Effects of Primary Care Team Social Networks on Quality of Care and Costs for Patients With Cardiovascular Disease

Marlon P. Mundt; Valerie J. Gilchrist; Michael F. Fleming; Larissa I. Zakletskaia; Wen-Jan Tuan; John W. Beasley

doi:10.1370/afm.1754

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
Download figure
Open in new tab
Download powerpoint
Figure 1
Primary care team communication networks about patient care.

CM = clinic manager; LT = laboratory technician; MA = medical assistant; MD = physician; MR = medical receptionist; RN = registered nurse; RT = radiology technician.

Note: Symbol size proportional to number of connections.
Download figure
Open in new tab
Download powerpoint
Figure 2
Structural equation model of team social networks and quality of care for patients with cardiovascular disease (n = 31 primary care teams, n = 7,457 patients).

CVD = cardiovascular disease; ED = emergency department.

Notes: Pathway coefficients between boxes denote the standardized change in end point outcome variable associated with a 1 SD increase in lead predictor variable. By multiplying pathway coefficients between structure, process, and outcome measures, the analysis estimates team social network impact on health care utilization. As an example, for every 1 SD increase in face-to-face interaction density, urgent care visits, emergency department visits, and hospital days decrease by 0.562 (0.816*0.689), 0.420 (0.816*0.515), and 0.380 (0.816*0.466) SDs, respectively.

^aP = <.001.

^bP = <.01.

View popup

Table 1

Study Sample (n = 155) Face-to-Face In-Degree Centrality

Characteristic	No. (%)	Mean (SD)	Range^a
Sex
Male	8 (5.2)	7.2 (2.7)	3–11
Female	147 (94.8)	10.8 (4.0)	2–25
Position
Physician	20 (12.9)	7.2 (2.3)	3–12
NP/PA	7 (4.5)	7.3 (2.0)	5–11
Clinic manager	6 (3.9)	10.8 (2.8)	8–15
RN	30 (19.4)	12.4 (4.3)	4–25
LPN/MA	29 (18.7)	13.0 (2.7)	9–19
Medical receptionist	38 (24.5)	11.8 (3.0)	5–17
Laboratory/radiology technician	23 (14.8)	8.2 (4.4)	2–18
Other (scribe/phlebotomist)	2 (1.2)	4.5 (0.7)	4–5
Years at clinic
≤1	30 (19.4)	10.1 (4.2)	2–15
1 to 3	43 (27.7)	12.4 (4.4)	4–25
3 to 6	29 (18.7)	9.4 (3.5)	2–18
6 to 10	16 (10.3)	11.1 (3.6)	4–1
>10	37 (23.9)	9.5 (3.1)	4–15
Full-time employment
≤50%	23 (14.8)	7.3 (3.4)	2–14
>50% to 75%	28 (18.1)	9.2 (3.5)	4–17
>75%	104 (67.1)	11.8 (3.7)	4–25

LPN = licensed practical nurse; MA = medical assistant; NP = nurse practitioner; PA = physician assistant; RN = registered nurse.
Note: Count of incoming communication ties (ie, the number of other team members who report communicating with the individual about patient care on a daily basis).
↵a Minimum-maximum.

View popup

Table 2

Primary Care Team Characteristics and Outcomes for Patients With Cardiovascular Disease

Characteristic	Value Mean (SD)	Range^a
Primary care teams (n = 31)
Social networks variable
Team social network density	0.53 (0.12)	0.36–0.77
Team social network centralization	0.29 (0.10)	0.13–0.43
Face-to-face interaction density	0.45 (0.11)	0.31–0.69
Face-to-face interaction centralization	0.28 (0.08)	0.15–0.41
EHR communication density	0.31 (0.08)	0.19–0.46
EHR communication centralization	0.24 (0.07)	0.11–0.43
Team climate (TCI-14) score
Shared vision (scale: 0–16)^b	12.8 (0.4)	11.6–13.6
Psychological safety (scale: 0–16)^b	11.4 (1.1)	8.6–12.3
Task orientation (scale: 0–12)^b	8.5 (0.5)	7.5–9.3
Innovation support (scale: 0–12)^b	7.9 (0.6)	6.3–8.8
Team characteristics
Team size, No.	22.2 (7.5)	12–28
Staff turnover in previous 12 mo, %	19	…
Patient panel outcomes in past 12 mo
Team patients (n = 7,457)
With controlled LDL cholesterol (<100 mg/dL), %	39	…
With controlled blood pressure (<130/80 mm Hg), %	71	…
Health care use per patient in patient panel
Urgent care visits, No.	0.10 (0.44)	…
Emergency department visits, No.	0.24 (1.20)	…
Hospital days, No.	0.65 (4.44)	…
Medical costs, US$	1,241 (7,538)	…

EHR = electronic health record; LDL = low-density lipoprotein; TCI-14 = Team Climate Inventory.
↵a Minimum-maximum.
↵b Where higher scores indicate better team climate.

View popup

Table 3

Hierarchical Modeling of the Association Between Team Social Networks and Outcomes for Panel of Patients With Cardiovascular Disease

Variable	BP Control OR (95% CI)	LDL Cholesterol Control OR (95% CI)	UC Visits RR (95% CI)	ED Visits RR (95% CI)	Hospital Days RR (95% CI)	Cost ($)β (SE)
Reduced form model
Team social network density^a	1.14 (0.98–1.33)	1.13 (0.97–1.31)	0.28^b (0.19–0.40)	0.58^c (0.42–0.80)	0.64^b (0.51–0.80)	−516^b (129)
Team social network centralization^a	1.00 (0.84–1.19)	0.94 (0.80–1.11)	3.07^b (2.09–4.51)	1.70^c (1.23–2.35)	1.47^c (1.14–1.89)	519^c (143)
Full model^d
Team social network density^a	1.15 (0.99–1.34)	1.14 (1.00–1.31)	0.95 (0.55–1.66)	0.98 (0.50–1.89)	0.62^b (0.50–0.77)	−556^b (115)
Team social network centralization^a	1.03 (0.85–1.25)	0.93 (0.79–1.08)	1.20 (0.79–1.81)	1.33 (0.83–2.13)	1.45^c (1.09–1.94)	506^c (155)

BP = blood pressure; ED = emergency department; LDL = low-density lipoprotein; OR = odds ratio; RR = rate ratio; SE = standard error; UC = urgent care.
Note: There were 31 primary care teams and 7,457 patients who had cardiovascular disease.
↵a Team social network interaction related to daily face-to-face and/or EHR communication connections about patient care in a team.
↵b P <.001.
↵c P <.01.
↵d Patient-level covariates entered in the full model were sex, age, age squared, race/ethnicity, insurance, and comorbidity.

View popup

Table 4

Three-Level Hierarchical Model of the Association Between Face-to-Face and EHR Communication Networks and Outcomes for Patients With Cardiovascular Disease (n = 31 teams, n = 7,457 patients)

Variable	BP Control Adj OR (95% CI)	LDL Cholesterol Control Adj OR (95% CI)	UC Visits Adj RR (95% CI)	ED Visits Adj RR (95% CI)	Hospital Days Adj RR (95% CI)	Cost (US$) Adj β (SE)
Primary care team social networks
Face-to-face interaction density	1.19 (0.89–1.59)	1.17^a (1.00–1.37)	0.34^a (0.12–0.98)	0.28^b (0.12–0.63)	0.63 (0.40–1.00)	−594^a (240)
EHR communication density	0.96 (0.71–1.29)	0.85^a (0.72–1.00)	1.64 (0.71–3.81)	2.34^a (1.10–4.97)	1.010. (63–1.60)	60 (247)
Patient characteristics
Male	1.52^c (1.23–1.87)	1.67^c (1.51–1.88)	0.79^b (0.67–0.93)	0.94 (0.85–1.04)	0.97 (0.91–1.03)	118 (180)
Age	0.99^b (0.99–1.00)	1.01^c (1.01–1.02)	0.96^c (0.95–0.97)	0.98^c (0.97–0.99)	0.99 (0.98–1.00)	−24^b (8)
Age-squared (÷10)	1.00 (0.99–1.01)	0.97^c (0.96–0.98)	1.00 (0.99–1.01)	1.01^c (1.00–1.01)	1.00 (0.99–1.01)	−1 (3)
Non-Hispanic White	0.58 (0.33–1.01)	1.22 (0.88–1.69)	0.64^b (0.49–0.84)	0.68^c (0.56–0.81)	1.52^c (1.27–1.83)	205 (540)
White Hispanic	2.01 (0.97–4.18)	1.48 (0.91–2.39)	1.16 (0.78–1.73)	1.17 (0.87–1.57)	0.50^c (0.34–0.74)	−453 (828)
Black	0.72 (0.49–1.07)	0.86 (0.56–1.32)	1.79^a (1.05–3.06)	1.89^a (1.03–3.47)	0.80 (0.32–2.00)	−160 (705)
Private insurance	4.71^c (2.44–9.12)	1.69^c (1.36–2.10)	1.88^b (1.27–2.79)	0.71^c (0.59–0.85)	0.46^c (0.42–0.51)	−1,135^b (359)
Medicaid	2.64^a (1.04–6.73)	1.40 (0.98–1.99)	3.18^c (2.03–4.97)	1.45^b (1.15–1.84)	0.81^a (0.69–0.95)	−588 (581)
Medicare	3.65^c (1.90–7.04)	2.24^c (1.81–2.79)	1.52^a (1.01–2.29)	1.51^c (1.28–1.78)	0.84^c (0.77–0.91)	−817^a (364)
Charlson comorbidity index	1.14^c (1.09–1.20)	1.23^c (1.20–1.26)	1.06^b (1.02–1.10)	1.32^c (1.29–1.34)	1.36^c (1.34–1.37)	690^c (47)

Adj OR = adjusted odds ratio; Adj RR = adjusted rate ratio; Adj β = adjusted regression coefficient; BP = blood pressure; ED = emergency department; EHR = electronic health record; LDL = low-density lipoprotein; SE = standard error; UC = urgent care.
Notes: Confidence intervals rounded to 2 significant digits. Statistically significant intervals for OR and RR do not contain 1.
↵a P <.05.
↵b P <.01.
↵c P <.001.

Additional Files

Figures
Tables

The Article in Brief

Effects of Primary Care Team Social Networks on Quality of Care and Costs for Patients With Cardiovascular Disease

Marlon P. Mundt , and colleagues

Background Cardiovascular disease (CVD) is the leading cause of death and disease in the US. Primary care teams, which provide support and share responsibilities for patient care, offer a unique opportunity to improve quality and lower medical costs for patients with CVD. This study evaluates the association between primary care team communication, interaction, and coordination (i.e., social networks), quality of care, and costs for patients with CVD.

What This Study Found Primary care teams that are more interconnected, less centralized, and have a shared team vision are better positioned to deliver high-quality cardiovascular disease care at a lower cost. Teams with more members reporting daily interactions with a greater number of team members show better quality of care, with a 38 percent reduction in hospital days and, on average, $516 less spent per patient in the previous 12 months. The study findings suggest that teams with more daily face-to-face interactions have fewer urgent care and emergency department visits and $594 less spent in medical costs per patient in the previous 12 months.

Implications

In this study, a team's shared vision of goals and expectations mediates the relationship between social network structures and patient quality of care outcomes.