Intended for healthcare professionals

Quality Improvement Reports

Using rapid-cycle quality improvement methodology to reduce feeding tubes in patients with advanced dementia: before and after study

BMJ 2004; 329 doi: (Published 26 August 2004) Cite this as: BMJ 2004;329:491
  1. Carol Monteleoni, coordinator of speech-language pathology (cmonteleoni{at},
  2. Elizabeth Clark, director of division of geriatric medicine1
  1. 1 Lenox Hill Hospital, 100 East 77th Street, New York, New York 10021, USA
  1. Correspondence to: C Monteleoni
  • Accepted 3 June 2004


Problem Despite lack of evidence that enteral feeding tubes benefit patients with dementia, and often contrary to the wishes of patient and family, patients with dementia who have difficulty swallowing or reduced food intake often receive feeding tubes when hospitalised for an acute illness.

Design We conducted a retrospective chart review of all patients receiving percutaneous endoscopic gastrostomy or jejunostomy tubes between March and September 2002. QI interventions including a palliative care consulting service and educational programmes were instituted. We conducted a second chart review for all patients receiving feeding tubes between March and September 2003.

Setting 652 bed urban acute care hospital.

Key measures for improvement We measured the number of feeding tubes placed in patients with dementia, the number of feeding tubes placed in patients with dementia capable of taking food by mouth, and the number of feeding tubes placed in patients with dementia with an advance directive stating the wish to forgo artificial nutrition and hydration.

Strategies for change Medical and allied health staff received educational programmes on end of life care and on feeding management of patients with dementia. A palliative care consulting team was established.

Effects of change After the interventions, the number of feeding tubes placed in all patients and in patients with dementia was greatly reduced.

Lessons learnt Multidisciplinary involvement, including participation by the administration, was essential to effect change in practice. The intensive focus on a particular issue and rapid change led to “culture shift” within the hospital community. The need to establish unified goals of care for each patient was highlighted.

Background A growing body of research over the past decade has questioned the utility of placing feeding tubes (percutaneous endoscopic gastrostomy (PEG) or jejunostomy) in patients with advanced dementia.1 Studies have found no evidence that feeding tubes in this population prevent aspiration,2 3 prolong life,46 improve overall function,7 or reduce pressure sores.8 Additionally, the quality of life of a patient with advanced dementia can be adversely affected when a feeding tube is inserted. The patient may require wrist restraints to prevent pulling on the tube1 3 or may develop cellulitis at the gastrostomy site, develop decubitus ulcers,1 be deprived of the social interaction and pleasure surrounding meals,9 10 and require placement in a nursing home. Unfortunately, many doctors are unfamiliar with this literature or face barriers—attitudinal, institutional, or imposed by the healthcare industry—to applying its findings to their practice.11 Thus feeding tubes are placed in patients who will not benefit from this intervention and whose quality of life in the terminal stage of their illness will be adversely affected. With the expected increase of elderly people with dementia,12 a great change in doctors' knowledge, attitudes, and practice is necessary to prevent even greater numbers of patients receiving this futile treatment.


  • Contributors CM and EC participated in project design, data collection, data analysis, palliative care case consultations, in-service instruction, and other staff education in palliative care, and writing of this report. Julie Wityk participated in data collection and served as liaison with the PCQuIC faculty. Nancy Mooney, Tamar Kotz, and Barrie Guise participated in data collection. Walter Ettinger assisted with project design and served as the project's senior advocate within the institution. John Rapoport and Arthur Blank assisted with data analysis. CM is guarantor.

  • Funding Participation in the PCQuIC collaborative was supported by grants from Lenox Hill Hospital and the United Hospital Fund. These grants subsidised attendance at learning sessions and access to PCQuIC faculty for guidance throughout the course of the activity.

  • Competing interests None declared.

  • Accepted 3 June 2004
View Full Text

Log in

Log in through your institution


* For online subscription