Who Knows What the Outcome May Be?
A number of recent articles in the lay press have dealt with attempts to utilize outcomes research to validate a "one best way" to perform health-care procedures. In orthodontics, consider the problem of trying to find a one best way to treat a Class II malocclusion. You could fill the seats of a ballpark with Class II malocclusions, and no two would be identical. You could line them up in numerical order throughout the range for any one characteristic and find many were identical for that characteristic. For example, there are many people who have the same SNA measurement. But now select another characteristic, and you have a whole new lineup.
One approach to identifying better outcomes has been through twin studies. Identical twins with similar malocclusions are treated in two different ways and the results compared. However, although identical twins may be identical in many ways, they are not identical in all their characteristics and measurements. They are not identical physiologically or psychologically. One twin may be cooperative and the other may not. Moreover, the treatment provided for one may be more difficult to cooperate with than that provided for the other.
Another method of comparing outcomes is to treat the left and right sides of the same individual differently and compare the results--for example, retract the left cuspid with a spring and the right one with an elastic, and compare how long it takes on each side. It is not possible to narrow the range of such experiments to one independent variable. This is precisely what makes clinical research frequently unscientific. Not invalid necessarily, but unscientific. In fact, considering the variables inherent in the samples, much of clinical research is qualitative and speculative. Not useless, but requiring a healthy amount of caution. Dressing up twin studies or outcome studies with some trappings of a scientific experiment doesn't make them any more valid.
What are some of the variables that affect orthodontic outcomes?
1. The diagnosis.
2. The nature of the malocclusion:
a. The complexity of the malocclusion--the number of corrections that can be treated at the same time and the number that may need to be treated separately.
b. The severity of the malocclusion.
3. The appliance used.
4. The direction and magnitude of forces applied.
5. The duration of force application:
a. The extent of patient cooperation.
b. Whether the patient's efforts were applied properly.
c. Force degradation within the appliance.
6. The nature of the force--continuous or intermittent.
7. The amount, direction, and duration of growth.
8. Individual cellular response to force application.
9. Complex cellular interactions.
10. The muscular balance and the position of the muscular attachments.
11. Physiologic age.
12. Genetic influences.
13. Interferences:
a. Occlusal.
b. Friction within the appliance.
c. Deleterious habits.
14. The time between appliance adjustments.
15. The skill of the operator.
16. Broken and canceled appointments.
17. Loss and breakage of appliances.
18. Missing and misshapen teeth.
19. Tooth-size discrepancies.
20. Pathology.
21. The experience and intuition of the orthodontist.
22. Systemic illness during growth.
23. Compromises that are sensible or necessary.
24. Separating relapse from physiologic change.
25. Others?
Many of these variables are not measurable or predictable. In many instances, there might not even be agreement on a one best diagnosis, much less a one best treatment.
Another aspect of the search for the one best outcome is: Who is trying to answer the question? Is it the orthodontist? The patient? The insurance company? A patient might favor the fastest treatment time and the best appearance for the teeth that show. An insurance company might opt for the cheapest acceptable result. Indeed, one approach to searching for the best outcomes has been to retrospectively study insurance claims data. However, even if this were successful, it might only validate the restrictions third parties have selected to produce the best third-party income and not necessarily the best health-care outcome.
What outcome is the most desirable? Perfect occlusion? Optimum function? Optimum periodontal health? Best appearance? Fastest treatment time? Most satisfied patient? Longest stability? Orthodontists are likely to say, "All of the above", because orthodontists have a predilection to hit home runs. We feel that we cannot in all conscience aim for any result but the best occlusion in the best location. We cannot accept a 50% or a 90% correction as a goal. We would opt to treat for two or even three years to try for 100% as against achieving 80% or 90% in six months. We feel that less than that would be an unstable failure. But would all patients or all third parties agree with a prolonged 100% outcome?
Those interested in outcomes research apparently hope that it will point the way to simpler, less expensive ways of doing things. Depending on how the research is conducted and who conducts it, however, outcomes research could well validate more complicated and more expensive procedures. The basic problem with trying to find the one best way to treat every condition is in the assumption that there is one best way. The best outcome might not even be an optimum result due to physical, physiological, mental, and psychological limitations.
Whether an orthodontic result is the best outcome for a given case is often decided retrospectively. That does not mean that there would be universal agreement on that judgment, or that it could have been made prospectively.