Document Type


Degree Name



Dept. of Public Health & Preventive Medicine


Oregon Health & Science University


Context: Between January 2005 and December 2007 the unemployment rate in Clark County Washington was 6.4 ± 0.8%.1 Between January 2009 and December 2011, following the Great Recession, 13.4 ± 1.3% of residents were unemployed and over 56,000 Clark County residents (13.4 ± 0.8%) were without health insurance. 2 To address the needs of the newly unemployed, PeaceHealth Medical Center created a pilot program to provide primary care to the region’s uninsured. The goal was to improve the overall health of the population and reduce their dependence on emergency and urgent care. The program was called “Direct Primary Care” (DPC).

Objective: This study tests whether providing primary care to the uninsured using the DPC model is associated with reduced utilization of Emergency Department (ED) and Inpatient (IP) services, and to identify patient characteristics associated with these effects. Reduced rates of ED and IP services are used as a proxy to indicate improved health of the community and better management of chronic conditions.

Methodology / Study Population: This longitudinal study involves 433 standard and 221 scholarship members with no previous history of insurance, enrolled in PeaceHealth’s DPC program for at least 3 months during the study period. Predictor variables collected from patient billing records include number and type of chronic conditions and select socioeconomic factors. Hospital and clinic encounters were collected from January 1, 2010 (to generate baseline health service rates prior to DPC enrollment) through December 31, 2013 or upon membership cancellation (to generate utilization rates while enrolled in DPC). A multivariate mixed effects negative binomial regression model was used to characterize the association between patient characteristics and chronic conditions with encounter rates, comparing the pre-intervention period to the enrollment period and with each patient serving as his/her own control. The model was tested for possible effect modification by membership type.

Results: This study shows enrollment in DPC reduces members’ utilization of ED and IP services, presumably by managing their chronic conditions. Standard members are associated with a 46 % (95% CI: 13 - 66%) reduction in ED encounters and a 48 % reduction (95% CIRR: 0.25 – 1.06) in IP encounters. As expected, these reductions were accompanied by increased use of OP services with enrollment in DPC associated with and an 82 % [15 – 191%] increase in OP encounters. Scholarship members were associated with a 15 % (95% CIRR: 0.52 – 1.41) reduction in ED encounters, 68 % (95% CI: 42 – 83%) reduction in IP encounters, and these were offset by a 3.3 fold (95% CI: 1.79 – 6.11 fold) increase in OP encounters after enrolling in DPC. The length of DPC membership influences the strength of these effects. In addition to improving the health of the community, such reductions are predicted to bring significant cost savings to the PeaceHealth hospital system.




School of Medicine

Available for download on Tuesday, June 19, 2018