Of the 46 invited practices, three urban clinics (23%) and 7 private practices (30%) declined. This resulted in 13 participating urban clinics and 23 participating, non-urban private practices. There were no statistically signficant differences in practice size, overall structure, or clinician demographics between participating practices and non-participating practices. The questionnaire response rate for clinicians (physicians and other clinicians) was 94% (138 out of 147) and for office staff (registered nurses, office managers, licensed practical nurses, medical assistants, receptionists, and billing personnel) was 92% (247 out of 269) (). There were no statistically significant differences between clinicians in urban clinics and private practices with respect to age, sex, race, full/part-time status, and years since obtaining highest degree.
Demographics of Participants
Participating practices ranged in size and varied in patient demographics. Urban clinics by definition were located in areas where at least 20% of households had incomes below the federal poverty level with a population density of greater than 1000 people per square mile and served largely a low-income, Medicaid population. The remaining (privately owned, non-urban) practices served largely a commercially insured population (percent of patients with public insurance 72 ± 22% in urban clinics vs 29 ± 19% in private practices, mean ± SD). None of the private practices served rural areas. The overall staff-to-clinician FTE ratio was 1.7 ± 0.8 (range 0.4 - 4.0). There was no difference in the staff-to-clinician ratio between urban clinics and private practices (1.54 ± 0.87 vs 1.77 ± 0.78) (p=0.42). However, compared with private practices, urban clinics had a higher percentage of non-physician clinicians compared with physicians (63% vs 20%, p=0.01). On average, physicians saw four more patients per half-day clinic session than the other clinicians, and physicians in private practices saw four more patients per clinic session than their urban counterparts.
Practice Characteristics and Easy Breathing Enrollment Rates
The distribution of Easy Breathing enrollment rates was skewed with values ranging from 0 to 661 enrolled children/FTE/year (mean = 125 enrolled children/FTE/year; median = 86). Practice size (the total number of clinician and staff FTEs) did not affect any of the three quality measures. Higher practice activity (the number of appointments/hour or the number of patients seen per week) was associated with higher enrollment rates (p=0.05) but not with higher use of a written asthma treatment plan or greater adherence to the guidelines for using anti-inflammatory therapy.
Organizational Characteristics and Their Relation to Quality Measures
The distributions of the organizational characteristics were roughly symmetric. Mean scores for the 9 scales of the PCOQ in the private practices were higher than scores for urban clinics with one exception (). There was no difference in decision authority between urban clinics and private practices. The differences in mean PCOQ scores for urban compared with private practices were observed both for clinicians and for staff in both locations. Scores, however, for physicians were similar to the scores reported by the non-physician clinicians regardless of location.
Comparison of Scores on Organizational Characteristics in Urban Clinics and Private Practices
At the level of the practice, when controlling for the percent public insurance, clinician type (MD vs other clinician), and staff-to-clinician ratio, higher scores on communication timeliness, decision authority and job satisfaction were associated with higher Easy Breathing enrollment rates () and the completion of a written asthma treatment plan. In addition, perceived effectiveness related to meeting family needs was associated with higher Easy Breathing enrollment rates. Scores on other organizational characteristics were not associated with higher enrollment rates or with the creation of a written asthma treatment plan. As predicted, organizational characteristics were not associated with use of anti-inflammatory therapy and in particular the use of ICS in accordance with national guidelines.
Organizational Attributes and Easy Breathing Enrollment and Distribution of a Written Asthma Treatment Plan
Variations in Organizational Characteristics Between Staff and Clinicians
On average, clinicians reported higher scores than staff for communication within the group (mean difference = 0.36 points, p<0.0001), communication timeliness (0.16 points, p=0.014), conflict resolution (0.21 points, p=0.01), and decision authority (0.53 points, p<0.0001), whereas staff reported higher scores than clinicians for perceived effectiveness in meeting family needs (0.25 points, p<0.0001), and in job-related psychological demands (0.23 points, p=.016). The differences between staff and clinician responses in the organization attributes were not associated with the number of children enrolled in Easy Breathing or the use of anti-inflammatory therapy.