Review of the literature concerning pressure ulcers in the intensive care setting. DATA SOURCE AND STUDY SELECTIONS: Computerized databases (Medline from 1980 until 1999 and CINAHL from 1982 until 1999). The indexing terms for article retrieval were: "pressure ulcers", "pressure sores", "decubitus", and "intensive care". Nineteen articles met the selection criteria, and seven more were found from the references of these articles. One thesis was also analyzed. Data on prevention, incidence, and costs of pressure ulcers in ICU patients are scarce. Overall there are no conclusive studies on the identification of pressure ulcer risk factors. None of the existing risk-assessment scales was developed especially for use in ICU patients. It is highly questionable to what extent these scales can be used in this setting as they are not even reliable in "standard care". The following risk factors might play a role in pressure ulcer development: duration of surgery and number of operations, fecal incontinence and/or diarrhea, low preoperative protein and albumin concentrations, disturbed sensory perception, moisture of the skin, impaired circulation, use of inotropic drugs, diabetes mellitus, too unstable to turn, decreased mobility, and high APACHE II score. The number of patients per study ranged from 5 from 638. The definition of "pressure ulcer" varied widely between authors or was not mentioned. Meaningful comparison cannot be made between the various studies because of the use of different grading systems for pressure ulcers, different methods of data collection, different (or lack of) population characteristics, unreported preventive measures, and the use of different inclusion and exclusion criteria. There is a need for well-conducted studies covering all these aspects.