Evidence-Based Treatment Planning: Assessment of Risk, Prognosis, and Expected Treatment Outcomes
The intent of this chapter is to frame the context for decision making in dentistry. We begin with a perspective on how dental treatment planning decisions have typically been made in the past and a discussion of the apparent limitations of that process. Recognizing that there is often inadequate knowledge at both the professional level and the patient-specific level upon which to base our treatment decisions, there is a call for sound clinical investigations to support clinical decision making. The emergence of evidence-based dentistry has significant implications for both the dental treatment planning process for individual patients and the design of parameters for decision making in the profession of dentistry. The importance of evidence-based dentistry is reinforced throughout this book in the What’s the Evidence? boxes.
Health promotion and disease prevention have become a focus of all the health sciences. Programs and practices put into place to promote these goals should be evidence based and should also include a careful assessment of disease risk and treatment outcomes. An analysis of both disease prognosis and treatment prognosis is also integral to this process. These three concepts risk assessment, outcomes assessment, and prognosis determination will all be defined and described, and their relevance to dental treatment planning will be discussed. The findings from research in these areas can be expected to both improve future dental treatment planning and have a positive impact on the oral health of patients. In the meantime, however, the concepts themselves offer the practicing dentist a systematic way to organize the process of assessing clinical problems and solutions. In addition, these concepts will provide a useful framework for presenting and discussing treatment options with the patient. In summary then, the purposes of this chapter are: to discuss the kinds of information necessary to help patients make informed decisions, to review some related areas of dentistry in which progress has been made, and to provide a template for dentist and patient decision making as more information becomes available. This chapter provides the foundation for the detailed process of treatment planning, which is delineated in Chapter 3 and then applied repeatedly throughout the subsequent chapters of this textbook.
Traditionally, dental students have been taught that creating a treatment plan for a patient incorporates a stepwise process: first, a thorough evaluation of the patient is conducted; then a diagnosis or problem list is developed; and finally, a plan for a series of treatments is constructed. Certainly this model has merit. Its rationale and virtues are discussed at length in other chapters, and it is the basis for the treatment planning process described throughout this text. But the model has shortcomings. One fundamental problem is how little the model has been implemented by practicing dentists. A second concern relates to the realization that the model is too simple and must be broadened to meet the diverse needs of patients and the profession.
As the dentist plans treatment for his or her patients, the traditional stepwise model may not be followed. Instead, a tooth condition or problem is evaluated and the dentist makes an immediate recommendation to the patient about what should be done to solve the problem.1 This is certainly an efficient way for the practitioner to gain a measure of consent from the patient to begin treatment. In this scenario, however, a clearly articulated diagnosis is often not made, and even in those cases in which the dentist makes a mental judgment about the rationale for treatment, the diagnosis may not be explicitly stated to the patient. Thus the patient remains relatively uninformed about the nature of the problem and the rationale for a particular treatment.
It is also unlikely in this situation that the patient will be presented with options from which the best treatment (for the patient) can be selected. Even when options are presented, the offerings tend to be perfunctory, with the patient given minimal information from which to make a well-reasoned, thoughtful decision. Given the time pressures of a daily dental practice, these omissions can evolve to become the routine rather than the exception. The patient, who remains uninformed about diagnoses and treatment options, however, is ill prepared to provide informed consent for treatment. Obviously, this can be both unwise and hazardous from a risk management perspective (see Chapter 4). The need to achieve fully informed consent is a central theme of this chapter and this text.
The second significant problem points to a defi-ciency not of application, but rather with the scope of the traditional stepwise model. Although when used to its fullest, the model helps ensure that the dentist considers and the patient is informed about all diagnoses and treatment options, the model does not include the potential or capability for any weighting of the relative benefits of the various treatment options. To accomplish this weighting and to provide a rational comparison of the options, more information is needed by the dentist.
Few dentists have difficulty making treatment recommendations for their patients. Typically, those recommendations are based on what the dentist learned in dental school and what he or she has gleaned from continuing education activities, reading dental journals, and discussions with peers. In addition, the dentist may be exposed to other, less objective, and often empirical sources of information. For instance, product salespersons or dental supply representatives who are selling products or promoting techniques may not have valid supporting evidence for the rationale and efficacy of their products. For many dentists, the most important basis on which they make their treatment planning recommendations is their own personal experience with a specific approach or technique.2 The wise dentist will recognize the limitations and hazards of this approach. In the absence of scientific scrutiny, old, sometimes misguided, approaches have been perpetuated, while new and untried (by the practitioner) approaches are rejected out of hand. When the dentist fails to offer and make available a complete range of treatment options to patients, or to accurately characterize the viability of each option, the quality of care provided to patients is diminished.
The reality at this time is that many treatment decisions in dentistry must be made in an environment of uncertainty. Our ability to make an accurate diagnosis, realistically predict outcomes of treatment, and delineate with precision the course of the disease in the presence or the absence of treatment is in many cases limited. That being the case, it behooves the dentist to place the ethical principle of nonmaleficence, or “do no harm,” in a position of preeminence. Stated another way, when the reasons for intervening are not conclusive or compelling or when the risks of “no treatment” do not have clear and demonstrable significance, then conservative therapy or no therapy should usually be recommended over aggressive therapy. The nature of the disease process certainly has a bearing on this analysis. Where there is diagnostic and/or treatment uncertainty but the disease has significant morbidity or mortality, as with oral cancer, aggressive intervention is generally warranted. On the other hand, when there is similar diagnostic uncertainty (as with incipient dental caries in a patient at low risk for new caries), and the short-term probability of negative sequelae (fracture, pulpal disease, or periapical disease) is low, then intervening slowly and conservatively is more professionally reasonable.
It has been well established that dentists frequently differ with one another on plans for treatment.2,3 When several dentists examine the same patient under the same conditions, they often disagree about the substance of the plan and on which teeth and surfaces need to be restored. Arguably, this is not necessarily a problem. If different practitioners could demonstrate with comparable positive outcomes measures that their individual plans were equally effective, there would be no reason for concern. But this optimistic scenario is unlikely. Presumably, one treatment plan would be found to have a better outcome than the others if all could be followed over an extended period of time. In reality, several plans may yield acceptable results, while a few plans would definitely be inferior or incorrect. The appropriate goal then should be to identify the inappropriate plans. Before discussing ways to achieve this goal, the reader should have a clear understanding of why clinicians disagree.
Dentists may disagree about diagnoses for a patient. These differences sometimes exist at the professional level. For example, historically, there has been significant disagreement among dentists as to how temporomandibular disorders (TMDs) should be assessed, diagnosed, and managed. Differences can exist on a patient level. Unfortunately, some dentists have underdiagnosed the occurrence of periodontal disease in their patients which has in some instances become the cause for malpractice litigation. Differences can also exist at a patient-, tooth-, or surface-specific level.2,4 Different practitioners examining the same patients frequently differ in their diagnoses regarding caries and restoration defects. There are multiple reasons for these differences the information base collected by each dentist may differ; the interpretations may differ; and the diagnostic options considered by each dentist also may differ.
Even with conditions as pervasive as dental caries, our diagnostic tests are imperfect. Bader and Shugars in their systematic review of dental caries detection methods found a mean sensitivity of 59 and a mean specificity of 72 for visual detection of occlusal carious lesions irrespective of lesion size; a mean sensitivity of 39 and a mean specificity of 94 for visual-tactile detection of occlusal carious lesions irrespective of lesion size; and a mean sensitivity of 50 and a mean specificity of 87 for radiographic detection of proximal carious lesions irrespective of lesion size.5 Newer diagnostic techniques can improve those numbers, but may still have notable error rates. In one study of the detection of occlusal carious lesions, the laser fluorescent measuring device (Diagnodent) yielded improved sensitivity (94%) but lower specificity (82%) when compared with expert examiners using conventional diagnostic techniques.6 Statistical measures of diagnostic accuracy (e.g., sensitivity/specificity ratio; percent of false positives) are also not ideal for commonly used clinical and radiographic caries detection methods. False-positive diagnoses for caries are particularly troubling as they may lead to unnecessary treatment of the patient.
If accurate disease- and patient-specific risk assessment either is not available or is not used by the dentist, treatment plans that take risk factors or indicators into account will not be developed and as a result may vary widely among dentists. Failure to properly assess a patient’s risk for caries, for example, can lead to improper treatment recommendations, restorative overtreatment, or failure to control the Streptococcus mutans infection.
Prognosis is the forecast of the probable course or outcome of a disease; or a prediction of the probability of success of a treatment. In the absence of accurate patient-, disease-, tooth-, and treatment-specific prognoses, treatment planning depends on individual clinical experience, and the benchmark for success becomes “what works in my hands.” Unfortunately, this is an unstable base on which to attempt to build consensus on treatment planning. The lack of evidence-based prognosis determination leads to errors in planning for the individual patient and impairs professionwide attempts to establish treatment parameters.
Outcomes measures provide a quantifiable and standardized method for comparing treatments. This is especially helpful in relation to oral conditions, such as TMDs for which there have historically been many different and sometimes conflicting treatment modalities. Unfortunately, for many dental procedures, outcomes data are not available. Even when available, however, many practitioners do not choose to make use of them. In either case, dentists have no dependable method of predicting which treatment is most reliable and likely to function the longest. Attempts to develop professionwide treatment parameters have been slow to emerge. As a result, the individual dentist is placed in the uncomfortable position of making judgments based primarily on empirical wisdom drawn from what has worked best in the past in his or her own practice.
It has been confirmed that when several dentists each independently examine the same patient under controlled conditions, each may interpret findings and the patient’s wishes regarding treatment differently.7 There are probably several plausible explanations for why this variability occurs. The dentist may be making assumptions about the patient’s desires and listening selectively, or the dentist may have a preconceived idea about what the ideal treatment should be and then may present that plan in a more favorable light.
If the reader accepts the premise that more and better evidence will assist the dentist and patient in making sound treatment decisions, achieving more completely informed consent, and delivering and receiving higher quality care, then the question becomes, “What information is needed to achieve these ends?” A series of parallel questions can be developed that, when addressed, will meet the needs of patients and practitioners. They are framed here from the patient’s perspective, but each can also be asked from the perspective of the dentist as the care provider for a specific patient. Each patient then has the right to ask and deserves to receive answers to questions such as:
Unfortunately, these pieces of information are rarely available for any particular treatment option not to mention for all possible options in a given clinical situation. Thus in many instances, the dentist must recommend treatment based on empirical information and personal experience, and the patient must make a corresponding decision based on that information.
The concept of evidence-based decision making is now well established in all of the health sciences. It has been defined as “the integration of best research evidence with clinical expertise and patient values.”8 In essence, it entails the view that clinical decisions should be based on scientific principles and that treatment regimens must be tried, tested, and proven worthy by accurate, substantiated, and reproducible studies. Ideally, any treatment method, whether in dentistry or medicine, should be supported by controlled, blinded, prospective longitudinal studies. The adjoining What’s the Evidence? box discusses in some detail how the dental practitioner can locate, evaluate, and derive clinically relevant information from the literature upon which to base appropriate treatment recommendations and plans.
The importance of evidence-based decision making is a recurring theme throughout this chapter. Evidence-based dentistry (EBD) has application to treatment planning on several levels. At the most basic level, research can sometimes provide compelling guidance to the patient and practitioner on the treat versus not to treat question. In other situations, for example, when several different viable alternatives are being weighed, it can provide the basis for moving to a specific decision. The application of EBD must be also be tempered by an understanding of its limitations, however. The strength of the evidence needs to be considered as it is factored into the decision making. The stronger the evidence, the more seriously it should be weighed and conversely, the weaker the evidence, the more other factors should drive the decision making.9
The growing importance of EBD in contemporary dental treatment planning cannot be overemphasized. Indeed, it can be anticipated that, in the future, all treatment planning will be based on such sound scientific principles and the body of knowledge that emerges to affirm or disavow the efficacy of various dental treatments.
Not all patients are equally likely to develop a particular disease. Some patients, because of heredity, environment, diet, personal habits, systemic health, medications, and other factors, are more likely than others to develop and continue to suffer from certain conditions. Those patients who have that innate predisposition or who engage in behaviors known to promote a particular disease or condition are described as at risk. This differs from the epidemiologic definition of “at risk.” In epidemiology, anyone who could potentially get the condition is “at risk.” Individuals who could not get the condition are “not at risk.” Edentulous patients, for example, would not be at risk for caries development; but everyone who has at least one natural tooth is at risk for caries development. This distinction is important in determining the denominator for incidence and prevalence estimates. But in both realms, clinical and epidemiologic, someone who is more likely to get the condition is “at high risk.”
Identifiable conditions that, if present, are known to be associated with a higher probability of the occurrence of the disease are designated as risk indicators. Risk factors are a subset of risk indicators for which there is a demonstrable causal biologic link between the factor and the disease. Risk factors are best confirmed by longitudinal studies during which patients with the hypothesized risk factor are evaluated over sufficient time to determine whether they (the patients) do or do not develop the specific disease or problem in question. Risk indicators may be identified by taking a cross section, or sample, of individuals and looking for instances of the risk indicator and the disease occurring together.10 Many risk indicators, although not yet confirmed by longitudinal study as risk factors, are useful in dental treatment planning.
Although risk and causality may be linked, they are not the same. Some risk indicators do have a causal relationship with the disease. For example, a diet that is heavily weighted with refined sugars constitutes a risk factor and is also a direct cause of dental caries. Other risk indicators can help identify individuals at risk, but do not themselves cause the disease process. For example, adolescents and the elderly are known to be at greater risk than other age groups for developing dental caries, but age per se is not a direct cause of the decay in an individual. Similarly, although growing up in an area that does not have fluoridated water represents a risk indicator for dental caries, the lack of fluoridation is not itself a cause of caries. Instead, the lack of exposure to systemic fluorides makes the individual more vulnerable to the combination of multifactorial issues that lead to the caries infection and subsequent demineralization process.
Another categorization that is particularly useful in a dental setting is the differentiation between mutable and immutable risk factors or risk indicators. Mutable risk indicators (such as diet, oral self care, smoking, poorly contoured restorations) are those that can be changed, and immutable risk indicators (such as age, socioeconomic status, and fluoride history) are those that cannot. The dental team can and should employ any and all reasonable interventions that have the potential to mitigate or eliminate mutable risk indicators. In the case of immutable risk indicators, however, the value of their identification may be limited to risk assessment and prescribing preventive therapy which can in themselves be useful tools in health promotion and oral disease pre/>