It is widely accepted that the uptake of research findings by practitioners is unpredictable, yet until they are adopted, advances in technology and clinical research cannot improve health outcomes in patients. Despite extensive research there is limited knowledge of the processes by which changes occur and ways of measuring the effectiveness of change of practice. The overall aim of this thesis was to investigate aspects of an educational intervention in clinical endodontic routines and new instrumentation techniques in a Swedish County Public Dental Service. Special reference was made to the establishment of changed behaviour in practice, the process of change, and the clinical effects. Although a high level of competence in root canal treatment procedures is required in general dental practice, a number of Swedish studies have revealed inadequate root-fillings quality and associated periapical inflammation in general populations. It is suggested that the adoption of the nickel-titanium rotary instrumentation (NiTiR) technique would improve the cleaning and shaping of root canals and the quality of the root-filling. However, there is limited knowledge of the effectiveness of the technique when applied in general dental practice. In two of four consecutive studies, the subjects were employees of a county Public Dental Service. The aim was to investigate the rate of adoption of clinical routines and the NiTiR technique: the output, and the qualitative meaning of successful change in clinical practice. In the other two studies the aim was to investigate treatment effect and the cost-effectiveness of root canal treatment in a general population: the outcome. Four hundred employees (dentists, dental assistants, administrative assistants and clinical managers) of a Swedish County Public Dental Service were mandatorily enrolled in an educational and training program over two years. Change of practice was investigated in a post-education survey. The NiTiR technique was adopted by significantly more dentists in the intervention county compared to a control county (77% and 6% respectively). Dentists in the intervention county completed root canal instrumentation in significantly fewer sessions than the dentists in the control county. Eight in-depth interviews, two with each participant, (dentist, dental assistant, receptionist, clinical manager), were strategically selected for a phenomenological analysis. Four factors were identified as necessary for successful change: 1) disclosed motivation, 2) allowance for individual learning processes, 3) continuous professional collaboration, and 4) a facilitating educator. A random sample of 850 performed root canal treatments was used for a study of treatment outcome; 425 before and 425 after the education and adoption of the NiTiR technique. Root-filling quality, periapical status and tooth survival were assessed on radiographs taken at treatment and at follow-up, > or = 4 years later. Apical periodontitis was found in 34% of the teeth root-filled before the education compared to 33%, after. After the education, root-filling quality improved significantly, tooth survival was significantly higher, however, without a subsequent improvement in success rate post-education; 68% vs. 67%. A micro-costing model was used to calculate the costs of root canal instrumentation, pre- and post-education, in the same sample used in the study of treatment outcome. Costs were lower post-education: by SEK 264 for teeth with one canal and SEK 564 for teeth with three or more canals. A reason for lower costs was that the NiTiR technique dominated after the education and required significantly fewer instrumentation sessions. A cost-minimization analysis disclosed that root canal treatments undertaken post-education were more cost-effective. In conclusion, there was only a partial relationship between output and outcome. Although root-filling quality improved significantly, the study did not show any association between the more frequent use of NiTiR and an improvement in remaining teeth with normal periapical status or success rate. However, the use of NiTiR was more cost-effective. These results are in accordance with previous findings of the so called efficacy-effectiveness gap in clinical practice: a high output is not predictive of a high outcome. The overall conclusion to be drawn from these studies is that further research is warranted to identify factors associated with improvement of the quality of endodontic care. The general interpretation of the findings of these implementation studies is as important as the effects of the change in endodontic instrumentation: a clinically relevant and applicable intervention, introduced by experienced expertise under allowing learning and collaborating circumstances, disclosed clinicians' motivation and facilitated implementation. The finding of qualitative differences between the questionnaire responses and the in-depth interviews suggest that a critical approach is warranted when comparing surveys and qualitative methods aimed at investigating qualitative experiences of change, due to their different epistemological premises.