Finkelstein JA, Ng M, Lozano P, Shulruff R, Soumerai SB, Inui TS, Weiss KB; Association for Health Services Research. Meeting.
Abstr Book Assoc Health Serv Res Meet. 1999; 16: 342-3.
Department of Ambulatory Care and Prevention, Harvard Medical School and Harvard Pilgram Health Care, Boston, MA 02215, USA.
RESEARCH OBJECTIVE: National consensus guidelines for the management of childhood asthma have existed since 1991, but implementation by physicians has been slow. We sought to determine current practice patterns for diagnosis and referral in primary care settings for children, and to determine the impact of physician specialty and managed care delivery system type (staff model vs. network) on patterns of care. STUDY DESIGN: We surveyed by mail pediatricians (PEDs) and family physicians (FPs) in three large managed care organizations in Seattle, Chicago, and Boston, during the spring and summer of 1998 in conjunction with the Pediatric Asthma Care PORT (Patient Outcomes Research Team). In 2 of the 3 sites both FPs and PEDs provide primary care to children. Two of the 3 sites include practices with staff-model and network managed care arragements. Domains of inquiry included familarity with National Asthma Education and Prevention Program (NAEPP) Expert Panel Report 2 (EPR2) and self-reported practice in the areas of diagnostic tests or maneuvers were dichotomized as ordered always/most of the time or some of the time/rarely when considering a new diagnosis of asthma. Respondents were presented with 12 possible indications for specialist referral and asked if the would "usually" seek consultation. All percentages refer to bivariate analyses tested for significance with chi square tests. Odds ratios and 95% confidence intervals are reported from logistic regression models including specialty, MCO type, site, and years in practice. PRINCIPAL FINDINGS: A total of 429 surveys were returned by 709 eligible physicians (61%). Though 92% had heard of the guidelines, only 78% reported having read them. Of those, 50% reported them very useful and 49% somewhat useful. In bivariate analyses, FPs were significantly more likely than PEDs to report use of diagnostic tests such as spirometry (40% v 12%), chest x-rays (38% v. 24%), skin testing (8% v 3%), sinus films (10% v 1%), and office peak flow measurement (96% v. 68%) ( all p<0.01), but were less likely to recommend a home trial of daily peak flow measurement (36% v. 54%, p<0.01). After control for cofounders, FPs were more likely to order spirometry (OR=5.9, CI: 2:4, 14.6) and less likely to order a home trial of daily peak flow measurement (OR=0.3, CI:0.1, 0.5). In bivariate analysis, for 5 of the 11 possible indications for referral, FPs were more likely than PEDs to refer (p<0.01). After control for cofounders, FPs remained more likely than PEDs ro refer patients who had: a single hospitalization (49% v. 14%; OR=2.8, CI: 1.3,6.3), 2-3 ER visits in one year (78% v. 40%; OR=5.1, CI: 2.5, 10.5), need for daily meds in a child <3 years (95% v. 64%; OR=7.8, CI: 2.7, 23), and 2 bursts of oral steriods in one year (68% v. 30%; OR=2.9, CI: 1.4, 6.0). The latter was among the listed indications for referral in EPR2. After control for cofounder, the only independent effects of MCO type were the staff model physicians were more likely than others to ude office peak flow (52% v. 47%; CI: 2.1, 10.3) and a homee trial of daily peak flow (52% v. 47%; OR=2.4, CI: 1.3, 4.3) for diagnosis. In multivariate models, years in practice was significantly associated with higher rates of specialist referral CONCLUSIONS: Patterns of self-reported diagnostic testing and referral differ substantially between FPs and PEDs, but MCO type appeared to have little effect. Some indications for referral recommended in EPR2 had low adherance rates among pediatricians. Years in practice may also be an important predictor of physician referral practices. IMPLICATIONS FOR POLICY, DELIVERY, OR PRACTICE: Further work should analyze outcomes of childhood asthma care by physicians of varying specialties so that information on "best practices" can be effectively transferred between them. Managed care organizations can use evidence on self-reported practice to determine which areas of guidelines should be the focus of future quality improvement efforts. End-user feedback will also be useful to guideline developers to determine which recommendations should be reconsidered in future revisions.
Publication Types:
Keywords:
- Asthma
- Boston
- Chicago
- Child
- Data Collection
- Demography
- Humans
- Managed Care Programs
- Physicians, Family
- Primary Health Care
- Referral and Consultation
- Specialties, Medical
- Spirometry
- economics
- hsrmtgs
Other ID:
UI: 102194092
From Meeting Abstracts