DOI: 10.1097/01.prs.0000179356.79256.a7
,
,
PMID: 16217509
Issn Print: 0032-1052
Publication Date: 2005/10/01
Facebook
Twitter
LinkedIn
Email
 |
Checking for direct PDF access through Ovid | |
Excerpt
Formalizing the decision-making process reduces potential bias and provides a process that is transparent to both residents and patients. This process has been used successfully in elective hysterectomy selection, to increase the number of patients proceeding to hysterectomy via a vaginal as opposed to an abdominal route, with benefits in terms of improved surgical outcome and reduced costs.2 Since plastic surgery commonly involves complex decisions that take into account not only potential surgical outcomes but also patients’ expectations, there is a clear potential for using decision trees to formalize the process of selection and provision of information for different conditions. The basic concepts used in making a formal decision tree are illustrated in this editorial using the example of reduction mammaplasty.3
There are four main steps to decision analysis:
Plastic surgeons may recognize the opportunity for error in assigning values to the probabilities in a decision tree. They reason that the technique encourages decision making based on small differences in expected values that are estimates at best. The defense against this concern, which also has been recognized by decision analysts in business and engineering, is the technique known as sensitivity analysis. In sensitivity analysis, a risk assessment is made for the best and worst cases based on evidence from the literature and the surgeon’s own experiences, and defined in terms of quality-of-life outcomes. This can be used to test the potential value of a decision for different scenarios to see to what extent a choice can be justified. The patient then has the information on which to either decline or accept the risk.
Bilateral breast reduction is a commonly performed procedure with good evidence for successful outcome in terms of functional improvement, notably relief of back and cervical pain. However, as with all surgical procedures, there are associated risks, and plastic surgery generally is expensive compared with other forms of surgery, although recent evidence has suggested a clear psychological benefit for some patients despite the greater cost relative to other surgical treatments.4,5 According to evidence from the United States and the United Kingdom, public demand for this operation has increased while funding has decreased.6 This increasing pressure means that it is important to provide clearly defined evidence of benefit, particularly for procedures that are life enhancing rather than life saving. Formal decision-making methods may be a useful way forward.
Using the above-described conventions of decision analysis, a decision tree was created (Fig. 1). A square box represents a decision node, and each line emanating from a decision node represents a possible action. Events under the control of chance are represented diagrammatically in the decision tree by a circular chance node. Each line represents a possible outcome, and associated with each line is the probability of the outcome occurring.
In the decision tree shown in Figure 1, these probabilities are based on evidence. A single patient had one or many outcomes for which the average value of the all outcomes was calculated.
