Provider views on childhood obesity management in primary care settings: a mixed methods analysis

We conducted four focus group discussions in December 2013 among pediatric providers in the Children’s Primary Care Medical Group (CPCMG), a large pediatric network in San Diego County. CPCMG has 20 offices in San Diego County and southern Riverside County with over 110 physicians, servicing a wide range of racial and socioeconomic regions. In conjunction with the CPCMG leadership, four sites in four demographically different regions of San Diego were identified to participate in the focus group to ensure that a wide range of practice experiences were represented. These areas included Poway, La Jolla, Escondido, and Chula Vista, and consisted of 26.1%–38.9% OW/OB children, 15.7%–58.2% Hispanic, and 5.9%–23.2% of families living below the federal poverty level. All pediatric providers (physicians and nurse practitioners) at each site were invited to participate in the focus group which occurred in their office during the lunch hour. Twenty-two providers (physicians and nurse practitioners) participated. Lunch was provided, but no other means of compensation was offered. The study was approved by the Institutional Review Board at the University of California, San Diego.

Focus group discussions were conducted in English and consisted of 2 to 10 providers. While themes started to converge by the second and third focus group, four focus groups were conducted (one at each site) to ensure that all demographic locations would be represented. Before the start of each discussion, providers completed the informed consent process and a short (10-item) questionnaire assessing their age, demographic characteristics, medical training, and training in obesity management. Focus group discussions were conducted by a researcher trained in qualitative methods and conducting focus groups (KR). A facilitator was present to take notes and assist with follow-up questions. A semi-structured interview guide was developed by the research team and based on researcher expertise, experience of the CPCMG leadership, and empirical literature. Topics included: current attitudes and behaviors towards weight management in the office; barriers to addressing weight in the office setting; and factors or solutions that would increase providers’ abilities to manage weight in the office (e.g., resources, staffing, electronic tools). The focus group leader followed this guide and allowed participants to openly express their opinions. Since focus group members worked together in the office, numbers were not assigned to protect anonymity. However, it was reiterated to the group that all responses would remain confidential and anonymous during the coding and reporting process. Focus group discussions were 45 min in length. Discussions were audio-taped, and the tapes transcribed verbatim for analysis. Transcriptions were entered into qualitative software (Atlas.ti Version 7.5.11 (2016, Scientific Software Development GmbH, Berlin)) for analysis.

Qualitative methods were used to analyze the focus group discussions [22]. Transcripts were independently coded by two authors (KR and SK). First, one investigator read all transcribed focus groups and applied the principles of microanalysis, [23] an in-depth analysis of the text to generate initial themes and create a preliminary coding scheme. The second investigator (SK) then attempted to apply the initial coding scheme to each transcript. The investigators met to refine the coding scheme and discuss new emergent themes using the constant comparison method [23]. Consensus was reached on the definition and application of each theme. Codes were associated with segments of dialogue based on a priori themes (i.e., questions asked in the focus group) or emergent themes (i.e., central ideas from the data). Different codes could be applied to the same segments of the transcript. Atlas.ti Version 7.5.11 (2016, Scientific Software Development GmbH, Berlin) was used to organize codes and their subcategories. SAS v9.4 (SAS Institute Inc., Cary, NC) was used to obtain descriptive statistics of the sample.

Providers were asked on a 4-point Likert-type scale how important it was to address weight with their OW/OB patients (1 = not important, 4 = very important). They were also asked to indicate what they felt their role should be in the management of pediatric obesity and check all the responses that were appropriate. Providers were then asked to report how effective they were in the behavioral management of obesity (4-point Likert-type scale, 1 = not very effective, 4 = very effective). Based on the focus group discussion, a list of obesity management choices were offered and providers were asked to rank their choices from most desirable (1) to least desirable (7). Choices ranged from group visits for families during regular office hours to a health educator or nurse who would engage in behavioral weight loss management in the office (see Table 3 for full list of options). Providers were then asked about specific electronic medical record (EMR) changes that they would like to see to help them with obesity management in their clinic setting. Providers were allowed to choose as many options as they deemed suitable. Providers were also asked about their knowledge of community resources and how they would like to administer surveys. Demographic characteristics (sex, years in practice), type of medical training, and prior obesity training (CME courses) were ascertained at the end. Descriptive analysis, including means and frequencies, was conducted using SAS v9.4 (SAS Institute Inc., Cary, NC).

A total of 22 providers participated in the focus groups. The majority (68.2%) were female, 81.8% Caucasian, and 77.3% physicians. One-third (36.3%) of the group had been practicing for 10 years or less, while 40.9% had been practicing for 11–20 years. Half of the group (50%) had never participated in childhood obesity training in the past, while 31.8% had participated in 1–5 obesity-related CME courses.

The majority of barriers fell into four categories: provider-level/personal barriers, practice-based/systems barriers, parent-level barriers, and environmental barriers (Table 1). Among provider-level barriers, the most commonly reported issues were lack of knowledge and confidence around obesity management, particularly around effective means of communication. The majority of providers were not comfortable with offering weight management advice and delivering it in a meaningful and effective manner. Many providers reported being nervous about addressing this topic with their patients and parents for fear of offending them.

The most common practice-based or systems level barriers included lack of time, poor training, and lack of resources in the office and community (Table 1). Providers expressed interest in learning more about what strategies to use to help children lose weight and how to use techniques like motivational interviewing in the office setting. However, lack of time limited their ability to engage in these discussions. Providers also expressed uncertainty regarding community resources, existing treatments, and where to refer their patients. Providers reported that they “didn’t know where else to send [their patients],” and expressed a desire to have a “simple way of knowing…where or even what’s available [in the community].”

Lack of parent motivation and readiness to address obesity was the third most commonly reported barrier. Providers felt that motivating parents to engage in treatment was difficult, particularly if parents did not recognize that their child was overweight (Table 1). Finally, providers identified several environmental-level issues related to access, specifically a lack of treatment options for their patients and an inability or unwillingness to travel among parents in order to participate in treatment.

When discussing possible solutions, providers’ responses centered around 2 major themes: systems-level changes in the office and specific work-flow changes, many of which involved the EMR (Table 2). When discussing systems-level changes, many providers requested additional support in the form of a clinical educator or nutritionist to help deliver effective obesity management strategies in the office. Providers also thought that if this care could be provided in the office, it would be beneficial. Of note, providers were open to addressing obesity in the primary care setting if there was a team approach. However, if education or counseling could not be done in the office, providers also identified the need for a referral coordinator to help manage these patients.

Providers also expressed an interest in having a clear obesity management guideline or best practice guideline within the EMR to help them streamline their care and ensure consistency between providers. In particular, providers wanted guidelines on what labs to order, what referrals to make, and when to bring a patient back for a weight check or additional counseling (Table 2). They were also interested in having evidence-based handouts, links to online resources, and a directory of active community programs and resources. Efforts to make documentation easier (e.g., in the form of smart sets, smart forms, obesity management templates, drop-down menus, and “dot phrases” or text expanders) were also important to providers. Given the intensity of the conversation needed for obesity management, these forms would help to standardize the interaction, but also decrease the documentation burden of an already busy encounter.

Based on the focus group results, a 30-item questionnaire was developed to gather the views of the entire network of providers. Of the provider survey respondents (n = 42), 52.5% were female, 90% were physicians, and 72.5% were 11–20+ years post training. All of the providers thought it was either very important (83.3%) or important (16.7%) for them to address the issue of obesity with their patients. When asked what their role should be in obesity management, the majority indicated that they should inform parents of their child’s weight status, discuss nutrition and physical activity recommendations, conduct laboratory tests, and refer to subspecialty or other weight management services (Fig. 1). However, only 23.8% indicated that providers should be the ones to conduct the behavioral management of obesity. Furthermore, only 23.8% reported that they were “effective” or “very effective” at the behavioral management of obesity. Nevertheless, 90.5% reported that they were interested in learning more about this type of management.

Additional questions were asked to determine what changes would help them streamline their ability to conduct obesity management in the clinic setting. The most highly ranked options were a trained health educator or physician champion who would engage in brief behavior change counseling and obesity management in the office (Fig. 2).

Finally, providers were asked to indicate how the EMR could help them better manage overweight children. Almost 2/3 indicated that they would like a Best Practice Advisory (BPA) to appear if the child had a BMI ≥ 85th percentile. Among those who said yes, 80.8% preferred a smart set or algorithm that they could “opt” to use to help them with subsequent management; less than 20% preferred to have a “hard stop” that forced them to order labs, make referrals, and provide handouts.

There were several options within this smart set that providers wanted (Table 3). First, providers suggested there should be a mechanism in the chart to help them recognize that the child’s BMI was in the OW/OB range, preferably highlighted in the progress note. They also identified a number of tools they thought would be helpful, such as a guide for what labs to order, appropriate nutrition and physical activity assessments, and a list of appropriate diagnosis codes. Templated progress notes would help to streamline the documentation process, and standardized handouts and referral links were requested. Providers were open to the idea of administering a survey to families to help gather more information regarding obesity-related behaviors. They indicated that they would like the medical assistant (MA) to provide this survey to the family to complete in the office (70%), or they would give the family the survey to complete at home and have them return to a follow-up visit to discuss their responses (67.5%). Of note, 90.2% of providers indicated that they did not have adequate knowledge of the available resources in the community to help their families with weight loss, and many preferred a link in the EMR to help them learn about these resources.