Since providers are likely to exert a major influence in the choice of PVV schedule, the purpose of our study was to describe the current awareness and practices of Ohio family physicians and pediatricians in implementing the 1997 recommendations for administration of PVV. The following hypotheses were tested: (1) There is a difference between family physicians and pediatricians in their awareness of the 1997 recommendations for the administration of PVV; (2) there are differences in practices between family physicians and pediatricians in their implementation of PVV recommendations; and (3) several factors (parent choice, additional cost of IPV, practice preference, difficulty obtaining IPV, liability concerns, risk of VAPP from OPV, increased number of injections per visit, and lack of combination DTaP-HIB vaccine for young infants) are associated with physicians’ practices in administering poliovirus vaccine.
Methods
Study Design
We used a descriptive survey research design. An 18-item questionnaire was specifically designed for our study. Along with basic demographic information, physicians were asked if they provided immunizations to children as part of their practice. They were asked if they followed the Centers for Disease Control and Prevention (CDC), American Academy of Family Physicians, or American Academy of Pediatrics protocol for the administration of immunizations. Additionally, physicians were asked if they were aware of the 1997 recommendations regarding PVV. They were asked to select which of the various options for administering PVV they usually recommend in their practice. The options included: (1) all OPV; (2) all IPV; (3) sequential: 2 IPV plus 2 OPV, (4) all of the above; or (5) other. They were then asked to select which of the immunization options they most frequently administer in their practice. Physicians were asked how these options were presented to the child’s parent or caregiver. The physician could select 1 or more of the following responses: (1) personally discuss each option with parent or caregiver; (2) office nurse or medical assistant discusses each option with the parent or caregiver; or (3) parent or caregiver reads information provided in handouts from the American Academy of Pediatrics or CDC vaccine information sheets. Finally, physicians were asked to indicate which factors influenced the actual administration of PVV in their practice: parent choice, additional cost, practice preference, difficulty obtaining IPV, liability concerns, risk of VAPP, increased number of injections per visit with IPV, or lack of a combination DTaP-HIB vaccine for infants.
Sample
A sample of 480 Ohio family physicians and pediatricians was randomly selected to participate. Membership lists from the Ohio Academy of Family Physicians (OAFP) and the Ohio Chapter of the American Academy of Pediatrics (OAAP) were used to identify potential participants. Approximately 2000 family physicians, excluding residents and students, are included on the OAFP mailing list, and approximately 900 pediatricians, excluding subspecialists, residents and students, are included on the OAAP list. Both lists included approximately 90% of all physicians eligible for their professional society’s membership. Simple random sampling was conducted using a random numbers table.
We estimated that 130 subjects per group would be required to perform the regression analyses (10 subjects per factor plus 50 additional subjects).17 A sample size of 130 per group was determined to be adequate to detect a moderate effect size (.3) for the differences between the 2 groups using chi-square analysis with adequate power (.90) at a = .05 (1-tailed). Family physicians were oversampled by 20% to account for previous research suggesting that approximately 20% of family physicians limit their care to adults only. Pediatricians were also oversampled by 20%, because the OAAP mailing list included some subspecialists that could not be clearly distinguished from the general pediatricians. A total of 480 questionnaires were originally sent in anticipation of at least a 60% response rate. We replaced 90 of the original subjects selected from the mailing list database with randomly selected subjects because of incorrect mailing addresses or retired or deceased members. Prior approval was obtained from the investigators’ Institutional Review Board.
The questionnaire was pilot tested among faculty and resident physicians from the investigators’ respective residency programs. After revision, the final questionnaire was mailed in January 1998. The mailing included a brief introduction to the study and a request that the physician complete the questionnaire and return it in the enclosed preaddressed, prestamped envelope within a 3-week period. A reminder postcard was sent approximately 2 weeks after the initial mailing. A follow-up questionnaire was sent to physicians who had not responded by the deadline. Responses to the survey were anonymous.
Data Analysis
All data were scanned directly into a database and analyzed using SPSS-PC 8.0, a computer software package (SPSS Inc; Chicago, Ill). Data were carefully examined for nonrandom missing values. Comparability of the 2 physician groups was examined for all demographic and dependent variables. A logistic regression model was developed to identify the independent effect of factors most likely to influence the family physicians’ and pediatricians’ practices in administering PVV.