Building the state of the art
The state of the art is what a device's safety and performance are argued against, and the base the clinical evaluation is built on. Under the MDR it is what is currently medically accepted, not what is most advanced. Establishing it takes a description of the condition and of the treatments and devices already in use, a literature search of its own, and a conclusion that the Summary of Safety and Clinical Performance can reuse.
What the state of the art is
The state of the art is a fundamental input to clinical evaluation, and one of its most important yet misunderstood concepts. It has also been called the most important part of the clinical evaluation process. Under the MDR it is what is medically acceptable, that is, what is currently accepted for similar devices and alternative therapies, and the whole clinical evaluation is built on it. The Oxford English Dictionary defines the term as belonging or relating to the latest or most sophisticated stage of technical development, having or using the latest techniques and equipment. In the MDR's sense, though, the state of the art is not the most advanced technological information available on similar devices. It is not necessarily the most technologically advanced solution, but embodies what is currently and generally accepted as good practice in technology and medicine. It covers existing treatments, competing technologies, clinical practice and, as a very important part, expected outcomes.
The state of the art in the clinical evaluation
Establishing the state of the art is a crucial aspect of the clinical evaluation that the EU MDR 2017/745 requires not only to obtain CE marking but also to maintain it. For one writer of clinical evaluation reports under the MDR, the state of the art is an extremely strategic part of the report: its kickoff, with enormous regulatory implications for every statement written after it. That writer goes into great detail in it.
Building the state of the art starts from the medical background and then moves to similar devices and alternative therapies, to find out, for example, which scores doctors use to evaluate such devices and which thresholds medical practice accepts. Those scores and thresholds can then be used to evaluate the device, which is the role and the most important part of a state of the art. The state of the art states the performance and safety thresholds of similar devices, from which what is accepted for the device is planned, and the device's validation is planned in the clinical evaluation plan. The clinical evaluation report then justifies that the device is safe because the criteria evaluated accord with the state of the art and lie in the same safety interval as other similar devices.
In the clinical evaluation report, the state-of-the-art section establishes a baseline for how the device compares with existing therapies for the specific medical condition that falls under its intended use. State of the art in this sense means that the device is currently clinically accepted. The section has its own scope, with clearly described goals, a methodology for the literature search, results and conclusions, and it is outlined in the clinical evaluation plan and focused. The state of the art plays an essential role in determining a device's development strategy and features prominently in both the clinical evaluation plan and the clinical evaluation report.
What the state of the art covers
The state of the art needs to cover two main topics: the clinical background of the medical conditions that the device under evaluation targets, and a description of the historical and alternative treatments together with the clinical evidence on benchmark devices.
The clinical background describes the natural course and consequences of the medical conditions concerned. Documenting their clinical forms, stages and severities is advisable, and the frequency of the condition should be described in the general population and in subpopulations defined by age, gender, ethnicity, genetic predisposition and the like. The section should also describe the condition to be treated, provide epidemiology data, explain how the disease is classified and managed, justify the choice of clinical endpoints and identify potential clinical hazards. The current-knowledge section is a broad description and assessment of the epidemiology of the disease, of its diagnosis and pathology, including disease classification, of treatment guidelines, and of the objectives and endpoints used in clinical investigations, and the state of the art is determined once all of this has been reviewed.
The questions the section answers include what the gold-standard treatment is, what other treatments are available, including medical, surgical and other alternative forms of treatment for the target condition, what the advantages and disadvantages of these treatments are in different patient groups, what the benefit-risk profile of other devices and treatments is, and how the subject device compares with the state of the art. Information about competitor products and equivalent devices should also be obtained, and any knowledge gaps identified. Among other questions, the current-knowledge section should address what disease the device is intended to treat; how the condition is currently treated, for example by other devices or by surgical, pharmaceutical or non-medical treatments; which treatments are suitable for which patients; whether the available treatments leave problems or unmet clinical needs; and what treatments are in development.
For the alternative treatments and benchmark devices, MEDDEV 2.7/1 revision 4 lists the topics the review should cover:
- a description of the available therapeutic, management and diagnostic options, with their historical context and developments
- a summary of the advantages and disadvantages of the options, with their benefit-risk profiles and limitations in relation to the clinical forms, stages and severities of the conditions and to the different target populations
- a description of the benefits and risks, by nature, extent, probability, duration and frequency, and of the acceptability of undesirable side effects and other risks, including the nature, severity, probability and duration of acceptable harm
- hazards due to substances and technologies that could be relevant to the device under evaluation, with the mechanisms of harm and the clinical aspects of minimising and managing side effects and other risks
- the types of users
- diverging opinions of professionals on the use of the different medical options
- unmet medical needs
The literature search
A literature search is required to determine the state of the art for the subject device, and it is conducted separately from the systematic search that appraises that device, which is the search for clinical data on its safety and performance. To cover the clinical background and the alternative treatments and benchmark devices, a documented literature review should be performed, in which the pertinent literature is selected and then analysed, evaluated and reported. Because the search for the state of the art is intended to define it, its terms should be broad and its time frame recent, for example up to two years. Another account also calls for broad terms and a short time frame, but one of generally two to five years back from the current date. The results of the two searches are documented separately. The bibliography of the state of the art should be kept separate from that of the literature review, even where the two overlap.
The MDR's requirements for the state of the art are almost non-existent. It mentions the state of the art several times but gives no concrete description of how to establish one, and suggests only that establishing it is to present the context of the medical condition the device targets. MEDDEV 2.7/1 revision 4 recommends a brief summary and justification of the search strategy used to retrieve information on current knowledge and the state of the art, including the sources used, the search questions, the search terms, the selection criteria applied to the output of the search, the quality control measures, the results, the number and type of literature found to be pertinent, and the appraisal criteria used. The International Medical Device Regulators Forum's 2019 technical document on clinical evaluation gives further information on the state-of-the-art literature review. Establishing a state of the art that complies with the MDR is complex, requires a strict methodology in stages of data identification, appraisal and analysis, and demands time, experience and expertise. Technically it requires what researchers routinely do, gathering, analysing and reporting scientific data from the literature, but one writer who came from academia needed much practice, and many phone calls with medical writers specialised in the field, to master every aspect of the methodology.
There is no "magic formula" for writing a state of the art: every device is different, so the approach and methodology will vary. Some key points are common to every state of the art: an objective approach, a literature search adapted to its context, such as the type of medical condition and the quantity and type of data available, and a justified methodology, for example in the choice of database, the quality control and the inclusion and exclusion criteria.
What the search covers
The literature on the clinical background of a condition can be vast, and the prevalence of a disease can change significantly with time, so it is important that the search targets the latest studies, to reflect only current knowledge of the condition. For the clinical background, the search covers the most up-to-date epidemiology and aetiology studies and, on the internet, the most up-to-date evidence-based standards, guidance documents and expert opinions on the condition. For the alternatives, it covers historical treatments, the conservative, surgical and medicinal options available to the target population, and existing and benchmark devices. It searches the internet for the most up-to-date applicable standards, guidance documents and expert opinions on the current standards of care, and searches the clinical safety and performance outcomes of those devices in an unbiased and systematic way. Practice and consensus guidelines, health technology assessment reports, systematic review databases such as Cochrane, and the websites and registries of competent authorities can be useful starting points for writing the state-of-the-art section.
Guidelines and standards
Public health agencies, such as the World Health Organization and the European and American centres for disease control and prevention, and medical societies or associations specialised in a field or condition, such as the European Hematology Association, the European Glaucoma Society and the American Cancer Society, publish and update evidence-based guidelines regularly. These guidelines and standards typically contain the latest knowledge on a medical condition and are the highest-quality source of references. Besides the clinical background, they can also cover most of the topics required for the part of the state of the art on alternative treatments and benchmark devices. One writer's very first step in preparing a state of the art is to find the latest evidence-based guidelines and standards. That writer treats guidelines published in the last five years as up to date, uses and cites only the latest version, and usually finds them with search terms such as "diabetes AND (guidance OR guideline OR standard)"; they can also be found on the websites of the agencies, societies and associations. Some documents carry guideline, guidance or standard in their title without being officially written by a medical body, so it is important to check the authors' affiliation as a quality control.
Review articles
Guidelines do not exist for every medical condition, nor do those that exist cover every aspect of one, and other references, such as review articles, can then be considered. Review papers can be found by searching a scientific literature database such as PubMed for the name of the condition, and a selection criterion can be added to target review papers only; to find specific data on a condition's prevalence or current standard of care, the search can be narrowed by adding the term "prevalence" or "standard of care". For the treatments, one writer uses straightforward terms such as "[medical condition] AND treatment" with a selection criterion for reviews. It is important to make sure that review articles are free of bias, which is why their authors' affiliations should be checked for potential conflicts of interest. It is crucial not to rely on a single review article, since it may not cover all the information, so data from several should be combined to describe the range of existing alternative treatments and their benefit-risk profiles. An additional search under a treatment's name can give more on that treatment, and ClinicalTrials.gov is a useful tool for finding new clinical trials in a therapeutic area that use more recently developed devices.
Clinical evidence on benchmark devices
Clinical evidence on benchmark devices can be found in clinical studies, meta-analyses or systematic reviews where the devices are well established. A systematic literature review should gather all the favourable and unfavourable clinical evidence on the safety and performance of the benchmark devices identified. Its search terms should be broad enough to capture all the clinical evidence on a benchmark device but not so generic that they return unrelated hits, and a published clinical study most often names the device's brand in its abstract, but not always. The search period can start from the date the device was commercialised, so that every safety and performance outcome reported in the literature is gathered. All titles and abstracts of the results should be pre-screened, and one writer excludes at this stage the results that do not refer to the benchmark device or are not related to its safety or performance; the full documents of the results that remain should then be obtained for appraisal.
The labels, instructions for use and implant registries of benchmark devices should be searched as well, where available, to establish their benefit-risk profiles. They can be found on the manufacturer's website or by internet searches.
Appraising the state of the art
In one practitioner's view, the clinical background in the state of the art need not be appraised extensively, because it is not used to assess the device directly. In that view, the appraisal method in MEDDEV 2.7/1 revision 4 and the International Medical Device Regulators Forum's technical document applies mainly to clinical data on the device under evaluation, similar devices and other devices, and is guidance that can be adapted. The same practitioner puts into the state of the art every guideline that public agencies and medical agencies or associations have published on the medical condition, as the highest level of clinical evidence. For review articles and epidemiology studies, the practitioner tries to detect any potential bias, such as subjective statements in favour of a treatment, and inadequate disclosure of information. It is also important to check the authors' conflicts of interest, to ensure that an unbiased clinical background is presented.
Writing the state-of-the-art section
The state-of-the-art section of a clinical evaluation report places the device in the context of the relevant treatments; it often feels forced and out of place, and it should be linked back to the rest of the document. In one example from the state-of-the-art section of a report on a guidewire device, the focus is stenting, not any one stent, and the reason for each development stands beside its claim: balloon-expandable stents are now used for nearly all coronary stent procedures because of the technical limitations of self-expanding stents and their tendency to provoke greater neointimal hyperplasia, drug-eluting stents were designed to inhibit neointimal hyperplasia because restenosis is a significant problem with bare-metal stents, and the risk of very late events with permanent metallic stents led to bioresorbable scaffolds.
The analysis should report every aspect of the clinical background described in the sources identified, and the review of the alternative and historical treatments should reflect every aspect those sources cover. It is essential to report a quantitative analysis of each clinical outcome reported for the benchmark devices. Similar outcomes should be reported in the state of the art and in the section on the clinical evidence of the device under evaluation, so that the device's safety and performance can be assessed against the evidence on the benchmark devices. One writer tries to add as much quantitative data as possible, to make the analysis objective, and typically reports the benchmark evidence in a table built on the PICO criteria: populations, diseases or conditions; interventions; comparator groups or controls; and outcomes or endpoints. That writer summarises only the main information in each column, without overpopulating the table, to give a clear view of the clinical studies on the benchmark device. It is important to detail every relevant outcome or endpoint used in each study and to give the key results, ideally all in the same units, so that they can be compared between benchmark devices and, most importantly, with the device under evaluation. Where a study gives a result in a different unit, that writer reports it as published and, when possible, adds a conversion to a more commonly used unit. A conclusion after the table should give the number and types of studies and the key outcomes or endpoints, and that writer also notes there how homogeneous the studies are, which helps draw scientifically sound conclusions.
Reuse in the Summary of Safety and Clinical Performance
Two writers prepare the state of the art for reuse in the Summary of Safety and Clinical Performance while writing the clinical evaluation report. One either includes a summary of the state of the art at the end of the report or prepares it knowing that it will be used later. The other has begun to write a conclusion to the state-of-the-art section of the report so that the summary can use that conclusion alone. That writer finds this quite helpful, and the habit was likened to tidying up the document and the writer's ideas before moving forward.