The survey used for this study was designed to gather information regarding adolescent wellness and barriers to health care services. The online survey, known as the Perceptions of Adolescent Health (PATH) survey, was developed by Harris Interactive Incorporated with involvement from the National Foundation of Infectious Diseases (NFID) and Pfizer Inc. The survey had three distinct populations of focus: adolescents (13-17 years old), parents of adolescents and health care providers (HCPs). HCPs included physicians, nurse practitioners, physician assistants and licensed practical nurses in the fields of family medicine, general practice, internal medicine and pediatrics. For the purpose of this study, we used responses from providers who identified themselves as pediatricians or FMPs as these physician groups provide the majority of primary care to adolescents [
6]. The nature of this study allowed for waiving of written informed consent which, along with study protocols, was approved by the Institutional Review Board at Indiana University-Purdue University, Indianapolis.
Survey data were obtained from December 2012 to January 2013. Survey respondents were required to meet certain criteria, i.e. time working directly in patient care, patients seen per month, and exposure to adolescent patients (13-17 years old). Subsequently, the data were weighted by Harris Interactive to be representative of the U.S. population as a whole. Descriptions of development and weighting have previously been described [
12].
Predictor variables
The survey collected data on provider demographics (gender, age, residency graduation year) as well as demographics of each provider’s patient population (patient race, patient age and patient reminders for well visits). Providers were asked about their familiarity with recommendations for adolescent health care and whether they believed well visits should occur annually or less frequently. The endorsement of annual adolescent well visits was the primary outcome for this study.
This online survey examined provider beliefs about the perceived importance of specified clinical discussions to the overall health of adolescent patients. Provider opinions regarding the level of knowledge held by adolescents about their own health as well as how proactive adolescents are in maintaining said health were also assessed. Lastly, the survey examined provider behaviors surrounding discussions with parents and adolescents regarding general health and wellness.
Provider age was subsequently collapsed into those less than fifty years old or greater than or equal to fifty years old. In addition, residency training completion was categorized into those who graduated prior to 2000 and those who graduated residency in or after 2000. This classification was based on the change in ACGME requirements for adolescent medicine specific training in Pediatrics.
Physicians reported on adolescent knowledge and proactivity separately using a 4-point likert scale ranging from “not at all knowledgeable/proactive” to “very knowledgeable/proactive”. For analyses, these variables were transformed into bivariate variables represented by “not at all to somewhat knowledgeable/proactive” and “knowledgeable/proactive to very knowledgeable/proactive”.
Physicians reported on how important they believed discussions about 11 topics were to the overall well-being of adolescent patients. These topics included weight, nutrition/diet, exercise, vaccines, issues at home, issues at school, self-image, sexual health, substance use, mental health and coping with stress, and again, importance was rated using a 4-point scale from “not at all important” to “very important.” Prior to analyses, responses were summed to create a Discussion Belief Scale (DBS). Scores ranged from 9 to 33 with a mean of 26.13 and a standard deviation of 5.80.
Provider discussion behaviors with adolescents and parents were assessed separately. Physicians were asked whether they regularly talk with adolescents and parents of adolescents about 12 topics including weight, nutrition/diet, exercise, vaccines, issues at home, issues at school, self-image, sexual health, substance use, mental health, prescriptions/medications and coping with stress. Responses were summed to create a Discussion Behaviors with Parents Scale (DBPS) and a Discussion Behaviors with Adolescents Scale (DBAS). Scores for the DBPS ranged from 0 to 12 with a mean of 5.95 and standard deviation of 3.38. Scores for the DBAS ranged from 0 to 12 as well with a mean of 9.68 and a standard deviation of 3.03.
Analysis
Bivariate comparisons (using Chi-Square or Analysis of Variance) were conducted between those FMPs who endorsed an annual visit (n = 164) as compared to those who did not (n = 57). Significant predictors (p < .05) were combined into two multivariate logistic regression models.
We used these two multivariate logistic regression models to separately model the influence of the DBPS and the DBAS (on the endorsement of annual well visits). All models controlled for patient race, patient age (proportion of 13 to 17 years olds in the provider’s practice), and the DBS. All analyses were performed using SPSS, 22 and STATA, 12.0.