Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 May 13.
Published in final edited form as: J Ambul Care Manage. 2025 Feb 24;48(2):84–94. doi: 10.1097/JAC.0000000000000520

Methodology of a Social Network Survey in Primary Care Practices With Medical Home Attributes

Justinna Dixon 1, Eleanor Turi 2,3,4, Madeline Pollifrone 5, Kyle Featherston 6, Jianfang Liu 7, Grant Martsolf 8, Lusine Poghosyan 9,10
PMCID: PMC12070469  NIHMSID: NIHMS2072938  PMID: 39961063

Abstract

Objective:

To provide an overview of the methods of a social network survey used to collect data in primary care practices on team structures, compositions, and social networks (eg, support, communication).

Methods:

A cross-sectional sociometric social network survey in 23 primary care practices with medical home attributes in New York and Pennsylvania was conducted. All primary care providers (ie, physicians, nurse practitioners, physician assistants), clinical staff (eg, registered nurses, social workers, and nutritionists, etc), and administrative staff (eg, practice managers, office assistants, etc) in each practice were eligible to participate in the online survey. All practice members received an online survey asking them to identify who they work with during a typical week. Data were collected about the characteristics of respondent’s relationships with each identified team member, perceived team effectiveness, and respondent-level job satisfaction, intent to leave, burnout, and demographics. We calculated the overall response rate and compared it among practices with different practice sizes and respondent roles using chi-square tests and independent samples t-tests.

Results:

The survey yielded an average response rate of 51.8%, with a range between 21.8% and 82.3% across the practices. Three hundred ninety-four primary care providers, 222 clinical staff, and 173 administrative staff responded to the survey. Respondents (n = 792) were more likely to be from smaller practices and be administrative or clinical staff rather than providers.

Conclusion:

This social network survey of primary care providers and staff used innovative approaches to collect data on team structures beyond traditional methods. Despite declining response rates in healthcare provider surveys, our survey reached high response rates in several practices. Future studies should find ways to engage primary care providers and staff in survey research.

Keywords: healthcare workforce, primary care, social network analysis, survey methods, team-based care

INTRODUCTION

The demand for accessible, high-quality primary care services in the United States is growing due to the aging population, increasing chronic disease burden, and insurance coverage expansion (Ansah Chiu, 2022; Brown et al., 2021). The heightened complexity and volume of patients challenge the primary care system’s ability to meet the need for coordinated, comprehensive, and continuous care (Blumenthal, 2024; Boersma et al., 2020; Davis et al., 2022). As a result, innovative care models are needed to support primary care practices and providers to deliver care. One such model is the medical home model—a system-level approach to improve the quality of primary care services (O’Dell, 2016). Primary care practices with medical home models of care emphasize providing team-based, comprehensive, patient-centered, accessible, and safe primary care (Agency for Healthcare Research and Quality, 2022). Policymakers, professional organizations, and other stakeholders have encouraged primary care practices to adopt the principles of medical homes (National Committee on Quality Assurance, 2024). Team-based care is a major component of the medical home model. Broadly, it is defined as at least two health providers who work collaboratively with patients and their caregivers to accomplish shared goals within and across settings to achieve coordinated, high-quality care (National Academies of Sciences, E., and Medicine; Health and Medicine Division; Board on Health Care Services; Committee on Implementing High-Quality Primary Care, 2021). Well-functioning healthcare teams can improve the coordination, efficiency, effectiveness, and value of primary care (National Academies of Sciences, E., and Medicine; Health and Medicine Division; Board on Health Care Services; Committee on Implementing High-Quality Primary Care, 2021).

Ample evidence exists to support the effectiveness of team-based primary care (Kong Bodenheimer, 2022). However, the definitions of teams in primary care are often vague and inconsistent (Li et al., 2023; Tandan et al., 2024). Previous studies have focused on interactions between primary care providers such as physicians and nurse practitioners or other patient-facing staff and have used inconsistent definitions of collaboration and teamwork (Kuo et al., 2021; Mundt Zakletskaia, 2019; Poghosyan et al., 2020; Tandan et al., 2024). These past studies did not focus on non-clinical staff, such as administrators, schedulers, and receptionists, who play key roles in primary care practices and contribute to their overall performance. Furthermore, past studies mostly use traditional survey research methods, conceptualize team membership from the perspective of one team member, and have relied on participant memory to describe team assignments and roles (Du et al., 2020; Everett et al., 2022). To fill this gap in the literature, innovative and robust methods are needed to understand team dynamics from the perspectives of all members of primary care teams.

Team members are connected to each other not only through their formal roles but also through their informal connections, called social networks (Kilduff Tsai, 2003). Through these networks, team members communicate, seek advice, share support, and trust one another to carry out their job-related tasks (Cross Parker, 2004). Key components of social networks are actors, the individuals in the network, and ties, the relations between the actors. Social networks can be egocentric (based on the perspective of a single actor) or sociometric (bounded by a characteristic(s) that defines or limits the network). The connections between team members, known as ties, facilitate these interactions (Borgatti et al., 2009). Social networks are important for effective teamwork in primary care practices and determine how team members exchange resources and perform tasks. They are also important for patient outcomes, such as patient satisfaction and fall prevention (Hu et al., 2021; Poghosyan et al., 2016). Overall, understanding primary care team compositions and social networks is essential to identifying and designing high-performing primary care teams to deliver high-quality patient care and improve patient outcomes.

Thus, we conducted a cross-sectional sociometric network survey in 23 primary care practices with medical home attributes in two states to collect data from all primary care providers and staff to understand primary care team composition, social networks, and potentially their impact on team members and patient outcomes. Social network approaches differ from traditional social and health science research methods, yielding evidence of a team’s dynamic nature (Smit et al., 2020). The purpose of this paper is to provide an overview of the social network survey methods used to collect data from all primary care practice members, including primary care providers (ie, physicians, nurse practitioners, physician assistants), clinical staff (eg, registered nurses, social workers, and nutritionists, etc), and administrative staff (eg, practice managers, office assistants, etc). We describe our survey methods, sampling frame, data collection procedures, nonresponse analysis, and overall response rates for the social network survey. We also describe the challenges of conducting a large-scale social network survey in primary care.

METHODS

Design

This study assessed team compositions and social networks in primary care practices. We applied a sociometric social network design in which all actors (providers and staff) were included in the sample, and networks were bounded by the practice’s employee roster (Scott, 1988).

Setting

We sampled 23 primary care practices with medical home attributes within three institutions in New York (n = 13) and Pennsylvania (n = 10). The senior leadership at these institutions identified primary care practices that implemented medical home principles to varying degrees and served patient populations aged 18 and older. We received support from key administrative stakeholders, including medical directors in each practice, before data collection. Survey administration took place one institution at a time from November 2021 to August 2022.

Participants

We surveyed all providers and staff in the sampled 23 primary care practices in New York and Pennsylvania. Practice administrators provided the study team with rosters of all practice members. These rosters included the individual’s name, role, and email address. All primary care providers (ie, physicians, nurse practitioners, physician assistants), clinical staff (eg, registered nurses, social workers, and nutritionists, etc), and administrative staff (eg, practice managers, office assistants, etc) in each practice were eligible to participate in the survey. We contacted all eligible practice members using their official clinic email addresses.

Survey questionnaire

The first question on the survey questionnaire in Supplemental Digital Content, Appendix 1 (http://links.lww.com/JACM/A161). asked respondents to self-identify from the roster and consent to participate. Then, to identify team composition, the respondents were instructed to select individuals from the practice roster with whom they interact during a typical week. The names of all practice members were listed on the online survey. We grouped all providers and staff based on occupation to help respondents more easily identify individuals by role. Respondents were provided with the Agency for Healthcare Research and Quality (AHRQ)’s definition of a primary care team: “a group of primary care practice personnel who identify as members of a team and who work together to provide care for a panel of patients.” After identifying who they interact with during a typical week, respondents answered a series of questions about each identified practice member, including whether or not they considered them members of their team (ie, team membership), how long they have worked together, and on their social networks including frequently respondents communicated and sought out the identified colleagues for knowledge, work-related support, social support, advice, problem-solving, and trust during a typical week. We focused on team membership and social networks as they are critically important for team performance (Chow Chan, 2008). Response options were based on a 5-point Likert scale ranging from “Never” to “Always.” After completing the items about their team members, the respondents were asked to rate the effectiveness of their teams. They also reported their own burnout levels, intentions to leave their current positions, job satisfaction, and demographic information.

In addition to these survey questions, one practice administrator from each clinic completed questions about the structural capabilities of their practice, such as the availability of guideline reminder systems or disease registries. These items have been used by researchers and were associated with quality of care and patient outcomes in primary care practices (Hovsepian et al., 2023; Martsolf et al., 2018).

Data collection procedures

Survey administration occurred between November 2021 and August 2022 and was conducted in one institution at a time. Survey questionnaires were developed and administered using the online data collection software Qualtrics, with both mobile phone and desktop compatibility (Qualtrics, 2020). For each of the 23 practices, the survey was populated with that practice’s employee roster; there were no other differences between the surveys across practices. Invitations to participate in the study were sent to practice members through email. The lead practice administrator for each practice was also asked to send an approved email to alert practice members to the survey and encourage participation. Respondents were assured that practice-level leadership would not have access to their responses. This study received Institutional Review Board approval.

We employed an adapted Dillman approach to maximize the response rate (Dillman et al., 2014). Following the initial email, reminder emails were sent at one-, two-, and three-week intervals. Reminder emails were sent to any practice member who had not completed the survey unless they refused to participate, emails were undeliverable, or they no longer worked at the practice. Four weeks after the initial email was sent, the survey data collection was closed. Respondents were offered a $40 prepaid gift card to complete the survey.

After the initial survey was closed, a random selection of 50% of nonrespondents in each practice was selected for a phase two follow-up. These nonrespondents were sent another email with the survey link. The follow-up survey closed one week after the email was sent. Phase two respondents were offered an increased incentive of $60 for participation.

At the end of the data collection period, we assigned each practice member a final disposition: “respondents” were those who consented to participate in the survey, “eligible nonrespondents” refused consent to the survey, and “ineligible nonrespondents” either self-identified as being ineligible (those who no longer work at that practice) or were deemed ineligible (emails sent to their work email address were undeliverable).

Nonresponse analysis

To assess for nonresponse bias, we conducted a nonresponse analysis. We compared practice roles and practice sizes between respondents and eligible nonrespondents from the rosters. We then compared the demographic characteristics of those who responded to the survey during the first wave of data collection (phase one respondents) to those who responded to the survey during the follow-up period (phase two respondents). We assessed differences in practice role, practice size, demographics, burnout, and perceived team effectiveness. We used chi-square tests (categorical variables) and independent samples t-tests (continuous variables) to identify significant differences between the groups. All analyses were conducted using SAS software (SAS, 2013).

RESULTS

Survey response rate

The final response rate of all eligible practice members across all practices was 51.8%. This response rate was calculated using the American Association for Public Opinion Research (AAPOR) formula two (American Association for Public Opinion Research, 2023). A total of 792 of the 1635 contacted practice members responded to the survey. Of the 843 who did not respond to the survey, 106 (6.2%) were deemed ineligible. Response rates varied widely between practices, with the lowest response rate being 21.8% and the highest response rate being 82.3%. Practice sizes ranged from 16 to 238 providers and staff. The mean number of providers and staff in all practices was 128.8 (SD = 86.5). The median number of providers and staff in all the practices was 104. Response rates, practice size, and number of primary care providers by practice are reported in Table 1.

Table 1.

Primary Care Practice Sizes and Practice-Level Response Rates

Practice Total Number of Practice Providers and Staff Number of Primary Care Providers* Total Response Rate
New York practices
1 238 184 50.0%
2 128 82 53.9%
3 80 32 55.0%
4 94 36 60.6%
5 68 38 70.6%
6 26 9 65.4%
7 80 55 60.0%
8 30 12 70.0%
9 39 26 43.6%
10 54 30 33.3%
11 145 119 33.1%
12 206 150 47.1%
13 32 27 50.0%
Pennsylvania practices
14 34 12 73.5%
15 17 6 82.3%
16 22 7 54.5%
17 16 6 56.2%
18 26 9 42.3%
19 55 36 21.8%
20 44 21 50.0%
21 38 13 73.7%
22 28 7 82.1%
23 29 8 58.6%
Total 1529 927 51.8%
Open in a new tab
*

Primary care providers were defined as physicians, physician trainees, nurse practitioners and physician assistants.

Nonresponse analysis

We used data from the administrator-supplied practice rosters to compare practice roles and practice size among survey respondents and eligible nonrespondents. Survey respondents were from smaller practices compared to those who did not respond to the survey (P = .003). The mean practice size in practices where respondents worked was 122.6 compared to 135.4 for nonrespondents. Additionally, there were significant differences in response rates among different categories of practice roles (P < .001). Administrative staff (68.7%) and clinical staff (64.8%) had higher rates of responding to the survey compared to primary care providers (42.7%). Among primary care providers, there were significant differences in response rates by type (P = .002). Nurse practitioners had the highest response rates (62.8%), while physicians had the lowest (36.9%). Results of the nonresponse analysis between respondents and eligible nonrespondents are reported in Table 2.

Table 2.

Characteristics of Survey Respondents and Eligible Nonrespondents

Variable Survey Respondents (n = 792) Eligible Nonrespondents (N = 737) P Value
Practice size (mean, SD) * 122.6 (86.9) 135.4 (85.5) .003
Practice rolea <.0001
Administrative staff 173 (68.7%) 79 (31.3%)
Patient financial advisors 49 (71.1%) 20 (28.9%)
Clinical staff 222 (65.1%) 119 (34.9%)
Behavioral health and social workers 48 (70.6%) 20 (29.4%)
Medical assistants 91 (72.2%) 35 (27.8%)
Registered nurses 53 (57.6%) 39 (42.4%)
Other 3 (33.3%) 6 (66.7%)
Primary care providers 394 (42.5%) 533 (57.5%) <.001
Nurse practitioners 32 (62.8%) 19 (37.3%)
Physician assistants 7 (53.9%) 6 (46.1%)
Physician trainees 227 (43.9%) 289 (56.1%)
Physicians 128 (36.9%) 219 (63.1%)
Open in a new tab
*

Independent samples t-tests were calculated.

a

Chi-square tests were calculated.

In total, 456 primary providers and staff were contacted during the phase two follow-up with 117 responding to the survey. There were significant differences in practice size between phase one respondents and phase two respondents (P = .002). Phase one respondents were from smaller practices (118.6, SD = 86.6) compared to phase two respondents (145.3, SD = 85.8). There were also significant differences in practice roles between phase one and phase two respondents (P < .001); approximately 21% of primary care providers responded during phase two compared to 9% of clinical staff and 8% of administrative staff. There were no significant differences in primary care provider type between phase one and phase two respondents.

There were no significant differences between phase one and two respondents regarding self-reported race, age, gender, marital status, perceived team effectiveness, and burnout. However, there were significant differences in educational level between phase one respondents and phase two respondents (P = .003). A higher proportion (10%) of those with a Master’s, PhD, or other professional degree responded to the survey during the phase two follow-up compared to 3% of those with bachelor’s degrees or less. The results of this analysis are reported in Table 3. Due to partially completed survey responses among phase two respondents, there was greater than 20% missing data for measures of age, race, gender, marital status, and team effectiveness.

Table 3.

Characteristics of Phase One Survey Respondents and Phase Two Respondents

Variable Phase One Respondents (n = 675) Phase Two Respondents (n = 117) P Value
Practice size (mean, SD) * 118.6 (86.6) 145.3 (85.8) .002
Staff rolea <.0001
 Administrative staff 159 (91.9%) 14 (8.1%)
 Patient financial advisor 49 (100%) 0 (0%)
 Clinical staff 202 (90.8%) 20 (9.2%)
Behavioral health and social workers 42 (87.5%) 6 (12.5%)
 Medical assistants 85 (93.4%) 6 (6.6%)
 Registered nurses 47 (88.7%) 6 (11.3%)
 Other 3 (100%) 0 (0%)
 Providers 311 (79.1%) 83 (20.9%) .21
 Nurse practitioners 28 (87.5%) 4 (12.5%)
 Physician assistants 7 (100%) 0 (0%)
 Physician trainees 180 (79.3%) 47 (20.7%)
 Physicians 96 (75%) 32 (25%)
Racea .80
 Asian 74 (92.5%) 6 (7.5%)
 Black 114 (94.2) 7 (5.8%)
 White 293 (92.7%) 23 (7.3%)
 Other race/multiple races 82 (95.4%) 4 (4.6%)
Gendera .89
 Female 466 (93.2%) 34 (6.8%)
 Male 122 (93.1%) 9 (6.9%)
 Other 3 (100%) 0 (0%)
Age (years)a .43
 21–30 153 (94.4%) 9 (5.6%)
 31–36 151 (90.9%) 15 (9.1%)
 37–51 141 (95.3%) 7 (4.7%)
 51 or older 138 (93.2%) 10 (6.8%)
Marital statusa .28
 Married 266 (94.3%) 16 (5.7%)
 Other marital status 317 (92.2%) 27 (7.8%)
Education levela .003
 Less than an associate degree 123 (96.9) 4 (3.1%)
 Associate or Bachelor’s degree 157 (96.9%) 5(3.1%)
 Post-secondary education (Masters, Doctorate, professional degree) 309 (90.1%) 34 (9.9%)
Perceived team effectivenessa .37
 1–5 (not effective) 46 (90.2%) 5 (9.8%)
 6–10 (effective) 542 (93.5%) 38 (6.5%)
Burnouta .08
 No burnout 393 (94.5) 23 (5.5%)
 Symptoms of burnout 196 (90.7%) 20 (9.3%)
Open in a new tab
*

Independent samples t-tests were calculated.

a

Chi-square tests were calculated.

DISCUSSION

We report the methods of an innovative social network survey of primary care practice providers and staff conducted in two states. We fielded an online survey distributed to 1635 members, with four outreach emails over four weeks. After the initial outreach period, half of the non-respondents were recruited for a phase two follow-up. Ultimately, our survey yielded a 51.8% response rate across 23 practices. Practice-level survey response rates ranged from 22% to 82%.

Our survey’s response rate was comparable to previous cross-sectional surveys of healthcare providers (Brtnikova et al., 2018; Ericson et al., 2023). Previous social network surveys of healthcare providers have yielded similar or higher response rates (De Brún McAuliffe, 2018; Mundt Zakletskaia, 2019; Sullivan et al., 2019). The response rates across these network studies range from 53% to 97%. However, these previous surveys have typically targeted smaller networks of individuals and were also conducted prior to the COVID-19 pandemic. Our analysis showed that primary care providers, such as physicians and nurse practitioners, were the least likely to respond to the survey. Furthermore, physicians were the least likely to respond among all primary care providers. These lower response rates align with a broader trend of declining response rates among providers during and after the COVID-19 pandemic. At the time of our survey administration, healthcare providers were inundated with invitations to participate in survey research, contributing to higher rates of respondent fatigue and lower response rates (de Koning et al., 2021; Gnanapragasam et al., 2022).

Our analysis comparing phase one and phase two respondents found no differences regarding demographics, burnout, or team effectiveness. However, there were significant differences in respondent education level, practice role, and practice size. Literature suggests that survey respondents who respond in later phases of data collection efforts are more likely to have characteristics similar to those people who did not respond to the survey at all (Gummer Struminskaya, 2020), thus we compared the characteristics of phase one respondents and phase two respondents. We found similar characteristics between nonrespondents and phase two respondents. For example, both nonrespondents and phase two respondents were more likely to be primary care providers and come from larger practices. Overall, our nonresponse analysis provides vital information about potential systematic bias stemming from significant differences between respondents and nonrespondents.

Our survey provides valuable insights into the challenges of surveying primary care providers and staff. Many of our sampled participants were deemed ineligible due to leaving their practice between the time we obtained the clinic roster and survey administration. This reflects the high turnover rate of healthcare professionals during the COVID-19 pandemic and corroborates literature that reports high turnover rates among ambulatory care staff during this time (Frogner Dill, 2022). The majority (75%) of ineligible nonrespondents who had work contact information that was repeatedly undeliverable were identified from practice-supplied rosters as primary care providers, signaling high turnover rates within this population.

Strengths, future research, and limitations

A strength of the study is that response rates for several practices were greater than 75%, allowing for adequate data to perform social network analysis on team structures, compositions, and networks in future work (Borgatti et al., 2024). We achieved an acceptable response rate despite declining response rates among providers by collaborating with practice administrative leaders and having multiple outreach attempts. For the first time, we have rich data from all primary care providers and staff—including medical assistants, social workers, care managers, and others—in practice to determine team composition and networks. Obtaining diverse perspectives is critically important in future team design efforts, as primary care practices face an overburdened workforce. Future research will focus on better understanding primary care team compositions, structures, social networks, and how they impact team member and patient outcomes.

Our study has limitations. In addition to general primary care practices, we included one obstetrics-gynecology and one pediatric practice in our sample. These practices may differ in organizational structure and team dynamics from other primary care practices. Furthermore, our analysis between survey respondents and eligible nonrespondents was limited to data from the administrator-supplied rosters. Therefore, we could not compare these groups on demographics or other potentially important factors. Our survey respondents and eligible nonrespondents may have differed in other unmeasured ways. For example, as a result of our partnership with practice administration, team members in practices with more engaged leadership may have been more likely to respond to the survey. Additionally, our analysis between phase one and phase two respondents was limited as there was greater than 20% missing data for demographic characteristics for phase two respondents. However, among our final analytic sample of those who completed the survey, there was less than 10% missing data. Lastly, our sample is limited to primary care practices in New York and Pennsylvania, thus limiting generalizability.

CONCLUSIONS

Team-based care is a cornerstone of primary care, and collecting data to study teams is vital. This online social network survey of primary care providers and staff used innovative approaches to collect data on team structures. We collected data from all providers and staff in primary care practices to determine team compositions and structure. Our data will allow us to analyze social networks in primary care settings and provide vital information about the characteristics and dynamics of primary care teams in future research. This evidence will be key to ultimately promoting team-based primary care and improving patient care and outcomes.

Supplementary Material

Supplementary Appendix

Key Points.

  • Increasing demand for primary care services has resulted in an emphasis on team-based care models. There are wide variations across primary care teams, yet little is known about their composition and structure.

  • Previous social network surveys of primary care teams have targeted singular practices and primarily focused on patient-facing staff.

  • This study details the methods of an innovative social network survey of 23 primary care practices that aimed to describe team structure and composition.

  • These data will add valuable information about primary care team dynamics from the perspectives of both primary care providers and clinical and administrative staff.

Acknowledgments

J.D.’s effort was supported by a National Institutes of Health, National Institute for Nursing Research T32 training grant (T32 NR014205, Poghosyan, Stone, PI/MPI) Comparative and Cost-Effectiveness Research Training for Nurse Scientists at Columbia University. E.T. is funded by the National Institutes of Health, National Institute of Mental Health T32 training grant (T32MH109433) and the National Clinician Scholars Program. This work was supported by the Agency for Healthcare Research and Quality under award number5R01HS025937-05. The authors report no conflicts of interest.

Contributor Information

Justinna Dixon, School of Nursing, Columbia University, New York City, New York.

Eleanor Turi, National Clinician Scholars Program, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; Center for Mental Health, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; Leonard Davis Institute of Health Economics, University of Pennsylvania, Philadelphia, Pennsylvania.

Madeline Pollifrone, School of Nursing, Columbia University, New York City, New York.

Kyle Featherston, School of Nursing, Columbia University, New York City, New York.

Jianfang Liu, School of Nursing, Columbia University, New York City, New York.

Grant Martsolf, School of Nursing, University of Pittsburgh, Pittsburgh, Pennsylvania.

Lusine Poghosyan, School of Nursing, Columbia University, New York City, New York; Department of Health Policy and Management, Mailman School of Public Health, Columbia University, New York City, New York.

REFERENCES

  1. Agency for Healthcare Research and Quality. (2022). Defining the PCMH. Author. https://www.ahrq.gov/ncepcr/research/care-coordination/pcmh/define.html
  2. American Association for Public Opinion Research. (2023). Standard definitions final dispositions of case codes and outcome rates for surveys. https://aapor.org/response-rates/
  3. Ansah JP Chiu CT (2022). Projecting the chronic disease burden among the adult population in the United States using a multi-state population model. Front Public Health, 10, 1082183. doi: 10.3389/fpubh.2022.1082183 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Blumenthal DM (2024). Increasing clinical complexity—implications for care delivery, payment models, and the health care workforce. JAMA Internal Medicine, 184(2), 192–193. doi: 10.1001/jamainternmed.2023.7407 [DOI] [PubMed] [Google Scholar]
  5. Boersma P, Black LI, Ward BW (2020). Prevalence of multiple chronic conditions among US adults, 2018. PrevChronicDis, 17, E106. doi: 10.5888/pcd17.200130 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Borgatti SP, Agneessens F, Johnson JC, Everett MG (2024). Analyzing social networks (Third ed). SAGE. [Google Scholar]
  7. Borgatti SP, Mehra A, Brass DJ, Labianca G (2009). Network analysis in the social sciences. Science, 323(5916), 892–895. doi: 10.1126/science.1165821 [DOI] [PubMed] [Google Scholar]
  8. Brown EA, White BM, Jones WJ, Gebregziabher M, Simpson KN (2021). Measuring the impact of the affordable care act medicaid expansion on access to primary care using an interrupted time series approach. Health Research Policy and Systems, 19(1), 77. doi: 10.1186/s12961-021-00730-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Brtnikova M, Crane LA, Allison MA, Hurley LP, Beaty BL, Kempe A (2018). A method for achieving high response rates in national surveys of U.S. primary care physicians. PLoS One, 13(8), e0202755. doi: 10.1371/journal.pone.0202755 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chow WS Chan LS (2008). Social network, social trust and shared goals in organizational knowledge sharing. Information Management, 45(7), 458–465. doi: 10.1016/j.im.2008.06.007 [DOI] [Google Scholar]
  11. Cross RL Parker A (2004). The hidden power of social networks: Understanding how work really gets done in organizations. Harvard Business Press. [Google Scholar]
  12. Davis AC, Chen A, Osuji TA, Chen J, Gould MK (2022). Defining complex patient populations: Implications for population size, composition, utilization, and costs. Journal of General Internal Medicine, 37(2), 351–358. doi: 10.1007/s11606-021-06815-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. De Brún A McAuliffe E (2018). Recruitment of healthcare staff to social network studies: A case study exploring experiences, challenges, and considerations. International Journal of Environmental Research and Public Health, 15(12), 2778. doi: 10.3390/ijerph15122778 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. de Koning R, Egiz A, Kotecha J, Ciuculete AC, Ooi SZY, Bankole NDA … … Kanmounye US (2021). Survey fatigue during the COVID-19 pandemic: An analysis of neurosurgery survey response rates. Front Surg, 8, 690680. doi: 10.3389/fsurg.2021.690680 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dillman DA, Smyth JD, Christian LM (2014). Internet, phone, mail, and mixed-mode surveys: The tailored design method. Wiley. https://books.google.com/books?id=fhQNBAAAQBAJ [Google Scholar]
  16. Du S, Wiegmann D, Beasley J, Steege L, Wetterneck T (2020). Defining team membership in primary care: Qualitative analysis. IISE Transactions on Healthcare Systems Engineering, 10(4), 251–260. doi: 10.1080/24725579.2020.1800869 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ericson A, Bonuck K, Green LA, Conry C, Martin JC, Carney PA (2023). Optimizing survey response rates in graduate medical education research studies. Family Medicine, 55(5), 304–310. doi: 10.22454/FamMed.2023.750371 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Everett CM, Docherty SL, Matheson E, Morgan PA, Price A, Christy J … … Jackson GL (2022). Teaming up in primary care: Membership boundaries, interdependence, and coordination. Jaapa, 35(2), 1–10. doi: 10.1097/01.JAA.0000805840.00477.58 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Frogner BK Dill JS (2022). Tracking turnover among health care workers during the COVID-19 pandemic: A cross-sectional study. JAMA Health Forum, 3(4), e220371. doi: 10.1001/jamahealthforum.2022.0371 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gnanapragasam SN, Hodson A, Smith LE, Greenberg N, Rubin GJ, Wessely S (2022). COVID-19 survey burden for health care workers: Literature review and audit. Public Health, 206, 94–101. doi: 10.1016/j.puhe.2021.05.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gummer T Struminskaya B (2020). Early and late participation during the field period: Response timing in a mixed-mode probability-based panel survey. Sociological Methods Research, 52(2), 909–932. doi: 10.1177/0049124120914921 [DOI] [Google Scholar]
  22. Hovsepian VE, Sadak T, Schlak AE, Liu J, Poghosyan L (2023). The association between primary care practices’ structural capabilities and hospitalizations among persons living with dementia. Journal of Applied Gerontology, 42(7), 1414–1423. doi: 10.1177/07334648231155444 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hu H, Yang Y, Zhang C, Huang C, Guan X, Shi L (2021). Review of social networks of professionals in healthcare settings—where are we and what else is needed? Globalization Health, 17(1), 139. doi: 10.1186/s12992-021-00772-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kilduff M Tsai W (2003). Social Networks and Organizations. SAGE Publications Ltd. doi: 10.4135/9781849209915 [DOI] [Google Scholar]
  25. Kong M Bodenheimer T (2022). Primary care urgently needs support for interprofessional teams. The Permanente Journal, 26(4), 100–103. doi: 10.7812/TPP/22.105 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kuo Y-F, Agrawal P, Chou L-N, Jupiter D, Raji MA (2021). Assessing association between team structure and health outcome and cost by social network analysis. Journal of the American Geriatrics Society, 69(4), 946–954. doi: 10.1111/jgs.16962 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Li M, Tang H, Liu X (2023). Primary care team and its association with quality of care for people with multi-morbidity: A systematic review. BMC Primary Care, 24(1), 20. doi: 10.1186/s12875-023-01968-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Martsolf GR, Ashwood S, Friedberg MW, Rodriguez HP (2018). Linking structural capabilities and workplace climate in community health centers. Inquiry, 55, 46958018794542. doi: 10.1177/0046958018794542 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Mundt MP Zakletskaia LI (2019). Professional communication networks and job satisfaction in primary care clinics. Annals Fam Med, 17(5), 428–435. doi: 10.1370/afm.2442 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. National Academies of Sciences, E., and Medicine; Health and Medicine Division; Board on Health Care Services; Committee on Implementing High-Quality Primary Care; 2021. Designing interprofessional teams and preparing the future primary care workforce. In Robinson SK, Meisnere M, Phillips RL Jr. McCauley L, Eds. Implementing high-quality primary care: Rebuilding the foundation of health care. doi: 10.17226/25983 [DOI] [PubMed] [Google Scholar]
  31. National Committee on Quality Assurance. (2024). Patient-centered medical home recognition. Author. https://www.ncqa.org/employers/ncqa-programs-of-interest-to-employers/patient-centered-medical-home-recognition/
  32. O’Dell ML (2016). What is a patient-centered medical home? Mo Med, 113(4), 301–304. [PMC free article] [PubMed] [Google Scholar]
  33. Poghosyan L, Ghaffari A, Liu J, Friedberg MW (2020). Physician-nurse practitioner teamwork in primary care practices in New York: A cross-sectional survey. Journal of General Internal Medicine, 35(4), 1021–1028. doi: 10.1007/s11606-019-05509-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Poghosyan L, Lucero RJ, Knutson AR, M WF, Poghosyan H (2016). Social networks in health care teams: Evidence from the United States. Journal Health Organization and Management, 30(7), 1119–1139. doi: 10.1108/jhom-12-2015-0201 [DOI] [PubMed] [Google Scholar]
  35. Qualtrics. (2020). Qualtrics LLC. https://www.qualtrics.com
  36. SAS. (2013). SAS version 9.4. SAS Institute. [Google Scholar]
  37. Scott J (1988). Social network analysis. Sociology, 22(1), 109–127. [Google Scholar]
  38. Smit LC, Dikken J, Schuurmans MJ, de Wit NJ, Bleijenberg N (2020). Value of social network analysis for developing and evaluating complex healthcare interventions: A scoping review. BMJ Open, 10(11), e039681. doi: 10.1136/bmjopen-2020-039681 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sullivan P, Saatchi G, Younis I, Harris ML (2019). Diffusion of knowledge and behaviours among trainee doctors in an acute medical unit and implications for quality improvement work: A mixed methods social network analysis. BMJ Open, 9(12), e027039. doi: 10.1136/bmjopen-2018-027039 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tandan M, Dunlea S, Cullen W, Bury G (2024). Teamwork and its impact on chronic disease clinical outcomes in primary care: A systematic review and meta-analysis. Public Health, 229, 88–115. doi: 10.1016/j.puhe.2024.01.019 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Appendix
NIHMS2072938-supplement-Supplementary_Appendix.pdf (428.2KB, pdf)

RESOURCES