Broad enthusiasm exists among hospitals for participation in Meaningful Use. However, many hospitals have a long road ahead to implement the advanced systems required for the program.
Objectives:
To update the status of electronic health record (EHR) adoption in US hospitals and assess their readiness for “Meaningful Use” (MU).
Study Design:
We used data from the 2010 American Hospital Association Annual Information Technology Survey. The survey was first conducted in 2007 and is made available both online and through the mail to all non-federal acute-care hospitals in the United States.
Methods:
We measure the percentages of applicable hospitals that have adopted “basic” and “comprehensive” EHRs as defined in previous literature. Additionally, we report the percentage of hospitals planning to apply for MU in the near term, and assess hospitals’ readiness for the program and how readiness varies by key characteristics.
Results: We received responses from 2902 hospitals (64% of all non-federal acute-care hospitals). More than 15% have adopted at least a “basic” EHR, representing nearly 75% growth since 2008. Approximately two-thirds plan to apply for MU before 2013; however, only 4.4% had implemented each of the “core” MU functionalities we measured. Hospitals closer to achieving MU are more likely to be larger non-profits (P <.001) and vary by other key characteristics. Certain functionalities included in MU, such as computerized provider order entry, electronic generation of quality measures, and electronic access to records for patients are proving more challenging to implement for all hospitals.
Conclusions:
Broad enthusiasm exists among hospitals for participation in MU. However, adoption will have to accelerate above its current pace for readiness to match intention. Gaps in adoption show bringing all hospitals along is the key policy challenge.
(Am J Manag Care. 2011;17(12 Spec No.):SP117-SP124)
The data show adoption of electronic health records among hospitals has increased steadily in recent years. In addition, the vast majority of hospitals intend to apply for “Meaningful Use” (MU) in the near term, although challenges still remain:
Over the past 2 years, the federal government has initiated an unprecedented effort to reform the healthcare system and promote the “Meaningful Use” (MU) of health information technology (HIT). Federal standards have been developed to define MU (the criteria for receiving financial incentives under Health Information Technology for Economic and Clinical Health [HITECH] Act) and to encourage providers to implement electronic health records (EHRs). Regular assessments of the readiness of the nation’s hospitals to achieve MU are critical, given that the incentives are front-loaded: hospitals that are slow to adopt and meaningfully use EHRs will receive considerably smaller amounts of incentives (with future federal subsidies highly unlikely). Further, Congress has authorized penalties for providers who are not meaningful users before 2015, raising the ante for hospitals and other providers to seriously consider the adoption and use of these systems.
The early data on EHR adoption among US hospitals have been sobering. In 2008, approximately 9% of US hospitals had a basic EHR,1 a number that grew by just 3% in 2009.2 However, despite the ongoing policy interest in this area, the most recent published data on adoption come from a survey administered well before the delineation of the MU rules, and the questions within it were not well aligned with the MU criteria. In 2010, we modified the national hospital EHR survey administered by the American Hospital Association to ask a series of specific questions pertaining to MU.
Using the 2010 survey data, which was administered during the time that the MU rules were being finalized, we sought to answer 3 questions: (1) What proportion of US hospitals have a basic or comprehensive EHR or could meet MU in 2010, the last year prior to the onset of incentives? (2) How many hospitals intend to apply for MU and how realistic is it that they are likely to be able to meet the standards? and (3) What are the major barriers holding hospitals back from adoption and from meaningfully using EHRs?
METHODS
We used data from the American Hospital Association annual surveys of HIT adoption from 2008, 2009, and 2010. The approach to the 2008 and 2009 surveys has been described previously and our approach in 2010 was very similar. 1,2 The American Hospital Association sent the survey, as a supplement to its annual survey, to the chief executive officer (CEO) of each short-term acute-care hospital in the United States, asking the CEO to assign the survey to the most knowledgeable member of his or her staff (typically the chief information officer or equivalent). All non-respondents received multiple mailings and phone calls in follow-up to try to achieve a high response rate. The 2010 survey was in the field between May and October 2010. Of note, 60% of our respondents completed the survey prior to the publication of the final rule defining MU, meaning that they were reporting their intention to apply for MU based on the proposed rule and their view of how the final rule might differ from it. Respondents were given the option of responding online or by mail; as with the 2 previous surveys, the vast majority completed their responses online.
Survey Content
We updated the 2010 survey instrument to better assess hospital readiness and plans to meet the criteria for MU.3 Under the MU regulation, non-federal acute-care hospitals are eligible for Medicare and Medicaid payment incentives when they meet specified goals for information technology use.4 To qualify for Stage 1 of the program, hospitals must meet 14 “core” requirements and no fewer than 5 of 10 capabilities chosen from a “menu.” We updated the instrument to incorporate many of the additional functionalities included in the MU criteria. We also measured whether hospitals planned to apply for the program, and, if so, in what year. The questions and response categories used are included in the eAppendices A-C.
Measures of EHR Use
We used definitions of a “basic” and “comprehensive” EHR that were previously developed by a consensus panel of HIT experts.1 A “basic” EHR is defined as the full implementation in at least 1 clinical unit of computerized system(s) for patient demographics, physician notes, nursing assessments, patient problem and medication lists, laboratory and radiologic reports, diagnostic test results, and order entry for medications. A “comprehensive” EHR must include all the functionalities of a “basic” system and 14 additional functionalities.1 In addition, a “comprehensive” record requires full implementation in all clinical units. Although many functions included in these definitions are requirements of MU, others are not. Thus, these measures do not fully capture hospital readiness for the incentive program. Therefore, we constructed an additional measure to assess hospital adoption of “core” criteria. Of the 14 “core” criteria that apply to hospitals, we had reasonable proxies for 12 (eAppendices A-C). The survey also did not include items measuring possession of an EHR that has been certified through a new federal process or use of specific standards for collecting and transmitting data, which are both requirements to be eligible for incentives.
Because the MU criteria require that electronic functions support the care of a certain percentage of patients, we considered a positive response of either “full implementation in at least 1 clinical unit” or “full implementation in all clinical units” as achieving the MU criteria. The standard of implementing in “at least 1 unit” is too generous for the bar set by MU, while requiring implementation in “all units” is too restrictive. Therefore, we present the proportion that had met all 12 criteria in “at least a single unit” as our MU upper- bound estimate and, in an attached appendix, show the proportion that met all 12 criteria in “all units” as our lowerbound estimate.
Statistical Analysis
Consistent with previous years, we compared the characteristics of non-federal acute-care hospitals responding to the survey to all hospitals in that category. Not surprisingly, we found modest but statistically significant differences (Table 1). As a result, we weighted all results for potential non-response using commonly used methods.5
We calculated the proportion of US hospitals that had adopted “basic” and “comprehensive” EHRs as of 2010 and plotted the trends in these proportions from 2008 to 2010. We used the same approach to calculate the proportion of US hospitals that had adopted all 12 of the “core” functionalities of MU that we had been able to capture on the survey. To assess how far along other hospitals were, we also calculated the proportion that had adopted 9 to 11 of the core functions, those that adopted 5 to 8, those that had adopted 1 to 4, and those that had adopted none of the 12 functions.
To assess hospitals’ future plans to apply for MU, we examined the proportion that reported planning to apply in 2011 or 2012, those planning to apply in 2013 or later, and those unsure about whether or when they might apply. We examined the responses to this question both overall, as well as stratified by the number of core functions available.
Finally, we categorized the cohort of US hospitals into 2 groups: (1) those that were further along and may be able to meet MU in 2011 or 2012, and (2) those that were highly unlikely to meet MU in the first 2 years of the program. We defined those hospitals further along who had 9 or more of the “core” functions fully implemented in at least 1 unit and refer to them as “closer to MU” hospitals and those with fewer than 9 functions “farther from MU.” For these 2 cohorts of hospitals, we assessed the functionalities they were still missing and the barriers they identified to meeting MU.
RESULTS
The American Hospital Association received responses from 3635 hospitals. After excluding federal hospitals, those located outside of the 50 states and the District of Columbia, and non-medical or surgical institutions, 2902 hospitals remained for analysis (64% of non-federal acute-care hospitals). This was comparable to the response rates in 2008 (63%) and 2009 (69%). Because there were modest but significant differences between responders and the entire census of US hospitals (Table 1), as described above, all results were statistically adjusted for potential non-response.
Changes in Adoption of EHRs in US Hospitals
We found that in 2010, the proportions of non-federal acute-care hospitals reporting the presence of “basic” and “comprehensive” EHR systems continued to increase compared with previous years. In 2010, 3.6% of non-federal acutecare US hospitals had a comprehensive EHR, compared with 2.7% in 2009 and 1.5% in 2008. The proportion of hospitals reporting just a “basic” system was 11.5%, compared with 9.2% in 2009 and 7.2% in 2008 (Figure 1). Consistent with the prior studies,1,2 characteristics of hospitals associated with a greater likelihood of having at least a basic EHR included larger size, being a non-profit or public (as opposed to for-profit), being a major or minor teaching hospital, being a member of a system, or being located in an urban area (P <.001 for all comparisons; see Table 1).
Hospital Readiness for MU
Figure 2
eAppendices A-C
Table 2
We found only 4.4% of eligible US hospitals had each of the 12 available “core” functionalities of MU fully implemented in at least 1 unit (). This declined to 2.8% using “full implementation in all clinical units” as the standard (). An additional 42.6% of hospitals had 9 or more of the functions in place in at least 1 unit (see last column, ). Nearly 25% of hospitals had less than a third of the 12 “core” functions implemented in even a single unit of the hospital.
We found that 66% of eligible US hospitals planned to apply for federal MU incentives in 2011 or 2012 (Table 2, last row). More than 3 in 4 of these hospitals reported they intended to apply for incentives administered both through Medicare and the Medicaid programs while 22% intended to apply for the Medicare incentives only. Of hospitals that were intending to apply in 2011 or 2012, 5.7% had all 12 core functions implemented in at least 1 unit while an additional 48.2% had 9 to 11 core functions implemented in at least 1 unit (Table 2, second column). Hospitals intending to apply in 2011 or 2012 were further along in their adoption of the MU functions than those intending to apply in 2013 or those unsure about applying for MU.
Characteristics of Hospitals With More MU Functions
Table 3
Hospitals that had adopted 9 or more “core” functions were,compared with hospitals who had adopted fewer core functions, larger, located in the Northeast, and less often forprofit (). They were also more likely to be major teaching hospitals (30% vs 18%, P<.001), members of hospital systems (60% vs 53%, P <.001), and with coronary care units (38% vs 26%, P <.001) than other hospitals. These hospitals with more MU functions in place were less likely to be located in a rural part of the country (37% vs 51%, P <.001) than other hospitals.
Functions Missing From More Advanced Hospitals
Table 4
Of the hospitals with 9 or more core functions in place, or “closer to MU” hospitals, 32% could automatically generate numerators and denominators for quality measurement, 56% were able to electronically provide patients access to their own records, 59% had implemented computerized provider order entry (CPOE) in at least 1 unit, and 79% had electronic problem lists. The “farther from MU” hospitals, with 8 or fewer core functions, were missing these key functions far more frequently (see ).
“Closer to MU” hospitals differed from the “farther from MU” hospitals in the difficulties they identified to meeting MU. The “closer to MU” hospitals were more likely to report the following as challenges to achieving MU: generating problem lists (16% vs 10%) and automated quality measures (31% vs 20%). Conversely, they were less likely to report CPOE (44% vs 55%) or clinical decision support implantation as a challenge (21% vs 28%; see Table 4).
DISCUSSION
We found modest but significant increases in EHR adoption among non-federal acute-care hospitals in 2010, with just over 15% of acute-care hospitals in the United States having at least a basic EHR. The increase between 2009 and 2010 was similar, when the adoption rate of a basic EHR increased from 9% to 12%. While most hospitals reported that they intend to apply for MU, only 4.4% of US hospitals could meet all 12 of the 14 “core” measures, pointing to the difficult road ahead. While a substantial minority of hospitals had 9 to 11 of the “core” measures implemented, many lacked CPOE, arguably the most difficult technology implementation challenge.6,7 Finally, a sizable minority of hospitals had very few of the functions needed. Small hospitals, those located in the South, and for-profit institutions lagged behind others in their adoption of functions needed to meet MU. It is unclear why there are differences by region or ownership and understanding the factors that underlie these differences will be helpful in ensuring equitable adoption.
These data, the last set before incentives begin to be distributed, represent the national baseline for assessing the impact of HITECH. They suggest that US hospitals continue to be on a slow but incremental path toward adopting EHRs. The challenges they face are real: EHR systems are expensive, often disruptive to implement, and there can be significant pushback from physicians and other healthcare providers. However, the benefits of these EHR systems are considerable: they can dramatically reduce medication errors, increase adherence to evidence-based care processes, and, if implemented well, can lead to overall gains in patient outcomes.8-10 The current reimbursement system, which primarily rewards hospitals for volume over quality, creates little financial incentive for hospitals to adopt these systems. Whether the MU regulations can overcome these barriers by providing direct incentives to providers for adopting and meaningfully using EHRs needs to be monitored closely.
Our findings of high level of enthusiasm, tempered with low levels of adoption of the core functions, underscore these sets of challenges for US hospitals. Although most providers report that they wish to become meaningful users of EHRs, the data show it will be challenging for hospitals to meet the timelines of MU at the current rate of adoption. Even with a potential delay of Stage 2 of MU, whether all the hospitals that report a desire to apply for MU will be able to achieve it before 2014 is unclear.
In an attempt to accelerate the adoption rate, the HITECH Act makes a series of aids available (beyond the incentives) to help providers adopt EHRs. These include Regional Extension Centers—which are funded to provide technical assistance to vulnerable providers—and modest support for small rural hospitals that often lack the technical knowledge to effectively implement these systems. There are also statebased health information exchange efforts aimed to promote broad-based clinical data exchange among providers. This should increase the value of EHRs by allowing clinicians to obtain more complete data on their patients and communicate efficiently with other providers.
Efforts to reform the delivery system will depend critically on HIT to facilitate and coordinate care. The Centers for Medicare & Medicaid Services (CMS) recently announced the final rules around Accountable Care Organizations (ACOs) and while it did not explicitly require that hospitals within ACOs be meaningful users of EHRs, these ACOs will have a far more difficult time managing the care of their patient population without broad-based EHR adoption. The CMS is required by the Affordable Care Act to reduce payments for certain types of readmissions. It will be significantly more difficult to manage discharges and prevent readmissions without electronic clinical data sharing withambulatory providers. Finally, CMS recently announced their value-based purchasing program. While the efforts in the first couple of years are very modest (only putting at risk 1% of the total payments with estimates that most hospitals should get a majority of those bonuses), future value-based purchasing programs that require more robust performance across a wide variety of metrics are likely to spur many hospitalsto adopt EHRs. If the Medicare and Medicaid EHR incentive programs do not adequately support adoption of EHRs, it will make these other reform efforts substantially more difficult to achieve.
There are limitations to our data. Although 64% of hospitals responded to the survey and we adjusted our findings to account for differences between respondents and nonrespondents, non-response bias could still have affected our results. We suspect, however, that non-responders were likely to be those hospitals with fewer HIT systems, therefore leading to an overestimate of the EHR capability of US hospitals. Second, although we tried to estimate how many hospitals would have met the MU requirements in 2010, for the small number of hospitals who wish to receive Medicaid incentives only, they do not need, in the first year, to meet MU. They just have to show that they are in the process of adopting, implementing, and upgrading to a certified EHR. Third, we used cut-points to assess hospitals’ progress toward meeting MU. However, these cut-points do not account for the fact that some functions are easier to adopt than others. Although we tried to address this issue by examining which functions hospitals were missing, even among subgroups (ie, hospitals with 5-8 core functions) there is surely some heterogeneity in the challenges hospitals will face to meet MU. Finally, our measures of MU functions do not map exactly to the full set of criteria: the requirements for meeting MU are more numerous and specific than what can be determined from a survey.
Taken as a whole, our findings demonstrate that there has been a significant but very modest uptick in EHR adoption by hospitals. While two-thirds of hospitals indicate they plan to achieve MU in the near term, the data also show only a small minority have the advanced systems needed to get there. How many hospitals achieve MU and when they achieve it will be influenced by a series of factors, including the commitment of the institutions, the readiness of the technology, and the effectiveness of federal programs and related incentives. Over the long term, as more data emerge about the quality and efficiency gains that are possible with EHRs, we expect that hospitals will follow through on plans to implement systems that benefit their patients and their bottom line.
Acknowledgment
The survey is conducted in part through a grant from the Office of the National Coordinator for Health Information Technology; however, no funding was provided for this analysis.
Author Affiliations: From Harvard School of Public Health (AKJ), VA Boston Healthcare System (AKJ), Boston, MA; Department of Health and Human Services (MFB, MBB), Office of the National Coordinator for Health Information Technology, Washington, DC; Mathematica Policy Research (CD), Cambridge MA; American Hospital Association (MSJ, PDK), Chicago, IL; Massachusetts General Hospital (EGC), Boston, MA.
Funding Source: None.
Author Disclosures: Dr Jha reports receiving grants from the Department of Health and Human Services. The other authors (MFB, CD, MSJ, PDK, EGC, MBB) report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Authorship Information: Concept and design (AKJ, CD, MSJ, PDK, EGC, MBB); acquisition of data (AKJ, MFB, MSJ, PDK, MBB); analysis and interpretation of data (AKJ, MFB, CD, PDK, MBB); drafting of the manuscript (AKJ, MFB, MSJ, PDK, MBB); critical revision of the manuscript for important intellectual content (CD, MSJ, PDK, EGC, MBB); statistical analysis (MFB, PDK, MBB); obtaining funding (PDK, EGC, MBB); administrative, technical, or logistic support (AKJ, MFB, PDK); supervision (AKJ, PDK, MBB), and survey expertise (EGC).
Address correspondence to: Ashish K. Jha, MD, MPH, Harvard School of Public Health, Department of Health Policy and Management, 677 Huntington Ave, Boston, MA 02115-6096. E-mail: ajha@hsph.harvard.edu.
1. Jha AK, DesRoches CM, Campbell EG, et al. Use of electronic health records in U.S. hospitals. N Engl J Med. 2009;360(16):1628-1638.
2. Jha AK, DesRoches CM, Kralovec PD, Joshi MS. A progress report on electronic health records in U.S. hospitals. Health Aff (Millwood). 2010;29(10):1951-1957.
3. US Department of Health and Human Services. Centers for Medicare & Medicaid Services. Medicare and Medicaid Programs; Electronic Health Record Incentive Program, 2010. https://www.cms.gov/EHRIncentivePrograms/. Accessed December 5, 2011.
4. Buntin MB, Jain SH, Blumenthal D. Health information technology: laying the infrastructure for national health reform. Health Aff (Millwood). 2010;29(6):1214-1219.
5. Woodbridge JM. Inverse probability weighted estimation for general missing data problems. J Econometrics. 2007;141(2):1281-1301.
6. Koppel R, Metlay JP, Cohen A, et al. Role of computerized physician order entry systems in facilitating medication errors. JAMA. 2005; 293(10):1197-1203.
7. Mollon B, Chong J Jr, Holbrook AM, Sung M, Thabane L, Foster G. Features predicting the success of computerized decision support for prescribing: a systematic review of randomized controlled trials. BMC Med Inform Decis Mak. 2009;9:11.
8. Amarasingham R, Plantinga L, Diener-West M, Gaskin DJ, Powe NR. Clinical information technologies and inpatient outcomes: a multiple hospital study. Arch Intern Med. 2009;169(2):108-114.
9. Kaushal R, Jha AK, Franz C, et al. Return on investment for a computerized physician order entry system. J Am Med Inform Assoc. 2006; 13(3):261-266.
10. Reckmann MH, Westbrook JI, Koh Y, Lo C, Day RO. Does computerized provider order entry reduce prescribing errors for hospital inpatients? a systematic review. J Am Med Inform Assoc. 2009;16(5): 613-623.
New AI Tool Identifies Undiagnosed PNH in Health Records
October 30th 2024The machine learning model shows promise in detecting paroxysmal nocturnal hemoglobinuria (PNH) by assessing electronic health records (EHR) data, potentially transforming the diagnostic landscape for rare diseases.
Read More
Overhauling Quality Measurement in the US: Measure What Matters
October 30th 2024As the US charts its course through the next political era, it is crucial that we boldly allocate resources and prioritize what truly impacts patients. When faced with complexity, feasibility concerns, or entrenched norms, we must proclaim: “It’s the outcomes, stupid.”
Read More
No Free Lunch: The Misaligned Incentives of the American Health Care System
October 30th 2024The author highlights reasons why we have not seen substantial cost savings in the health care industry and why future efforts are likely to continue to see forceful pushback, as well as offers potential solutions.
Read More