This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Medical Informatics, is properly cited. The complete bibliographic information, a link to the original publication on http://medinform.jmir.org/, as well as this copyright and license information must be included.
Pressures to contain health care costs, personalize patient care, use big data, and to enhance health care quality have highlighted the need for integration of evidence at the point of care. The application of evidence-based medicine (EBM) has great promise in the era of electronic health records (EHRs) and health technology. The most successful integration of evidence into EHRs has been complex decision tools that trigger at a critical point of the clinical visit and include patient specific recommendations. The objective of this viewpoint paper is to investigate why the incorporation of complex CDS tools into the EMR is equally complex and continues to challenge health service researchers and implementation scientists. Poor adoption and sustainability of EBM guidelines and CDS tools at the point of care have persisted and continue to document low rates of usage. The barriers cited by physicians include efficiency, perception of usefulness, information content, user interface, and over-triggering. Building on the traditional EHR implementation frameworks, we review keys strategies for successful CDSs: (1) the quality of the evidence, (2) the potential to reduce unnecessary care, (3) ease of integrating evidence at the point of care, (4) the evidence’s consistency with clinician perceptions and preferences, (5) incorporating bundled sets or automated documentation, and (6) shared decision making tools. As EHRs become commonplace and insurers demand higher quality and evidence-based care, better methods for integrating evidence into everyday care are warranted. We have outlined basic criteria that should be considered before attempting to integrate evidenced-based decision support tools into the EHR.
Pressures to contain health care costs, personalize patient care, use of big data, and enhance health care quality have highlighted the need for integration of evidence at the point of care [
With the onset of Health Information Technology for Economic and Clinical Health Act and Meaningful Use initiatives in 2009, researchers have been hopeful that health technology will be the solution to bringing EBM to the point of care. Substantial investments of funding, intellect, and energy have yielded an array of EHRs and electronic clinical decision support (CDS) tools to improve patients’ quality of care and reduce inappropriate use of critical resources.
The most successful integration of evidence into EHRs has been complex decision tools that trigger at a critical point of the clinical visit and include patient specific recommendations. In contrast, most of the CDS tools being launched are uni-dimensional and not incorporated into the physicians’ workflow. For the purpose of this article, we have designated these forms of evidence integration as “flat reminders”: one-dimensional alerts that are typically triggered by one or two EHR components such as an element of patient history [
Complex, multidimensional forms of CDS are patient-specific, provide specific recommendations for rapid frontline decision making, and therefore have had a greater impact on patient outcomes and resource utilization. CPRs are forms of complex CDS. Based on real-time patient data points such as medical history, physical examination, and laboratory data, CPRs are EBM based algorithms that are able to personalize the patient’s diagnosis, prognosis, and likely response to treatment [
Incorporating complex CDS tools, such as CPRs, into an EHR holds great promise for finally realizing their potential by standardizing their application, reinforcing their application, and documentation. The caveat is that incorporation of complex CDS tools into the EMR is equally complex and continues to challenge health service researchers and implementation scientists. Poor adoption and sustainability of CDS tools at the point of care has persisted and continues to have low rates of usage [
Over the past five years, our research team has been working to improve the integration and adoption of complex CDS tools. Similar to the EBM cycle, we see CDS integration as a step wise process of: identifying a clinical problem, reviewing the evidence, usability testing of the tool, integration and deployment of the tool into the EHR, incorporation of shared decision making, and continuous monitoring and maintenance for sustained effectiveness (
Five steps of evidence-based practice. Evidence-based medicine: EBM.
Flat versus dynamic clinical decision support tools. PCP: primary care provider.
Steps to complex clinical decision support tool integration. CPRs: clinical prediction rules.
During our research, we have encountered challenges that have repeatedly emerged. In this paper, we propose strategies to overcome those challenges to the integration of complex CDS in order to improve EHR-embedded CDS tools adoption rates, patient outcomes, and resource utilization [
Building on the traditional EHR implementation frameworks, we review keys strategies for successful CDSs: (1) the quality of the evidence, (2) the potential to reduce unnecessary care, (3) ease of integrating evidence at the point of care, (4) the evidence’s consistency with clinician perceptions and preferences, (5) incorporating bundled sets or automated documentation, and (6) shared decision-making tools.
The first consideration to successful adoption of CDS tools is assessing the quality of the evidence. This may seem an obvious step, but this critical element is often overlooked, with inaccurate assumptions about evidence and impeding the hoped-for results. Therefore, a formal process of evaluating and grading the evidence of the CDS prior to integration is critical. In order for the CDS tool to have a significant impact on health care outcomes, it must be based on high-quality evidence [
The quality of CPRs is determined by how well they have been validated and tested. CPRs are typically developed in a three-step process: (1) derivation of the rule and creation of a model, usually with a retrospective database; (2) validation of the rule, in which the model is tested, preferably in a prospective fashion in several different sites to demonstrate that it is transportable and stable; and (3) impact analysis, when the rule’s impact on clinical behavior is assessed (
Several risk-stratification tools with poor and unclear levels of evidence are currently in wide use, for example, the Modified Early Warning Scoring and the pulmonary embolism rule criteria rule for pulmonary embolism. The rules have been derived, but haven’t shown consistency in prospective validations performed in various clinical settings [
Development and testing of a clinical prediction rule.
A second consideration we identified as critical is the potential for the evidence to have a significant effect on health care delivery. Historically, CDS and CPR tools have often been introduced in clinical areas plagued by overuse of diagnostic tests or treatments. CPRs aim to accurately identify patients at very low risk who can possibly forgo further testing and those at high risk who can be prioritized for further diagnostic tests or immediate treatment. If the goal of evidence integration is to reduce unnecessary testing or treatment in low risk populations, then estimating how many patients fall into the low risk category will help give an accurate measure of the potential impact of a prediction rule. Our experience has been that estimating this risk will allow you to weigh the potential impact to the work/resources need to build and implement an EHR based CDS tool.
For example, CPRs that guide clinicians through the complex process of risk stratification usually shift the distribution of patients from higher to lower risk. With a CPR such as the strep or pneumonia rule, shifting patients from medium to low risk could reduce orders for antibiotics, which are recommended for patients with medium or high risk; if this constitutes a large proportion of patients, the CPR will have a substantial impact on public health implications (antibiotic resistance) and reduce unnecessary usage of antibiotics.
A key to both clinician adoption of CDS tools at the point of care and successful integration is how it easy it is to meld the CPR into workflow and how the patient specific data are entered into the tool. Some CDS tools are extremely complex, requiring multiple data points that may not be automatically integrated into the CDS tool and thus require manual entry. In some practice settings, the EHR may not automatically interface with required data such as x-ray results in the emergency departments or rapid point-of-care test results in primary care clinics. Busy clinicians are unlikely to adopt tools that require them to manually derive, obtain, or enter data. Examples of overly complex tools are the Pneumonia Severity Index (PSI) used in emergency departments to help providers decide to admit patients. The rule has over 20 data elements [
A related issue to ease of integration is “triggering” of the tool. To be truly effective, CDS tools need to be an active (automatic) trigger and seamlessly integrated into the flow of care. Providers should not have to activate the decision support, but rather be automatically offered it in the appropriate setting and related to the appropriate patient. Certain phrases or orders and combinations can act as trigger points in the EMR, such as chief complaint or diagnosis. For a trigger to be successful, it needs to trigger accurately (when truly needed) and not be overly sensitive. In our study on using a CPR for pneumonia, entering cough in the chief complaint section was one method of automatically triggering the complex decision support tool. Ideally, decision support is triggered infrequently and is targeted to the specific condition where it can most assist the provider. If there is no method for accurate triggering, the decision tool may not be effective in changing clinicians’ behavior.
Clinicians are most likely to adopt decision support tools or act on evidence guidelines that align with their predispositions about care. Literature suggests providers’ understand the value of CPRs and state they utilize them in decision making, but CPR adoption rates continue to be low and vary across CPRs [
CPRs that can stratify risk and have a corresponding management plan or diagnostic testing, which can be streamlined and bundled into order sets, will likely have more buy in by physicians, leading to higher usage and therefore larger impact on patient outcomes. The largest incentive we have witnessed through our usability testing is how the CPR and CDS tools can streamline clinical practice instead of impeding and slowing it. By incorporating order sets or automated documentation in progress notes of the EMR and automated documentation in progress notes of the EMR, physicians see the CDS as a facilitator rather than a burden. Therefore, we work to develop CDS tools that offer some incentive to using the tool. In our models, we embedded patient education material in both English and Spanish for patients to take home [
The final piece of completing the evidence cycle, which has yet to be sufficiently studied, is the integration of shared decision making when it’s appropriate and as long as it’s based on the best available evidence. Shared decision making (SDM) is becoming an integral part of patient centered care and is seen as a method to improve patient-clinician communication [
As EHRs become commonplace and insurers demand higher quality and evidence-based care, better methods for integrating evidence into everyday care are warranted. We have outlined basic criteria that should be considered before attempting to integrate evidenced-based decision support tools into the EHR. First and foremost, this process emphasizes a critical appraisal of the quality of the evidence behind the decision support. Second, CDS tools should be evaluated for their ability to perform and impact clinical care through assessments of providers’ perception of utility. Finally, usability testing and integration into workflow need to be thoroughly evaluated prior to attempts to integrate evidence. Evaluation of the evidence and usability testing, however, are often lacking in research design, implementation methodology, and training of researchers in this area. If the federal government, EHR vendors, or health care institutions do not support research in these areas, the integration of successful CDS tools will continue to lag in creating change in patient outcomes. At this critical juncture of widespread EHRs and pressure to bend the cost curve, incentives to help industry, government, and academic health centers to support these research areas is urgent.
clinical decision support
clinical prediction rules
evidence-based medicine
electronic health records
Ottawa Ankle Rule
Pneumonia Severity Index
shared decision making
Thrombolysis in Myocardial Infarction
None declared.