easier to send electronic PDFs at repeated intervals than it was to scan, print multiple copies, put them in folders, box them up, send by post, and then try to find the right page three months later. 

But technology has also created new problems. Electronic disclosure has removed the friction of postage — so far more gets sent, far more frequently, and much of it is noise. Files arrive at 1.8GB. PDFs crash readers. Electronic health records, extracted from EHR systems that were designed to be navigated on-screen, are flattened into documents that are practically unusable — repetitive boilerplate, computer-generated attestations, and occasional critical facts buried in the middle of hundreds of pages of template output. In some respects, this is harder to navigate than handwritten notes. 

It is a form of steganography — useful information hidden in plain sight, invisible because of the sheer volume surrounding it. You cannot see the wood because of the number of trees that have been sent. 

What has not changed: I still conduct face-to-face claimant assessments, and I expect this to remain the case. Much of my work involves physical examination and direct observation of behaviour. This is not amenable to video consultation, and it is one area where technology is not going to substitute for presence. 

The challenges that are structural, not technological 

Delays in obtaining reports are a persistent feature of this field. I and my colleagues are typically working full-time in clinical practice — often covering vacancies, managing on-call commitments, running research alongside. Expert witness work is fitted around this. It is not, for most of us, the day job. It is important work, but it occupies irregular hours, often evenings and weekends. 

Finding the right expert is harder than it looks. A clinician may have the right job title but not the relevant experience for the specific question at hand. They may have impeccable clinical standing but no familiarity with CPR Part 35 or with how their opinion will be tested in court. The converse also happens: someone with impressive credentials in a related area is instructed on questions that sit outside their actual expertise. 

In my field, questions about breach of duty require the expert to know what current practice was at the time of the alleged breach — not what it is now. That demands someone who was active as a practitioner at that date. Experience and currency are both necessary, and they are not the same thing. 

Credibility in this work requires ongoing involvement in professional practice. A case in its early stages may not come to hearing for several years. The knowledge base moves on. A clinician who has retired from active practice may be addressing questions about current standard of care from a position that is increasingly out of date. Maintaining credibility means staying in the field, not just drawing on prior experience. 

Increasing complexity of evidence is real, though it is partly a product of the electronic garbage problem. Two competent experts can read the same authoritative literature and reach opposite conclusions, each believing it supports their view. The scientific literature only resolves so many questions. In areas of genuine uncertainty, the quality and logic of the reasoning matters as much as the conclusions reached. 

What makes a report worth reading 

Readable. Clear. Not oversimplified when the underlying question is complicated. Opinions linked directly to facts or assumed facts. A description of where the range of expert opinion lies, and an explanation of where this opinion sits within that range and why. 

The job is not to help one side win. It is to assist the court. That distinction matters in practice, not just in principle. Going into a conference with a view that will not survive cross-examination serves nobody, least of all the solicitor who has spent years developing the case on that basis. 

I am not there to have my tummy tickled by the instructing solicitor, nor to tickle theirs. My opinion is for the court. If the case is not winnable from the medical evidence, it is better for everyone to know that early rather than late. 

Strong expert opinion is reproducible in a specific sense: another competent expert, given the same facts and the same literature, following the same logical steps, should arrive at the same place. If they cannot, the reasoning needs to be more explicit. Demonstrating that logic — tracing the opinion back to its evidential foundations — is what distinguishes an opinion that holds up from one that does not. 

Technology and where it is, and is not, useful 

I have not yet significantly changed my practice due to AI, primarily because of data privacy concerns. The tools I use are still largely separate from one another: email, cloud document transfer, Teams and Zoom for conferences, semantic search for literature review. They are functional but disconnected. 

Semantic search — search that understands intent rather than matching exact keywords — is genuinely useful for scientific literature work. I use it in my research practice. Whether it will find its way into clinical expert witness work depends heavily on how data governance issues are resolved. 

Technology is a driver for efficiency and reduced friction, but it is not all one way. The explosion of electronic disclosure has created its own friction. More is not the same as better. A purpose-built tool that helps manage the volume, provenance, and traceability of evidence would be more useful than an AI tool that tries to synthesise an opinion. The hard problem is not opinion generation. It is information management. 

The administrative burden on an expert running a practice — tracking case progress, managing disclosure versions, reconstructing context on cases that go dormant for a year then suddenly reactivate, keeping on top of conference scheduling — is substantial and mostly invisible to the instructing parties. It does not require AI to improve. It requires well-designed practice management software that reflects how this work actually operates, rather than generic case management tools built for solicitor firms. 

That gap is, in my view, the most tractable near-term problem. And it is currently unsolved. 

Looking ahead 

Whether demand for specialist expertise increases will depend partly on whether the court system can process cases more quickly. Court diary capacity is currently a significant constraint. Faster case resolution would increase throughput, but that is not primarily a function of expert availability — it is a court capacity problem. 

The direction of travel for technology in this field is clear. Electronic signatures, remote hearings, cloud-based disclosure management, and eventually some form of AI-assisted document review will all become more standard. But the fundamental problems — getting paid on time, managing the asymmetry between a large solicitor firm and a sole-practitioner expert, dealing with the signal-to-noise ratio of modern electronic disclosure — are not technology problems. They are structural ones, and technology on its own will not resolve them. 

What I would like to see is simple: purpose-built tools that fit the actual workflow of an expert witness practice. Not adapted CRM systems. Not generic legal case management software. Something designed around the reality of managing 50 or 100 cases at different stages simultaneously, across a multi-year lifecycle, often as a single practitioner with a part-time PA, while continuing to maintain an active clinical career. 

The work is complex, the cases are long, and the administrative layer is invisible to everyone except the expert carrying it. That is not a complaint. It is a description of a problem that has a solution.