Utilization Patterns in an Urban Direct Primary Care Model: A Case Study

Introduction

While accessible, robust primary care is widely recognized as essential in improving patient outcomes and strengthening the healthcare system, ¹ traditional delivery models often present significant barriers to achieving these goals. Direct Primary Care (DPC), a relatively new primary care practice model, has risen in response to these challenges. DPCs are a membership-based model where patients pay a flat, recurring monthly fee to receive unlimited comprehensive primary services, including regular checkups, routine care, and preventative care. ² Aiming to ‘uncouple’ primary care from health insurers, DPCs allow patients to alleviate previous pressures and limitations associated with traditional fee-for-service (FFS) and insurance models. ³ As a bottom-up, physician-driven approach to healthcare, ⁴ DPCs aim to increase transparency and to satisfy the quadruple aim of lower cost and better patient experience, health outcomes, and clinician experience. ⁵

In the view of many, including the American Academy of Family Physicians, DPC’s 2-party structural change can contribute to bending the cost curve while also eliminating barriers to direct physician access associated with insurance-centered healthcare delivery models. Early work suggests that DPC increases clinician time with patients, ⁶ decreases physician burn-out, ⁷ while reducing overhead costs ³ and administrative burden. Another suggested benefit of the DPC model is that clinicians can introduce new services to their patients without prior authorization from insurance companies. In the past 6 years, DPCs have experienced exponential growth among clinical practices, exceeding 1000% ⁸ with over 2000 practices in the United States across 48 states. ⁹ This widespread growth was significantly seen following the COVID-19 pandemic, with HintOS, a management system for DPC practices, expanding its presence from 20 states in 2019 to 40 states by 2021. ¹⁰ The global direct primary care market generated $55.8 billion in 2022 and is predicted to generate $82.5 billion by 2032. ¹¹

Though it offers a variety of potential benefits, gaps remain in our understanding of utilization patterns in DPC practices. For example, some have suggested that DPC may exacerbate already existing healthcare disparities by decreasing access to those who are unable to afford the periodic fees. ¹² The DPC concept also raises the questions of whether DPC is the preferred healthcare model for patients themselves and whether it promotes patient-clinician interactions. Working with a DPC clinic in the Houston area, this study seeks to characterize DPC in a large urban setting and examine patient interaction patterns with DPC providers. This work contributes to the ongoing discourse on DPC by exploring the role of subscription duration in patient engagement and assessing whether DPC fosters the continuity necessary for effective patient-clinician relationships in an urban healthcare environment.

Methods

This is a retrospective cohort study. Deidentified data were obtained from a DPC clinic in the greater Houston area that emphasizes holistic patient care. The data encompassed demographic characteristics and appointment information for patients seen between January 1, 2023, and January 31, 2024. Data included demographic characteristics (patient age, biological sex, and race), appointment history (length of time in months in DPC and visit month/year), frequency of patient-clinic contacts (emails, telemedicine, chart interactions, and SMS), DPC administrative details (subscription cost and subscription duration), and patient geographic residence (urban vs suburban vs rural). Descriptive statistics (mean, standard deviation, frequency, and percentage) were used to summarize the data. Two separate linear regression models examined adjusted relationships between demographic characteristics and frequency of patient-clinic interactions, as depicted by the following independent outcomes of interest: (1) chart interactions defined as interactions initiated by patients and/or clinicians, and (2) appointments made: defined as in-person clinical consults or audio-visual calls. These counts were normalized by adjusting observed use to a 12-month period. For subscriptions shorter or longer than 1 year, the total number of uses was divided by the number of active months and multiplied by 12 to yield an annualized utilization rate.

Due to the deidentified nature of the data, the requirements for informed consent and IRB approval were waived by an independent institutional review board affiliated with the authors’ institution.

Results

Table 1 shows the summary characteristics of the patients at the DPC clinic. Approximately 21% of the sample were aged 19 years or under, 27% were 20 to 39 years, 38% were 40 to 64 years, and 14% were 65 years or older. The sample comprised 77% females, and racial and ethnic subgroups included 21% Non-Hispanic White, 71% Non-Hispanic Black, and 8% Other. For monthly subscription costs, 13% of patients paid $50 or less, 4% paid between $51 and $100, 30% paid $110, 9% paid $135, and 44% paid more than $125 monthly. Regarding DPC subscription duration, 25% of patients had been subscribed for 0 to 5 months, 25% for 6 to 12 months, 25% for 13 to 20 months, and 25% for over 21 months. Although none of the patients reported living in a rural area, 21% resided in suburban areas and 79% in urban areas.

OPEN IN VIEWER

Table 1.

Unique DPC Patient Characteristics at Baseline (n = 112).

Variable	Total
	N (%)
Age
≤19	23 (20.5)
20-39	30 (26.8)
40-64	43 (38.4)
>65	16 (14.3)
Sex
Female	86 (76.8)
Male	26 (23.2)
Race/ethnicity
Non-Hispanic White	23 (20.5)
Non-Hispanic Black	79 (70.5)
Other	10 (9.0)
Monthly subscription cost
≤$50	15 (13.4)
$51-$100	4 (3.6)
$110	33 (29.5)
$125	10 (8.9)
>$125	50 (44.6)
Subscription duration (months)
0-5	27 (25.6)
6-12	26 (24.8)
13-20	26 (24.8)
≥21	26 (24.8)
Patient geographic residence
Rural	0 (0.00)
Suburban	24 (21.4)
Urban	88 (78.5)

Average Count of Chart Interactions and Appointments Made Among DPC Patients

Table 2 shows the average count of chart interactions and appointments made (including in-person and audio-visual calls) across a 12-month period. Patients under a DPC clinician had, on average, 13 chart interactions and 7 appointments annually. Patients aged 19 years or under had a yearly average of 9 chart interactions and 5 appointments; patients aged 20 to 39 years had a yearly average of 11 chart interactions and 6 appointments; patients aged 40 to 64 years had a yearly average of 15 chart interactions and 8 appointments; patients aged 64 years or older had a yearly average of 19 chart interactions and 8 appointments. Female patients had a yearly average of 11 chart interactions and 6 appointments, while male patients had a yearly average of 14 chart interactions and 7 appointments.

OPEN IN VIEWER

Table 2.

Average Count of Chart Interactions and Appointments Made among DPC Patients (n = 112).

Variable	Chart interactions	Appointments made (in-person and audio-visual calls)
Overall	13.29	6.89
Age
≤19	9.30	5.10
20-39	10.96	5.96
40-64	14.71	7.80
>65	18.75	8.38
Sex
Female	11.20	5.88
Male	13.95	7.20
Race/Ethnicity
Non-Hispanic White	15.23	7.76
Non-Hispanic Black	13.00	6.76
Other	11.00	5.75
Monthly subscription cost
≤$50	10.47	5.60
$51-$100	7.50	5.00
$110	15.88	8.19
$125	20.50	7.60
>$125	11.11	6.33
Subscription duration (months)
0-5	4.89	2.78
6-12	12.42	6.81
13-20	16.23	8.42
≥21	19.96	9.69
Patient geographic residence
Suburban	13.99	7.17
Urban	10.83	5.87

Patients identifying as Non-Hispanic White had a yearly average of 15 chart interactions and 8 appointments; those identifying as Non-Hispanic Black had a yearly average of 13 chart interactions and 7 appointments; those identifying as other had a yearly average of 11 chart interactions and 6 appointments. Regarding monthly subscription costs, patients paying $50 or less had a yearly average of 11 chart interactions and 6 appointments; those paying between $51 and $100 had a yearly average of 8 chart interactions and 5 appointments; those paying $110 had a yearly average of 16 chart interactions and 8 appointments; those paying $125 had a yearly average of 21 chart interactions and 8 appointments; and those paying more than $125 had a yearly average of 11 chart interactions and 6 appointments. For DPC subscription length, patients with subscriptions to a DPC clinician for 5 months or less, had a yearly average of 5 chart interactions and 3 appointments; patients with subscriptions between 6 and 12 months had a yearly average of 12 chart interactions and 7 appointments; patients with subscriptions between 13 and 20 months had a yearly average of 16 chart interactions and 8 appointments; patients with subscriptions for 21 months or longer had a yearly average of 20 chart interactions and 10 appointments. With regards to patient geographic residence, patients living in suburban areas had a yearly average of 14 chart interactions and 7 appointments, while patients living in urban areas had a yearly average of 11 chart interactions and 6 appointments.

Multivariate Analysis

Table 3 shows the regression results of the multivariate analysis conducted for chart interactions, and appointments made across a 12-month period, respectively. Compared to those enrolled for 0 to 5 months, those enrolled for longer periods had significantly higher chart interactions (6-12 months β = 7.41, P < .001; 13-20 months β = 94 8.35, P < .001; ≥21 months β = 12.70, P < .001). There was also a similar dose-response increase for appointments made (6-12 months β = 3.93, P < .001; 13-20 months β = 5.07, P < .001; ≥21 months β = 6.27, P < .001). No other patient characteristics were found to have a significant influence on DPC utilization.

OPEN IN VIEWER

Table 3.

Regression Model of Yearly Chart Interactions and Appointments Made among DPC Patients (n= 112).

Variable	Chart interactions					Appointments made (in-person and audio-visual calls)
	Coef.	95% CI			P	Coef.	95% CI			P
Age
≤19	Ref					Ref
20-39	2.04	−4.28	—	8.36	0.52	1.27	−1.91	—	4.45	0.43
40-64	4.25	−1.58	—	10.09	0.15	1.82	−1.11	—	4.77	0.22
≥65	4.26	−2.19	—	10.70	0.19	1.79	−1.45	—	5.04	0.28
Sex
Female	Ref.					Ref.
Male	2.24	−0.65	—	5.13	0.13	0.03	−1.43	—	1.48	0.97
Race/ethnicity
Non-Hispanic White	Ref.					Ref.
Non-Hispanic Black	−0.72	−3.75	—	2.31	0.64	−0.33	−1.85	—	1.20	0.67
Other	−1.89	−7.02	—	3.24	0.47	−0.60	−3.18	—	1.98	0.65
Monthly subscription cost
≤$50	Ref.					Ref.
$51-$100	−2.32	−13.37	—	8.72	0.68	0.63	−4.93	—	6.18	0.82
$110	1.56	−5.40	—	8.51	0.66	0.57	−2.93	—	4.07	0.75
$125	0.27	−7.68	—	8.22	0.95	0.69	−3.31	—	4.69	0.73
>$125	−0.38	−6.83	—	6.07	0.91	0.55	−2.70	—	3.80	0.74
Subscription duration (months)
0-5	Ref.					Ref.
6-12	7.41	3.79	—	11.04	<0.001	3.93	2.10	—	5.75	<0.001
13-20	8.35	4.50	—	12.20	<0.001	5.07	3.13	—	7.01	<0.001
≥21	12.70	8.82	—	16.58	<0.001	6.27	4.31	—	8.22	<0.001
Patient geographic residence
Suburban	Ref.					Ref.
Urban	−2.93	−5.98	—	0.12	0.06	−1.14	−2.68	—	0.40	0.14

Discussion

In this study examining chart interactions and appointments made by DPC patients, our results suggest this novel setting to be characterized by high utilization and consistent use across patient characteristics. Importantly, our findings demonstrate high utilization compared to traditional FFS clinics reported in the literature. For example, while the average annual visit rate was 1.5 visits per person in 2015, ¹³ and varied between 0.7 and 2.5 telemedicine visits in Medicare settings, ¹⁴ DPC patients exhibited an average of 6.89 in-person or audio-visual appointments per year. While limited research has examined patient interactions in DPC settings, these findings suggest that DPCs may facilitate more patient interactions. We posit that the subscription-based nature of the DPC model allows patients to seek care for even minor health concerns, enabling proactive health management. This high utilization reflects DPC’s ease of care continuity, a principle that has been linked to higher satisfaction rates, cost-effective care, and better healthcare outcomes. ¹⁵

The finding of a dose-response increase in DPC utilization as subscription duration increased is note-worthy. Compared to the current literature, which has long reported disparities in chart interactions, telemedicine, and in-person appointments by age, sex, and race/ethnicity,^16,17 our results found none of these demographic characteristics to be associated with DPC utilization. These results potentially tie the DPC model with equity, potentially through building and maintaining the patient-clinician relationship ¹⁸ although research is needed in this area, considering our small sample size.

This study has several limitations that should be considered when interpreting the findings. First, the data are derived from a single DPC organization, which may limit generalizability to other DPC practices with varying patient demographics, geographic locations, or operational models. The single-site design and lack of a comparator group (i.e. non-DPC practice) makes it difficult to fully isolate the effects of the DPC model and may restrict generalizability. Second, we relied on retrospective data, which inherently risks missing contextual information about patient behaviors or external factors influencing utilization. Additionally, the observational nature of this research limits our ability to establish causation between the subscription-based DPC model and the observed high utilization. Future studies should address these limitations by incorporating prospective designs, diverse DPC settings, and direct patient outcome comparisons to explore the impact of this innovative care model further.

In conclusion, this study highlights the distinct advantages of the DPC model in fostering high utilization of primary care. Notably, the absence of disparities in utilization across age, sex, and race/ethnicity potentially positions DPC as a model that advances health equity; however, future studies are needed to examine this relationship. Furthermore, the dose-response relationship between subscription duration and utilization emphasizes the importance of trust and ongoing communication in strengthening patient-clinician relationships. These results provide insights into the DPC model’s potential to transform healthcare delivery, emphasizing its value in creating a more accessible and patient-centered care environment.