Objectives: Rural hospitals face significant barriers to adoption of advanced-stage electronic medical records (EMRs), which may translate to an unexplored disparity for children in rural hospitals. Our objective was to determine whether children hospitalized in rural settings are less likely to be cared for using advanced-stage EMRs.
Methods: We merged the 2009 Healthcare Cost and Utilization Project Kids Inpatient Dataset with the 2009 Healthcare Information and Management Systems Society database. Logistic regression determined the independent relationship between receiving care in a rural hospital and advanced-stage EMRs.
Results: A total of 430 055 (9.3%) of the 4 605 454 pediatric discharges were rural. Logistic regression analysis determined that even when an extensive list of various patient and hospital characteristics are accounted for, rurality continues to be a strong predictor of a child’s care without advanced-stage EMRs (odds ratio 0.3; 95% confidence interval, 0.2–0.5).
Conclusions: Children hospitalized in a rural hospital are less than half as likely to be treated using advanced-stage EMRs. A focus of government and hospital policies to expand the use of EMRs among rural hospitals may reduce this child health care disparity.
The widespread adoption of electronic medical records (EMRs) to improve the safety, efficiency, and ultimately the cost of the US health care system has been endorsed by the Institute of Medicine since 2001.1 Additionally, the federal government offers financial incentives (known as meaningful use incentives) for EMR adoption and meaningful use through the Healthcare Information Technology for Economic and Clinical Health (HITECH) Act. Approximately 60 million people live in rural communities in the United States, and research suggests patients in the rural health care environment are at risk for receiving inferior health care compared with nonrural populations in regard to access and quality.2 Rural residents often depend on small rural hospitals as their primary source of medical care. Therefore, rural hospitals are an important part of the health care system for all rural residents, including children needing hospitalization. However, rural hospitals face many significant and at times unique barriers when considering widespread adoption of newer innovations in health care such as EMRs. Rural hospitals’ barriers to EMR adoption include a lack of financial resources, unclear return on investment, larger financial impact of costly EMRs on smaller hospital budgets (economies of scale), and a paucity of information technology staff available to implement and maintain EMRs in rural areas.3,4
Potential quality benefits of EMR adoption range from basic process improvements in safety, such as improved legibility of handwritten orders, to more complex process improvements, such as efficient data transfer to ensure coordination of care across various health care settings. The latter could be beneficial to children with a medical home in the rural setting who need intermittent subspecialty care from urban tertiary centers. Early research focusing on the safety benefits of EMR adoption was performed primarily in large urban academic centers and suggested EMRs may decrease the frequency of medication errors and adverse drug events for children.5–7 More recent research suggests that the proposed benefits of EMRs for measures of efficiency and cost in children’s health care have not yet been achieved.8,9 Despite the need for more research to elucidate EMRs’ effectiveness in improving safety and efficiency specifically for children and rural hospitals, EMRs are widely accepted as having the potential to improve quality of care in all hospitals.3,10
Residents of rural communities are more likely to live in poverty, be uninsured, and report poorer health status.11,12 This disparity in health status is probably due, in part, to poorer access to various health care services in rural settings.13 Access limitations combined with poverty and lack of insurance can lead to worse health outcomes.14,15 Amid many possible strategies to improve quality of care in rural settings, one that has received recent attention is the use of EMRs.3,10
US hospitals that care for children are on the path to widespread adoption of EMRs.16 The most recent report on EMR adoption suggests that 39% of free-standing children’s hospitals and 23% of all hospitals that care for children use advanced stages of EMRs in the care of hospitalized children.17 However, rural hospitals differ from nonrural hospitals in their ability to adopt costly innovations such as EMRs.10,17,18 In a previously published study using a hospital-level analysis of the Healthcare Cost and Utilization Project Kids Inpatient Dataset (HCUP KID) 2009 and the Healthcare Information and Management Systems Society (HIMSS) 2009 database, our group reported that among hospitals that care for children that have advanced EMRs, only 20% were rural, whereas 40% were urban nonteaching and 40% were urban teaching, suggesting a hospital-level disparity.19 The potential disparity at the level of the individual child hospitalized in the rural environment as a result of a lower adoption rate of EMRs among rural hospitals remains unexplored. The purpose of this study is to determine whether children in the inpatient rural health care environment are cared for less frequently with advanced-stage EMRs and to quantify any disparity that exists.
The 2009 HCUP KID20 was used to identify and describe children discharged from hospitals across the United States. HCUP KID 2009 included weighted discharge data from 44 states. HCUP KID is released every 3 years with significant delay, and the 2009 release is the most current available version. Because the 2009 HCUP KID is the most recent available patient-level data, the 2009 HIMSS database21 was used to identify hospitals’ use of EMR components and stages. The HIMSS data have been used in previous health service research, including a previous publication by this study team.19 As previously noted by Hillestad et al,16 the HIMSS data “represent a broad canvassing of acute care hospitals, chronic care facilities, and ambulatory practices on their adoption and plans to adopt various health IT components.”
Hospitalizations of children ≤21 years of age in 2009 from HCUP KID were analyzed. Patient characteristics included age (<1 year, 1–3 years, 4–5 years, 6–10 years, 11–15 years, and 16–21 years), hospital birth (“y/n” indicates in-hospital birth), race (white, black, Hispanic, and other), and expected payer (Medicaid or Medicare, private or health maintenance organization, self-pay or other) at time of discharge. Severity of illness for a patient is categorized using the 3M All Patient Refined Diagnosis Related Groups (APR-DRGs) classification of illness severity that is present in HCUP KID.22
The HIMSS 2009 database was used to identify hospitals’ use of EMR. EMR use was classified on the hospital level according to previous studies as stage 0 (no automation/no adoption), stage 1 (automation of ancillary services: radiology information systems, laboratory information systems, pharmacy information systems, and clinical data repository), stage 2 (stage 1 plus automation of nursing workflow: nursing documentation and electronic medication administration records), or stage 3 (advanced EMR including stage 2 plus computerized physician order entry and clinical decision support).23 Hospitals were classified as advanced EMR if they reported use of all stage 3 functions. This model for classifying EMR stage has been used in previous health service research studies.19,24 Additionally, a benefit of this model is that functions required for advanced stage 3 status are also required to compete for meaningful use incentives.25 For these reasons, this 4-stage model was chosen over other models with 8 stages. HIMSS provides information about the level of implementation of EMR applications, including not automated, contracted/not yet installed, installation in progress, and live and operational. For the purpose of this study we included only applications that were reported as live and operational. Previous research suggests that nonrespondents to surveys on information technology are more similar to nonadopters than adopters; therefore, consistent with previous research, if a hospital does not report use of an EMR component as live and operational, it is assumed that the hospital is a nonuser.26
Primary Independent Variable
The primary independent variable for these analyses is a child receiving inpatient care in a rural hospital environment. Since 2004, the rural designation in HCUP KID has been determined by the Core Based Statistical Area. Hospitals residing in counties with a Core Based Statistical Area type of metropolitan area were considered urban, whereas hospitals with a Core Based Statistical Area type of micropolitan or noncore are were classified as rural.27 HCUP KID combines the rural–urban designation with teaching status. Hospitals are categorized as urban teaching, urban nonteaching, or rural. For the purposes of these analyses we considered a child’s care to be either in a rural environment or nonrural environment (urban teaching or urban nonteaching hospital).
Hospital and Environmental Characteristics
The analysis included the following hospital characteristics from the HCUP KID data: hospital bed size (small, medium, and large, based on location and teaching status) and location of hospital (Northeast, South, Midwest, and West regions).
Bivariate analysis on categorical variables was performed with χ2. Logistic regression was used to determine independent predictors of a child receiving care with advanced-stage EMR. All analyses were performed with SAS software (version 9.3; SAS Institute, Inc, Cary, NC).
A total of 430 055 (9.3%) of the 4 605 454 discharges were rural. The patients included in our analyses include a diverse range of children across the United States, as demonstrated in Table 1. Children cared for in the rural environment are less likely to be cared for by stage 3 EMR and more likely to be cared for by stage 1 or 2, as demonstrated in Fig 1. In the rural environment, the largest percentage of children are cared for with stage 1 EMR (39%), as compared with children in nonrural hospitals, who are most often cared for with stage 3 EMR (34%).
In bivariate analyses, receiving care in a rural environment was significantly associated with newborn (P < .001), expected payer (P < .001), severity of illness (P < .001), bed size (P = .001), region (P < .001), and EMR stage (P < .001). This emphasizes the importance of multivariable analysis to determine whether rurality is independently associated with EMR among hospitalized children or whether this finding is secondary to confounding factors.
To assess the influence of bed size on the relationship between rurality and EMR stage, we conducted a bivariate analysis among rural hospitals between bed size and advanced EMR. The rates of advanced-stage EMR for small, medium, and large hospital bed size were 12.5%, 10.0%, and 15.3%, respectively (P = .31).
The multivariable analysis presented in Table 2 shows that after newborn status, APR-DRG severity, bed size, region, and a potential interaction between bed size and rurality are accounted for, rurality continues to be a strong predictor of a child’s care without advanced-stage EMR (odds ratio 0.3; 95% confidence interval, 0.2–0.5).
The results of this study show that children who receive care in rural hospitals are less than half as likely to be cared for with an advanced-stage EMR as children in nonrural hospitals. This finding persists when other potential confounders are accounted for. These results should be considered in the context of other disparities children in rural areas face, such as higher rates of poverty, lack of insurance, and poorer access to care.11,12
The Institute of Medicine report Crossing the Quality Chasm suggested in 2001 that implementing information technology tools could improve quality in the US health care system.1 Subsequent policy changes occurred through HITECH (enacted in February 2009) to incentivize the use of EMR tools to improve the safety and efficiency of US health care. These incentives offer up to $2 million for hospitals to implement certified EMR tools and demonstrate improvement in outcomes, and they are scheduled to last for 5 years within Medicare. After that time penalties for not using EMR will begin for hospitals serving Medicare patients, which emphasizes the federal government’s focus on implementation of EMRs. It has been argued that the funds within Medicare are limited, and therefore financially incentivizing one institution is ultimately removing resources from other institutions.28 Despite the differences between Medicaid and Medicare policies, because rural hospitals often serve both children and adults, Medicare policies are most relevant. These policies may have a disproportionately negative effect on rural hospitals that are already struggling to reach advanced-stage EMR status.
Barriers to successful adoption of advanced-stage EMR in rural hospitals have been described.3,4,29 A survey among chief information officers from rural hospitals in Florida suggests that these people recognize the need to implement EMR tools to reduce medical errors and promote patient safety, but they noted financial barriers to successful implementation.30 Critical access hospital designation may alleviate some of the financial barriers to adoption rural hospitals face. Of the 819 rural hospitals in this HCUP KID analysis, 449 (55%) are considered critical access hospitals. However, it remains conceivable that the financial incentives available for meaningful use of EMR through HITECH may contribute to a child health disparity by channeling funds to hospitals that have more resources or already have implemented EMRs, ultimately removing resources from other institutions.28,31
Previous research has shown that the realized benefits of EMR adoption in rural hospitals may not match the anticipated benefits.3 Because rural hospitals often have less capital to invest in EMR purchase, implementation, and maintenance, it is critical that measurable and attainable benefits are demonstrated. Mills et al3 interviewed 15 leaders of rural hospitals in Iowa that had adopted EMRs. The most commonly cited motivation for adoption was to improve efficiency, access, and quality. It is also generally thought that EMR adoption is necessary to remain financially viable and competitive. With limited financial resources and without clear and immediate financial or efficiency benefits, rural hospitals will continue to face significant barriers to advanced-stage EMR adoption.
Meaningful use incentives and HITECH legislation have led to adoption rates that illustrate the existing disparities in resources at rural and nonrural hospitals. Over time, this may worsen the EMR disparity children in rural hospitals face, as illustrated by this study. In other countries that used policies to encourage adoption of EMRs, the policies were considered to have worsened the rate of adoption in certain subsets of hospitals and contributed to a digital divide.32 A recent study in the United States shows a similar phenomenon.33 The divide may be considered a natural or appropriate stage in adoption given the limited research on EMRs in the rural setting and limited funds to invest in such research. But Because the goal is widespread adoption and the federal government has begun to leverage incentives, with financial penalties over time, the disparate care noted in this article may highlight the need for research to support EMR use in the rural setting and policies to limit the financial impact on rural hospitals. One potential way for rural hospitals to offset the cost of EMR adoption is attaining critical access designation. Hospitals may be eligible for critical access designation from Medicare by meeting specific criteria, including being located in a rural area, having ≤25 inpatient beds, having 24-hour emergency care services, and being ≤35 miles away from the nearest hospital. This designation allows cost-based reimbursement. Attaining critical access designation is an example of a policy that may alleviate some financial barriers by facilitating access to financial resources for EMR adoption for the subset of hospitals that qualify.34
There are several limitations to this study. In 2009, HCUP KID represented only 44 of the 50 states and a portion of the states that report information to HCUP KID do not report information on the hospital level because of state policies. The information on discharges from these hospitals cannot be included in our analyses because we are unable to describe the use of EMRs in these hospitals. Second, HIMSS data represent only reported use of EMRs and do not measure the degree to which any hospital uses the technology effectively throughout the organization. Third, other factors may contribute to the use or nonuse of EMRs among rural hospitals that we are unable to detect with this analysis. Although the strength of our analysis is that we represent a national sample of children, these more detailed hospital factors will be important for future research. Lastly, our study uses the most recent HCUP KID patient-level data but represents only the year the HITECH Act was enacted. Although the HITECH Act may broaden or narrow this divide over time, our analysis emphasizes the effect of hospital-level factors on access to care with advanced EMRs for children in the rural setting. Repeating this analysis when new datasets are available will shed light on the effects of the HITECH Act.
Children cared for in the inpatient rural environment are half as likely to receive care with advanced-stage EMRs. When policies are developed to increase rates of EMR adoption, it will be critical to consider the unique barriers rural hospitals face in regard to adopting EMR to ensure that the highest quality of care is provided to all hospitalized children, regardless of location.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- All Patient Refined Diagnosis Related Groups
- electronic medical record
- Healthcare Information Technology for Economic and Clinical Health
- HCUP KID
- Healthcare Cost and Utilization Project Kids Inpatient Dataset
- Healthcare Information and Management Systems Society
- 1.↵Institute of Medicine (U.S.). Committee on Quality of Health Care in America. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press; 2001.
- 2.↵United States Census Bureau. 2010. Census Urban and Rural Classification and Urban Area Criteria.
- King WJ,
- Paice N,
- Rangrej J,
- Forestell GJ,
- Swartz R
- Holdsworth MT,
- Fichtl RE,
- Raisch DW,
- et al
- DeNavas-Walt C,
- Proctor BD,
- Smith JC
- Nawal Lutfiyya M,
- Bhat DK,
- Gandhi SR,
- Nguyen C,
- Weidenbacher-Hoper VL,
- Lipsky MS
- Hillestad R,
- Bigelow J,
- Bower A,
- et al
- 20.↵AHRQ. HCUP Kids’ Inpatient Database (KID). Healthcare Cost and Utilization Project (HCUP). 2009. Rockville, MD: Agency for Healthcare Research and Quality. Available at: www.hcup-us.ahrq.gov/kidoverview.jsp.
- 21.↵HIMSS Analytics SM Database (formerly the Dorenfest IHDS + Database), 2009 release.
- Hughes J
- Furukawa MF
- 27.↵Healthcare Cost and Utilization Project. https://www.hcup-us.ahrq.gov/db/vars/hosp_locteach/kidnote.jsp. Accessed February 4, 2014.
- Jha AK,
- DesRoches CM,
- Shields AE,
- et al
- Diana ML,
- Kazley AS,
- Ford EW,
- Menachemi N
- 34.↵Centers for Medicare and Medicaid Services. www.cms.gov/Center/Provider-Type/Critical-Access-Hospitals-Center.html. Accessed February 4, 2014.
- Copyright © 2014 by the American Academy of Pediatrics