Michael Silver, MS

The proportion of US doctors using electronic health records (EHRs) has risen sharply in response to the federal Meaningful Use (MU) program, which incentivizes EHR adoption. To track consumer perceptions of EHRs during this period, we conducted a national telephone survey annually for 3 consecutive years, from 2011 to 2013, corresponding with the early years of MU.

The survey used random digit-dial sampling on a dual frame of landline and cell phone numbers in the continental United States, but was not otherwise stratified by geographic region, race, or other variables. Because our primary goal was to identify relationships between variables and EHR attitudes, we constructed post hoc survey weights to align all 3 samples with each other. Relationships between sociodemographics and EHR questions were assessed with logistic regression models using the survey weights. Cross-year comparisons were conducted with χ2 tests and Cochran-Armitage tests for linear trend.

Between 2011 and 2013, the proportion of respondents with a doctor who used an EHR rose from 64% to 71%. In 2011, 64% endorsed the belief that EHRs would improve healthcare quality, dropping to 62% by 2013. Simultaneously, the proportion concerned about the effects of EHRs on privacy dropped from 48% to 41%. Consumers whose doctors used EHRs were generally more likely to believe EHRs would improve healthcare quality and less concerned about privacy risks than those whose doctors did not use EHRs. 

Overall, we conclude that during the early years of the MU program, exposure to EHRs increased while confidence in the benefits of EHRs and concerns about privacy risks became less marked. The subset of people exposed to EHRs via their physicians continued to have more positive attitudes toward them than those without that exposure. These attitudinal trends may be linked to increased familiarity with health information technology.

		Between 2011 and 2013—the early years of the Meaningful Use electronic health record (EHR) incentive program&mdash;surveys found a drop in the number of consumers who believed that EHRs would improve healthcare quality, as well as in the number who expressed concern about the effects of EHRs on privacy. 

		Consumers whose doctors used EHRs were generally more likely to believe EHRs would improve healthcare and less concerned about privacy risks than those whose doctors did not use EHRs.

Since the passage of the Health Information Technology for Economic and Clinical Health Act (HITECH) in 2009, the federal government has invested billions of dollars to promote the adoption of the electronic health record (EHR) through the Meaningful Use program (MU).

 The long-term goal of this program is to improve the quality and safety of healthcare.

 Since the launch of MU, the prevalence of EHRs has increased dramatically, with about 72% of ambulatory physicians and 44% of acute care hospitals using some form of EHR as of 2012.

National and regional surveys conducted before and in the early years of MU generally suggest that consumers believe EHRs are likely to improve healthcare.

 We recently found that most consumers think that EHRs will improve healthcare, and positive opinions were even more common among those who said their doctors use EHRs.

 However, a relatively large percentage of respondents (18%) did not know whether their physicians used EHRs.

Despite this overall positive perception of EHRs, consumers express strong concerns about the security of their electronic medical data.

 Privacy concerns are generally associated with lower confidence that EHRs will improve healthcare

 and with an increased level of concern about personal medical data breaches.

 Agaku and colleagues recently reported that about 12% of patients have withheld health information from doctors out of concern about the privacy of electronic data.

 Privacy concerns are also common with regard to health information technologies (ITs) that support the electronic exchange of medical records between healthcare providers (ie, health information exchange).

 Our objective was to track national consumer perceptions of quality effects and privacy risks associated with EHRs over a 3-year period (2011-2013), corresponding with the early years of MU.

The Cornell National Social Survey is an annual random digit-dial telephone survey conducted by the Cornell Survey Research Institute. Every year, the sample size of 1000 provides a margin of error of ± 3.1 percentage points. The Cornell University Institutional Review Board approved the study and respondents provided oral consent.

Each year’s sample is a random digit-dial sampling conducted on a dual frame of landline and cell phone numbers in the continental United States, but not otherwise stratified by geographic region, race, or other variables. The proportion of cell phone numbers is calculated from county-level data on prevalence of cell phone-only households. Listed and unlisted numbers are included in the sample list, but known business, disconnected, and nonhousehold numbers are excluded. When the telephone is answered, the adult with the most recent birthday is interviewed—a technique that ensures each adult has an equal chance of selection.

Researchers may submit potential questions for inclusion in the survey, which are competitively reviewed by the Cornell Survey Research Institute. Three questions about EHRs were included for 3 consecutive years: 2011, 2012, and 2013. The questions, adapted from previous surveys

1. An electronic medical record is a computer-based version of the patient’s medical record. Does your doctor use an electronic medical record for you? (Answer options: yes/no/not sure/have no doctor.)

2. If doctors used electronic medical records instead of paper records, how do you think that would affect the quality of medical care? (Answers were based on a 5-point Likert scale from “greatly improve it” to “greatly worsen it”; midpoint at “no effect.”)

3. If doctors used electronic medical records instead of paper records, how do you think that would affect the privacy and security of medical information? (Answers were based on a 5-point Likert scale from “greatly improve it” to “greatly worsen it”; midpoint at “no effect.”)

No additional explanation was provided for the questions.

For the definition of an EHR, we deliberately used the label “electronic medical record” and wording similarly used by the California Healthcare Foundation so that our findings would be comparable with their 2010 survey, which reported that about 26% of consumers had physicians using an EHR.

The survey was conducted in English only, and data collection closed in November or December of each year. Survey administration took approximately 20 to 25 minutes.

Exploratory comparisons of our samples with national census data suggested that respondents were representative in terms of gender distribution, age, and employment status, but were somewhat more educated, affluent, white, and non-Hispanic, most likely due to differential response rates in different populations. Because our primary goal was not to estimate strength of support for EHRs nationally, but rather to identify relationships between variables and EHR attitudes, we opted not to construct post hoc survey weights to bring the sample closer to national distributions on all demographics. Instead, we constructed post hoc survey weights to align all 3 samples with each other. This ensures that year-to-year differences cannot be attributed to random variation in demographics from sample to sample. The normalized survey weights corrected for slight differences in the distribution of education and income from year to year. Two individuals were excluded for missing demographics, leaving a total sample size of 2998 over the 3 years. We report the weighted frequencies and percents. Relationships between sociodemographics and EHR questions were assessed with logistic regression models using the survey weights. Cross-year comparisons were conducted with χ

 tests and Cochran-Armitage tests for linear trend. Analyses were conducted in SAS version 9.3 (SAS Institute, Cary, North Carolina).

Each year, 67.1% to 71.4% of eligible respondents reached by telephone consented to participate (the “cooperation rate” according to American Association of Public Opinion Research definitions

). The samples were representative of the US population in terms of age, sex, and employment, but whites, non-Hispanics, and well-educated respondents were somewhat overrepresented (

The proportion reporting a doctor using an EHR rose over time (63.9% in 2011 to 71.1% in 2013; 

). The proportion endorsing the belief that EHRs would improve healthcare quality decreased modestly from 64% in 2011 to 61.5% in 2013, while the percentage of those who believed EHRs would worsen healthcare quality rose from 7% to 8.7% (

). The proportion believing EHRs would have no effect on quality rose from 25.5% to 29.6% over the same time period (test for linear trend, 

 = .009). The proportion expressing concerns that EHRs would worsen privacy and security decreased from 47.5% in 2011 to 41.2% in 2013 (test for linear trend, 

). More than one-fifth of respondents believed that EHRs would actually improve privacy and security.

In every year, the belief that EHRs would improve healthcare quality was more common among wealthier individuals, those who were employed, and those who used the Internet daily (

). In addition, this belief was significantly less common among individuals concerned that EHRs would worsen privacy and security. Respondents who reported having a doctor that uses an EHR were more likely to believe EHRs would improve healthcare quality, although the effect narrowly missed statistical significance in 2013. In the first 2 years of the survey, college graduates were less concerned about privacy and security, but in the final year they became more likely than other respondents to express concerns (

In addition to college graduates, concerns about the effect of EHRs on privacy and security were consistently less common among individuals whose physician used an EHR. However, this concern was significantly more common among individuals who identified themselves as politically conservative.

During a 3-year period in which Americans became more likely to report seeing doctors who use EHRs, there were modest but statistically significant decreases in both optimism about EHR effects on quality and concerns about EHR-related privacy and security risks. Familiarity with a doctor using an EHR was generally positively associated with the belief that EHRs would improve healthcare quality as well as with less concern about EHR risks to privacy. Our 3-year trend results show an association between EHR exposure and perception of quality benefits that is consistent with prior national surveys over shorter timelines.

As national discourse about the positive and negative aspects of EHRs continues, consumers may be forming their opinions on the basis of a variety of sources of information. Both positive and negative perspectives have been publicized in the media; in addition, physicians may talk to their patients about their enthusiasm for or frustration with EHRs. Finally, more consumers have had a personal experience with a doctor using an EHR—some of whom may have felt that the computer made the visit more impersonal (as shown in a Veterans Administration patient survey

), whereas others may have noticed improvements in physician communication skills (demonstrated in a small study of young physicians

Despite significant overall decreases in concerns about privacy and security, the perceived risks in these areas were negatively associated with the perceived benefits of EHRs. Thus, consumers who were more concerned with EHR-related privacy and security risks were less likely to perceive EHRs as likely to benefit quality of care. It is logically possible for quality benefits to be independent of privacy risks—for example, an EHR could support more personalized care, and yet it could also be hacked&mdash;but it is well-established that perceived benefits of almost any activity tend to decrease as its perceived risks grow.

 As a result, it is likely that privacy concerns may be one of the causes of lower perceived benefits and, therefore, explicitly addressing consumer privacy concerns would lead to a more positive view of EHRs overall.

Awareness of medical data breaches has likely risen since 2009 when federal regulations first mandated that large breaches be reported to news media and to the HHS.

 The number of data breaches publicly reported recently reached 1000, affecting nearly 32 million Americans.

 Contact from healthcare organizations or news stories about these breaches (eg, the 

) could be raising public awareness of threats to privacy and security. It is also noteworthy that Edward Snowden’s revelations about National Security Agency (NSA) surveillance of electronic communication

 occurred in June 2013, several months before the third year of data collection. It is possible that prominent news stories such as these could be linked to the change we observed in the opinions of college graduates, who were less likely to be concerned about EHR privacy threats in 2011 and 2012, but became significantly more likely to be concerned the following year. Additionally, a 2014 Harris poll has linked increased caution online to awareness of NSA digital surveillance.

Nevertheless, in spite of these privacy concerns by many consumers, a substantial minority of respondents (up to 21% in our data) perceived health IT as a way to improve privacy and security.

 These consumers may believe electronic protections to be stronger than the protections for paper records or they may have had negative experiences with paper records.

Random digit-dial surveys have limitations, including sampling coverage and nonresponse bias.

 One advantage of dual-frame methodology is the inclusion of cell phone households, which is particularly more inclusive of younger demographics than a landline-only survey.

 The sampling strategy was random, so we did not apply survey weights in order to match demographic characteristics of the country. As a result, although our sample was fairly representative in gender, age, and employment, our respondents were somewhat more educated and affluent, more likely to be white, and less likely to be Hispanic than the national population—most likely because of differential response rates. The questions in the survey have not been validated. The use of the term “electronic medical record” in the survey does not reflect the national consensus that “electronic health record” is the most accurate term for describing contemporary IT with certain capacities for interoperability. On the other hand, it is not clear that the general public makes a distinction between these terms, and we sought to minimize any confusion by providing a brief definition of the term&mdash;“an electronic medical record is a computer-based version of the patient’s medical record”&mdash;in the question stem. Lastly, the study design was cross-sectional, so association does not imply causation.

During the launch and early years of the MU program, Americans became more likely to report having a physician who used an EHR. Individuals whose doctors used EHRs tended to have more positive opinions of EHRs. Nevertheless, over the 3-year study time frame, both confidence in the benefits of EHRs and concerns about privacy risks became less marked overall, a pair of trends that might be attributed to increased familiarity with health IT. Opinions on electronic privacy threats remain split, with a plurality expressing privacy concerns about EHRs —a concern generally linked to lower perceived benefit&mdash;along with a substantial minority believing that EHRs will improve medical privacy and security.

 Department of Healthcare Policy and Research (JSA), Weill Cornell Medical College (JR, MS, RK), New York, NY; Stanford School of Medicine (SB), Palo Alto, CA.

 The survey was funded by the Cornell University Office of the Provost. Dr Ancker has received funding from an AHRQ grant (K01HS021531).

 The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

 Concept and design (JSA, RK, SB); acquisition of data (JSA); analysis and interpretation of data (JSA, RK, SB, JER, MS); drafting of the manuscript (JSA, RK, SB, JER, MS); critical revision of the manuscript for important intellectual content (JSA, RK, SB, JER); statistical analysis (JSA, SB, MS); and supervision (JSA).

 Jessica S Ancker, MPH, PhD, 425 E. 61st St, Ste 301, New York, NY 10065. E-mail: jsa7002@med.cornell.edu.

Articles

During the early years of the "meaningful use" program, surveys found decreases in both optimism and concerns about electronic health records.

Trends in Public Perceptions of Electronic Health Records During Early Years of Meaningful Use

Federal policies are incentivizing hospitals and providers to adopt and meaningfully use electronic health records (EHRs). Nursing homes are not eligible for incentives. However, understanding health information technology (HIT) adoption among nursing homes will be critical to developing HIT policies for this sector. Our objective was to assess the pace of EHR adoption, changes in computerized function adoption, and participation in health information exchange by New York state nursing homes over time.

We used a repeated, cross-sectional study design.

We surveyed all New York state nursing homes between February and May 2013, comparing results to the same survey administered in 2012.

We received responses from 472 of 630 nursing homes (74.9%). Rates of EHR adoption increased from 48.6% to 56.3% (

= .03). Participation in health information exchange remained unchanged (54.5% to 55.3%, 

= .8). The top barriers to EHR adoption cited were: a) the initial cost of HIT investment (67.9%, n = 133), b) lack of technical IT staff (46.4%, n = 91), and c) lack of fiscal incentives (45.8%, n = 88). Comparing nursing homes with EHRs in 2012 to nursing homes with EHRs in 2013, the availability of many types of computerized functionalities significantly increased, although no gains were seen for order entry or clinical tools.

While some gains are being made by nursing homes, HIT adoption generally lags behind that of other sectors. Public policy focusing on building HIT infrastructure is essential to ensure that nursing homes keep up with other healthcare segments.

Am J Manag Care. 2014;20(11 Spec No. 17):eSP53-eSP59

Although federal policies are incentivizing health information technology (HIT) adoption by hospitals and providers, nursing homes are excluded. Few large-scale or longitudinal assessments of HIT adoption in this sector have been published. We assessed electronic health record (EHR) adoption and health information exchange participation among New York state nursing homes over time.

		There was a 7.7 percentage point increase in rates of EHR adoption between 2012 and 2013.

		Rates of health information exchange remained stagnant.

		HIT engagement in this sector lags behind other healthcare sectors.

		Understanding HIT adoption in this sector is critical to guide policy development.

Recent federal policy has focused upon building the nation’s health information technology (HIT) infrastructure. Through the Electronic Health Record (EHR) Incentive Program, payments are available to hospitals and providers who meaningfully use EHRs.

 This program, with total incentive payments estimated between $14 billion and $27 billion, has resulted in tremendous increases in adoption in both sectors.

Unlike hospitals and providers, the long-term care sector has been excluded from federal incentive programs. While only a few large-scale studies have assessed rates of EHR adoption among nursing homes—and reported rates vary widely&mdash;it is believed this sector lags behind.

 The barriers to adoption cited include the high cost of purchasing and maintaining EHRs, implementation and training challenges, and lack of evidence for return on investment.

However, for those facilities that invest in EHRs, there are many reported benefits, including improved information access, more accurate documentation, increased adherence to evidence-based guidelines, and improved employee satisfaction and retention.

Currently, there are more than 1.5 million residents in long-term care facilities—a number expected to increase as the elderly population grows.

These patients are medically complex, have high medical costs, and are frequently transferred to acute care hospitals. Indeed, a 2011 report found that nursing homes transferred 25% of their Medicare residents annually to hospitals at a cost of $14.3 billion.

As a result, this sector arguably has an urgent need to keep pace with the HIT adoption occurring in other sectors—a sentiment underscored by the Office of the National Coordinator for Health Information Technology.

In 2012, we surveyed all New York state (NYS) nursing homes to assess their level of EHR adoption and health information exchange (HIE) participation.

NYS leads the nation in state-based HIT investments, primarily through the Health Care Efficiency and Affordability Law for New Yorkers (HEAL NY) Capital Grants Program.

While this program does not specifically provide funding to nursing homes, a primary focus is community-wide HIT investment.

The objective of this follow-up survey was to assess how rates of EHR adoption and HIE participation have changed over time. We also sought to identify characteristics associated with adoption during this time period, changes in computerized functions available, and barriers to EHR adoption. To our knowledge, this is the first such large-scale, repeated cross-sectional assessment. Given the rapidly changing HIT landscape, closely monitoring the progress of nursing homes can provide valuable data to inform HIT policies.

This study was conducted as part of the HEAL NY program evaluation process, led by researchers from HEAL NY’s designated evaluation entity, the Health Information Technology Evaluation Collaborative. Our survey instrument was a novel questionnaire developed with guidance from experts in HIT and HIE, as well as leaders from 3 key NYS long-term care associations: Continuing Care Leadership Coalition, LeadingAge New York, and New York State Health Facilities Association.

Survey questions remained consistent between 2012 and 2013, allowing for longitudinal comparison. We obtained Institutional Review Board approval from Weill Cornell Medical College.

Our survey consisted of 10 questions and assessed EHR implementation, automation of 23 key functionalities, HIE participation, and barriers to EHR implementation.

We surveyed all 630 NYS nursing homes between February and May 2013. Surveys were e-mailed to each nursing home administrator by Cornell Survey Research Institute, whose expertise is providing survey research support services. The 3 long-term care associations alerted administrators via electronic newsletters.

We collected information about nursing home location, bed size, hospital system membership, ownership, and status as a continuing care retirement community from the CMS Nursing Home Compare database.

Continuing care retirement communities are generally privately owned and combine independent living, assisted living, and nursing home facilities. We speculated that their rates of EHR adoption and HIE participation may be higher than that of other nursing homes in order to better standardize care and facilitate exchange of information between facilities.

To categorize location, we used the catchment areas of the NYS Department of Health regional long-term care offices (Capital District, Central, Western, Hudson Valley, New York City, and Long Island).

We divided bed size into 4 categories, with each representing approximately one-fourth of total nursing homes: less than 100, 100 to 159, 160 to 239, and 240 or more. We categorized ownership into privately owned for-profit, privately owned nonprofit, and publicly owned. Hospital affiliation, chain ownership, and status as a continuing care retirement community were dichotomized into yes/no variables.

We compared respondent and nonrespondent characteristics using Pearson χ2 tests or Fisher’s exact tests. We evaluated whether an EHR was implemented either fully or partially (counting those facilities as adopters) and whether the facility participated in HIE. We used descriptive statistics to assess the availability of computerized functionalities among nursing homes with EHRs, comparing 2012 and 2013 results. For facilities participating in HIE, we assessed the type and directionality of data exchanged. We compared results between 2012 and 2013 using 2 sample tests of proportions.

We also analyzed the relationships between EHR adoption and HIE participation with key nursing home characteristics (eg, location, bed size, membership in a hospital system, ownership, and status as a continuing care retirement community), using logistic regression to isolate the predictive effects of each characteristic. We examined barriers to EHR adoption among nursing homes without EHRs, using χ2 or Fisher’s exact tests.

Lastly, we compared nursing homes that adopted EHRs between 2012 and 2013 to those that did not, and nursing homes that began participating in HIE between 2012 and 2013 to those that did not, in order to identify characteristics associated with HIT acquisition. We again used χ2 or Fisher’s exact tests to describe differences. We used SAS 9.3 (Cary, North Carolina) for all analyses.

We received responses from 472 of 630 nursing homes in 2013 (74.9%) (

). There were no significant differences between respondents and nonrespondents.

There was a 7.7 percentage point increase in adoption between 2012 and 2013, from 48.6% to 56.3% (n = 264) (

= .03). In 2013, among the 119 nursing homes planning to implement EHRs, 22.7% (n = 27) planned to implement within 12 months, 37.8% (n = 45) within 13 to 24 months, 20.2% (n = 24) in 2 years or more, and 19.3% (n = 23) were unsure of the timeline. Only 11.7% (n = 55) reported no EHR implementation plans, similar to 2012 (11.4%).

As of 2013, among EHR adopters, the functionalities most likely to be all electronic were minimum data set reporting, financial management, and patient demographics (

). Computerized provider order entry was available in 56.9% (n = 148) of facilities, and clinical notes were available in 51.4% (n = 133). Clinical decision support was only available in 8.2% (n = 21). Between 2012 and 2013, the availability of many types of computerized functionalities significantly increased among nursing homes with EHRs, particularly with regard to documentation and results viewing (Table 2). However, there was no significant change in the availability of medication or other type of order entry or clinical tool functions.

As of 2013, 50.5% (n = 233) of nursing homes exchanged key clinical information electronically within their care system and 27.2% (n = 125) exchanged information electronically outside their care system. Overall, the rate of HIE in 2013 was 55.3% (n = 256). There was no significant change in rates of HIE from 2012 to 2013 (54.5% to 55.3%, 

= .8). Similar to 2012, the most common data exchange partners were pharmacies, hospitals, and laboratories (

). Comparing facilities engaged in HIE in 2012 and 2013, there were no significant differences with regard to data exchange partners or the level of bidirectionality of data exchange.

Characteristics Associated With EHR Adoption and Participation in HIE

Only being part of a nursing home chain was significantly associated with having an EHR (odds ratio [OR] = 1.9; 95% CI, 1.0-3.4) (

). Being affiliated with a hospital (OR = 2.4; 95% CI, 1.2-4.8), having an EHR (OR = 2.2; 95% CI, 1.5-3.2), and being a private nonprofit nursing home versus a for-profit nursing home (OR = 1.8; 95% CI, 1.3-2.7) were all significantly associated with HIE participation.

We also looked specifically at the 141 nursing homes that did not have EHRs in 2012 to identify any characteristics that distinguished the 2013 adopters (n = 45) from those who remained nonadopters (n = 96). We found no difference in facility characteristics between these 2 groups; however, among recent adopters we saw a significant increase in the use of many available computerized functionalities, a change we did not see among nonadopters. This was true for 16 of 23 functionalities, including 4 of 6 administrative functionalities, all documentation functionalities, all order entry functionalities, and consult viewing.

Similarly, we looked specifically at the 129 nursing homes that did not participate in HIE in 2012. We compared characteristics of those that participated in HIE in 2013 (n = 60) to nonparticipants both years (n = 69). We found that more of the facilities that newly engaged in HIE were medium-size (

 = .02). In addition, there was significantly greater adoption in those facilities of several computerized functionalities—specifically medical history documentation (

= .04), and electronic laboratory results viewing (

The top barriers to EHR adoption among nursing homes without an EHR were stated to be: a) initial cost of HIT investment (67.9%, n = 133), b) lack of technical IT staff (46.4%, n = 91), and c) lack of fiscal incentives (45.8%, n = 88). Significantly more nursing homes without an EHR identified these as major barriers compared with those that had an EHR (

To our knowledge, this study represents the only large-scale study to track EHR adoption and HIE participation by nursing homes over time. Our results show that there was a 7.7 percentage point increase in EHR adoption among NYS nursing homes between 2012 and 2013, and rates of HIE participation remained stagnant. These results provide important information about the pace of HIT adoption in nursing homes that can help guide policy discussions. Our results suggest that nursing homes are not keeping pace with the achievements in HIT acquisition seen among office-based providers and hospitals. National data from the CDC show that adoption of any EHR system (defined similarly to our study) by office-based physicians increased from 71.8% to 78.4% between 2012 and 2013.20 While this is also a 6.6 percentage point change, rates of overall EHR adoption by providers are much higher. Between 2011 and 2012, the adoption of EHR grew from 51.7% to 71.8%.2 Prior to the start of incentive payments, rates of adoption were increasing only 4% per year.

Among hospitals nationally, rates of adoption of at least a basic EHR system have nearly tripled since 2010, increasing by 15 percentage points from 2012 to 2013.

While this comparison differs slightly in that we assessed adoption of any EHR system, not utilizing the stricter criteria for a basic EHR system as defined by the Office of National Coordinator for Health Information Technology in the above report, it is worth noting that 87% of all hospitals in 2013 reported receiving at least 1 meaningful use incentive payment (and thus, by definition, must have an EHR).

 Prior to 2010, rates of EHR adoption were increasing only 3% per year among US hospitals. Also of note, in NYS, we did a similar cross-sectional survey of hospitals assessing adoption of any EHR and participation in HIE, and found that as of 2012, 97% of hospitals had adopted an EHR and 79% were engaging in HIE with other partners (manuscript in press).

Among nursing homes with EHRs, the proportion of computerized functionalities available significantly increased for many functions between 2012 and 2013. Unfortunately, 2 areas where this did not occur were order entry and clinical tools. These areas are generally more difficult to implement; however, their use is likely necessary to achieve maximal quality and safety benefits. Such gains have already been seen in the nursing home setting. For example, several dozen nursing homes in California reduced their rates of pressure ulcers by 42% to 55% through use of clinical decision support embedded in EHRs.

Our data suggest that the pace of HIT adoption will be much slower in healthcare sectors not receiving financial incentives, such as nursing homes. Indeed, 2 of the top barriers to adoption identified in our survey were the initial cost of HIT investment and lack of fiscal incentives. This is supported by research in other sectors excluded from the EHR Incentive Program. For example, a recent national study examining EHR adoption among non-acute care hospitals not eligible for federal incentives found that only 6% of long-term care hospitals, 4% of rehabilitation hospitals, and 2% of psychiatric hospitals have at least a basic EHR system.

By comparison, among office-based physicians eligible for EHR incentives, top barriers to adoption often center around work flow challenges, lack of interoperability of EHR systems, and the costs of purchasing and maintaining EHRs.

In contrast to EHR adoption, there was no change in HIE participation between 2012 and 2013. Among the nursing homes we surveyed, 50.5% engaged in HIE with providers within their care system and 27.2% engaged in HIE with providers outside their care system. By comparison, among hospitals nationally as of 2012, 65% were participating in HIE with hospitals inside their organization and 58% were exchanging data with providers outside their organization.

Similar to hospitals, having an EHR system significantly increases the likelihood of HIE participation in our sample of nursing homes.

Meaningful use of Stage 2 criteria requires that eligible hospitals who transition patients to another care setting provide a care summary, electronically or via exchange, for a percentage of patients. Perhaps this is helping to facilitate some HIE between nursing homes and hospitals.

 However, there are many challenges to HIE beyond lack of technology, including required collaboration between competitors and lack of sustainable business models.

These may explain why EHR adoption is increasing, even if slowly, while HIE rates remain unchanged.

This study examines adoption of electronic health records and participation in health information exchange by New York state nursing homes over time.

Michael Silver, MS

Articles

Trends in Public Perceptions of Electronic Health Records During Early Years of Meaningful Use

Trending Health Information Technology Adoption Among New York Nursing Homes