Is a Team-Based Approach in Primary Care Worth It?

Author(s)
Lily Van, PharmD
Courtney Davis, PharmD, BCACP

Reviewed By
Kathleen (Katy) Pincus, PharmD, BCPS, BCACP
Debra Barnette, PharmD, BCPS, BCACP

Citation
David MJ, Alyna CT, Kevin NH, Zhonghe L, Sara SJ, Meredith RB. Association of team-based primary care with health care utilization and costs among chronically ill patients. JAMA Intern Med 2019; 179 (1): 54-61. doi:10.1001/jamainternmed.2018.5118

The Problem

 

A team-based approach to patient care is well established in acute care settings, but not as widely adopted in primary care settings.  Ambulatory care pharmacists (and others including psychologists, social workers, physical therapists, and dieticians) often strive to integrate into the team-based model to assist with managing therapeutic regimens and optimize patient care. Working within a team could have a positive impact on the efficiency of visits, quality of care, workload, job satisfaction, and patient satisfaction. Previous studies in acute-care hospital settings reveal positive outcomes, but there have been mixed results in primary care settings.^1,2,3  Will the extra time, effort, and money necessary to change to a collaborative team-based approach be worth it? Will it positively impact health care utilization, quality, and cost?

 

What’s Known

 

In 2001, the Institute of Medicine (IOM) described six aims for changing the healthcare system that many organizations now use when creating integrated team-based approaches to care.^4,5 The IOM called for comprehensive health services provided by at least two healthcare professionals working collaboratively with patients, family members, and caregivers. This approach has demonstrated reduction in mortality and hospital length of stay as well as functional improvements among older adults, but it is unclear if this is a cost-effective approach for all patients or whether specific patient populations would benefit most.^6,7

 

The Academic Innovations Collaborative (AIC) was initiated by the Harvard Medical School in 2012 and sought to transform primary care education and practice through implementing changes within academic medical centers and their associated community-based teaching sites in the Boston area.⁸ This collaborative focused on building interprofessional teams, managing populations, and empowering patients.⁹ Each practice developed aims for the site, attended thrice yearly educational sessions, implemented methods to improve the quality of care, and met periodically to share challenges and successes with similar practices. Practices were transformed from usual care to team-based care through changes in practice configuration to form one or more clinical care teams that were made up of a mix of physicians, nurse practitioners, physician assistances, medical residents, nurses, medical assistants, social workers, and non-clinical staff, with an average of 14 members per team. ¹⁰ Practices implemented daily 15-minute team huddles, increased care coordination, and employed a population management system that would allow for follow-up with patients who required screenings.⁸

 

What’s New

 

A recently published research study examined the outcomes from the AIC initiative.  A total of 18 AIC practice sites (intervention group) were identified and matched to 76 academically affiliated comparator sites (comparison group). See Table 1. Patient identification for each site was systematically conducted using the National Provider Identification (NPI) numbers associated with medical claims from the Massachusetts All Payer Claims Database (APCD) for patients younger than 65 years enrolled in either a commercial insurance plan or Medicaid. Differences in the claims data compared 1 year pre-implementation (2011-2012) to 1 year post- AIC implementation (2014 – 2015). The primary outcomes were cost and patient-level utilization rates, including outpatient visits, hospitalizations, emergency department visits, ambulatory care-sensitive hospitalizations, ambulatory care-sensitive emergency department visits (ie, events that may be potentially preventable).

 

TABLE 1

Descriptive Characteristics of the Study Sample
Variable	IPW Weighted, %
	Comparison	Intervention
Unique Patients, No.	401,573	138,113
³ 2 Comorbidities	43.8%	43.3%
Age 19-64, y	79.4%	79.8%
Medicaid	21.1%	19.7%
Mean enrollment, mo.	11.7	11.7

Patients who received care in the intervention practices were similar at baseline to those in the comparator group, with the exception of slightly more Medicaid patients received care at a comparison practice. Due to the inherent differences in patient characteristics between sites, propensity scores were estimated for assignment to an AIC practice and a logit model with binary outcome of AIC assignment was combined with covariates for sex, age group, Medicaid status, and duration of enrollment in primary plan.  This information was used to calculate inverse probability of treatment weights (IPTW). This was done to balance the selection of observable characteristics. After this inverse probability weighting, observable patient characteristics between intervention and comparison groups were considered well balanced. Patients that had two or more chronic conditions were classified as “chronically ill.”

 

TABLE 2

Difference-in-Difference Regression Results
Result	Difference-in-Difference
	Difference-in-Difference (95% CI)	Effect as of intervention baseline relative to comparison, %	p value
³ 2 Chronic Conditions
Outpatient visits	89.8	1.90	0.56
Inpatient hospitalizations	-48.4	-18.60	0.03
Ambulatory-sensitive hospitalizations	-20.1	-17.10	0.27
ED visits	-244.5	-25.20	0.008
Ambulatory-sensitive ED visit	-145.9	-36.00	0.009
< 2 Chronic Conditions
Outpatient visits	319.2	9	<0.001
Inpatient hospitalizations	25.6	36	0.003
Ambulatory-sensitive hospitalizations	8.5	50	0.02
ED visits	81.1	13	0.08
Ambulatory-sensitive ED visit	14.8	9	0.20

Among the 322,408 participants, overall, the AIC intervention group experience a 7.4% increase (p<0.001) in annual outpatient visits relative to the comparison group. However, among patients with two or more chronic conditions the AIC intervention group experienced a statistically significant 18.6% reduction in hospitalizations and a 36.7% reduction in ambulatory care-sensitive emergency department visits relative to the comparison group. In contrast, in the healthier subset who had only 1 or no chronic illness, the AIC group experienced a statistically significant increase in outpatient visits (9%), hospitalizations (36%), and ambulatory care sensitive hospitalizations (50%).  However, the authors indicate that these data were difficult to interpret because the parallel trends assumption was not met. 

 

Our Critical Appraisal

 

The large patient population followed across 6 academic medical centers with multiple payer groups increases the generalizability of the results. Each intervention site received training and ongoing support throughout the study period along with a platform that allowed for the sharing of ideas, which may have improved consistency across sites. Moreover, the data captured through the APCD included all medical service costs, regardless of where care was provided.  Thus, this database gives a robust picture of cost-related outcomes.

 

There are several notable limitations.  First, drug costs were excluded.  Second, patients over 65 years and those with Medicare coverage were excluded, including patients who were dually eligible for Medicare and Medicaid.  Many patient-specific characteristics which may have a direct bearing on the results were not included in the analysis. Additionally, the quality of patient care was not assessed – the analysis was based solely on claims data. Since data was collected based on the quantity rather than the quality of care, it is not possible to determine how a team-based approach led to reductions in hospitalizations and ED visits.

 

While descriptions of the AIC practices has been described in previous publications, it would be helpful to have more details regarding care coordination and team dynamics within these sites so others can potentially duplicate this model. There was no mention of clinical pharmacists as a team member, which seems surprising given these were practices affiliated with academic medical centers. Lastly, the composition of the interprofessional teams were different and evolved over time, which could be seen as a potential limitation.

 

The Bottom Line

 

The concept of working collaboratively within a team has been promoted in authoritative consensus statements and adopted in many healthcare organizations. This study found that team-based care in primary care settings may be most beneficial in patients with two or more chronic conditions and may lead to fewer hospitalizations and emergency department visits. However, in healthier patients, a team-based approach might actually lead to increased utilization, including acute care visits and hospitalizations. While improvements in health care utilization among those with chronic illnesses was observed relatively quickly, within one year of implementation, the long-term impact remains unknown. We need to look at the cost-effectiveness of this model on a larger scale, in a variety of organizations and patient populations.  While the results are promising, this study still leaves us wondering about the impact of team-based care on the quality of care, patient satisfaction, and provider satisfaction – none of which were measured in this study. While this study isn’t about pharmacists participating on primary care teams, it suggests that a team-based model, which certainly could and perhaps should include pharmacists, is worth it in patients under 65 who have multiple chronic diseases.

 

The Key Points

• Implementation of a team-based approach for primary care in outpatient clinics affiliated with academic medical centers increased office visit utilization.
• A subset of patients with two or more chronic diseases had significantly fewer hospitalizations, emergency department visits, and ambulatory care-sensitive emergency department visits when a team-based care approach was used.
• Further studies that assess both quality and satisfaction are needed to determine the true impact of team-based care models in primary care settings.

NOTE:  This program will be available for recertification credit through the American Pharmacists Association (APhA) Ambulatory Care Review and Recertification Program.  To learn more, visit https://www.pharmacist.com/ambulatory-care-review-and-recertification-activities.