For anyone that’s sick, understanding what the likely outcome is one of the first questions they want the answer to. Yet despite the millions of dollars poured into health research every year actually predicting outcomes is a surprisingly inexact science.
This week the PROGnosis RESearch Strategy (PROGRESS) Group, comprised of investigators from multiple countries who participated in three workshops (which also included one editor each from the BMJ and PLOS Medicine) and discussions, publish four papers in PLOS Medicine and the BMJ, (itself an unusual collaboration between journals ) that aim to better define the field of prognosis research.
Prognosis research is defined in the series as the investigation of the “relations between future outcomes (endpoints) among people with a given baseline health state (startpoints) in order to improve health”.
Prognosis research has yet to be identified as a distinct field of research and a major issue in this area is the lack of cohesion for methodology, analysis, and data reporting. Currently, there are a multitude of disparate studies claiming to have identified factors or models that predict future clinical outcomes in relation to current diagnostic and treatment practice. In a single clinical field there may be several studies that identify different factors or models for the same outcomes, but do not attempt to build on or even compare previously identified factors and models. In addition, individual studies are often underpowered or not replicated or validated in other populations. A good example of this can be found in a study by Malats and colleagues that analyzed 168 reports and concluded that “after 10 years of research [including over 10 000 patients], evidence is not sufficient to conclude whether changes in P53 act as markers of outcome in patients with bladder cancer.” In reality, there are very few prognostic studies with sufficiently robust methods to be clinically informative; common methodological problems include the use of retrospective study designs, study biases, and selective reporting of significant findings.
In the papers published this week the researchers provide a four point framework for prognostic research to take the field further:
(1) The course of health related conditions in the context of the nature and quality of current care (fundamental prognosis research)
(2) Specific factors (such as biomarkers) that are associated with prognosis (prognostic factor research)
(3) The development, validation, and impact of statistical models that predict individual risk of a future outcome
(prognostic model research)
(4) The use of prognostic information to help tailor treatment decisions to an individual or group of individuals with similar characteristics (stratified medicine research).
There are many reporting guidelines designed to help researchers improve study design and reporting of clinical research outcomes (e.g. STROBE, CONSORT, PRISMA, STARD). These four PROGRESS papers extend earlier work outlining prognosis models in clinical practice published in the BMJ in 2009, Ten steps towards improving prognosis research published in the BMJ in 2010, and a related set of recommendations that were developed specifically for prognosis research related to tumor markers (REporting recommendations for tumour MARKer prognostic studies, REMARK). The PROGRESS group’s recommendations highlight the importance of study registration, a published protocol, ideally a prospective approach, and an appropriate statistical analysis plan. In addition, the PROGRESS papers provide a broader overview of examples of prognosis research from cancer, cardiovascular disease, musculoskeletal disorders, trauma, and beyond to provide a more cohesive approach to research practices in the field of prognosis research and will help prognosis researchers to enforcing the high standards required in other fields.
The PROGRESS series has been published as a collaborative effort between the PROGRESS Group, PLOS Medicine and BMJ. The first and fourth papers are published in BMJ and second and third papers are published in PLOS Medicine and can be accessed in the links below: