SCIENTIFIC EVIDENCE AND RESEARCH IN PRIMARY-CARE

The key areas of scientific research in general internal medicine are (1) prevention; (2) the natural history of common illnesses; (3) impro ving the outcomes and efficiency of the health care system and (4) orp han diseases. Disease prevention is at the top of the list because of the enormous role preventable causes play in morbidity and mortality, above all tobacco. Research in this field is difficult because it touc hes such questions as individual behaviour and personal choice, Resear ch in the natural history of common illnesses is critical to informed patient decision making. Recent studies show that procedures thought t o be safe bear a high percentage of complications, when viewed from th e generalist's point of view: high incidence of strokes after elective coronary bypass surgery; higher mortality rates among patients having had pulmonary catheterization; high incidence of incontinence and imp otence after transurethral resection of the prostate. A third area for research in primary care is how to improve outcome and efficiency thr ough improvements in the health care delivery system. This field touch es the problem of unnecessary surgical interventions and inappropriate prescription of antibiotics. Orphan diseases in this context are cond itions no speciality wants to study, such as dementia and low back pai n. The most important obstacle for research in the field of general in ternal medicine is funding. It is much easier to be funded for researc h in high profile conditions, like heart disease, cancer and AIDS. A s econd barrier to research relates to the role of special interest grou ps in influencing not only funding but also policy. Important examples were the pressure on consensus conference decisions on the role of sp inal fusion surgery for low back pain and on the question whether wome n between 40 and 50 should have annual mammography. For generalist res earch to be fruitful it is of outmost importance to have an adequate i ntellectual infrastructure, i.e. support by epidemiologists, biostatis ticians, economists and research methodologists.