Example of questions from the Decision Quality Instrument © for breast cancer surgery
16.3 Interventions to Improve Decisions
16.3.1 Why Intervene?
Reports of large geographic practice variations in surgery have raised questions about the quality of decisions about those procedures. For example, the rate of breast reconstruction varies fivefold across the country, raising the question of how patients and their surgeons decide who gets breast reconstruction. For some conditions, patients have reported high levels of regret about surgery, and for others, patients have reported wishing they had known more about their options. In an effort to reduce unwarranted practice variations and improve decision making, patient decision aids have been developed.
16.3.2 Decision Aids
A patient decision aid is defined by three primary components – provision of information, clarification of patient values, and preparation of the patient for interaction with the provider. The most common decision aid format is video, but a decision aid can consist of a piece of paper, a booklet, poster board, or website (Fig. 16.2). Decision aids are generally intended to be used by the patient prior to the provider visit. They can be used at home or in the health care setting just before the visit. They are intended to serve as an adjunct and not a replacement to patient-provider communication.
Video and paper decision aids from the Informed Medical Decisions Foundation
Over 80 randomized controlled trials of decision aids have been conducted, including many trials of decision aids for surgical decisions. Decision aid trials have found that decision aids were associated with higher patient knowledge, reduced decisional conflict, and greater satisfaction with decisions. In the subset of trials that measured treatment choice as a primary outcome, decision aids were associated with fewer invasive treatments and less surgery. Most studies of decision aids have been efficacy trials and not effectiveness studies, in which decision aids would be evaluated in clinical practice or the “real world” setting. A recent population-based study of decision aid implementation in a large HMO found lower rates of hip and knee replacement surgery . Recent state-level policies encouraging the use of decision aids may facilitate opportunities for studying decision aid effectiveness.
16.4 Criteria to Evaluate Decision Aids
The International Patient Decision Aid Standards (IPDAS) Collaboration has developed and published standards for evaluating the quality of decision aids. These include criteria for: the development process, how probabilities are presented, the use of patient testimonials, how decision aids are disseminated on the internet, and addressing health literacy. The IPDAS criteria have generally met wide acceptance in the medical decision-making research community and should be considered in any study to evaluate the quality of decision aids.
16.5 Opportunities for Surgeon Scientists
Surgeons have unique opportunities to contribute to medical decision-making research. Unlike many medical decisions, most surgical decisions take place at discrete times that are readily identifiable. For example, if you wanted to study satisfaction with decisions in patients undergoing surgery for bladder cancer, it would be feasible to identify dates of surgical consultations and dates of surgery, which could be opportunities for objective measurement or intervention. Similarly, the treatment options and potential outcomes for surgical decisions tend to be discrete, lending them amenable to study.
Few surgical decisions have been well-studied, leaving major opportunities for junior investigators to make a contribution and develop their research expertise. The ideal decision for a young surgeon to investigate is one for which some evidence exists, but for which clinicians disagree about ideal management or patients vary in preference for the procedure. For example, the decision about sphincter preservation in rectal cancer surgery draws on a growing body of evidence about efficacy. Surgeons differ in their judgment, however, about who is a candidate for sphincter preservation, and patients differ in how they feel about living with an ostomy. Many medical decision-making researchers who have methodological expertise are actively seeking opportunities to collaborate with surgeons who have front-line experience, clinical insight, and access to patients. Surgeon investigators who develop their own skills in medical decision-making research methods could build their academic career by creating a unique niche in surgical decision-making research.
16.6 What Is Decision-Analytic Modeling?
Decision science is used throughout many disciplines and is focused on assigning value and degree of uncertainty about a given choice. Based on defined assumptions, the best available evidence, and specified outcomes, you can identify the “optimal” strategy using decision analysis which is essentially computer-simulated decision-making or “modeling”. A key strength of decision analysis is the ability to apply logic to complex decisions, making it ideally suited for health care decision-making. In addition to assessing traditional medical endpoints, such as mortality or recurrence, decision analytic models can also factor in additional outcomes like cost, patient preference, quality of life, and quality of care. For instance, you can assess whether integrating a decision aid into clinical care affects clinical or quality of life outcomes. Modeling allows a provider to identify and quantify the trade-offs incurred with a specific intervention.
There are a number of types of analyses utilized depending on the perspective (i.e. patient, health-care provider, or payer) and the value (i.e. quality of life, survival, cost). One useful method is cost-effectiveness analysis (CEA). CEA is a form of comparative effectiveness research that assigns costs or resources to each competing strategy. As shown in Fig. 16.3, strategies that have a comparatively low cost and better outcome (bottom right) are preferred. It differs from a cost-benefit analysis in that the benefits or health consequences are not strictly monetary. Computer modeling or simulation is used to perform these, often complex, analyses.
Cost-effectiveness map, showing trade-offs between cost (y-axis) and effectiveness (x-axis)
16.7 Why Use Decision Analysis in Surgical Research?
16.7.1 Challenges in Surgical Research
The body of surgical literature has been criticized for being of poor quality – both in data and study design. In fact, few “surgical” publications are considered level I evidence . Many studies are underpowered, uncontrolled, and biased. This is due in part to the fact that (1) surgical diseases are less common and have heterogeneous populations, (2) there is variability in surgeon technique and institutional practices – making generalizability challenging, and (3) there is a general lack of formal epidemiologic and statistical training among surgeons. Historically, a surgeon’s “effort” was primarily clinical and surgical research was primarily focused on basic science. Until recently, surgeons were not supported in pursuing advanced training in statistics and epidemiology – vital to the development of a quality research program – and do not have the time or funding to run large-scaled trials.