Page 12
the authors interviewed key informants and visited schools, clinics, and/or training facilities. The reports issued by Bradlaw et al. (1951) and Dunning (1972) are entirely impressionistic. In fact, in neither case did the authors attempt to evaluate any work first-hand. Both, however, came away with positive impressions of the work being done by dental nurses (as therapists were then known) in the School Dental Service (SDS). Indeed, according to Bradlaw et al., “[t]he dentists we met were in arms at any suggestion of clinical shortcomings which we deliberately suggested to test opinion.” The other 3 observational reports, while having empirical elements, do not provide truly rigorous evaluations of technical competence. Though in each case the authors observed restorations in children, the judging criteria tended to be subjective and/or were not fully reported. Both Fulton (1951) and Friedman (1972) concluded that the quality of the work performed by school dental nurses was high. Friedman stated that “[h]aving seen the product first hand, I can attest to the adequacy of training.” He also maintained that, based on a review of x-rays, he was unable to tell the difference between restorations placed by nurses and those placed by dentists. A report by Gruebbel (1950), while ostensibly empirical, contains sufficient shortcomings, not to mention clear biases, that it must be considered observational. Gruebbel judged the amalgam restorations placed by school dental nurses to be “mediocre”. In fact, he found nearly 30% to be defective. He appears, however, to have assumed that all of the restorations he examined were placed by nurses, though there is reason to believe that a large number had actually been placed by dentists (Saunders, 1951). Gruebbel also remarked repeatedly on the negative implications of New Zealand’s “socialist” system for both society and the dental profession. And, in direct contrast to the contemporaneous reports of Fulton and Bradlaw et al., Gruebbel stated that “a large number of dentists” had concerns about the SDS. This observation was vehemently disputed by the Director of Dental Hygiene (Saunders, 1951). While our ability to draw firm implications from the preceding studies may be weak, randomized controlled experiments provide more convincing evidence on the effectiveness of a given treatment. Five studies from the 1970s, 4 conducted in the United States and one in the Netherlands, were designed this way. Each tested whether dental hygienists could be taught to prepare and place restorations as well as dentists. The study conducted in the Netherlands is the hardest to assess, since the only Englishlanguage article discussing it focused on examiner variability in assessing the quality of restorations, rather than on the quality of the restorations, per se. The data used, however, came from a study in which hygienists were trained to prepare and restore adults’ teeth; their work was compared with that of both dental students and private practice dentists. Based on the data reported, the hygienists appeared to have performed at least as well as the dental students, and perhaps better than the dentists (Swallow et al., 1978). At around the same time, 3 U.S. universities (Howard, Iowa, and Kentucky), as well the Forsyth Dental Center in Massachusetts, began pilot programs to train dental hygiene students (or, in the case of Forsyth, recently graduated hygienists with some practice experience) to prepare and place restorations. Their work was compared with that of dental students, or, in the case of Forsyth, with that of practicing dentists. All of the pilots had dentists pre-screen patients who were then randomly assigned to a practitioner; all used outside examiners to conduct blind evaluations, according to specific, set criteria. All 4 concluded that hygienists performed as well as dentists (Powell et al., 1974; Spohn et al., 1976; Sisty et al., 1978; Lobene, 1979). Nearly every individual comparison resulted in no qualitative difference between the two practitioner groups; in those few instances when statistically