How will a Biostatistical Expert Witness enhance your
medical malpractice, insurance, criminal or forensic case?
Initial planning stages:
We can conduct scoping reviews, systematic reviews and meta-analysis of existing medical and psychological literature.
Scoping review of the available evidence allows us to make new connections and draw together emerging evidence to further inform the focus of systematic reviews. They enable the identification of key factors and characteristics relevant to the situation or diagnosis that may not have been previously entertained by clinicians.
Systematic reviews allow us to collect and examine existing clinical evidence from peer reviewed journal articles in order to answer specific questions pertinent to the case.
Once a systematic review has been performed, meta-analysis of existing medical literature allows us to produce statistical estimates as to the likelihood of particular outcomes occurring given a particular circumstance.
An example of this may be to critically draw from numerous existing clinical research studies into a particular disease/misadventure in order to produce a likelihood of that disease/misadventure being produced due to an environmental exposure, given extenuating circumstances relevant to the present case. This way you get a more powerful estimate than any one study can provide. It also allows the incorporation of additional features to the estimate that might not be available in the existing research.
Statistical modelling using discriminative and generative models:
We can use existing clinical data sources such as from available clinical data sets or drawn from a range of sources, we can produce detailed statistical models, such as regression models, that predict: The likelihood of a specific outcome occurring due to a specified circumstance, or range of circumstances over time. These circumstances may include delayed, missed or misdiagnosis, injury or misadventure due to medical treatment,
The level of suffering, complications or incapacity that can be expected to occur as a result of a particular medical incident. Medical incidents might include surgical misadventure, delayed, missed or mis-diagnosis,
In some situations we can take innovative and highly sophisticated approaches to informing your case, such as neural network analysis, Bayesian generative models such as dynamic causal models or other latent variable models to uncover connections between variables relevant to the case.
We have the advantage of constant exposure to as well as design of a broad spectrum of the latest clinical research studies, as well as being in a superior position to evaluate them statistically.
Why is a Biostatistics Expert such a valuable asset to your team?
We keep abreast of a wide array of the latest in medical and psychological research and clinical practice. This enables us to see the wood for the trees, give a unique perspective and make connections that clinical practitioners may have missed.
We also have access to potentially hundreds of clinical researchers within highly specific fields which may be able to assist with cutting edge perspectives from which to inform our statistical models.
As each situation is different we are keen to discuss the details and develop dynamic, bespoke approaches to enhancing your case. We can combine multiple approaches to cover different aspects of your case, where appropriate. Our statistical models are not simply limited to the clinical or medical context. We can inform on criminal and forensic matters among others.
A peer-reviewed research paper mentioned in a previous blog post, found that up to 78% peer-reviewed research studies published the medical field contain statistical errors. As these published journal articles are what doctors use to keep up to date with the latest clinical knowledge and evaluate new treatment approaches, these figures are incredibly problematic. Without the specialised expertise in biostatistics it is impossible to discern these errors and be truly critical in evaluating the research findings of these studies. This is why relying on soely a medical doctor expert witness may be limiting your case as compared to a combination of both a clinician and biostatistician. While clinical practitioners have the daily lived experience of patient, practice and disease, biostatisticians have the capacity to synthesise this experience mathematically and impartially.
Recent research has shown that when a machine learning algorithm was fed Electrocardiogram (EEG) data from patients that doctors deemed perfectly healthy, the algorithm was able to predict death within one year due to cardiac problems with .85 accuracy, where .50 indicates chance. When doctors reviewed the EEG readings of dead patients post prediction, they were in many cases not able to identify patterns indicating as to why the patient had died. This shows the power of maths to predict outcomes and the advantage of current methods compared to medical opinion alone.