Article Text

Download PDFPDF

Time for a prospective study to evaluate the Amber Care Bundle
  1. David C Currow1 and
  2. Irene Higginson2
  1. 1Discipline, Palliative and Supportive Services, Flinders University, Adelaide, South Australia, Australia
  2. 2Department of Palliative Care and Rehabilitation, King's College, London, UK
  1. Correspondence to Professor David Currow, Discipline, Palliative and Supportive Services, Health Sciences Building, Repatriation General Hospital, Daws Road, Daw Park, SA 5041, Australia; david.currow{at}, David.Currow{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Amber Care Bundle

Any new intervention will have benefits and burdens, some of which will be predictable and others unpredictable. Therefore, it is imperative that all new interventions are evaluated within a stringent framework to understand both the benefits and harms. The predictable effects must be measured prospectively, and the population who experience any adverse outcomes characterised in order to refine the use of the intervention. Once harms are identified, if they are unable to be mitigated then they can have risk management plans put in place or focus the intervention only on specific sub-populations who do not experience identified harms. Unpredictable harms need to be sought vigilantly and characterised in frequency and magnitude.

This approach is commonly applied to new medicines and increasingly to non-pharmacological treatments including new and modified systems of care. Unfortunately, robust evaluations of such complex interventions such as systems of care, care pathways, bundles, new services and quality improvement programmes are rare in any area of health care, including palliative care. Importantly, complex interventions require as much or more evaluation because their implementation is affected by context. Contextual factors, in this case, include resources and local key opinion leaders, staff training prior to implementation and knowledge of each clinician.1

Such a process was not put in place before the widespread uptake of the Liverpool Care Pathway.2 Despite having strong face validity, the outcomes from the programme were never properly evaluated in most jurisdictions in which the Pathway had been put in place. Opportunities for learning from the best of the pathway and ways of avoiding associated harms were lost. There is now a unique opportunity to evaluate rigorously the outcomes of a new clinical tool—the Amber Care Bundle.3 ‘Bundles’ were developed in 2001 by the Institute for Healthcare Improvement (IHI) in the USA. The ‘bundle’ concept is different from a care pathway in that a care bundle combines a small number of specific components or actions to improve care.4 The first care bundle from IHI aimed to improve critical care by enhancing teamwork and communication in multidisciplinary teams. The Amber Care Bundle was developed at Guys and St Thomas's Hospital UK, in a collaboration between palliative care and general medicine. It is designed for the times when clinicians are uncertain whether a patient may recover and are concerned that they may have a prognosis now limited to months. This Bundle “encourages staff, patients and families to continue with treatment in the hope of a recovery, while talking openly about people's wishes and putting plans in place should the worst happen.” It consists of four elements:

  • talking to the person and their family to let them know that the healthcare team has concerns about their condition;

  • establishing their preferences and wishes;

  • deciding together how the person will be cared for should their condition get worse; documenting a medical plan; and

  • agreeing these plans with all of the clinical team looking after the person.

The Amber Care Bundle has shown promise in pilot work but before widespread uptake, the Bundle needs formal prospective evaluation to ensure that it delivers the patient, clinician and health systems benefits hoped for, and that any harms are quantified, understood and managed. This especially so because pilot work (J Koffman, Personal Communication, October 2013) has suggested variability in its implementation.

Although such health services research costs money, the cost of not doing this type of research is unacceptable. No matter how good the idea, it is a false economy not to evaluate new programmes before there are policy-mandated multi-million pound implementation projects. Such evaluation has to have demonstrated improved outcomes in well-designed, prospective and, preferably, randomised studies before widespread uptake.

The Medical Research Council Guidance on evaluating complex interventions provides the perfect framework for this, especially now that these recommendations have been refined specifically for palliative and end of life care in the MORECare Statement.5

Evidence-based practice is a concept that has widespread applicability. Evidence-based policy is lagging seriously behind. After implementation, the central investment in evaluation is not over. As a programme is rolled out, its uptake and application into diverse settings requires continued evaluation. This can only happen when there is a sustained commitment by funders to ongoing assessment of net effects—both benefits and harms.



  • Competing interests None.

  • Provenance and peer review Not commissioned; internally peer reviewed.