Of the 1640 physicians contacted by mail, 16 letters were returned by the post office as undeliverable, and 107 physicians were no longer in practice, no longer in primary care (e.g., hospitalists), deceased, or unknown at the practice, leaving 1517 eligible physicians. Of these, 1015 physicians returned surveys for a response rate of 67%. Participation was higher among family physicians (68.9%) than internists (64.8%; P < .01), and among board certified (68.9%) than non-board certified (60.7%; P < .01) physicians. Participants and nonparticipants did not vary by age (mean = 50.7 years), length of time in practice (mean = 23 years), domestic vs. international training, or geographic region. Among participating physicians, 81.9% received their medical degree in the U.S. and 79.5% were board certified. The modal practice pattern for physicians in the sample was an independent practice (51%) in a suburban setting (45.3%). The next most common practice setting was to be based in a large corporate health system (25.5%) in an urban setting (31.2%).
Of the 1015 physicians who completed the survey, 115 did not immunize adults in their practices, leaving 900 physicians in the sample. Of these 900, 20 did not respond to questions specific to influenza vaccination. Thus, the analytic sample consisted of 880 physicians who provided information on SOPs for influenza vaccination and PPV. Consistent use of SOPs for pneumococcal vaccination in the absence of similar use of SOPs for influenza vaccination was rare; in fact, only one respondent in the sample fell into this category. By contrast, SOPs for influenza vaccination without standing orders for PPV were common. Given this distribution, we established three SOP groups: no consistent use of SOPs for either vaccination, including the one physician reporting SOP use for PPV only (n = 502, 57.0%), consistent use of SOPs for influenza vaccination only (n = 175, 19.9%), and consistent use of SOPs for both vaccinations (n = 203, 23.1%). Forty-three percent of the sample reported SOPs for influenza vaccination (with or without concomitant SOP for PPV). Physicians in the three groups did not differ in age, length of time in practice, U.S. medical training, board certification status, or specialty.
Differences among the three SOPs groups are shown in Table . Across the three groups, some correlates increased uniformly as use of SOPs increased, for example, awareness of ACIP recommendations and/or Medicare regulations, belief that SOPs enhance adult vaccination rates, having an immunization champion in the practice, agreement that SOPs are effective in boosting vaccination coverage, access to an electronic medical record (EMR), and having staff who are open to innovation and who work well together. Training level and number of clinical assistants were significantly related to use of SOPs, with practices reporting a greater staff to clinician ratio and those with more highly trained assistants (PA/CRNP or RN/LPN) more often reporting consistent use of SOPs. The proportion reporting 2 helpers per provider was 69.3% in the consistent influenza/PPV SOP group vs. 56.3% in the consistent influenza only group and 46.3% in the no consistent SOP group. Pairwise differences across these groupings were significant in post hoc comparisons. Number of physicians in a practice was not a significant correlate of SOP group.
Characteristics of Physicians/Practices Reporting No Consistent Use of Standing Orders Programs (SOPs), Consistent Use of SOPs for Influenza Vaccine Only, and Consistent Use of SOPs for Both Influenza Vaccine and PPV*
To investigate correlates of SOPs in more detail, multinomial logistic regression models were developed to examine the entire set of correlates. These models allow a view of how physicians in each of the two SOP groups differ from the group not using SOPs. Table shows odds ratios and 95% confidence intervals associated with each correlate in a model that includes variables shown earlier in Table , adjusted for physician age and length of time in practice.
Practice Characteristics Associated with Consistent Use of Standing Orders Programs (SOPs) for Influenza Vaccination Only or Influenza and PPV, Relative to No Consistent Use of SOPs in Multinomial Logistic Regression
In the final adjusted model, consistent use of SOPs limited to influenza vaccine was independently and significantly associated with awareness of ACIP recommendations and/or Medicare regulations (odds ratio [OR] 2.35, 95% confidence interval [CI] 1.60-3.46), agreement that SOPs are an effective way to boost vaccination coverage (OR 2.97, 95% CI 1.89-4.67), family medicine specialty (OR 1.63, 95% CI 1.10-2.42), greater amount of help available to physicians (access to 2 assistants [OR 2.75, 95% CI 1.31-5.79]; access to 1 assistant [OR 2.30, 1.23-4.30], relative to the group with only 1 assistant for 2+ providers).
Consistent use of SOPs for both influenza vaccine and PPV was associated with the same set of factors but additional factors as well. Shared correlates included awareness of ACIP recommendations and/or Medicare regulations (OR 4.46, 95% CI 2.91-6.85), agreement that SOPs are an effective way to boost vaccination coverage (OR 3.5, 95% CI 2.14-5.71), family medicine specialty (OR 1.27, 95% CI 0.89-1.91, N.S.), and a greater staff to clinician ratio (OR 2.22, 95% CI 1.09-4.54). Correlates associated only with combined influenza and PPV SOPs included a variety of practice-level factors: practice openness to change and innovation (OR 2.15, 95% CI 1.33-3.47), strong practice teamwork (OR 2.78, 95% CI 1.49-5.21), access to an electronic medical record (OR 1.90, 95% CI 1.22-2.96), presence of an immunization champion in the practice (OR 1.94, 95% 1.27-2.98), and access to nurse/physician assistant staff as opposed to medical assistants alone (OR 1.49, 95% CI, 0.99-2.24, p = .054).
The two SOP groups were directly compared in Table . Physicians reporting consistent use of SOPs for both immunizations were more likely to be aware of ACIP recommendations and/or Medicare regulations (OR 1.95, 95% CI 1.18-3.24). Practice-level factors distinguishing consistent use of SOPs for both vaccines included strong practice teamwork (OR 2.24, 95% CI 1.10-4.54), having an immunization champion on site (OR 1.67, 95% CI 1.01-4.54), and access to nurse/physician assistant staff as opposed to medical assistants (OR 2.21, 95% CI 1.31-3.45).
Practice Characteristics Associated with Consistent Use of SOPs for Both Influenza Vaccine and PPV* Relative to Influenza Vaccine Only in Logistic Regression
Finally, we examined the number of physician and practice characteristics associated with adoption of SOPs and the prevalence of reported SOPs for either influenza vaccine or PPV. We summed across the 12 characteristics analyzed in the regression models and plotted adoption of SOPs against this index (Figure ). In this sample of physicians, few respondents reported 0 or 1 characteristic (n = 3) and few reported all 12 (n = 3). Physicians reported a median of six characteristics associated with adoption. Figure shows that adoption of SOPs is as low as 10.5% when only 1 or 2 factors are present and as high as 76.9% in the presence of all 11 or 12 factors.
Uploaded separately. Prevalence of SOPs Relative to Number of Physician and Practice Characteristics Associated with Adoption.