Professional Variability in Decision Making in Modern Dentistry: A Pilot Study

A widespread belief among patients is that treatment decisions first and foremost depend on the patient's disease. This perception seems to be based on the assumption that for each and every clinical condition there is unambiguous evidence about the effectiveness of alternative therapies and that this information is the most authoritative criterion when treatments are recommended and undertaken by medical providers. However, this may not always be the case in dentistry.

The supposed wonderland of perfect information may be obscured by the fact that available evidence about treatment alternatives is often very limited, and, thus, no clear treatment guidelines exist. One such clinical scenario refers to single posterior bounded edentulous spaces (BES). Such conditions do not just make up a few rare cases; their treatment has been reported to account for about 7% of annual dental expenses.^1,2  Treatment of BES is likely to be of continuing relevance as the population ages, and it has been suggested that despite a general decline in complete edentulism throughout the past decades, the frequency of partial edentulism has remained relatively stable.³  In terms of treatment, it has often been argued that a missing tooth should be replaced to avoid arch collapse as a result of movement of adjacent and unopposed teeth.^4,5  However, a recent systematic review suggests that occlusal changes in BES after tooth loss are often limited; therefore, tooth replacement should not necessarily be regarded as the mainstay of therapy.⁶  Given the absence of clear treatment guidelines for BES, however, dentists' recommendations may vary with respect to parameters that do not solely relate to patient characteristics.

First, many dentists may believe they have no option but to abide by information that has been passed from colleague to colleague. And as dentists may specialize in one or another subdiscipline, each of which may favor different lines of action, treatment recommendations may vary considerably across disciplines. For example, this sort of variation has previously been reported for treatment of periodontally compromised teeth, and different dental practice scopes have frequently been observed to be associated with different treatment preferences.^7-10  Second, the active dental profession comprises recently graduated dentists and dentists with decades of treatment experience. Yet, depending on the level of knowledge and experience, treatment recommendations may vary. Differences in levels of competency may also vary between dental students at different stages of education.^11-15  Third, the settings within which dentists work may differ. Many providers are self-employed and bear the cost risk of a small or medium-sized business. Other providers are employed (eg, in public institutions like a university hospital). Previous evidence, for example, suggests that different reimbursement arrangements can influence treatment decisions.^16-18  This may be another reason why treatment recommendations may depend on factors that do not solely relate to a patient's clinical condition.

So far, little information exists about the extent to which such dentist-related factors influence treatment recommendations for BES. Therefore, the purpose of the present study was to explore the extent to which treatment recommendations vary among dentists with different levels of experience, different fields of specialization, and different work settings.

Using a standardized questionnaire form (Table 1 and a clinical vignette case for characterizing a clinical decision scenario (described in the next section), an anonymous survey about treatment recommendations was conducted among dentists and students of the dental clinics at the University of Heidelberg and among dentists working in private practices in June and July 2010. After the clinical vignette was presented by one person in a standard slide-show format to all lecture-attending respondents, the form was completed by dentists affiliated with the departments of prosthodontics (n=24; response rate=100%), conservative dentistry (n=20; response rate=91%), maxillofacial surgery (n=17; response rate=71%), and orthodontics (n=10; response rate=91%). The survey was also completed by students in the first (n=57; response rate=90%) and third year (n=31; response rate=84%) of clinical training who had already completed the preclinical years of education. The form was also completed by private practitioners who attended a lecture that was part of a training course for dentists specializing in endodontics (Gesellschaft fur Endodontie Bonn e.V.) (n=55; response rate=92%). The survey was conducted anonymously by all respondents. No time limit was set for filling out the questionnaire.

The vignette described a previously reported case of a 59-year-old woman who is seeking advice regarding missing lower first molars in both sides of the lower jaw.¹⁹  The patient has been missing these teeth since childhood for reasons unknown, has no treatment preferences (neither for nor against tooth replacement nor any other type of therapy), and is willing to follow any recommendation given by the dentist. The dental condition has been stable for many years, and the patient has had no recent or current signs of discomfort; pain; or esthetic, functional, or other limitation. The patient is described as health conscious and very cooperative. An orthopantomogram, pictures of the clinical situation, and diagnostic plaster models were provided (Figures 1 through 6). The full description of the presented clinical vignette case is shown in Table 2.

For both sides of the lower jaw, the proportion of respondents recommending various treatment alternatives was computed. Response categories indicating identical treatment within both mandibular sides were recorded such that the according recommendation was imputed to the separate categories for both left and right side. We used χ² tests (p<0.05) to detect statistically significant differences in treatment recommendations between (1) University hospital dentists, dentists working in private practice, and students; (2) clinical practitioners within different university departments; and (3) students in different years of clinical education. Statistical analysis was performed by one scientist (SL) who was not involved in conceptualizing and carrying out the survey and was, thus, independent and blinded with respect to the data-generating process. All data analyses were carried out with the software package STATA/SE 12.0 (StataCorp, College Station, Texas).

Table 3 shows the frequency of different treatment approaches as recommended by clinical practitioners, private practitioners, and dental students. Bridgework (left) was most frequently recommended by students in their first year. Bridgework (right) was chosen most often by respondents from the prosthodontics department. Implant placement (left) was most frequently recommended by students in their first year. Implant placement (right) was chosen most often by respondents from the maxillofacial department. Tooth recontouring (ie, direct composite buildups to close gaps)²⁰  was most frequently recommended by students in their first clinical year. Orthodontic treatment was suggested most often by respondents from the maxillofacial department. Among University hospital dentists, a monitoring-only approach was most frequently recommended by dentists from the conservative dentistry department, followed by dentists from the prosthodontics, orthodontics, and maxillofacial departments (see Figure 7).

Table 4 shows χ² statistics from tests for differences in treatment recommendations among hospital practitioners, private practitioners, and students. When comparing University hospital dentists with students, treatment with bridgework (left), implant placement (left), and recontouring (right) were found to be recommended significantly more often by students, whereas a monitoring approach was recommended significantly more often by dentists working in the clinics. When comparing dentists working in the clinics with private practitioners, treatment with bridgework (left) was recommended significantly more often by private practitioners. When comparing private practitioners with students, treatment with implant placement (left), recontouring (right), and orthodontic treatment (left) were recommended significantly more often by students, whereas a monitoring approach was recommended significantly more often by private practitioners.

Table 5 shows χ² statistics from tests for differences in treatment recommendations among dentists from different departments within the university hospital. No significant difference was found between the conservative dentistry and the prosthodontics departments. When comparing dentists working in the conservative dentistry department with their colleagues working in the maxillofacial department, implant placement and orthodontic treatment were recommended significantly more often in the maxillofacial department, whereas a monitoring approach was recommended significantly more often in the conservative dentistry department. When comparing dentists working in the conservative dentistry department with their colleagues working in the orthodontics department, bridgework (left) was recommended significantly more often by the orthodontists. When comparing dentists working in the prosthodontics department with their colleagues working in the maxillofacial department, implant placement and orthodontic treatment were recommended significantly more often in the maxillofacial department, whereas a monitoring approach was recommended significantly more often in the conservative dentistry department. When comparing dentists working in the prosthodontics department with their colleagues working in the orthodontics department, recontouring (right) was recommended significantly more often by the orthodontists. When comparing dentists working in the maxillofacial department with their colleagues working in the orthodontics department, implant placement (left) was recommended significantly less often by the orthodontists.

As shown in Table 6, implant placement (left), recontouring (right), and orthodontic treatment (left) were recommended significantly more often by students in their first clinical year, whereas a monitoring approach was recommended significantly more often by students in their third clinical year.

The present study revealed substantial variability regarding treatment recommendations with respect to dentists' level of experience, area of specialization, and institutional setting in which they work. First, our findings appear to give evidence for a tendency toward less invasive treatment being recommended with increasing level of treatment experience. Dental students were more likely to recommend implant placement (left) and recontouring (right) than practicing dentists. Students were also more likely to recommend bridgework than clinical practitioners and more likely to recommend orthodontic treatment than private practitioners (left). Dentists were more likely to recommend monitoring than students. Among students, implant placement (left), recontouring (right), and orthodontic treatment (left) were recommended more often by students in their first year, whereas monitoring was recommended more often by students in their third year. It should be noted tooth recontouring is a specific component of the first-year curriculum.

Second, it seems that the area of dentists' specialization also has a considerable impact on the type of recommended treatment. Although no differences were detected between the conservative dentistry and prosthodontics department, dentists in the maxillofacial department were more likely to recommend implant and orthodontic treatment but less likely to recommend monitoring compared with colleagues from the prosthodontics and conservative dentistry departments. Orthodontists were more likely to recommend bridgework (left) than conservative dentists, more likely to recommend recontouring (right) than prosthodontists, and less likely to recommend implants (left) than colleagues from the maxillofacial department.

Third, our findings also hint at the relevance of settings in which dentists work. In particular, bridgework (left) was recommended more often by dentists working in private practice than by University hospital dentists. It may be tempting to speculate whether such a difference in treatment recommendations may be attributable to different financial incentives for private and clinical practitioners, yet identification of such a link was outside the scope of the present study.

Given that, to date, treatment guidelines for BES are not fully conclusive; the observed variations may not appear completely unexpected.²¹  Nevertheless, the extent of differences among various groups of respondents seems remarkable. As there was also great variation within different groups of respondents, this may hint at further uncertainties about which treatment approach would be the most adequate for BES. Unquestionably, the currently existing knowledge regarding clinical management of BES is insufficient to judge any of the proposed treatment approaches as entirely right or wrong. However, for clinical scenarios characterized by stable BES without any limitations perceived by the patient, previous evidence suggests that it is unlikely that the patient will incur a high risk of adverse health consequences if a monitoring approach is taken rather than immediate tooth replacement.⁶  In order to establish more differentiated treatment guidelines for BES that facilitate more patient-centered treatment decisions in the future, however, more research is needed that examines the potential of monitoring approaches. Moreover, there is a need for better understanding about how dental professionals change their clinical practice and about reviewing dental education programs because reviewing clinical guidelines alone may not necessarily lead to rapid implementation of altered therapeutic approaches.²² 

As with any other pilot study, the present investigation has its limitations. First, no tailored sample-size planning or other specific sampling procedures could be applied. Some of the variation in the reported results could have been influenced by an unfitted sample size or by sample selection bias. Second, the present study could not control for other potentially relevant covariates, such as, for example, a different age composition of self-employed dentists and their colleagues working in hospital. This could be relevant because it may complicate the distinction between age-related effects of experience and effects of different practice settings (private practice vs hospital). Third, the sample for private practitioners was recruited from dentists having a specific interest in endodontics and thus may not be considered representative of the entire professional community.

Nevertheless, the findings of the present study uniquely quantify the extent of intraprofessional variability in treatment recommendations as regards BES. The fact that the treatment patients receive seems to be strongly influenced by provider characteristics that do not directly relate to the patients' clinical condition may be a challenge. Patients usually seek treatment advice from dentists because of their supposed expert knowledge with regard to the necessity of clinical intervention and which treatment approach, if any, is the most clinically adequate. However, if the existing clinical evidence about therapeutic alternatives is unclear, it seems not unlikely that dentists seek to justify treatment approaches in the way in which they are most comfortable. This is not to say that dentists would only act in their own interest and would not respect the patient preferences, but if it is impossible to differentiate between the clinical appropriateness of two therapeutic alternatives and, thus, both are clinically justifiable, it seems plausible that providers prefer the one that best accommodates their own skill. Based on the supposed expert advice, patients may then shape their treatment preferences accordingly. Ultimately, however, this implies that one specific treatment is chosen over several others despite no evidence of better clinical effectiveness. This may not only raise concerns of inefficient resource use (if providers prefer more expensive treatments) but also of inconsistencies within the dental profession (if patients with identical clinical conditions receive largely different treatment recommendations).

All in all, the present study gives novel evidence for considerable intraprofessional variability in dental treatment recommendations for BES with respect to dentists' level of experience, area of specialization, and institutional setting in which they work. The extent of the observed variations seems highly remarkable. Treatment guidelines and dental education programs should thus be reviewed. Moreover, future randomized controlled clinical research examining the effectiveness of monitoring approaches will be helpful to further the development of more differentiated treatment guidelines for BES. In addition, more detailed investigations of factors relevant for dentists when making clinical decisions could take particular advantage of state-of-the art experimental techniques used in the behavioral sciences, such as discrete choice experiments.

This study was solely funded through the authors' institutions. No external funding was received.

This study was conducted in accordance with all the provisions of the local human subjects oversight committee guidelines and policies of Heidelberg University. The approval code for this study is 07/07/2011.

The authors of this manuscript certify that they have no proprietary, financial, or other personal interest of any nature or kind in any product, service, and/or company that is presented in this article.