As increasing numbers of people worldwide live with one or more health problems, the study of prognosis has never been more important. Prognosis research provides information crucial to understanding, explaining and predicting future clinical outcomes in people with existing disease or health conditions. It provides pivotal evidence to inform outcome prediction, clinical decision making, design and evaluation of stratified medicine (stratified care), and all stages of translational research from molecular biology to health policy.

What is prognosis research?

In clinical medicine, the term prognosis refers to the risk of future health outcomes in people with a given disease or health condition. Prognosis research is thus the investigation of the relations between future outcomes (‘endpoints’) among people with a given baseline health state (‘startpoint’) in order to improve health.

The study of prognosis has never been more important as globally more people are now living with one or more disease or health-impairing condition. Currently, prognosis research studies too often fall a long way short of the high standards required in other fields, such as therapeutic trials. Common failings include poor study design, retrospective and ‘data dredging’ investigations, incorrect statistical analysis, selective reporting and publication biases, and limited translational impact on patient care.

Better prognosis research is urgently needed to enhance the quality and translational impact of prognosis research findings for clinical practice and improved patient outcomes.

A framework for prognosis research

The PROGRESS Partnership has outlined a framework for prognosis research, published as a series with BMJ and PLoS Medicine (see  the Publications page for more details). This framework has four key elements:

1. Fundamental prognosis research

  • In specific clinical contexts, fundamental prognosis research aims to examine the average prognosis of patients, often called their ‘baseline risk’.
  • It provides initial answers to the question “What is the prognosis of people with a given disease?”, and so quantifies the impact/quality of current care, and motivates & prioritises further inquiry.
  • Example: In 2006 an average of about 15% of people aged > 65 admitted to a US hospital with a heart attack died within 30 days, compared with 19% in 1995.

2. Prognostic factor research

  • A prognostic factor is any measure that, among people with a given health condition, is associated with a subsequent clinical outcome.
  • Example: In breast cancer patients treated with tamoxifen, tumour grade is a prognostic factor: survival times are shorter in those with a higher grade.
  • Prognostic factors help define disease at diagnosis, inform clinical and therapeutic decisions, enhance the design & analysis of intervention trials, and help identify targets for new interventions that aim to modify disease course.
  • But…there are major limitations in current prognostic factor research, such as publication bias & inadequate replication of initial findings.

3. Prognostic model research

  • Prognostic models utilise multiple prognostic factors in combination to predict the risk of future clinical outcomes in individual patients.
  • A useful prognostic model provides accurate predictions that inform patients and their caregivers, supports clinical research, and allows for better informed decisions to improve patient outcomes.
  • Prognostic model research has three main phases: model development, external validation, & investigations of clinical impact.
  • Most publications describe model development; a small number report external validation studies; very few consider clinical impact or usefulness.

4. Stratified medicine research

  • Stratified medicine involves tailoring therapeutic decisions for specific, often biologically distinct individuals with the aim of maximising treatment benefit and reducing treatment-related harm.
  • A key part of stratified medicine research is to identify tests (such as biomarker levels or genotypes) that predict an individual’s response to treatment and enables clinicians to identify patients for whom treatment is (most) effective.
  • Example: Trastuzumab is most effective in positive, and not negative, HER-2 breast cancer patients.
  • The clinical use of such tests is small; current flaws in study design, analysis, and reporting can lead to spurious evidence either for or against a test.