| OCR Text |
Show mendation is updated, and a delay follows before the change is generally known. For example, AAFP recommended BSE then withdrew support based on new research. Physicians who cited AAFP may have been following the earlier recommendation, but now that AAFP no longer recommends BSE, their practices would align with the recommendations of other groups. Another possibility is that, for the physicians we identified who cited a single published recommendation, personal experience or discussions with colleagues may have been more influential than the literature, a fact which would not have been apparent in our analysis. Another explanation is that the recommendations of ACOG and/or ACS may have been distributed to all physicians more effectively than the recommendations of other organizations. Despite polarization toward individual specialty societies, physicians in all 3 specialties rated ACS highly (Table 3), indicating that it is a respected, neutral source, and may have more bearing on practice than physicians realized. ACOG distributes a compendium of technical bulletins and committee opinions to all ACOG fellows (graduates of residency programs in obstetrics & gynecology). It also publishes regular committee opinions and practice bulletins with recommendations on a variety of topics in a single binder. In everyday practice, a combination of these and other reasons is probably involved. Although the majority of physicians said they screen aggressively regardless of the recommendation they follow, a substantial minority of physicians who cited ACP and USPSTF actually followed the recommendation of each organization to begin mammography at age 50. At least for raam-mography, these physicians knew which recommendation they agreed with and they followed it. With CBE though, these physicians were not so conservative. USPSTF recommends beginning CBE at age 50, but only 2% of those who said they follow this recommendation actually started so late, indicating that they either were not aware of the USPSTF's recommendation for BSE, or were aware and chose not to follow it. Another interesting finding was that of the 220 physicians we identified who consider a single published recommendation as most influential, three-fourths of obstetrician-gynecologists, but only one-fourth of family physicians and one-fifth of internists, stated they follow the recommendation of their specialty society. Obstetrician-gynecologists tended to screen more aggressively, and were least likely to cite USPSTF as the recommendation that had influenced their screening practices. Obstetrician-gynecologists therefore appear to be a more homogeneous group than internists or family physicians. One possible reason is that within the different specialties there may be varying degrees of social bias to identify their own specialty society. Another explanation may be the fact that family physicians and internists were much more likely to identify USPSTF, whose recommendations are similar to those of AAFP and ACP. Finally, as mentioned above, ACOG distributes a compendium detailing its recommendations. Perhaps because they can find ACOG's recommendations in a consistent, easily accessible place, obstetrician-gynecologists both cite and follow their own specialty recommendation more so than the other specialties. Conflicting recommendations are confusing. Each organization reviews the available data at a given time and then publishes its recommendations. This results in several different recommendations that are published at different times, and therefore are not necessarily based on the same research. Furthermore, because so many recommendations exist, physicians may believe they are following the recommendations of an organization, unaware that an update has been issued. The use of the Internet may mean that recommendations can be updated more easily as new information becomes available, but it may also increase the number and variety of recommendations. Expanding on the recommendations of the AMA (2001), organizations should establish consensus recommendations with each other, update them as new data become available, and distribute them in a form that is simple and accessible to physicians. Study Limitations There are a number of possible limitations in this study: 1. The overall response rate was 38.8%. Hence, selection bias may be present. It is possible that this group of respondents may not represent all primary care physicians in Utah. However, the specialty distribution among respondents (54% family physicians, 24% internists and 22% obstetrician-gynecologists) was similar to the distribution among all physicians contacted (48% family physicians, 28% internists and 20% obstetrician-gynecologists). 2. Second, this study presumes that the screening practices indicated on the survey remain constant between patients and over time for a given patient. Utah's Health: An Annual Review Volume VIII 37 |
