EHR adopters vs. non-adopters: Impacts of, barriers to, and federal initiatives for EHR adoption

Abstract

While adoption of electronic health record (EHR) systems has grown rapidly, little is known about physicians' perspectives on its adoption and use. Nationally representative survey data from 2011 are used to compare the perspectives of physicians who have adopted EHRs with those that have yet to do so across three key areas: the impact of EHRs on clinical care, practice efficiency and operations; barriers to EHR adoption; and factors that influence physicians to adopt EHRs. Despite significant differences in perspectives between adopters and non-adopters, the majority of physicians perceive that EHR use yields overall clinical benefits, more efficient practices and financial benefits. Purchase cost and productivity loss are the greatest barriers to EHR adoption among both adopters and non-adopters; although non-adopters have significantly higher rates of reporting these as barriers. Financial incentives and penalties, technical assistance, and the capability for electronic health information exchange are factors with the greatest influence on EHR adoption among all physicians. However, a substantially higher proportion of non-adopters regard various national health IT policies, and in particular, financial incentives or penalties as a major influence in their decision to adopt an EHR system. Contrasting these perspectives provides a window into how national policies have shaped adoption thus far; and how these policies may shape adoption in the near future.

Keywords: Electronic health records; Health information technology; Health policy; National Ambulatory Medical Care Survey; Physician workflow.

Fig. 1

Adjusted percent of physicians’ agreement about overall EHR impacts by adoption status. Notes: All differences between adopters and non-adopters were significant (p < 0.01). Percentages were calculated with the use of multivariable logistic regression model. Variables included in the model were medical specialty (primary care vs. not primary care), age (under 50 and 50 years or older), the number of physicians in the practice (1–2, 3–10, 11 +, missing), ownership (physician/physician group owned, other/missing), region (Northeast, Midwest, South, West), and whether in a metropolitan statistical area (Yes, No). Missing was excluded (overall sample size n = 3180). Source: Authors’ analysis of the Physician Workflow study, 2011 (numbers are adjusted).

Fig. 2

Adjusted percent of physicians reporting major barriers to adopting an EHR by adoption status. Notes: All differences between adopters and non-adopters were significant (p < 0.01). Percentages were calculated with the use of multivariable logistic regression model. Variables included in the model were medical specialty (primary care vs. not primary care), age (under 50 and 50 years or older), the number of physicians in the practice (1–2, 3–10, 11 +, missing), ownership (physician/physician group owned, other/missing), region (Northeast, Midwest, South, West), and whether in a metropolitan statistical area (Yes, No). Missing was excluded (overall sample size n = 3180). Source: Authors’ analysis of the Physician Workflow study, 2011 (numbers are adjusted).

Fig. 3

Adjusted percent of physicians reporting health IT policies as a major influence on EHR adoption by adoption status. Notes: All differences between adopters and non-adopters were significant (p < 0.01). Percentages were calculated with the use of multivariable logistic regression model. Variables included in the model were medical specialty (primary care vs. not primary care), age (under 50 and 50 years or older), the number of physicians in the practice (1–2, 3–10, 11 +, missing), ownership (physician/physician group owned, other/missing), region (Northeast, Midwest, South, West), and whether in a metropolitan statistical area (Yes, No). Missing was excluded (overall sample size n = 3180). Source: Authors’ analysis of the Physician Workflow study, 2011 (numbers are adjusted).

Fig. 4

Adjusted percent of new and experienced EHR users reporting federal policies as a major influence in their decision to adopt an EHR system. ** Differences by experience with any EHR use was significant (p < 0.01) between new (1 year or less) and experienced (2 years or more) EHR users. Percentages were calculated with the use of multivariable logistic regression model on physicians who adopted an EHR system. Variables included in the model were medical specialty (primary care vs. not primary care), age (Under 50 and 50 years or older), the number of physicians in the practice (1–2, 3–10, 11 +, missing), ownership (physician/physician group owned, other, missing), region (Northeast, Midwest, South, West), and whether in a metropolitan statistical area (Yes, No). Missing was excluded (overall sample size n = 1783). Source: NAMCS Physician Workflow Survey, 2011.