by David W. Chambers, EdM, MBA, PhD
Practitioners are under an ethical obligation to experiment with a goal of continuous improvement.
There's a bargain between patients and dentists. The dentist promises to help the patient and avoid causing harm. In exchange, the patient agrees to follow the dentist's directions and to pay the fees. This is a tacit agreement sanctioned by society, and it is unethical for either party to renege.
But there are some problems interpreting this ethical imperative. There may even be an inconsistency between the admonition for beneficence (helping) and the admonition for nonmaleficence (not harming). Discomfort is an inevitable part of dental treatment, and the best way — the only way — to completely avoid causing harm through accident is to refrain from treatment entirely. There is also a limit on putting the patient's interests first. High on many patients' lists of desirable features in health care would be to have the work provided at no fee. Beneficence and nonmaleficence are relative principals.
Although providing benefit and avoiding harm are not absolute, they are certainly more than slogans. A patient has every right to be suspicious of a dentist who says, "I will try my best to help you and I will try my best to avoid harming you." In turn, the dentist would be justifiably suspicious of a patients who said, "I will take your advice when convenient and I hope to be able to pay you."
The way out of our ethical difficulty is to realize that beneficence and nonmaleficence are not isolated acts on the professional's part. Dental procedures take place in a context, and it is the whole context that matters, not the outcomes of individual procedures. The ethical imperative is to prepare to give optimal care and avoid putting the patient at unjustifiable risk.
The basketball coach Bobby Knight is reputed to have said something like, "The will to win is not worth much without the will to prepare to win." In fact, continuous professional growth, consistent with standards shared throughout the profession, is one of the ethical principals enumerated by the American College of Dentists.
There is more to the ethics of practice than choosing the best among alternative new ways to serve patients. There is also an obligation to stop using old ones that have been superseded. Dentists who say "I will stick with what I know has always worked because it is unethical to experiment on patients" are rationalizing their cop-out.
But that does raise a valid question about whether all outcomes-based practice, or all evidence-based dentistry, or even all research involving patients is ethically defensible. I have come to the position that both dentists and researchers have an obligation to experiment to improve dental care, but only certain approaches to these experiments are ethically valid. The rules for researchers have been developed in the research community. They are carefully monitored by institutional review boards at each research facility.
The rules for experimentation in dental practice have been given little attention. First, we must realize that experimenting on patients is a recognized phenomenon within the profession. State boards acknowledge that every dentist now practicing performed dental care on patients before they were licensed. It is part of the initial licensure procedure for unqualified individuals to perform irreversible procedures on patients with a nationwide expected failure rate of 20 percent.
Regardless of how much laboratory testing is done on pharmacological agents or materials, there must always be a first time for each dentist, and the first time is an experiment. A dentist who has used a conservative procedure, say an approach to periodontal therapy that was popular in the 1950s, is also experimenting on his or her patients. In this case, it is not a first for the procedure, it is a first for the procedure on that patient.
There are four criteria, all of which must be met to justify experimentation in dental practice — that is to say, to justify practice at all. The first condition is that whatever is done is performed for the sake of providing better care to patients. It is always a good idea to involve patients in these decisions. Through informed consent, they can participate in the treatment goals and methods.
The second criterion for experimentation is the standard of care. The range of experimentation, the way it is approached, and the materials and techniques experimented with must all be ones generally approved by professional colleagues. If one's colleagues are shocked by the novel nature of a practice, the ethical warning light starts flashing.
The third criterion is that there should be a reason for experimenting. It is not good enough to expose patients to risk in a random fashion or just because one thinks something might work. There are three reasons that can support efforts to improve practice effectiveness:
- Conversation with the best of one's colleagues
- Published literature
- Outcomes-Based Practices such as component search and variable search
The final criterion is that experimentation must be done carefully and reflectively. It is no good to try something in a sloppy fashion, to fail to observe all relevant factors and their results, or to adopt a change simply on someone's say-so. The last remark is powerful. If it is accepted that all innovation must be proven in the context of the practice where it is to be deployed, evidence-based dentistry by itself is inadequate.
All four criteria must be met. It should be noted that experimentation for the sake of practice efficiency, profitability, or personal gain are not mentioned among the criteria. It should also be noted that representations by manufactures are not listed among the three reasons justifying experimentation. The minimal obligation for the professional is to work diligently to improve capacity for high-quality oral health care and to verify the effectiveness of any new approach within the dental setting.
A critical difference between research and practice is the impossibility of the practitioner to avoid the challenge presented by the patient. Something must be done, even if that is to refer the patient to a person with more expertise. Researchers may choose not to study a problem, and they are under no obligation to see that someone else looks into it. There is also a difference having to do with the imperative for continual improvement. A researcher who uses old techniques is not very smart and may be unlikely to secure funding. But they have harmed no one in the process. The same is not true for dentists. Practitioners are under an ethical obligation to experiment with a goal of continuous improvement. Guidelines for this experimentation have been sketched in this article. Research, including evidence-based dentistry, cannot relieve the practitioner of this obligation. Outcomes have to be validated in the context of individual practices.
Before and after
There is a simple technique in outcomes-based practice that meets the criteria for verifying practice innovations, even when the innovation was not generated within the practice. The technique is called B&A — standing for before and after. The reason for changing one's practice, the reason for adopting a new approach, must always be that the outcomes from the procedure after the change are better than the outcomes before it.
The B&A procedure works like this: Two examples of the outcomes of the "before" procedure and two of the "after" procedure are identified. They are made as anonymous as possible and presented to a neutral party who is asked simply to arrange all of them in order from best to worst. There should be some discussion of the criteria, but this is not a precise activity and a whole range of considerations should be used.
If it happens that the two "after" outcomes (As) are ordered as best and the two "before" outcomes (Bs) are labeled worst, there is a two-thirds probability that the new approach is really better than the old one. Anything less than that means the dentist needs to do more work before making the new approach a standard routine. If six outcomes are rank-ordered and three As come before three Bs, the probability of a true improvement can be raised to 95 percent.
If a reasonable number of comparisons are made, the result need not be as neat as all of the As before all of the Bs. Simply count from both ends: all the As before reaching the first B, and, from the other end, all of the Bs before reaching the first A. If the combined number from both ends is four, the likelihood of a true improvement is 67 percent; if the combined number is 6, the likelihood is 95 percent; if the combined number is 8, the likelihood is 99 percent.
